diff --git a/cbits/README b/cbits/README new file mode 100644 --- /dev/null +++ b/cbits/README @@ -0,0 +1,139 @@ + + Expat, Release 2.2.1 + +This is Expat, a C library for parsing XML, written by James Clark. +Expat is a stream-oriented XML parser. This means that you register +handlers with the parser before starting the parse. These handlers +are called when the parser discovers the associated structures in the +document being parsed. A start tag is an example of the kind of +structures for which you may register handlers. + +Windows users should use the expat_win32bin package, which includes +both precompiled libraries and executables, and source code for +developers. + +Expat is free software. You may copy, distribute, and modify it under +the terms of the License contained in the file COPYING distributed +with this package. This license is the same as the MIT/X Consortium +license. + +Versions of Expat that have an odd minor version (the middle number in +the release above), are development releases and should be considered +as beta software. Releases with even minor version numbers are +intended to be production grade software. + +If you are building Expat from a check-out from the CVS repository, +you need to run a script that generates the configure script using the +GNU autoconf and libtool tools. To do this, you need to have +autoconf 2.58 or newer. Run the script like this: + + ./buildconf.sh + +Once this has been done, follow the same instructions as for building +from a source distribution. + +To build Expat from a source distribution, you first run the +configuration shell script in the top level distribution directory: + + ./configure + +There are many options which you may provide to configure (which you +can discover by running configure with the --help option). But the +one of most interest is the one that sets the installation directory. +By default, the configure script will set things up to install +libexpat into /usr/local/lib, expat.h into /usr/local/include, and +xmlwf into /usr/local/bin. If, for example, you'd prefer to install +into /home/me/mystuff/lib, /home/me/mystuff/include, and +/home/me/mystuff/bin, you can tell configure about that with: + + ./configure --prefix=/home/me/mystuff + +Another interesting option is to enable 64-bit integer support for +line and column numbers and the over-all byte index: + + ./configure CPPFLAGS=-DXML_LARGE_SIZE + +However, such a modification would be a breaking change to the ABI +and is therefore not recommended for general use - e.g. as part of +a Linux distribution - but rather for builds with special requirements. + +After running the configure script, the "make" command will build +things and "make install" will install things into their proper +location. Have a look at the "Makefile" to learn about additional +"make" options. Note that you need to have write permission into +the directories into which things will be installed. + +If you are interested in building Expat to provide document +information in UTF-16 encoding rather than the default UTF-8, follow +these instructions (after having run "make distclean"): + + 1. For UTF-16 output as unsigned short (and version/error + strings as char), run: + + ./configure CPPFLAGS=-DXML_UNICODE + + For UTF-16 output as wchar_t (incl. version/error strings), + run: + + ./configure CFLAGS="-g -O2 -fshort-wchar" \ + CPPFLAGS=-DXML_UNICODE_WCHAR_T + + 2. Edit the MakeFile, changing: + + LIBRARY = libexpat.la + + to: + + LIBRARY = libexpatw.la + + (Note the additional "w" in the library name.) + + 3. Run "make buildlib" (which builds the library only). + Or, to save step 2, run "make buildlib LIBRARY=libexpatw.la". + + 4. Run "make installlib" (which installs the library only). + Or, if step 2 was omitted, run "make installlib LIBRARY=libexpatw.la". + +Using DESTDIR or INSTALL_ROOT is enabled, with INSTALL_ROOT being the default +value for DESTDIR, and the rest of the make file using only DESTDIR. +It works as follows: + $ make install DESTDIR=/path/to/image +overrides the in-makefile set DESTDIR, while both + $ INSTALL_ROOT=/path/to/image make install + $ make install INSTALL_ROOT=/path/to/image +use DESTDIR=$(INSTALL_ROOT), even if DESTDIR eventually is defined in the +environment, because variable-setting priority is +1) commandline +2) in-makefile +3) environment + +Note: This only applies to the Expat library itself, building UTF-16 versions +of xmlwf and the tests is currently not supported. + +Note for Solaris users: The "ar" command is usually located in +"/usr/ccs/bin", which is not in the default PATH. You will need to +add this to your path for the "make" command, and probably also switch +to GNU make (the "make" found in /usr/ccs/bin does not seem to work +properly -- apparently it does not understand .PHONY directives). If +you're using ksh or bash, use this command to build: + + PATH=/usr/ccs/bin:$PATH make + +When using Expat with a project using autoconf for configuration, you +can use the probing macro in conftools/expat.m4 to determine how to +include Expat. See the comments at the top of that file for more +information. + +A reference manual is available in the file doc/reference.html in this +distribution. + +The homepage for this project is http://www.libexpat.org/. There +are links there to connect you to the bug reports page. If you need +to report a bug when you don't have access to a browser, you may also +send a bug report by email to expat-bugs@mail.libexpat.org. + +Discussion related to the direction of future expat development takes +place on expat-discuss@mail.libexpat.org. Archives of this list and +other Expat-related lists may be found at: + + http://mail.libexpat.org/mailman/listinfo/ diff --git a/cbits/amigaconfig.h b/cbits/amigaconfig.h deleted file mode 100644 --- a/cbits/amigaconfig.h +++ /dev/null @@ -1,32 +0,0 @@ -#ifndef AMIGACONFIG_H -#define AMIGACONFIG_H - -/* 1234 = LIL_ENDIAN, 4321 = BIGENDIAN */ -#define BYTEORDER 4321 - -/* Define to 1 if you have the `bcopy' function. */ -#define HAVE_BCOPY 1 - -/* Define to 1 if you have the header file. */ -#undef HAVE_CHECK_H - -/* Define to 1 if you have the `memmove' function. */ -#define HAVE_MEMMOVE 1 - -/* Define to 1 if you have the header file. */ -#define HAVE_UNISTD_H 1 - -/* whether byteorder is bigendian */ -#define WORDS_BIGENDIAN - -/* Define to specify how much context to retain around the current parse - point. */ -#define XML_CONTEXT_BYTES 1024 - -/* Define to make parameter entity parsing functionality available. */ -#define XML_DTD - -/* Define to make XML Namespaces functionality available. */ -#define XML_NS - -#endif /* AMIGACONFIG_H */ diff --git a/cbits/expat.h b/cbits/expat.h --- a/cbits/expat.h +++ b/cbits/expat.h @@ -95,7 +95,9 @@ /* Added in 2.0. */ XML_ERROR_RESERVED_PREFIX_XML, XML_ERROR_RESERVED_PREFIX_XMLNS, - XML_ERROR_RESERVED_NAMESPACE_URI + XML_ERROR_RESERVED_NAMESPACE_URI, + /* Added in 2.2.1. */ + XML_ERROR_INVALID_ARGUMENT }; enum XML_Content_Type { @@ -342,7 +344,7 @@ XML_EntityDeclHandler handler); /* OBSOLETE -- OBSOLETE -- OBSOLETE - This handler has been superceded by the EntityDeclHandler above. + This handler has been superseded by the EntityDeclHandler above. It is provided here for backward compatibility. This is called for a declaration of an unparsed (NDATA) entity. @@ -706,6 +708,7 @@ be called, despite an external subset being parsed. Note: If XML_DTD is not defined when Expat is compiled, returns XML_ERROR_FEATURE_REQUIRES_XML_DTD. + Note: If parser == NULL, returns XML_ERROR_INVALID_ARGUMENT. */ XMLPARSEAPI(enum XML_Error) XML_UseForeignDTD(XML_Parser parser, XML_Bool useDTD); @@ -729,19 +732,43 @@ to the XML_StartElementHandler that were specified in the start-tag rather than defaulted. Each attribute/value pair counts as 2; thus this correspondds to an index into the atts array passed to the - XML_StartElementHandler. + XML_StartElementHandler. Returns -1 if parser == NULL. */ XMLPARSEAPI(int) XML_GetSpecifiedAttributeCount(XML_Parser parser); /* Returns the index of the ID attribute passed in the last call to - XML_StartElementHandler, or -1 if there is no ID attribute. Each - attribute/value pair counts as 2; thus this correspondds to an - index into the atts array passed to the XML_StartElementHandler. + XML_StartElementHandler, or -1 if there is no ID attribute or + parser == NULL. Each attribute/value pair counts as 2; thus this + correspondds to an index into the atts array passed to the + XML_StartElementHandler. */ XMLPARSEAPI(int) XML_GetIdAttributeIndex(XML_Parser parser); +#ifdef XML_ATTR_INFO +/* Source file byte offsets for the start and end of attribute names and values. + The value indices are exclusive of surrounding quotes; thus in a UTF-8 source + file an attribute value of "blah" will yield: + info->valueEnd - info->valueStart = 4 bytes. +*/ +typedef struct { + XML_Index nameStart; /* Offset to beginning of the attribute name. */ + XML_Index nameEnd; /* Offset after the attribute name's last byte. */ + XML_Index valueStart; /* Offset to beginning of the attribute value. */ + XML_Index valueEnd; /* Offset after the attribute value's last byte. */ +} XML_AttrInfo; + +/* Returns an array of XML_AttrInfo structures for the attribute/value pairs + passed in last call to the XML_StartElementHandler that were specified + in the start-tag rather than defaulted. Each attribute/value pair counts + as 1; thus the number of entries in the array is + XML_GetSpecifiedAttributeCount(parser) / 2. +*/ +XMLPARSEAPI(const XML_AttrInfo *) +XML_GetAttributeInfo(XML_Parser parser); +#endif + /* Parses some input. Returns XML_STATUS_ERROR if a fatal error is detected. The last call to XML_Parse must have isFinal true; len may be zero for this call (or any other). @@ -878,11 +905,22 @@ entities is requested; otherwise it will return non-zero. Note: If XML_SetParamEntityParsing is called after XML_Parse or XML_ParseBuffer, then it has no effect and will always return 0. + Note: If parser == NULL, the function will do nothing and return 0. */ XMLPARSEAPI(int) XML_SetParamEntityParsing(XML_Parser parser, enum XML_ParamEntityParsing parsing); +/* Sets the hash salt to use for internal hash calculations. + Helps in preventing DoS attacks based on predicting hash + function behavior. This must be called before parsing is started. + Returns 1 if successful, 0 when called after parsing has started. + Note: If parser == NULL, the function will do nothing and return 0. +*/ +XMLPARSEAPI(int) +XML_SetHashSalt(XML_Parser parser, + unsigned long hash_salt); + /* If XML_Parse or XML_ParseBuffer have returned XML_STATUS_ERROR, then XML_GetErrorCode returns information about the error. */ @@ -904,6 +942,10 @@ the location is the location of the character at which the error was detected; otherwise the location is the location of the last parse event, as described above. + + Note: XML_GetCurrentLineNumber and XML_GetCurrentColumnNumber + return 0 to indicate an error. + Note: XML_GetCurrentByteIndex returns -1 to indicate an error. */ XMLPARSEAPI(XML_Size) XML_GetCurrentLineNumber(XML_Parser parser); XMLPARSEAPI(XML_Size) XML_GetCurrentColumnNumber(XML_Parser parser); @@ -941,9 +983,12 @@ /* Exposing the memory handling functions used in Expat */ XMLPARSEAPI(void *) +XML_ATTR_MALLOC +XML_ATTR_ALLOC_SIZE(2) XML_MemMalloc(XML_Parser parser, size_t size); XMLPARSEAPI(void *) +XML_ATTR_ALLOC_SIZE(3) XML_MemRealloc(XML_Parser parser, void *ptr, size_t size); XMLPARSEAPI(void) @@ -984,7 +1029,8 @@ XML_FEATURE_SIZEOF_XML_CHAR, XML_FEATURE_SIZEOF_XML_LCHAR, XML_FEATURE_NS, - XML_FEATURE_LARGE_SIZE + XML_FEATURE_LARGE_SIZE, + XML_FEATURE_ATTR_INFO /* Additional features must be added to the end of this enum. */ }; @@ -998,13 +1044,11 @@ XML_GetFeatureList(void); -/* Expat follows the GNU/Linux convention of odd number minor version for - beta/development releases and even number minor version for stable - releases. Micro is bumped with each release, and set to 0 with each - change to major or minor version. +/* Expat follows the semantic versioning convention. + See http://semver.org. */ #define XML_MAJOR_VERSION 2 -#define XML_MINOR_VERSION 0 +#define XML_MINOR_VERSION 2 #define XML_MICRO_VERSION 1 #ifdef __cplusplus diff --git a/cbits/expat_external.h b/cbits/expat_external.h --- a/cbits/expat_external.h +++ b/cbits/expat_external.h @@ -65,13 +65,27 @@ #endif #endif /* not defined XML_STATIC */ +#if !defined(XMLIMPORT) && defined(__GNUC__) && (__GNUC__ >= 4) +#define XMLIMPORT __attribute__ ((visibility ("default"))) +#endif /* If we didn't define it above, define it away: */ #ifndef XMLIMPORT #define XMLIMPORT #endif +#if defined(__GNUC__) && (__GNUC__ > 2 || (__GNUC__ == 2 && __GNUC_MINOR__ >= 96)) +#define XML_ATTR_MALLOC __attribute__((__malloc__)) +#else +#define XML_ATTR_MALLOC +#endif +#if defined(__GNUC__) && ((__GNUC__ > 4) || (__GNUC__ == 4 && __GNUC_MINOR__ >= 3)) +#define XML_ATTR_ALLOC_SIZE(x) __attribute__((__alloc_size__(x))) +#else +#define XML_ATTR_ALLOC_SIZE(x) +#endif + #define XMLPARSEAPI(type) XMLIMPORT type XMLCALL #ifdef __cplusplus @@ -79,7 +93,10 @@ #endif #ifdef XML_UNICODE_WCHAR_T -#define XML_UNICODE +# define XML_UNICODE +# if defined(__SIZEOF_WCHAR_T__) && (__SIZEOF_WCHAR_T__ != 2) +# error "sizeof(wchar_t) != 2; Need -fshort-wchar for both Expat and libc" +# endif #endif #ifdef XML_UNICODE /* Information is UTF-16 encoded. */ diff --git a/cbits/internal.h b/cbits/internal.h --- a/cbits/internal.h +++ b/cbits/internal.h @@ -71,3 +71,25 @@ #define inline #endif #endif + +#ifndef UNUSED_P +# ifdef __GNUC__ +# define UNUSED_P(p) UNUSED_ ## p __attribute__((__unused__)) +# else +# define UNUSED_P(p) UNUSED_ ## p +# endif +#endif + + +#ifdef __cplusplus +extern "C" { +#endif + + +void +align_limit_to_full_utf8_characters(const char * from, const char ** fromLimRef); + + +#ifdef __cplusplus +} +#endif diff --git a/cbits/macconfig.h b/cbits/macconfig.h deleted file mode 100644 --- a/cbits/macconfig.h +++ /dev/null @@ -1,53 +0,0 @@ -/*================================================================ -** Copyright 2000, Clark Cooper -** All rights reserved. -** -** This is free software. You are permitted to copy, distribute, or modify -** it under the terms of the MIT/X license (contained in the COPYING file -** with this distribution.) -** -*/ - -#ifndef MACCONFIG_H -#define MACCONFIG_H - - -/* 1234 = LIL_ENDIAN, 4321 = BIGENDIAN */ -#define BYTEORDER 4321 - -/* Define to 1 if you have the `bcopy' function. */ -#undef HAVE_BCOPY - -/* Define to 1 if you have the `memmove' function. */ -#define HAVE_MEMMOVE - -/* Define to 1 if you have a working `mmap' system call. */ -#undef HAVE_MMAP - -/* Define to 1 if you have the header file. */ -#undef HAVE_UNISTD_H - -/* whether byteorder is bigendian */ -#define WORDS_BIGENDIAN - -/* Define to specify how much context to retain around the current parse - point. */ -#undef XML_CONTEXT_BYTES - -/* Define to make parameter entity parsing functionality available. */ -#define XML_DTD - -/* Define to make XML Namespaces functionality available. */ -#define XML_NS - -/* Define to empty if `const' does not conform to ANSI C. */ -#undef const - -/* Define to `long' if does not define. */ -#define off_t long - -/* Define to `unsigned' if does not define. */ -#undef size_t - - -#endif /* ifndef MACCONFIG_H */ diff --git a/cbits/siphash.h b/cbits/siphash.h new file mode 100644 --- /dev/null +++ b/cbits/siphash.h @@ -0,0 +1,344 @@ +/* ========================================================================== + * siphash.h - SipHash-2-4 in a single header file + * -------------------------------------------------------------------------- + * Derived by William Ahern from the reference implementation[1] published[2] + * by Jean-Philippe Aumasson and Daniel J. Berstein. Licensed in kind. + * by Jean-Philippe Aumasson and Daniel J. Berstein. + * Minimal changes by Sebastian Pipping on top, details below. + * Licensed under the CC0 Public Domain Dedication license. + * + * 1. https://www.131002.net/siphash/siphash24.c + * 2. https://www.131002.net/siphash/ + * -------------------------------------------------------------------------- + * HISTORY: + * + * 2017-06-10 (Sebastian Pipping) + * - Clarify license note in the header + * - Address C89 issues: + * - Stop using inline keyword (and let compiler decide) + * - Turn integer suffix ULL to UL + * - Replace _Bool by int + * - Turn macro siphash24 into a function + * - Address invalid conversion (void pointer) by explicit cast + * - Always expose sip24_valid (for self-tests) + * + * 2012-11-04 - Born. (William Ahern) + * -------------------------------------------------------------------------- + * USAGE: + * + * SipHash-2-4 takes as input two 64-bit words as the key, some number of + * message bytes, and outputs a 64-bit word as the message digest. This + * implementation employs two data structures: a struct sipkey for + * representing the key, and a struct siphash for representing the hash + * state. + * + * For converting a 16-byte unsigned char array to a key, use either the + * macro sip_keyof or the routine sip_tokey. The former instantiates a + * compound literal key, while the latter requires a key object as a + * parameter. + * + * unsigned char secret[16]; + * arc4random_buf(secret, sizeof secret); + * struct sipkey *key = sip_keyof(secret); + * + * For hashing a message, use either the convenience macro siphash24 or the + * routines sip24_init, sip24_update, and sip24_final. + * + * struct siphash state; + * void *msg; + * size_t len; + * uint64_t hash; + * + * sip24_init(&state, key); + * sip24_update(&state, msg, len); + * hash = sip24_final(&state); + * + * or + * + * hash = siphash24(msg, len, key); + * + * To convert the 64-bit hash value to a canonical 8-byte little-endian + * binary representation, use either the macro sip_binof or the routine + * sip_tobin. The former instantiates and returns a compound literal array, + * while the latter requires an array object as a parameter. + * -------------------------------------------------------------------------- + * NOTES: + * + * o Neither sip_keyof, sip_binof, nor siphash24 will work with compilers + * lacking compound literal support. Instead, you must use the lower-level + * interfaces which take as parameters the temporary state objects. + * + * o Uppercase macros may evaluate parameters more than once. Lowercase + * macros should not exhibit any such side effects. + * ========================================================================== + */ +#ifndef SIPHASH_H +#define SIPHASH_H + +#include /* size_t */ +#include /* uint64_t uint32_t uint8_t */ + + +#define SIP_ROTL(x, b) (uint64_t)(((x) << (b)) | ( (x) >> (64 - (b)))) + +#define SIP_U32TO8_LE(p, v) \ + (p)[0] = (uint8_t)((v) >> 0); (p)[1] = (uint8_t)((v) >> 8); \ + (p)[2] = (uint8_t)((v) >> 16); (p)[3] = (uint8_t)((v) >> 24); + +#define SIP_U64TO8_LE(p, v) \ + SIP_U32TO8_LE((p) + 0, (uint32_t)((v) >> 0)); \ + SIP_U32TO8_LE((p) + 4, (uint32_t)((v) >> 32)); + +#define SIP_U8TO64_LE(p) \ + (((uint64_t)((p)[0]) << 0) | \ + ((uint64_t)((p)[1]) << 8) | \ + ((uint64_t)((p)[2]) << 16) | \ + ((uint64_t)((p)[3]) << 24) | \ + ((uint64_t)((p)[4]) << 32) | \ + ((uint64_t)((p)[5]) << 40) | \ + ((uint64_t)((p)[6]) << 48) | \ + ((uint64_t)((p)[7]) << 56)) + + +#define SIPHASH_INITIALIZER { 0, 0, 0, 0, { 0 }, 0, 0 } + +struct siphash { + uint64_t v0, v1, v2, v3; + + unsigned char buf[8], *p; + uint64_t c; +}; /* struct siphash */ + + +#define SIP_KEYLEN 16 + +struct sipkey { + uint64_t k[2]; +}; /* struct sipkey */ + +#define sip_keyof(k) sip_tokey(&(struct sipkey){ { 0 } }, (k)) + +static struct sipkey *sip_tokey(struct sipkey *key, const void *src) { + key->k[0] = SIP_U8TO64_LE((const unsigned char *)src); + key->k[1] = SIP_U8TO64_LE((const unsigned char *)src + 8); + return key; +} /* sip_tokey() */ + + +#define sip_binof(v) sip_tobin((unsigned char[8]){ 0 }, (v)) + +static void *sip_tobin(void *dst, uint64_t u64) { + SIP_U64TO8_LE((unsigned char *)dst, u64); + return dst; +} /* sip_tobin() */ + + +static void sip_round(struct siphash *H, const int rounds) { + int i; + + for (i = 0; i < rounds; i++) { + H->v0 += H->v1; + H->v1 = SIP_ROTL(H->v1, 13); + H->v1 ^= H->v0; + H->v0 = SIP_ROTL(H->v0, 32); + + H->v2 += H->v3; + H->v3 = SIP_ROTL(H->v3, 16); + H->v3 ^= H->v2; + + H->v0 += H->v3; + H->v3 = SIP_ROTL(H->v3, 21); + H->v3 ^= H->v0; + + H->v2 += H->v1; + H->v1 = SIP_ROTL(H->v1, 17); + H->v1 ^= H->v2; + H->v2 = SIP_ROTL(H->v2, 32); + } +} /* sip_round() */ + + +static struct siphash *sip24_init(struct siphash *H, const struct sipkey *key) { + H->v0 = 0x736f6d6570736575UL ^ key->k[0]; + H->v1 = 0x646f72616e646f6dUL ^ key->k[1]; + H->v2 = 0x6c7967656e657261UL ^ key->k[0]; + H->v3 = 0x7465646279746573UL ^ key->k[1]; + + H->p = H->buf; + H->c = 0; + + return H; +} /* sip24_init() */ + + +#define sip_endof(a) (&(a)[sizeof (a) / sizeof *(a)]) + +static struct siphash *sip24_update(struct siphash *H, const void *src, size_t len) { + const unsigned char *p = (const unsigned char *)src, *pe = p + len; + uint64_t m; + + do { + while (p < pe && H->p < sip_endof(H->buf)) + *H->p++ = *p++; + + if (H->p < sip_endof(H->buf)) + break; + + m = SIP_U8TO64_LE(H->buf); + H->v3 ^= m; + sip_round(H, 2); + H->v0 ^= m; + + H->p = H->buf; + H->c += 8; + } while (p < pe); + + return H; +} /* sip24_update() */ + + +static uint64_t sip24_final(struct siphash *H) { + char left = H->p - H->buf; + uint64_t b = (H->c + left) << 56; + + switch (left) { + case 7: b |= (uint64_t)H->buf[6] << 48; + case 6: b |= (uint64_t)H->buf[5] << 40; + case 5: b |= (uint64_t)H->buf[4] << 32; + case 4: b |= (uint64_t)H->buf[3] << 24; + case 3: b |= (uint64_t)H->buf[2] << 16; + case 2: b |= (uint64_t)H->buf[1] << 8; + case 1: b |= (uint64_t)H->buf[0] << 0; + case 0: break; + } + + H->v3 ^= b; + sip_round(H, 2); + H->v0 ^= b; + H->v2 ^= 0xff; + sip_round(H, 4); + + return H->v0 ^ H->v1 ^ H->v2 ^ H->v3; +} /* sip24_final() */ + + +static uint64_t siphash24(const void *src, size_t len, const struct sipkey *key) { + struct siphash state = SIPHASH_INITIALIZER; + return sip24_final(sip24_update(sip24_init(&state, key), src, len)); +} /* siphash24() */ + + +/* + * SipHash-2-4 output with + * k = 00 01 02 ... + * and + * in = (empty string) + * in = 00 (1 byte) + * in = 00 01 (2 bytes) + * in = 00 01 02 (3 bytes) + * ... + * in = 00 01 02 ... 3e (63 bytes) + */ +static int sip24_valid(void) { + static const unsigned char vectors[64][8] = { + { 0x31, 0x0e, 0x0e, 0xdd, 0x47, 0xdb, 0x6f, 0x72, }, + { 0xfd, 0x67, 0xdc, 0x93, 0xc5, 0x39, 0xf8, 0x74, }, + { 0x5a, 0x4f, 0xa9, 0xd9, 0x09, 0x80, 0x6c, 0x0d, }, + { 0x2d, 0x7e, 0xfb, 0xd7, 0x96, 0x66, 0x67, 0x85, }, + { 0xb7, 0x87, 0x71, 0x27, 0xe0, 0x94, 0x27, 0xcf, }, + { 0x8d, 0xa6, 0x99, 0xcd, 0x64, 0x55, 0x76, 0x18, }, + { 0xce, 0xe3, 0xfe, 0x58, 0x6e, 0x46, 0xc9, 0xcb, }, + { 0x37, 0xd1, 0x01, 0x8b, 0xf5, 0x00, 0x02, 0xab, }, + { 0x62, 0x24, 0x93, 0x9a, 0x79, 0xf5, 0xf5, 0x93, }, + { 0xb0, 0xe4, 0xa9, 0x0b, 0xdf, 0x82, 0x00, 0x9e, }, + { 0xf3, 0xb9, 0xdd, 0x94, 0xc5, 0xbb, 0x5d, 0x7a, }, + { 0xa7, 0xad, 0x6b, 0x22, 0x46, 0x2f, 0xb3, 0xf4, }, + { 0xfb, 0xe5, 0x0e, 0x86, 0xbc, 0x8f, 0x1e, 0x75, }, + { 0x90, 0x3d, 0x84, 0xc0, 0x27, 0x56, 0xea, 0x14, }, + { 0xee, 0xf2, 0x7a, 0x8e, 0x90, 0xca, 0x23, 0xf7, }, + { 0xe5, 0x45, 0xbe, 0x49, 0x61, 0xca, 0x29, 0xa1, }, + { 0xdb, 0x9b, 0xc2, 0x57, 0x7f, 0xcc, 0x2a, 0x3f, }, + { 0x94, 0x47, 0xbe, 0x2c, 0xf5, 0xe9, 0x9a, 0x69, }, + { 0x9c, 0xd3, 0x8d, 0x96, 0xf0, 0xb3, 0xc1, 0x4b, }, + { 0xbd, 0x61, 0x79, 0xa7, 0x1d, 0xc9, 0x6d, 0xbb, }, + { 0x98, 0xee, 0xa2, 0x1a, 0xf2, 0x5c, 0xd6, 0xbe, }, + { 0xc7, 0x67, 0x3b, 0x2e, 0xb0, 0xcb, 0xf2, 0xd0, }, + { 0x88, 0x3e, 0xa3, 0xe3, 0x95, 0x67, 0x53, 0x93, }, + { 0xc8, 0xce, 0x5c, 0xcd, 0x8c, 0x03, 0x0c, 0xa8, }, + { 0x94, 0xaf, 0x49, 0xf6, 0xc6, 0x50, 0xad, 0xb8, }, + { 0xea, 0xb8, 0x85, 0x8a, 0xde, 0x92, 0xe1, 0xbc, }, + { 0xf3, 0x15, 0xbb, 0x5b, 0xb8, 0x35, 0xd8, 0x17, }, + { 0xad, 0xcf, 0x6b, 0x07, 0x63, 0x61, 0x2e, 0x2f, }, + { 0xa5, 0xc9, 0x1d, 0xa7, 0xac, 0xaa, 0x4d, 0xde, }, + { 0x71, 0x65, 0x95, 0x87, 0x66, 0x50, 0xa2, 0xa6, }, + { 0x28, 0xef, 0x49, 0x5c, 0x53, 0xa3, 0x87, 0xad, }, + { 0x42, 0xc3, 0x41, 0xd8, 0xfa, 0x92, 0xd8, 0x32, }, + { 0xce, 0x7c, 0xf2, 0x72, 0x2f, 0x51, 0x27, 0x71, }, + { 0xe3, 0x78, 0x59, 0xf9, 0x46, 0x23, 0xf3, 0xa7, }, + { 0x38, 0x12, 0x05, 0xbb, 0x1a, 0xb0, 0xe0, 0x12, }, + { 0xae, 0x97, 0xa1, 0x0f, 0xd4, 0x34, 0xe0, 0x15, }, + { 0xb4, 0xa3, 0x15, 0x08, 0xbe, 0xff, 0x4d, 0x31, }, + { 0x81, 0x39, 0x62, 0x29, 0xf0, 0x90, 0x79, 0x02, }, + { 0x4d, 0x0c, 0xf4, 0x9e, 0xe5, 0xd4, 0xdc, 0xca, }, + { 0x5c, 0x73, 0x33, 0x6a, 0x76, 0xd8, 0xbf, 0x9a, }, + { 0xd0, 0xa7, 0x04, 0x53, 0x6b, 0xa9, 0x3e, 0x0e, }, + { 0x92, 0x59, 0x58, 0xfc, 0xd6, 0x42, 0x0c, 0xad, }, + { 0xa9, 0x15, 0xc2, 0x9b, 0xc8, 0x06, 0x73, 0x18, }, + { 0x95, 0x2b, 0x79, 0xf3, 0xbc, 0x0a, 0xa6, 0xd4, }, + { 0xf2, 0x1d, 0xf2, 0xe4, 0x1d, 0x45, 0x35, 0xf9, }, + { 0x87, 0x57, 0x75, 0x19, 0x04, 0x8f, 0x53, 0xa9, }, + { 0x10, 0xa5, 0x6c, 0xf5, 0xdf, 0xcd, 0x9a, 0xdb, }, + { 0xeb, 0x75, 0x09, 0x5c, 0xcd, 0x98, 0x6c, 0xd0, }, + { 0x51, 0xa9, 0xcb, 0x9e, 0xcb, 0xa3, 0x12, 0xe6, }, + { 0x96, 0xaf, 0xad, 0xfc, 0x2c, 0xe6, 0x66, 0xc7, }, + { 0x72, 0xfe, 0x52, 0x97, 0x5a, 0x43, 0x64, 0xee, }, + { 0x5a, 0x16, 0x45, 0xb2, 0x76, 0xd5, 0x92, 0xa1, }, + { 0xb2, 0x74, 0xcb, 0x8e, 0xbf, 0x87, 0x87, 0x0a, }, + { 0x6f, 0x9b, 0xb4, 0x20, 0x3d, 0xe7, 0xb3, 0x81, }, + { 0xea, 0xec, 0xb2, 0xa3, 0x0b, 0x22, 0xa8, 0x7f, }, + { 0x99, 0x24, 0xa4, 0x3c, 0xc1, 0x31, 0x57, 0x24, }, + { 0xbd, 0x83, 0x8d, 0x3a, 0xaf, 0xbf, 0x8d, 0xb7, }, + { 0x0b, 0x1a, 0x2a, 0x32, 0x65, 0xd5, 0x1a, 0xea, }, + { 0x13, 0x50, 0x79, 0xa3, 0x23, 0x1c, 0xe6, 0x60, }, + { 0x93, 0x2b, 0x28, 0x46, 0xe4, 0xd7, 0x06, 0x66, }, + { 0xe1, 0x91, 0x5f, 0x5c, 0xb1, 0xec, 0xa4, 0x6c, }, + { 0xf3, 0x25, 0x96, 0x5c, 0xa1, 0x6d, 0x62, 0x9f, }, + { 0x57, 0x5f, 0xf2, 0x8e, 0x60, 0x38, 0x1b, 0xe5, }, + { 0x72, 0x45, 0x06, 0xeb, 0x4c, 0x32, 0x8a, 0x95, } + }; + unsigned char in[64]; + struct sipkey k; + size_t i; + + sip_tokey(&k, "\000\001\002\003\004\005\006\007\010\011\012\013\014\015\016\017"); + + for (i = 0; i < sizeof in; ++i) { + in[i] = i; + + if (siphash24(in, i, &k) != SIP_U8TO64_LE(vectors[i])) + return 0; + } + + return 1; +} /* sip24_valid() */ + + +#if SIPHASH_MAIN + +#include + +int main(void) { + int ok = sip24_valid(); + + if (ok) + puts("OK"); + else + puts("FAIL"); + + return !ok; +} /* main() */ + +#endif /* SIPHASH_MAIN */ + + +#endif /* SIPHASH_H */ diff --git a/cbits/winconfig.h b/cbits/winconfig.h --- a/cbits/winconfig.h +++ b/cbits/winconfig.h @@ -17,6 +17,12 @@ #include #include + +#if defined(HAVE_EXPAT_CONFIG_H) /* e.g. MinGW */ +# include +#else /* !defined(HAVE_EXPAT_CONFIG_H) */ + + #define XML_NS 1 #define XML_DTD 1 #define XML_CONTEXT_BYTES 1024 @@ -26,5 +32,9 @@ /* Windows has memmove() available. */ #define HAVE_MEMMOVE + + +#endif /* !defined(HAVE_EXPAT_CONFIG_H) */ + #endif /* ndef WINCONFIG_H */ diff --git a/cbits/xmlparse.c b/cbits/xmlparse.c --- a/cbits/xmlparse.c +++ b/cbits/xmlparse.c @@ -1,6279 +1,6821 @@ /* Copyright (c) 1998, 1999, 2000 Thai Open Source Software Center Ltd See the file COPYING for copying permission. -*/ - -#include -#include /* memset(), memcpy() */ -#include - -#define XML_BUILDING_EXPAT 1 - -#ifdef COMPILED_FROM_DSP -#include "winconfig.h" -#elif defined(MACOS_CLASSIC) -#include "macconfig.h" -#elif defined(__amigaos4__) -#include "amigaconfig.h" -#elif defined(__WATCOMC__) -#include "watcomconfig.h" -#elif defined(HAVE_EXPAT_CONFIG_H) -#include -#endif /* ndef COMPILED_FROM_DSP */ - -#include "ascii.h" -#include "expat.h" - -#ifdef XML_UNICODE -#define XML_ENCODE_MAX XML_UTF16_ENCODE_MAX -#define XmlConvert XmlUtf16Convert -#define XmlGetInternalEncoding XmlGetUtf16InternalEncoding -#define XmlGetInternalEncodingNS XmlGetUtf16InternalEncodingNS -#define XmlEncode XmlUtf16Encode -/* Using pointer subtraction to convert to integer type. */ -#define MUST_CONVERT(enc, s) (!(enc)->isUtf16 || (((char *)(s) - (char *)NULL) & 1)) -typedef unsigned short ICHAR; -#else -#define XML_ENCODE_MAX XML_UTF8_ENCODE_MAX -#define XmlConvert XmlUtf8Convert -#define XmlGetInternalEncoding XmlGetUtf8InternalEncoding -#define XmlGetInternalEncodingNS XmlGetUtf8InternalEncodingNS -#define XmlEncode XmlUtf8Encode -#define MUST_CONVERT(enc, s) (!(enc)->isUtf8) -typedef char ICHAR; -#endif - - -#ifndef XML_NS - -#define XmlInitEncodingNS XmlInitEncoding -#define XmlInitUnknownEncodingNS XmlInitUnknownEncoding -#undef XmlGetInternalEncodingNS -#define XmlGetInternalEncodingNS XmlGetInternalEncoding -#define XmlParseXmlDeclNS XmlParseXmlDecl - -#endif - -#ifdef XML_UNICODE - -#ifdef XML_UNICODE_WCHAR_T -#define XML_T(x) (const wchar_t)x -#define XML_L(x) L ## x -#else -#define XML_T(x) (const unsigned short)x -#define XML_L(x) x -#endif - -#else - -#define XML_T(x) x -#define XML_L(x) x - -#endif - -/* Round up n to be a multiple of sz, where sz is a power of 2. */ -#define ROUND_UP(n, sz) (((n) + ((sz) - 1)) & ~((sz) - 1)) - -/* Handle the case where memmove() doesn't exist. */ -#ifndef HAVE_MEMMOVE -#ifdef HAVE_BCOPY -#define memmove(d,s,l) bcopy((s),(d),(l)) -#else -#error memmove does not exist on this platform, nor is a substitute available -#endif /* HAVE_BCOPY */ -#endif /* HAVE_MEMMOVE */ - -#include "internal.h" -#include "xmltok.h" -#include "xmlrole.h" - -typedef const XML_Char *KEY; - -typedef struct { - KEY name; -} NAMED; - -typedef struct { - NAMED **v; - unsigned char power; - size_t size; - size_t used; - const XML_Memory_Handling_Suite *mem; -} HASH_TABLE; - -/* Basic character hash algorithm, taken from Python's string hash: - h = h * 1000003 ^ character, the constant being a prime number. - -*/ -#ifdef XML_UNICODE -#define CHAR_HASH(h, c) \ - (((h) * 0xF4243) ^ (unsigned short)(c)) -#else -#define CHAR_HASH(h, c) \ - (((h) * 0xF4243) ^ (unsigned char)(c)) -#endif - -/* For probing (after a collision) we need a step size relative prime - to the hash table size, which is a power of 2. We use double-hashing, - since we can calculate a second hash value cheaply by taking those bits - of the first hash value that were discarded (masked out) when the table - index was calculated: index = hash & mask, where mask = table->size - 1. - We limit the maximum step size to table->size / 4 (mask >> 2) and make - it odd, since odd numbers are always relative prime to a power of 2. -*/ -#define SECOND_HASH(hash, mask, power) \ - ((((hash) & ~(mask)) >> ((power) - 1)) & ((mask) >> 2)) -#define PROBE_STEP(hash, mask, power) \ - ((unsigned char)((SECOND_HASH(hash, mask, power)) | 1)) - -typedef struct { - NAMED **p; - NAMED **end; -} HASH_TABLE_ITER; - -#define INIT_TAG_BUF_SIZE 32 /* must be a multiple of sizeof(XML_Char) */ -#define INIT_DATA_BUF_SIZE 1024 -#define INIT_ATTS_SIZE 16 -#define INIT_ATTS_VERSION 0xFFFFFFFF -#define INIT_BLOCK_SIZE 1024 -#define INIT_BUFFER_SIZE 1024 - -#define EXPAND_SPARE 24 - -typedef struct binding { - struct prefix *prefix; - struct binding *nextTagBinding; - struct binding *prevPrefixBinding; - const struct attribute_id *attId; - XML_Char *uri; - int uriLen; - int uriAlloc; -} BINDING; - -typedef struct prefix { - const XML_Char *name; - BINDING *binding; -} PREFIX; - -typedef struct { - const XML_Char *str; - const XML_Char *localPart; - const XML_Char *prefix; - int strLen; - int uriLen; - int prefixLen; -} TAG_NAME; - -/* TAG represents an open element. - The name of the element is stored in both the document and API - encodings. The memory buffer 'buf' is a separately-allocated - memory area which stores the name. During the XML_Parse()/ - XMLParseBuffer() when the element is open, the memory for the 'raw' - version of the name (in the document encoding) is shared with the - document buffer. If the element is open across calls to - XML_Parse()/XML_ParseBuffer(), the buffer is re-allocated to - contain the 'raw' name as well. - - A parser re-uses these structures, maintaining a list of allocated - TAG objects in a free list. -*/ -typedef struct tag { - struct tag *parent; /* parent of this element */ - const char *rawName; /* tagName in the original encoding */ - int rawNameLength; - TAG_NAME name; /* tagName in the API encoding */ - char *buf; /* buffer for name components */ - char *bufEnd; /* end of the buffer */ - BINDING *bindings; -} TAG; - -typedef struct { - const XML_Char *name; - const XML_Char *textPtr; - int textLen; /* length in XML_Chars */ - int processed; /* # of processed bytes - when suspended */ - const XML_Char *systemId; - const XML_Char *base; - const XML_Char *publicId; - const XML_Char *notation; - XML_Bool open; - XML_Bool is_param; - XML_Bool is_internal; /* true if declared in internal subset outside PE */ -} ENTITY; - -typedef struct { - enum XML_Content_Type type; - enum XML_Content_Quant quant; - const XML_Char * name; - int firstchild; - int lastchild; - int childcnt; - int nextsib; -} CONTENT_SCAFFOLD; - -#define INIT_SCAFFOLD_ELEMENTS 32 - -typedef struct block { - struct block *next; - int size; - XML_Char s[1]; -} BLOCK; - -typedef struct { - BLOCK *blocks; - BLOCK *freeBlocks; - const XML_Char *end; - XML_Char *ptr; - XML_Char *start; - const XML_Memory_Handling_Suite *mem; -} STRING_POOL; - -/* The XML_Char before the name is used to determine whether - an attribute has been specified. */ -typedef struct attribute_id { - XML_Char *name; - PREFIX *prefix; - XML_Bool maybeTokenized; - XML_Bool xmlns; -} ATTRIBUTE_ID; - -typedef struct { - const ATTRIBUTE_ID *id; - XML_Bool isCdata; - const XML_Char *value; -} DEFAULT_ATTRIBUTE; - -typedef struct { - unsigned long version; - unsigned long hash; - const XML_Char *uriName; -} NS_ATT; - -typedef struct { - const XML_Char *name; - PREFIX *prefix; - const ATTRIBUTE_ID *idAtt; - int nDefaultAtts; - int allocDefaultAtts; - DEFAULT_ATTRIBUTE *defaultAtts; -} ELEMENT_TYPE; - -typedef struct { - HASH_TABLE generalEntities; - HASH_TABLE elementTypes; - HASH_TABLE attributeIds; - HASH_TABLE prefixes; - STRING_POOL pool; - STRING_POOL entityValuePool; - /* false once a parameter entity reference has been skipped */ - XML_Bool keepProcessing; - /* true once an internal or external PE reference has been encountered; - this includes the reference to an external subset */ - XML_Bool hasParamEntityRefs; - XML_Bool standalone; -#ifdef XML_DTD - /* indicates if external PE has been read */ - XML_Bool paramEntityRead; - HASH_TABLE paramEntities; -#endif /* XML_DTD */ - PREFIX defaultPrefix; - /* === scaffolding for building content model === */ - XML_Bool in_eldecl; - CONTENT_SCAFFOLD *scaffold; - unsigned contentStringLen; - unsigned scaffSize; - unsigned scaffCount; - int scaffLevel; - int *scaffIndex; -} DTD; - -typedef struct open_internal_entity { - const char *internalEventPtr; - const char *internalEventEndPtr; - struct open_internal_entity *next; - ENTITY *entity; - int startTagLevel; - XML_Bool betweenDecl; /* WFC: PE Between Declarations */ -} OPEN_INTERNAL_ENTITY; - -typedef enum XML_Error PTRCALL Processor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr); - -static Processor prologProcessor; -static Processor prologInitProcessor; -static Processor contentProcessor; -static Processor cdataSectionProcessor; -#ifdef XML_DTD -static Processor ignoreSectionProcessor; -static Processor externalParEntProcessor; -static Processor externalParEntInitProcessor; -static Processor entityValueProcessor; -static Processor entityValueInitProcessor; -#endif /* XML_DTD */ -static Processor epilogProcessor; -static Processor errorProcessor; -static Processor externalEntityInitProcessor; -static Processor externalEntityInitProcessor2; -static Processor externalEntityInitProcessor3; -static Processor externalEntityContentProcessor; -static Processor internalEntityProcessor; - -static enum XML_Error -handleUnknownEncoding(XML_Parser parser, const XML_Char *encodingName); -static enum XML_Error -processXmlDecl(XML_Parser parser, int isGeneralTextEntity, - const char *s, const char *next); -static enum XML_Error -initializeEncoding(XML_Parser parser); -static enum XML_Error -doProlog(XML_Parser parser, const ENCODING *enc, const char *s, - const char *end, int tok, const char *next, const char **nextPtr, - XML_Bool haveMore); -static enum XML_Error -processInternalEntity(XML_Parser parser, ENTITY *entity, - XML_Bool betweenDecl); -static enum XML_Error -doContent(XML_Parser parser, int startTagLevel, const ENCODING *enc, - const char *start, const char *end, const char **endPtr, - XML_Bool haveMore); -static enum XML_Error -doCdataSection(XML_Parser parser, const ENCODING *, const char **startPtr, - const char *end, const char **nextPtr, XML_Bool haveMore); -#ifdef XML_DTD -static enum XML_Error -doIgnoreSection(XML_Parser parser, const ENCODING *, const char **startPtr, - const char *end, const char **nextPtr, XML_Bool haveMore); -#endif /* XML_DTD */ - -static enum XML_Error -storeAtts(XML_Parser parser, const ENCODING *, const char *s, - TAG_NAME *tagNamePtr, BINDING **bindingsPtr); -static enum XML_Error -addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId, - const XML_Char *uri, BINDING **bindingsPtr); -static int -defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *, XML_Bool isCdata, - XML_Bool isId, const XML_Char *dfltValue, XML_Parser parser); -static enum XML_Error -storeAttributeValue(XML_Parser parser, const ENCODING *, XML_Bool isCdata, - const char *, const char *, STRING_POOL *); -static enum XML_Error -appendAttributeValue(XML_Parser parser, const ENCODING *, XML_Bool isCdata, - const char *, const char *, STRING_POOL *); -static ATTRIBUTE_ID * -getAttributeId(XML_Parser parser, const ENCODING *enc, const char *start, - const char *end); -static int -setElementTypePrefix(XML_Parser parser, ELEMENT_TYPE *); -static enum XML_Error -storeEntityValue(XML_Parser parser, const ENCODING *enc, const char *start, - const char *end); -static int -reportProcessingInstruction(XML_Parser parser, const ENCODING *enc, - const char *start, const char *end); -static int -reportComment(XML_Parser parser, const ENCODING *enc, const char *start, - const char *end); -static void -reportDefault(XML_Parser parser, const ENCODING *enc, const char *start, - const char *end); - -static const XML_Char * getContext(XML_Parser parser); -static XML_Bool -setContext(XML_Parser parser, const XML_Char *context); - -static void FASTCALL normalizePublicId(XML_Char *s); - -static DTD * dtdCreate(const XML_Memory_Handling_Suite *ms); -/* do not call if parentParser != NULL */ -static void dtdReset(DTD *p, const XML_Memory_Handling_Suite *ms); -static void -dtdDestroy(DTD *p, XML_Bool isDocEntity, const XML_Memory_Handling_Suite *ms); -static int -dtdCopy(DTD *newDtd, const DTD *oldDtd, const XML_Memory_Handling_Suite *ms); -static int -copyEntityTable(HASH_TABLE *, STRING_POOL *, const HASH_TABLE *); - -static NAMED * -lookup(HASH_TABLE *table, KEY name, size_t createSize); -static void FASTCALL -hashTableInit(HASH_TABLE *, const XML_Memory_Handling_Suite *ms); -static void FASTCALL hashTableClear(HASH_TABLE *); -static void FASTCALL hashTableDestroy(HASH_TABLE *); -static void FASTCALL -hashTableIterInit(HASH_TABLE_ITER *, const HASH_TABLE *); -static NAMED * FASTCALL hashTableIterNext(HASH_TABLE_ITER *); - -static void FASTCALL -poolInit(STRING_POOL *, const XML_Memory_Handling_Suite *ms); -static void FASTCALL poolClear(STRING_POOL *); -static void FASTCALL poolDestroy(STRING_POOL *); -static XML_Char * -poolAppend(STRING_POOL *pool, const ENCODING *enc, - const char *ptr, const char *end); -static XML_Char * -poolStoreString(STRING_POOL *pool, const ENCODING *enc, - const char *ptr, const char *end); -static XML_Bool FASTCALL poolGrow(STRING_POOL *pool); -static const XML_Char * FASTCALL -poolCopyString(STRING_POOL *pool, const XML_Char *s); -static const XML_Char * -poolCopyStringN(STRING_POOL *pool, const XML_Char *s, int n); -static const XML_Char * FASTCALL -poolAppendString(STRING_POOL *pool, const XML_Char *s); - -static int FASTCALL nextScaffoldPart(XML_Parser parser); -static XML_Content * build_model(XML_Parser parser); -static ELEMENT_TYPE * -getElementType(XML_Parser parser, const ENCODING *enc, - const char *ptr, const char *end); - -static XML_Parser -parserCreate(const XML_Char *encodingName, - const XML_Memory_Handling_Suite *memsuite, - const XML_Char *nameSep, - DTD *dtd); -static void -parserInit(XML_Parser parser, const XML_Char *encodingName); - -#define poolStart(pool) ((pool)->start) -#define poolEnd(pool) ((pool)->ptr) -#define poolLength(pool) ((pool)->ptr - (pool)->start) -#define poolChop(pool) ((void)--(pool->ptr)) -#define poolLastChar(pool) (((pool)->ptr)[-1]) -#define poolDiscard(pool) ((pool)->ptr = (pool)->start) -#define poolFinish(pool) ((pool)->start = (pool)->ptr) -#define poolAppendChar(pool, c) \ - (((pool)->ptr == (pool)->end && !poolGrow(pool)) \ - ? 0 \ - : ((*((pool)->ptr)++ = c), 1)) - -struct XML_ParserStruct { - /* The first member must be userData so that the XML_GetUserData - macro works. */ - void *m_userData; - void *m_handlerArg; - char *m_buffer; - const XML_Memory_Handling_Suite m_mem; - /* first character to be parsed */ - const char *m_bufferPtr; - /* past last character to be parsed */ - char *m_bufferEnd; - /* allocated end of buffer */ - const char *m_bufferLim; - XML_Index m_parseEndByteIndex; - const char *m_parseEndPtr; - XML_Char *m_dataBuf; - XML_Char *m_dataBufEnd; - XML_StartElementHandler m_startElementHandler; - XML_EndElementHandler m_endElementHandler; - XML_CharacterDataHandler m_characterDataHandler; - XML_ProcessingInstructionHandler m_processingInstructionHandler; - XML_CommentHandler m_commentHandler; - XML_StartCdataSectionHandler m_startCdataSectionHandler; - XML_EndCdataSectionHandler m_endCdataSectionHandler; - XML_DefaultHandler m_defaultHandler; - XML_StartDoctypeDeclHandler m_startDoctypeDeclHandler; - XML_EndDoctypeDeclHandler m_endDoctypeDeclHandler; - XML_UnparsedEntityDeclHandler m_unparsedEntityDeclHandler; - XML_NotationDeclHandler m_notationDeclHandler; - XML_StartNamespaceDeclHandler m_startNamespaceDeclHandler; - XML_EndNamespaceDeclHandler m_endNamespaceDeclHandler; - XML_NotStandaloneHandler m_notStandaloneHandler; - XML_ExternalEntityRefHandler m_externalEntityRefHandler; - XML_Parser m_externalEntityRefHandlerArg; - XML_SkippedEntityHandler m_skippedEntityHandler; - XML_UnknownEncodingHandler m_unknownEncodingHandler; - XML_ElementDeclHandler m_elementDeclHandler; - XML_AttlistDeclHandler m_attlistDeclHandler; - XML_EntityDeclHandler m_entityDeclHandler; - XML_XmlDeclHandler m_xmlDeclHandler; - const ENCODING *m_encoding; - INIT_ENCODING m_initEncoding; - const ENCODING *m_internalEncoding; - const XML_Char *m_protocolEncodingName; - XML_Bool m_ns; - XML_Bool m_ns_triplets; - void *m_unknownEncodingMem; - void *m_unknownEncodingData; - void *m_unknownEncodingHandlerData; - void (XMLCALL *m_unknownEncodingRelease)(void *); - PROLOG_STATE m_prologState; - Processor *m_processor; - enum XML_Error m_errorCode; - const char *m_eventPtr; - const char *m_eventEndPtr; - const char *m_positionPtr; - OPEN_INTERNAL_ENTITY *m_openInternalEntities; - OPEN_INTERNAL_ENTITY *m_freeInternalEntities; - XML_Bool m_defaultExpandInternalEntities; - int m_tagLevel; - ENTITY *m_declEntity; - const XML_Char *m_doctypeName; - const XML_Char *m_doctypeSysid; - const XML_Char *m_doctypePubid; - const XML_Char *m_declAttributeType; - const XML_Char *m_declNotationName; - const XML_Char *m_declNotationPublicId; - ELEMENT_TYPE *m_declElementType; - ATTRIBUTE_ID *m_declAttributeId; - XML_Bool m_declAttributeIsCdata; - XML_Bool m_declAttributeIsId; - DTD *m_dtd; - const XML_Char *m_curBase; - TAG *m_tagStack; - TAG *m_freeTagList; - BINDING *m_inheritedBindings; - BINDING *m_freeBindingList; - int m_attsSize; - int m_nSpecifiedAtts; - int m_idAttIndex; - ATTRIBUTE *m_atts; - NS_ATT *m_nsAtts; - unsigned long m_nsAttsVersion; - unsigned char m_nsAttsPower; - POSITION m_position; - STRING_POOL m_tempPool; - STRING_POOL m_temp2Pool; - char *m_groupConnector; - unsigned int m_groupSize; - XML_Char m_namespaceSeparator; - XML_Parser m_parentParser; - XML_ParsingStatus m_parsingStatus; -#ifdef XML_DTD - XML_Bool m_isParamEntity; - XML_Bool m_useForeignDTD; - enum XML_ParamEntityParsing m_paramEntityParsing; -#endif -}; - -#define MALLOC(s) (parser->m_mem.malloc_fcn((s))) -#define REALLOC(p,s) (parser->m_mem.realloc_fcn((p),(s))) -#define FREE(p) (parser->m_mem.free_fcn((p))) - -#define userData (parser->m_userData) -#define handlerArg (parser->m_handlerArg) -#define startElementHandler (parser->m_startElementHandler) -#define endElementHandler (parser->m_endElementHandler) -#define characterDataHandler (parser->m_characterDataHandler) -#define processingInstructionHandler \ - (parser->m_processingInstructionHandler) -#define commentHandler (parser->m_commentHandler) -#define startCdataSectionHandler \ - (parser->m_startCdataSectionHandler) -#define endCdataSectionHandler (parser->m_endCdataSectionHandler) -#define defaultHandler (parser->m_defaultHandler) -#define startDoctypeDeclHandler (parser->m_startDoctypeDeclHandler) -#define endDoctypeDeclHandler (parser->m_endDoctypeDeclHandler) -#define unparsedEntityDeclHandler \ - (parser->m_unparsedEntityDeclHandler) -#define notationDeclHandler (parser->m_notationDeclHandler) -#define startNamespaceDeclHandler \ - (parser->m_startNamespaceDeclHandler) -#define endNamespaceDeclHandler (parser->m_endNamespaceDeclHandler) -#define notStandaloneHandler (parser->m_notStandaloneHandler) -#define externalEntityRefHandler \ - (parser->m_externalEntityRefHandler) -#define externalEntityRefHandlerArg \ - (parser->m_externalEntityRefHandlerArg) -#define internalEntityRefHandler \ - (parser->m_internalEntityRefHandler) -#define skippedEntityHandler (parser->m_skippedEntityHandler) -#define unknownEncodingHandler (parser->m_unknownEncodingHandler) -#define elementDeclHandler (parser->m_elementDeclHandler) -#define attlistDeclHandler (parser->m_attlistDeclHandler) -#define entityDeclHandler (parser->m_entityDeclHandler) -#define xmlDeclHandler (parser->m_xmlDeclHandler) -#define encoding (parser->m_encoding) -#define initEncoding (parser->m_initEncoding) -#define internalEncoding (parser->m_internalEncoding) -#define unknownEncodingMem (parser->m_unknownEncodingMem) -#define unknownEncodingData (parser->m_unknownEncodingData) -#define unknownEncodingHandlerData \ - (parser->m_unknownEncodingHandlerData) -#define unknownEncodingRelease (parser->m_unknownEncodingRelease) -#define protocolEncodingName (parser->m_protocolEncodingName) -#define ns (parser->m_ns) -#define ns_triplets (parser->m_ns_triplets) -#define prologState (parser->m_prologState) -#define processor (parser->m_processor) -#define errorCode (parser->m_errorCode) -#define eventPtr (parser->m_eventPtr) -#define eventEndPtr (parser->m_eventEndPtr) -#define positionPtr (parser->m_positionPtr) -#define position (parser->m_position) -#define openInternalEntities (parser->m_openInternalEntities) -#define freeInternalEntities (parser->m_freeInternalEntities) -#define defaultExpandInternalEntities \ - (parser->m_defaultExpandInternalEntities) -#define tagLevel (parser->m_tagLevel) -#define buffer (parser->m_buffer) -#define bufferPtr (parser->m_bufferPtr) -#define bufferEnd (parser->m_bufferEnd) -#define parseEndByteIndex (parser->m_parseEndByteIndex) -#define parseEndPtr (parser->m_parseEndPtr) -#define bufferLim (parser->m_bufferLim) -#define dataBuf (parser->m_dataBuf) -#define dataBufEnd (parser->m_dataBufEnd) -#define _dtd (parser->m_dtd) -#define curBase (parser->m_curBase) -#define declEntity (parser->m_declEntity) -#define doctypeName (parser->m_doctypeName) -#define doctypeSysid (parser->m_doctypeSysid) -#define doctypePubid (parser->m_doctypePubid) -#define declAttributeType (parser->m_declAttributeType) -#define declNotationName (parser->m_declNotationName) -#define declNotationPublicId (parser->m_declNotationPublicId) -#define declElementType (parser->m_declElementType) -#define declAttributeId (parser->m_declAttributeId) -#define declAttributeIsCdata (parser->m_declAttributeIsCdata) -#define declAttributeIsId (parser->m_declAttributeIsId) -#define freeTagList (parser->m_freeTagList) -#define freeBindingList (parser->m_freeBindingList) -#define inheritedBindings (parser->m_inheritedBindings) -#define tagStack (parser->m_tagStack) -#define atts (parser->m_atts) -#define attsSize (parser->m_attsSize) -#define nSpecifiedAtts (parser->m_nSpecifiedAtts) -#define idAttIndex (parser->m_idAttIndex) -#define nsAtts (parser->m_nsAtts) -#define nsAttsVersion (parser->m_nsAttsVersion) -#define nsAttsPower (parser->m_nsAttsPower) -#define tempPool (parser->m_tempPool) -#define temp2Pool (parser->m_temp2Pool) -#define groupConnector (parser->m_groupConnector) -#define groupSize (parser->m_groupSize) -#define namespaceSeparator (parser->m_namespaceSeparator) -#define parentParser (parser->m_parentParser) -#define ps_parsing (parser->m_parsingStatus.parsing) -#define ps_finalBuffer (parser->m_parsingStatus.finalBuffer) -#ifdef XML_DTD -#define isParamEntity (parser->m_isParamEntity) -#define useForeignDTD (parser->m_useForeignDTD) -#define paramEntityParsing (parser->m_paramEntityParsing) -#endif /* XML_DTD */ - -XML_Parser XMLCALL -XML_ParserCreate(const XML_Char *encodingName) -{ - return XML_ParserCreate_MM(encodingName, NULL, NULL); -} - -XML_Parser XMLCALL -XML_ParserCreateNS(const XML_Char *encodingName, XML_Char nsSep) -{ - XML_Char tmp[2]; - *tmp = nsSep; - return XML_ParserCreate_MM(encodingName, NULL, tmp); -} - -static const XML_Char implicitContext[] = { - ASCII_x, ASCII_m, ASCII_l, ASCII_EQUALS, ASCII_h, ASCII_t, ASCII_t, ASCII_p, - ASCII_COLON, ASCII_SLASH, ASCII_SLASH, ASCII_w, ASCII_w, ASCII_w, - ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, ASCII_o, ASCII_r, ASCII_g, - ASCII_SLASH, ASCII_X, ASCII_M, ASCII_L, ASCII_SLASH, ASCII_1, ASCII_9, - ASCII_9, ASCII_8, ASCII_SLASH, ASCII_n, ASCII_a, ASCII_m, ASCII_e, - ASCII_s, ASCII_p, ASCII_a, ASCII_c, ASCII_e, '\0' -}; - -XML_Parser XMLCALL -XML_ParserCreate_MM(const XML_Char *encodingName, - const XML_Memory_Handling_Suite *memsuite, - const XML_Char *nameSep) -{ - XML_Parser parser = parserCreate(encodingName, memsuite, nameSep, NULL); - if (parser != NULL && ns) { - /* implicit context only set for root parser, since child - parsers (i.e. external entity parsers) will inherit it - */ - if (!setContext(parser, implicitContext)) { - XML_ParserFree(parser); - return NULL; - } - } - return parser; -} - -static XML_Parser -parserCreate(const XML_Char *encodingName, - const XML_Memory_Handling_Suite *memsuite, - const XML_Char *nameSep, - DTD *dtd) -{ - XML_Parser parser; - - if (memsuite) { - XML_Memory_Handling_Suite *mtemp; - parser = (XML_Parser) - memsuite->malloc_fcn(sizeof(struct XML_ParserStruct)); - if (parser != NULL) { - mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem); - mtemp->malloc_fcn = memsuite->malloc_fcn; - mtemp->realloc_fcn = memsuite->realloc_fcn; - mtemp->free_fcn = memsuite->free_fcn; - } - } - else { - XML_Memory_Handling_Suite *mtemp; - parser = (XML_Parser)malloc(sizeof(struct XML_ParserStruct)); - if (parser != NULL) { - mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem); - mtemp->malloc_fcn = malloc; - mtemp->realloc_fcn = realloc; - mtemp->free_fcn = free; - } - } - - if (!parser) - return parser; - - buffer = NULL; - bufferLim = NULL; - - attsSize = INIT_ATTS_SIZE; - atts = (ATTRIBUTE *)MALLOC(attsSize * sizeof(ATTRIBUTE)); - if (atts == NULL) { - FREE(parser); - return NULL; - } - dataBuf = (XML_Char *)MALLOC(INIT_DATA_BUF_SIZE * sizeof(XML_Char)); - if (dataBuf == NULL) { - FREE(atts); - FREE(parser); - return NULL; - } - dataBufEnd = dataBuf + INIT_DATA_BUF_SIZE; - - if (dtd) - _dtd = dtd; - else { - _dtd = dtdCreate(&parser->m_mem); - if (_dtd == NULL) { - FREE(dataBuf); - FREE(atts); - FREE(parser); - return NULL; - } - } - - freeBindingList = NULL; - freeTagList = NULL; - freeInternalEntities = NULL; - - groupSize = 0; - groupConnector = NULL; - - unknownEncodingHandler = NULL; - unknownEncodingHandlerData = NULL; - - namespaceSeparator = ASCII_EXCL; - ns = XML_FALSE; - ns_triplets = XML_FALSE; - - nsAtts = NULL; - nsAttsVersion = 0; - nsAttsPower = 0; - - poolInit(&tempPool, &(parser->m_mem)); - poolInit(&temp2Pool, &(parser->m_mem)); - parserInit(parser, encodingName); - - if (encodingName && !protocolEncodingName) { - XML_ParserFree(parser); - return NULL; - } - - if (nameSep) { - ns = XML_TRUE; - internalEncoding = XmlGetInternalEncodingNS(); - namespaceSeparator = *nameSep; - } - else { - internalEncoding = XmlGetInternalEncoding(); - } - - return parser; -} - -static void -parserInit(XML_Parser parser, const XML_Char *encodingName) -{ - processor = prologInitProcessor; - XmlPrologStateInit(&prologState); - protocolEncodingName = (encodingName != NULL - ? poolCopyString(&tempPool, encodingName) - : NULL); - curBase = NULL; - XmlInitEncoding(&initEncoding, &encoding, 0); - userData = NULL; - handlerArg = NULL; - startElementHandler = NULL; - endElementHandler = NULL; - characterDataHandler = NULL; - processingInstructionHandler = NULL; - commentHandler = NULL; - startCdataSectionHandler = NULL; - endCdataSectionHandler = NULL; - defaultHandler = NULL; - startDoctypeDeclHandler = NULL; - endDoctypeDeclHandler = NULL; - unparsedEntityDeclHandler = NULL; - notationDeclHandler = NULL; - startNamespaceDeclHandler = NULL; - endNamespaceDeclHandler = NULL; - notStandaloneHandler = NULL; - externalEntityRefHandler = NULL; - externalEntityRefHandlerArg = parser; - skippedEntityHandler = NULL; - elementDeclHandler = NULL; - attlistDeclHandler = NULL; - entityDeclHandler = NULL; - xmlDeclHandler = NULL; - bufferPtr = buffer; - bufferEnd = buffer; - parseEndByteIndex = 0; - parseEndPtr = NULL; - declElementType = NULL; - declAttributeId = NULL; - declEntity = NULL; - doctypeName = NULL; - doctypeSysid = NULL; - doctypePubid = NULL; - declAttributeType = NULL; - declNotationName = NULL; - declNotationPublicId = NULL; - declAttributeIsCdata = XML_FALSE; - declAttributeIsId = XML_FALSE; - memset(&position, 0, sizeof(POSITION)); - errorCode = XML_ERROR_NONE; - eventPtr = NULL; - eventEndPtr = NULL; - positionPtr = NULL; - openInternalEntities = NULL; - defaultExpandInternalEntities = XML_TRUE; - tagLevel = 0; - tagStack = NULL; - inheritedBindings = NULL; - nSpecifiedAtts = 0; - unknownEncodingMem = NULL; - unknownEncodingRelease = NULL; - unknownEncodingData = NULL; - parentParser = NULL; - ps_parsing = XML_INITIALIZED; -#ifdef XML_DTD - isParamEntity = XML_FALSE; - useForeignDTD = XML_FALSE; - paramEntityParsing = XML_PARAM_ENTITY_PARSING_NEVER; -#endif -} - -/* moves list of bindings to freeBindingList */ -static void FASTCALL -moveToFreeBindingList(XML_Parser parser, BINDING *bindings) -{ - while (bindings) { - BINDING *b = bindings; - bindings = bindings->nextTagBinding; - b->nextTagBinding = freeBindingList; - freeBindingList = b; - } -} - -XML_Bool XMLCALL -XML_ParserReset(XML_Parser parser, const XML_Char *encodingName) -{ - TAG *tStk; - OPEN_INTERNAL_ENTITY *openEntityList; - if (parentParser) - return XML_FALSE; - /* move tagStack to freeTagList */ - tStk = tagStack; - while (tStk) { - TAG *tag = tStk; - tStk = tStk->parent; - tag->parent = freeTagList; - moveToFreeBindingList(parser, tag->bindings); - tag->bindings = NULL; - freeTagList = tag; - } - /* move openInternalEntities to freeInternalEntities */ - openEntityList = openInternalEntities; - while (openEntityList) { - OPEN_INTERNAL_ENTITY *openEntity = openEntityList; - openEntityList = openEntity->next; - openEntity->next = freeInternalEntities; - freeInternalEntities = openEntity; - } - moveToFreeBindingList(parser, inheritedBindings); - FREE(unknownEncodingMem); - if (unknownEncodingRelease) - unknownEncodingRelease(unknownEncodingData); - poolClear(&tempPool); - poolClear(&temp2Pool); - parserInit(parser, encodingName); - dtdReset(_dtd, &parser->m_mem); - return setContext(parser, implicitContext); -} - -enum XML_Status XMLCALL -XML_SetEncoding(XML_Parser parser, const XML_Char *encodingName) -{ - /* Block after XML_Parse()/XML_ParseBuffer() has been called. - XXX There's no way for the caller to determine which of the - XXX possible error cases caused the XML_STATUS_ERROR return. - */ - if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) - return XML_STATUS_ERROR; - if (encodingName == NULL) - protocolEncodingName = NULL; - else { - protocolEncodingName = poolCopyString(&tempPool, encodingName); - if (!protocolEncodingName) - return XML_STATUS_ERROR; - } - return XML_STATUS_OK; -} - -XML_Parser XMLCALL -XML_ExternalEntityParserCreate(XML_Parser oldParser, - const XML_Char *context, - const XML_Char *encodingName) -{ - XML_Parser parser = oldParser; - DTD *newDtd = NULL; - DTD *oldDtd = _dtd; - XML_StartElementHandler oldStartElementHandler = startElementHandler; - XML_EndElementHandler oldEndElementHandler = endElementHandler; - XML_CharacterDataHandler oldCharacterDataHandler = characterDataHandler; - XML_ProcessingInstructionHandler oldProcessingInstructionHandler - = processingInstructionHandler; - XML_CommentHandler oldCommentHandler = commentHandler; - XML_StartCdataSectionHandler oldStartCdataSectionHandler - = startCdataSectionHandler; - XML_EndCdataSectionHandler oldEndCdataSectionHandler - = endCdataSectionHandler; - XML_DefaultHandler oldDefaultHandler = defaultHandler; - XML_UnparsedEntityDeclHandler oldUnparsedEntityDeclHandler - = unparsedEntityDeclHandler; - XML_NotationDeclHandler oldNotationDeclHandler = notationDeclHandler; - XML_StartNamespaceDeclHandler oldStartNamespaceDeclHandler - = startNamespaceDeclHandler; - XML_EndNamespaceDeclHandler oldEndNamespaceDeclHandler - = endNamespaceDeclHandler; - XML_NotStandaloneHandler oldNotStandaloneHandler = notStandaloneHandler; - XML_ExternalEntityRefHandler oldExternalEntityRefHandler - = externalEntityRefHandler; - XML_SkippedEntityHandler oldSkippedEntityHandler = skippedEntityHandler; - XML_UnknownEncodingHandler oldUnknownEncodingHandler - = unknownEncodingHandler; - XML_ElementDeclHandler oldElementDeclHandler = elementDeclHandler; - XML_AttlistDeclHandler oldAttlistDeclHandler = attlistDeclHandler; - XML_EntityDeclHandler oldEntityDeclHandler = entityDeclHandler; - XML_XmlDeclHandler oldXmlDeclHandler = xmlDeclHandler; - ELEMENT_TYPE * oldDeclElementType = declElementType; - - void *oldUserData = userData; - void *oldHandlerArg = handlerArg; - XML_Bool oldDefaultExpandInternalEntities = defaultExpandInternalEntities; - XML_Parser oldExternalEntityRefHandlerArg = externalEntityRefHandlerArg; -#ifdef XML_DTD - enum XML_ParamEntityParsing oldParamEntityParsing = paramEntityParsing; - int oldInEntityValue = prologState.inEntityValue; -#endif - XML_Bool oldns_triplets = ns_triplets; - -#ifdef XML_DTD - if (!context) - newDtd = oldDtd; -#endif /* XML_DTD */ - - /* Note that the magical uses of the pre-processor to make field - access look more like C++ require that `parser' be overwritten - here. This makes this function more painful to follow than it - would be otherwise. - */ - if (ns) { - XML_Char tmp[2]; - *tmp = namespaceSeparator; - parser = parserCreate(encodingName, &parser->m_mem, tmp, newDtd); - } - else { - parser = parserCreate(encodingName, &parser->m_mem, NULL, newDtd); - } - - if (!parser) - return NULL; - - startElementHandler = oldStartElementHandler; - endElementHandler = oldEndElementHandler; - characterDataHandler = oldCharacterDataHandler; - processingInstructionHandler = oldProcessingInstructionHandler; - commentHandler = oldCommentHandler; - startCdataSectionHandler = oldStartCdataSectionHandler; - endCdataSectionHandler = oldEndCdataSectionHandler; - defaultHandler = oldDefaultHandler; - unparsedEntityDeclHandler = oldUnparsedEntityDeclHandler; - notationDeclHandler = oldNotationDeclHandler; - startNamespaceDeclHandler = oldStartNamespaceDeclHandler; - endNamespaceDeclHandler = oldEndNamespaceDeclHandler; - notStandaloneHandler = oldNotStandaloneHandler; - externalEntityRefHandler = oldExternalEntityRefHandler; - skippedEntityHandler = oldSkippedEntityHandler; - unknownEncodingHandler = oldUnknownEncodingHandler; - elementDeclHandler = oldElementDeclHandler; - attlistDeclHandler = oldAttlistDeclHandler; - entityDeclHandler = oldEntityDeclHandler; - xmlDeclHandler = oldXmlDeclHandler; - declElementType = oldDeclElementType; - userData = oldUserData; - if (oldUserData == oldHandlerArg) - handlerArg = userData; - else - handlerArg = parser; - if (oldExternalEntityRefHandlerArg != oldParser) - externalEntityRefHandlerArg = oldExternalEntityRefHandlerArg; - defaultExpandInternalEntities = oldDefaultExpandInternalEntities; - ns_triplets = oldns_triplets; - parentParser = oldParser; -#ifdef XML_DTD - paramEntityParsing = oldParamEntityParsing; - prologState.inEntityValue = oldInEntityValue; - if (context) { -#endif /* XML_DTD */ - if (!dtdCopy(_dtd, oldDtd, &parser->m_mem) - || !setContext(parser, context)) { - XML_ParserFree(parser); - return NULL; - } - processor = externalEntityInitProcessor; -#ifdef XML_DTD - } - else { - /* The DTD instance referenced by _dtd is shared between the document's - root parser and external PE parsers, therefore one does not need to - call setContext. In addition, one also *must* not call setContext, - because this would overwrite existing prefix->binding pointers in - _dtd with ones that get destroyed with the external PE parser. - This would leave those prefixes with dangling pointers. - */ - isParamEntity = XML_TRUE; - XmlPrologStateInitExternalEntity(&prologState); - processor = externalParEntInitProcessor; - } -#endif /* XML_DTD */ - return parser; -} - -static void FASTCALL -destroyBindings(BINDING *bindings, XML_Parser parser) -{ - for (;;) { - BINDING *b = bindings; - if (!b) - break; - bindings = b->nextTagBinding; - FREE(b->uri); - FREE(b); - } -} - -void XMLCALL -XML_ParserFree(XML_Parser parser) -{ - TAG *tagList; - OPEN_INTERNAL_ENTITY *entityList; - if (parser == NULL) - return; - /* free tagStack and freeTagList */ - tagList = tagStack; - for (;;) { - TAG *p; - if (tagList == NULL) { - if (freeTagList == NULL) - break; - tagList = freeTagList; - freeTagList = NULL; - } - p = tagList; - tagList = tagList->parent; - FREE(p->buf); - destroyBindings(p->bindings, parser); - FREE(p); - } - /* free openInternalEntities and freeInternalEntities */ - entityList = openInternalEntities; - for (;;) { - OPEN_INTERNAL_ENTITY *openEntity; - if (entityList == NULL) { - if (freeInternalEntities == NULL) - break; - entityList = freeInternalEntities; - freeInternalEntities = NULL; - } - openEntity = entityList; - entityList = entityList->next; - FREE(openEntity); - } - - destroyBindings(freeBindingList, parser); - destroyBindings(inheritedBindings, parser); - poolDestroy(&tempPool); - poolDestroy(&temp2Pool); -#ifdef XML_DTD - /* external parameter entity parsers share the DTD structure - parser->m_dtd with the root parser, so we must not destroy it - */ - if (!isParamEntity && _dtd) -#else - if (_dtd) -#endif /* XML_DTD */ - dtdDestroy(_dtd, (XML_Bool)!parentParser, &parser->m_mem); - FREE((void *)atts); - FREE(groupConnector); - FREE(buffer); - FREE(dataBuf); - FREE(nsAtts); - FREE(unknownEncodingMem); - if (unknownEncodingRelease) - unknownEncodingRelease(unknownEncodingData); - FREE(parser); -} - -void XMLCALL -XML_UseParserAsHandlerArg(XML_Parser parser) -{ - handlerArg = parser; -} - -enum XML_Error XMLCALL -XML_UseForeignDTD(XML_Parser parser, XML_Bool useDTD) -{ -#ifdef XML_DTD - /* block after XML_Parse()/XML_ParseBuffer() has been called */ - if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) - return XML_ERROR_CANT_CHANGE_FEATURE_ONCE_PARSING; - useForeignDTD = useDTD; - return XML_ERROR_NONE; -#else - return XML_ERROR_FEATURE_REQUIRES_XML_DTD; -#endif -} - -void XMLCALL -XML_SetReturnNSTriplet(XML_Parser parser, int do_nst) -{ - /* block after XML_Parse()/XML_ParseBuffer() has been called */ - if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) - return; - ns_triplets = do_nst ? XML_TRUE : XML_FALSE; -} - -void XMLCALL -XML_SetUserData(XML_Parser parser, void *p) -{ - if (handlerArg == userData) - handlerArg = userData = p; - else - userData = p; -} - -enum XML_Status XMLCALL -XML_SetBase(XML_Parser parser, const XML_Char *p) -{ - if (p) { - p = poolCopyString(&_dtd->pool, p); - if (!p) - return XML_STATUS_ERROR; - curBase = p; - } - else - curBase = NULL; - return XML_STATUS_OK; -} - -const XML_Char * XMLCALL -XML_GetBase(XML_Parser parser) -{ - return curBase; -} - -int XMLCALL -XML_GetSpecifiedAttributeCount(XML_Parser parser) -{ - return nSpecifiedAtts; -} - -int XMLCALL -XML_GetIdAttributeIndex(XML_Parser parser) -{ - return idAttIndex; -} - -void XMLCALL -XML_SetElementHandler(XML_Parser parser, - XML_StartElementHandler start, - XML_EndElementHandler end) -{ - startElementHandler = start; - endElementHandler = end; -} - -void XMLCALL -XML_SetStartElementHandler(XML_Parser parser, - XML_StartElementHandler start) { - startElementHandler = start; -} - -void XMLCALL -XML_SetEndElementHandler(XML_Parser parser, - XML_EndElementHandler end) { - endElementHandler = end; -} - -void XMLCALL -XML_SetCharacterDataHandler(XML_Parser parser, - XML_CharacterDataHandler handler) -{ - characterDataHandler = handler; -} - -void XMLCALL -XML_SetProcessingInstructionHandler(XML_Parser parser, - XML_ProcessingInstructionHandler handler) -{ - processingInstructionHandler = handler; -} - -void XMLCALL -XML_SetCommentHandler(XML_Parser parser, - XML_CommentHandler handler) -{ - commentHandler = handler; -} - -void XMLCALL -XML_SetCdataSectionHandler(XML_Parser parser, - XML_StartCdataSectionHandler start, - XML_EndCdataSectionHandler end) -{ - startCdataSectionHandler = start; - endCdataSectionHandler = end; -} - -void XMLCALL -XML_SetStartCdataSectionHandler(XML_Parser parser, - XML_StartCdataSectionHandler start) { - startCdataSectionHandler = start; -} - -void XMLCALL -XML_SetEndCdataSectionHandler(XML_Parser parser, - XML_EndCdataSectionHandler end) { - endCdataSectionHandler = end; -} - -void XMLCALL -XML_SetDefaultHandler(XML_Parser parser, - XML_DefaultHandler handler) -{ - defaultHandler = handler; - defaultExpandInternalEntities = XML_FALSE; -} - -void XMLCALL -XML_SetDefaultHandlerExpand(XML_Parser parser, - XML_DefaultHandler handler) -{ - defaultHandler = handler; - defaultExpandInternalEntities = XML_TRUE; -} - -void XMLCALL -XML_SetDoctypeDeclHandler(XML_Parser parser, - XML_StartDoctypeDeclHandler start, - XML_EndDoctypeDeclHandler end) -{ - startDoctypeDeclHandler = start; - endDoctypeDeclHandler = end; -} - -void XMLCALL -XML_SetStartDoctypeDeclHandler(XML_Parser parser, - XML_StartDoctypeDeclHandler start) { - startDoctypeDeclHandler = start; -} - -void XMLCALL -XML_SetEndDoctypeDeclHandler(XML_Parser parser, - XML_EndDoctypeDeclHandler end) { - endDoctypeDeclHandler = end; -} - -void XMLCALL -XML_SetUnparsedEntityDeclHandler(XML_Parser parser, - XML_UnparsedEntityDeclHandler handler) -{ - unparsedEntityDeclHandler = handler; -} - -void XMLCALL -XML_SetNotationDeclHandler(XML_Parser parser, - XML_NotationDeclHandler handler) -{ - notationDeclHandler = handler; -} - -void XMLCALL -XML_SetNamespaceDeclHandler(XML_Parser parser, - XML_StartNamespaceDeclHandler start, - XML_EndNamespaceDeclHandler end) -{ - startNamespaceDeclHandler = start; - endNamespaceDeclHandler = end; -} - -void XMLCALL -XML_SetStartNamespaceDeclHandler(XML_Parser parser, - XML_StartNamespaceDeclHandler start) { - startNamespaceDeclHandler = start; -} - -void XMLCALL -XML_SetEndNamespaceDeclHandler(XML_Parser parser, - XML_EndNamespaceDeclHandler end) { - endNamespaceDeclHandler = end; -} - -void XMLCALL -XML_SetNotStandaloneHandler(XML_Parser parser, - XML_NotStandaloneHandler handler) -{ - notStandaloneHandler = handler; -} - -void XMLCALL -XML_SetExternalEntityRefHandler(XML_Parser parser, - XML_ExternalEntityRefHandler handler) -{ - externalEntityRefHandler = handler; -} - -void XMLCALL -XML_SetExternalEntityRefHandlerArg(XML_Parser parser, void *arg) -{ - if (arg) - externalEntityRefHandlerArg = (XML_Parser)arg; - else - externalEntityRefHandlerArg = parser; -} - -void XMLCALL -XML_SetSkippedEntityHandler(XML_Parser parser, - XML_SkippedEntityHandler handler) -{ - skippedEntityHandler = handler; -} - -void XMLCALL -XML_SetUnknownEncodingHandler(XML_Parser parser, - XML_UnknownEncodingHandler handler, - void *data) -{ - unknownEncodingHandler = handler; - unknownEncodingHandlerData = data; -} - -void XMLCALL -XML_SetElementDeclHandler(XML_Parser parser, - XML_ElementDeclHandler eldecl) -{ - elementDeclHandler = eldecl; -} - -void XMLCALL -XML_SetAttlistDeclHandler(XML_Parser parser, - XML_AttlistDeclHandler attdecl) -{ - attlistDeclHandler = attdecl; -} - -void XMLCALL -XML_SetEntityDeclHandler(XML_Parser parser, - XML_EntityDeclHandler handler) -{ - entityDeclHandler = handler; -} - -void XMLCALL -XML_SetXmlDeclHandler(XML_Parser parser, - XML_XmlDeclHandler handler) { - xmlDeclHandler = handler; -} - -int XMLCALL -XML_SetParamEntityParsing(XML_Parser parser, - enum XML_ParamEntityParsing peParsing) -{ - /* block after XML_Parse()/XML_ParseBuffer() has been called */ - if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) - return 0; -#ifdef XML_DTD - paramEntityParsing = peParsing; - return 1; -#else - return peParsing == XML_PARAM_ENTITY_PARSING_NEVER; -#endif -} - -enum XML_Status XMLCALL -XML_Parse(XML_Parser parser, const char *s, int len, int isFinal) -{ - switch (ps_parsing) { - case XML_SUSPENDED: - errorCode = XML_ERROR_SUSPENDED; - return XML_STATUS_ERROR; - case XML_FINISHED: - errorCode = XML_ERROR_FINISHED; - return XML_STATUS_ERROR; - default: - ps_parsing = XML_PARSING; - } - - if (len == 0) { - ps_finalBuffer = (XML_Bool)isFinal; - if (!isFinal) - return XML_STATUS_OK; - positionPtr = bufferPtr; - parseEndPtr = bufferEnd; - - /* If data are left over from last buffer, and we now know that these - data are the final chunk of input, then we have to check them again - to detect errors based on that fact. - */ - errorCode = processor(parser, bufferPtr, parseEndPtr, &bufferPtr); - - if (errorCode == XML_ERROR_NONE) { - switch (ps_parsing) { - case XML_SUSPENDED: - XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); - positionPtr = bufferPtr; - return XML_STATUS_SUSPENDED; - case XML_INITIALIZED: - case XML_PARSING: - ps_parsing = XML_FINISHED; - /* fall through */ - default: - return XML_STATUS_OK; - } - } - eventEndPtr = eventPtr; - processor = errorProcessor; - return XML_STATUS_ERROR; - } -#ifndef XML_CONTEXT_BYTES - else if (bufferPtr == bufferEnd) { - const char *end; - int nLeftOver; - enum XML_Error result; - parseEndByteIndex += len; - positionPtr = s; - ps_finalBuffer = (XML_Bool)isFinal; - - errorCode = processor(parser, s, parseEndPtr = s + len, &end); - - if (errorCode != XML_ERROR_NONE) { - eventEndPtr = eventPtr; - processor = errorProcessor; - return XML_STATUS_ERROR; - } - else { - switch (ps_parsing) { - case XML_SUSPENDED: - result = XML_STATUS_SUSPENDED; - break; - case XML_INITIALIZED: - case XML_PARSING: - result = XML_STATUS_OK; - if (isFinal) { - ps_parsing = XML_FINISHED; - return result; - } - } - } - - XmlUpdatePosition(encoding, positionPtr, end, &position); - nLeftOver = s + len - end; - if (nLeftOver) { - if (buffer == NULL || nLeftOver > bufferLim - buffer) { - /* FIXME avoid integer overflow */ - char *temp; - temp = (buffer == NULL - ? (char *)MALLOC(len * 2) - : (char *)REALLOC(buffer, len * 2)); - if (temp == NULL) { - errorCode = XML_ERROR_NO_MEMORY; - return XML_STATUS_ERROR; - } - buffer = temp; - if (!buffer) { - errorCode = XML_ERROR_NO_MEMORY; - eventPtr = eventEndPtr = NULL; - processor = errorProcessor; - return XML_STATUS_ERROR; - } - bufferLim = buffer + len * 2; - } - memcpy(buffer, end, nLeftOver); - } - bufferPtr = buffer; - bufferEnd = buffer + nLeftOver; - positionPtr = bufferPtr; - parseEndPtr = bufferEnd; - eventPtr = bufferPtr; - eventEndPtr = bufferPtr; - return result; - } -#endif /* not defined XML_CONTEXT_BYTES */ - else { - void *buff = XML_GetBuffer(parser, len); - if (buff == NULL) - return XML_STATUS_ERROR; - else { - memcpy(buff, s, len); - return XML_ParseBuffer(parser, len, isFinal); - } - } -} - -enum XML_Status XMLCALL -XML_ParseBuffer(XML_Parser parser, int len, int isFinal) -{ - const char *start; - enum XML_Status result = XML_STATUS_OK; - - switch (ps_parsing) { - case XML_SUSPENDED: - errorCode = XML_ERROR_SUSPENDED; - return XML_STATUS_ERROR; - case XML_FINISHED: - errorCode = XML_ERROR_FINISHED; - return XML_STATUS_ERROR; - default: - ps_parsing = XML_PARSING; - } - - start = bufferPtr; - positionPtr = start; - bufferEnd += len; - parseEndPtr = bufferEnd; - parseEndByteIndex += len; - ps_finalBuffer = (XML_Bool)isFinal; - - errorCode = processor(parser, start, parseEndPtr, &bufferPtr); - - if (errorCode != XML_ERROR_NONE) { - eventEndPtr = eventPtr; - processor = errorProcessor; - return XML_STATUS_ERROR; - } - else { - switch (ps_parsing) { - case XML_SUSPENDED: - result = XML_STATUS_SUSPENDED; - break; - case XML_INITIALIZED: - case XML_PARSING: - if (isFinal) { - ps_parsing = XML_FINISHED; - return result; - } - default: ; /* should not happen */ - } - } - - XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); - positionPtr = bufferPtr; - return result; -} - -void * XMLCALL -XML_GetBuffer(XML_Parser parser, int len) -{ - switch (ps_parsing) { - case XML_SUSPENDED: - errorCode = XML_ERROR_SUSPENDED; - return NULL; - case XML_FINISHED: - errorCode = XML_ERROR_FINISHED; - return NULL; - default: ; - } - - if (len > bufferLim - bufferEnd) { - /* FIXME avoid integer overflow */ - int neededSize = len + (int)(bufferEnd - bufferPtr); -#ifdef XML_CONTEXT_BYTES - int keep = (int)(bufferPtr - buffer); - - if (keep > XML_CONTEXT_BYTES) - keep = XML_CONTEXT_BYTES; - neededSize += keep; -#endif /* defined XML_CONTEXT_BYTES */ - if (neededSize <= bufferLim - buffer) { -#ifdef XML_CONTEXT_BYTES - if (keep < bufferPtr - buffer) { - int offset = (int)(bufferPtr - buffer) - keep; - memmove(buffer, &buffer[offset], bufferEnd - bufferPtr + keep); - bufferEnd -= offset; - bufferPtr -= offset; - } -#else - memmove(buffer, bufferPtr, bufferEnd - bufferPtr); - bufferEnd = buffer + (bufferEnd - bufferPtr); - bufferPtr = buffer; -#endif /* not defined XML_CONTEXT_BYTES */ - } - else { - char *newBuf; - int bufferSize = (int)(bufferLim - bufferPtr); - if (bufferSize == 0) - bufferSize = INIT_BUFFER_SIZE; - do { - bufferSize *= 2; - } while (bufferSize < neededSize); - newBuf = (char *)MALLOC(bufferSize); - if (newBuf == 0) { - errorCode = XML_ERROR_NO_MEMORY; - return NULL; - } - bufferLim = newBuf + bufferSize; -#ifdef XML_CONTEXT_BYTES - if (bufferPtr) { - int keep = (int)(bufferPtr - buffer); - if (keep > XML_CONTEXT_BYTES) - keep = XML_CONTEXT_BYTES; - memcpy(newBuf, &bufferPtr[-keep], bufferEnd - bufferPtr + keep); - FREE(buffer); - buffer = newBuf; - bufferEnd = buffer + (bufferEnd - bufferPtr) + keep; - bufferPtr = buffer + keep; - } - else { - bufferEnd = newBuf + (bufferEnd - bufferPtr); - bufferPtr = buffer = newBuf; - } -#else - if (bufferPtr) { - memcpy(newBuf, bufferPtr, bufferEnd - bufferPtr); - FREE(buffer); - } - bufferEnd = newBuf + (bufferEnd - bufferPtr); - bufferPtr = buffer = newBuf; -#endif /* not defined XML_CONTEXT_BYTES */ - } - } - return bufferEnd; -} - -enum XML_Status XMLCALL -XML_StopParser(XML_Parser parser, XML_Bool resumable) -{ - switch (ps_parsing) { - case XML_SUSPENDED: - if (resumable) { - errorCode = XML_ERROR_SUSPENDED; - return XML_STATUS_ERROR; - } - ps_parsing = XML_FINISHED; - break; - case XML_FINISHED: - errorCode = XML_ERROR_FINISHED; - return XML_STATUS_ERROR; - default: - if (resumable) { -#ifdef XML_DTD - if (isParamEntity) { - errorCode = XML_ERROR_SUSPEND_PE; - return XML_STATUS_ERROR; - } -#endif - ps_parsing = XML_SUSPENDED; - } - else - ps_parsing = XML_FINISHED; - } - return XML_STATUS_OK; -} - -enum XML_Status XMLCALL -XML_ResumeParser(XML_Parser parser) -{ - enum XML_Status result = XML_STATUS_OK; - - if (ps_parsing != XML_SUSPENDED) { - errorCode = XML_ERROR_NOT_SUSPENDED; - return XML_STATUS_ERROR; - } - ps_parsing = XML_PARSING; - - errorCode = processor(parser, bufferPtr, parseEndPtr, &bufferPtr); - - if (errorCode != XML_ERROR_NONE) { - eventEndPtr = eventPtr; - processor = errorProcessor; - return XML_STATUS_ERROR; - } - else { - switch (ps_parsing) { - case XML_SUSPENDED: - result = XML_STATUS_SUSPENDED; - break; - case XML_INITIALIZED: - case XML_PARSING: - if (ps_finalBuffer) { - ps_parsing = XML_FINISHED; - return result; - } - default: ; - } - } - - XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); - positionPtr = bufferPtr; - return result; -} - -void XMLCALL -XML_GetParsingStatus(XML_Parser parser, XML_ParsingStatus *status) -{ - assert(status != NULL); - *status = parser->m_parsingStatus; -} - -enum XML_Error XMLCALL -XML_GetErrorCode(XML_Parser parser) -{ - return errorCode; -} - -XML_Index XMLCALL -XML_GetCurrentByteIndex(XML_Parser parser) -{ - if (eventPtr) - return parseEndByteIndex - (parseEndPtr - eventPtr); - return -1; -} - -int XMLCALL -XML_GetCurrentByteCount(XML_Parser parser) -{ - if (eventEndPtr && eventPtr) - return (int)(eventEndPtr - eventPtr); - return 0; -} - -const char * XMLCALL -XML_GetInputContext(XML_Parser parser, int *offset, int *size) -{ -#ifdef XML_CONTEXT_BYTES - if (eventPtr && buffer) { - *offset = (int)(eventPtr - buffer); - *size = (int)(bufferEnd - buffer); - return buffer; - } -#endif /* defined XML_CONTEXT_BYTES */ - return (char *) 0; -} - -XML_Size XMLCALL -XML_GetCurrentLineNumber(XML_Parser parser) -{ - if (eventPtr && eventPtr >= positionPtr) { - XmlUpdatePosition(encoding, positionPtr, eventPtr, &position); - positionPtr = eventPtr; - } - return position.lineNumber + 1; -} - -XML_Size XMLCALL -XML_GetCurrentColumnNumber(XML_Parser parser) -{ - if (eventPtr && eventPtr >= positionPtr) { - XmlUpdatePosition(encoding, positionPtr, eventPtr, &position); - positionPtr = eventPtr; - } - return position.columnNumber; -} - -void XMLCALL -XML_FreeContentModel(XML_Parser parser, XML_Content *model) -{ - FREE(model); -} - -void * XMLCALL -XML_MemMalloc(XML_Parser parser, size_t size) -{ - return MALLOC(size); -} - -void * XMLCALL -XML_MemRealloc(XML_Parser parser, void *ptr, size_t size) -{ - return REALLOC(ptr, size); -} - -void XMLCALL -XML_MemFree(XML_Parser parser, void *ptr) -{ - FREE(ptr); -} - -void XMLCALL -XML_DefaultCurrent(XML_Parser parser) -{ - if (defaultHandler) { - if (openInternalEntities) - reportDefault(parser, - internalEncoding, - openInternalEntities->internalEventPtr, - openInternalEntities->internalEventEndPtr); - else - reportDefault(parser, encoding, eventPtr, eventEndPtr); - } -} - -const XML_LChar * XMLCALL -XML_ErrorString(enum XML_Error code) -{ - static const XML_LChar* const message[] = { - 0, - XML_L("out of memory"), - XML_L("syntax error"), - XML_L("no element found"), - XML_L("not well-formed (invalid token)"), - XML_L("unclosed token"), - XML_L("partial character"), - XML_L("mismatched tag"), - XML_L("duplicate attribute"), - XML_L("junk after document element"), - XML_L("illegal parameter entity reference"), - XML_L("undefined entity"), - XML_L("recursive entity reference"), - XML_L("asynchronous entity"), - XML_L("reference to invalid character number"), - XML_L("reference to binary entity"), - XML_L("reference to external entity in attribute"), - XML_L("XML or text declaration not at start of entity"), - XML_L("unknown encoding"), - XML_L("encoding specified in XML declaration is incorrect"), - XML_L("unclosed CDATA section"), - XML_L("error in processing external entity reference"), - XML_L("document is not standalone"), - XML_L("unexpected parser state - please send a bug report"), - XML_L("entity declared in parameter entity"), - XML_L("requested feature requires XML_DTD support in Expat"), - XML_L("cannot change setting once parsing has begun"), - XML_L("unbound prefix"), - XML_L("must not undeclare prefix"), - XML_L("incomplete markup in parameter entity"), - XML_L("XML declaration not well-formed"), - XML_L("text declaration not well-formed"), - XML_L("illegal character(s) in public id"), - XML_L("parser suspended"), - XML_L("parser not suspended"), - XML_L("parsing aborted"), - XML_L("parsing finished"), - XML_L("cannot suspend in external parameter entity"), - XML_L("reserved prefix (xml) must not be undeclared or bound to another namespace name"), - XML_L("reserved prefix (xmlns) must not be declared or undeclared"), - XML_L("prefix must not be bound to one of the reserved namespace names") - }; - if (code > 0 && code < sizeof(message)/sizeof(message[0])) - return message[code]; - return NULL; -} - -const XML_LChar * XMLCALL -XML_ExpatVersion(void) { - - /* V1 is used to string-ize the version number. However, it would - string-ize the actual version macro *names* unless we get them - substituted before being passed to V1. CPP is defined to expand - a macro, then rescan for more expansions. Thus, we use V2 to expand - the version macros, then CPP will expand the resulting V1() macro - with the correct numerals. */ - /* ### I'm assuming cpp is portable in this respect... */ - -#define V1(a,b,c) XML_L(#a)XML_L(".")XML_L(#b)XML_L(".")XML_L(#c) -#define V2(a,b,c) XML_L("expat_")V1(a,b,c) - - return V2(XML_MAJOR_VERSION, XML_MINOR_VERSION, XML_MICRO_VERSION); - -#undef V1 -#undef V2 -} - -XML_Expat_Version XMLCALL -XML_ExpatVersionInfo(void) -{ - XML_Expat_Version version; - - version.major = XML_MAJOR_VERSION; - version.minor = XML_MINOR_VERSION; - version.micro = XML_MICRO_VERSION; - - return version; -} - -const XML_Feature * XMLCALL -XML_GetFeatureList(void) -{ - static const XML_Feature features[] = { - {XML_FEATURE_SIZEOF_XML_CHAR, XML_L("sizeof(XML_Char)"), - sizeof(XML_Char)}, - {XML_FEATURE_SIZEOF_XML_LCHAR, XML_L("sizeof(XML_LChar)"), - sizeof(XML_LChar)}, -#ifdef XML_UNICODE - {XML_FEATURE_UNICODE, XML_L("XML_UNICODE"), 0}, -#endif -#ifdef XML_UNICODE_WCHAR_T - {XML_FEATURE_UNICODE_WCHAR_T, XML_L("XML_UNICODE_WCHAR_T"), 0}, -#endif -#ifdef XML_DTD - {XML_FEATURE_DTD, XML_L("XML_DTD"), 0}, -#endif -#ifdef XML_CONTEXT_BYTES - {XML_FEATURE_CONTEXT_BYTES, XML_L("XML_CONTEXT_BYTES"), - XML_CONTEXT_BYTES}, -#endif -#ifdef XML_MIN_SIZE - {XML_FEATURE_MIN_SIZE, XML_L("XML_MIN_SIZE"), 0}, -#endif -#ifdef XML_NS - {XML_FEATURE_NS, XML_L("XML_NS"), 0}, -#endif -#ifdef XML_LARGE_SIZE - {XML_FEATURE_LARGE_SIZE, XML_L("XML_LARGE_SIZE"), 0}, -#endif - {XML_FEATURE_END, NULL, 0} - }; - - return features; -} - -/* Initially tag->rawName always points into the parse buffer; - for those TAG instances opened while the current parse buffer was - processed, and not yet closed, we need to store tag->rawName in a more - permanent location, since the parse buffer is about to be discarded. -*/ -static XML_Bool -storeRawNames(XML_Parser parser) -{ - TAG *tag = tagStack; - while (tag) { - int bufSize; - int nameLen = sizeof(XML_Char) * (tag->name.strLen + 1); - char *rawNameBuf = tag->buf + nameLen; - /* Stop if already stored. Since tagStack is a stack, we can stop - at the first entry that has already been copied; everything - below it in the stack is already been accounted for in a - previous call to this function. - */ - if (tag->rawName == rawNameBuf) - break; - /* For re-use purposes we need to ensure that the - size of tag->buf is a multiple of sizeof(XML_Char). - */ - bufSize = nameLen + ROUND_UP(tag->rawNameLength, sizeof(XML_Char)); - if (bufSize > tag->bufEnd - tag->buf) { - char *temp = (char *)REALLOC(tag->buf, bufSize); - if (temp == NULL) - return XML_FALSE; - /* if tag->name.str points to tag->buf (only when namespace - processing is off) then we have to update it - */ - if (tag->name.str == (XML_Char *)tag->buf) - tag->name.str = (XML_Char *)temp; - /* if tag->name.localPart is set (when namespace processing is on) - then update it as well, since it will always point into tag->buf - */ - if (tag->name.localPart) - tag->name.localPart = (XML_Char *)temp + (tag->name.localPart - - (XML_Char *)tag->buf); - tag->buf = temp; - tag->bufEnd = temp + bufSize; - rawNameBuf = temp + nameLen; - } - memcpy(rawNameBuf, tag->rawName, tag->rawNameLength); - tag->rawName = rawNameBuf; - tag = tag->parent; - } - return XML_TRUE; -} - -static enum XML_Error PTRCALL -contentProcessor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - enum XML_Error result = doContent(parser, 0, encoding, start, end, - endPtr, (XML_Bool)!ps_finalBuffer); - if (result == XML_ERROR_NONE) { - if (!storeRawNames(parser)) - return XML_ERROR_NO_MEMORY; - } - return result; -} - -static enum XML_Error PTRCALL -externalEntityInitProcessor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - enum XML_Error result = initializeEncoding(parser); - if (result != XML_ERROR_NONE) - return result; - processor = externalEntityInitProcessor2; - return externalEntityInitProcessor2(parser, start, end, endPtr); -} - -static enum XML_Error PTRCALL -externalEntityInitProcessor2(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - const char *next = start; /* XmlContentTok doesn't always set the last arg */ - int tok = XmlContentTok(encoding, start, end, &next); - switch (tok) { - case XML_TOK_BOM: - /* If we are at the end of the buffer, this would cause the next stage, - i.e. externalEntityInitProcessor3, to pass control directly to - doContent (by detecting XML_TOK_NONE) without processing any xml text - declaration - causing the error XML_ERROR_MISPLACED_XML_PI in doContent. - */ - if (next == end && !ps_finalBuffer) { - *endPtr = next; - return XML_ERROR_NONE; - } - start = next; - break; - case XML_TOK_PARTIAL: - if (!ps_finalBuffer) { - *endPtr = start; - return XML_ERROR_NONE; - } - eventPtr = start; - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (!ps_finalBuffer) { - *endPtr = start; - return XML_ERROR_NONE; - } - eventPtr = start; - return XML_ERROR_PARTIAL_CHAR; - } - processor = externalEntityInitProcessor3; - return externalEntityInitProcessor3(parser, start, end, endPtr); -} - -static enum XML_Error PTRCALL -externalEntityInitProcessor3(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - int tok; - const char *next = start; /* XmlContentTok doesn't always set the last arg */ - eventPtr = start; - tok = XmlContentTok(encoding, start, end, &next); - eventEndPtr = next; - - switch (tok) { - case XML_TOK_XML_DECL: - { - enum XML_Error result; - result = processXmlDecl(parser, 1, start, next); - if (result != XML_ERROR_NONE) - return result; - switch (ps_parsing) { - case XML_SUSPENDED: - *endPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: - start = next; - } - } - break; - case XML_TOK_PARTIAL: - if (!ps_finalBuffer) { - *endPtr = start; - return XML_ERROR_NONE; - } - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (!ps_finalBuffer) { - *endPtr = start; - return XML_ERROR_NONE; - } - return XML_ERROR_PARTIAL_CHAR; - } - processor = externalEntityContentProcessor; - tagLevel = 1; - return externalEntityContentProcessor(parser, start, end, endPtr); -} - -static enum XML_Error PTRCALL -externalEntityContentProcessor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - enum XML_Error result = doContent(parser, 1, encoding, start, end, - endPtr, (XML_Bool)!ps_finalBuffer); - if (result == XML_ERROR_NONE) { - if (!storeRawNames(parser)) - return XML_ERROR_NO_MEMORY; - } - return result; -} - -static enum XML_Error -doContent(XML_Parser parser, - int startTagLevel, - const ENCODING *enc, - const char *s, - const char *end, - const char **nextPtr, - XML_Bool haveMore) -{ - /* save one level of indirection */ - DTD * const dtd = _dtd; - - const char **eventPP; - const char **eventEndPP; - if (enc == encoding) { - eventPP = &eventPtr; - eventEndPP = &eventEndPtr; - } - else { - eventPP = &(openInternalEntities->internalEventPtr); - eventEndPP = &(openInternalEntities->internalEventEndPtr); - } - *eventPP = s; - - for (;;) { - const char *next = s; /* XmlContentTok doesn't always set the last arg */ - int tok = XmlContentTok(enc, s, end, &next); - *eventEndPP = next; - switch (tok) { - case XML_TOK_TRAILING_CR: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - *eventEndPP = end; - if (characterDataHandler) { - XML_Char c = 0xA; - characterDataHandler(handlerArg, &c, 1); - } - else if (defaultHandler) - reportDefault(parser, enc, s, end); - /* We are at the end of the final buffer, should we check for - XML_SUSPENDED, XML_FINISHED? - */ - if (startTagLevel == 0) - return XML_ERROR_NO_ELEMENTS; - if (tagLevel != startTagLevel) - return XML_ERROR_ASYNC_ENTITY; - *nextPtr = end; - return XML_ERROR_NONE; - case XML_TOK_NONE: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - if (startTagLevel > 0) { - if (tagLevel != startTagLevel) - return XML_ERROR_ASYNC_ENTITY; - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_NO_ELEMENTS; - case XML_TOK_INVALID: - *eventPP = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_ENTITY_REF: - { - const XML_Char *name; - ENTITY *entity; - XML_Char ch = (XML_Char) XmlPredefinedEntityName(enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (ch) { - if (characterDataHandler) - characterDataHandler(handlerArg, &ch, 1); - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - } - name = poolStoreString(&dtd->pool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!name) - return XML_ERROR_NO_MEMORY; - entity = (ENTITY *)lookup(&dtd->generalEntities, name, 0); - poolDiscard(&dtd->pool); - /* First, determine if a check for an existing declaration is needed; - if yes, check that the entity exists, and that it is internal, - otherwise call the skipped entity or default handler. - */ - if (!dtd->hasParamEntityRefs || dtd->standalone) { - if (!entity) - return XML_ERROR_UNDEFINED_ENTITY; - else if (!entity->is_internal) - return XML_ERROR_ENTITY_DECLARED_IN_PE; - } - else if (!entity) { - if (skippedEntityHandler) - skippedEntityHandler(handlerArg, name, 0); - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - } - if (entity->open) - return XML_ERROR_RECURSIVE_ENTITY_REF; - if (entity->notation) - return XML_ERROR_BINARY_ENTITY_REF; - if (entity->textPtr) { - enum XML_Error result; - if (!defaultExpandInternalEntities) { - if (skippedEntityHandler) - skippedEntityHandler(handlerArg, entity->name, 0); - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - } - result = processInternalEntity(parser, entity, XML_FALSE); - if (result != XML_ERROR_NONE) - return result; - } - else if (externalEntityRefHandler) { - const XML_Char *context; - entity->open = XML_TRUE; - context = getContext(parser); - entity->open = XML_FALSE; - if (!context) - return XML_ERROR_NO_MEMORY; - if (!externalEntityRefHandler(externalEntityRefHandlerArg, - context, - entity->base, - entity->systemId, - entity->publicId)) - return XML_ERROR_EXTERNAL_ENTITY_HANDLING; - poolDiscard(&tempPool); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - } - case XML_TOK_START_TAG_NO_ATTS: - /* fall through */ - case XML_TOK_START_TAG_WITH_ATTS: - { - TAG *tag; - enum XML_Error result; - XML_Char *toPtr; - if (freeTagList) { - tag = freeTagList; - freeTagList = freeTagList->parent; - } - else { - tag = (TAG *)MALLOC(sizeof(TAG)); - if (!tag) - return XML_ERROR_NO_MEMORY; - tag->buf = (char *)MALLOC(INIT_TAG_BUF_SIZE); - if (!tag->buf) { - FREE(tag); - return XML_ERROR_NO_MEMORY; - } - tag->bufEnd = tag->buf + INIT_TAG_BUF_SIZE; - } - tag->bindings = NULL; - tag->parent = tagStack; - tagStack = tag; - tag->name.localPart = NULL; - tag->name.prefix = NULL; - tag->rawName = s + enc->minBytesPerChar; - tag->rawNameLength = XmlNameLength(enc, tag->rawName); - ++tagLevel; - { - const char *rawNameEnd = tag->rawName + tag->rawNameLength; - const char *fromPtr = tag->rawName; - toPtr = (XML_Char *)tag->buf; - for (;;) { - int bufSize; - int convLen; - XmlConvert(enc, - &fromPtr, rawNameEnd, - (ICHAR **)&toPtr, (ICHAR *)tag->bufEnd - 1); - convLen = (int)(toPtr - (XML_Char *)tag->buf); - if (fromPtr == rawNameEnd) { - tag->name.strLen = convLen; - break; - } - bufSize = (int)(tag->bufEnd - tag->buf) << 1; - { - char *temp = (char *)REALLOC(tag->buf, bufSize); - if (temp == NULL) - return XML_ERROR_NO_MEMORY; - tag->buf = temp; - tag->bufEnd = temp + bufSize; - toPtr = (XML_Char *)temp + convLen; - } - } - } - tag->name.str = (XML_Char *)tag->buf; - *toPtr = XML_T('\0'); - result = storeAtts(parser, enc, s, &(tag->name), &(tag->bindings)); - if (result) - return result; - if (startElementHandler) - startElementHandler(handlerArg, tag->name.str, - (const XML_Char **)atts); - else if (defaultHandler) - reportDefault(parser, enc, s, next); - poolClear(&tempPool); - break; - } - case XML_TOK_EMPTY_ELEMENT_NO_ATTS: - /* fall through */ - case XML_TOK_EMPTY_ELEMENT_WITH_ATTS: - { - const char *rawName = s + enc->minBytesPerChar; - enum XML_Error result; - BINDING *bindings = NULL; - XML_Bool noElmHandlers = XML_TRUE; - TAG_NAME name; - name.str = poolStoreString(&tempPool, enc, rawName, - rawName + XmlNameLength(enc, rawName)); - if (!name.str) - return XML_ERROR_NO_MEMORY; - poolFinish(&tempPool); - result = storeAtts(parser, enc, s, &name, &bindings); - if (result) - return result; - poolFinish(&tempPool); - if (startElementHandler) { - startElementHandler(handlerArg, name.str, (const XML_Char **)atts); - noElmHandlers = XML_FALSE; - } - if (endElementHandler) { - if (startElementHandler) - *eventPP = *eventEndPP; - endElementHandler(handlerArg, name.str); - noElmHandlers = XML_FALSE; - } - if (noElmHandlers && defaultHandler) - reportDefault(parser, enc, s, next); - poolClear(&tempPool); - while (bindings) { - BINDING *b = bindings; - if (endNamespaceDeclHandler) - endNamespaceDeclHandler(handlerArg, b->prefix->name); - bindings = bindings->nextTagBinding; - b->nextTagBinding = freeBindingList; - freeBindingList = b; - b->prefix->binding = b->prevPrefixBinding; - } - } - if (tagLevel == 0) - return epilogProcessor(parser, next, end, nextPtr); - break; - case XML_TOK_END_TAG: - if (tagLevel == startTagLevel) - return XML_ERROR_ASYNC_ENTITY; - else { - int len; - const char *rawName; - TAG *tag = tagStack; - tagStack = tag->parent; - tag->parent = freeTagList; - freeTagList = tag; - rawName = s + enc->minBytesPerChar*2; - len = XmlNameLength(enc, rawName); - if (len != tag->rawNameLength - || memcmp(tag->rawName, rawName, len) != 0) { - *eventPP = rawName; - return XML_ERROR_TAG_MISMATCH; - } - --tagLevel; - if (endElementHandler) { - const XML_Char *localPart; - const XML_Char *prefix; - XML_Char *uri; - localPart = tag->name.localPart; - if (ns && localPart) { - /* localPart and prefix may have been overwritten in - tag->name.str, since this points to the binding->uri - buffer which gets re-used; so we have to add them again - */ - uri = (XML_Char *)tag->name.str + tag->name.uriLen; - /* don't need to check for space - already done in storeAtts() */ - while (*localPart) *uri++ = *localPart++; - prefix = (XML_Char *)tag->name.prefix; - if (ns_triplets && prefix) { - *uri++ = namespaceSeparator; - while (*prefix) *uri++ = *prefix++; - } - *uri = XML_T('\0'); - } - endElementHandler(handlerArg, tag->name.str); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - while (tag->bindings) { - BINDING *b = tag->bindings; - if (endNamespaceDeclHandler) - endNamespaceDeclHandler(handlerArg, b->prefix->name); - tag->bindings = tag->bindings->nextTagBinding; - b->nextTagBinding = freeBindingList; - freeBindingList = b; - b->prefix->binding = b->prevPrefixBinding; - } - if (tagLevel == 0) - return epilogProcessor(parser, next, end, nextPtr); - } - break; - case XML_TOK_CHAR_REF: - { - int n = XmlCharRefNumber(enc, s); - if (n < 0) - return XML_ERROR_BAD_CHAR_REF; - if (characterDataHandler) { - XML_Char buf[XML_ENCODE_MAX]; - characterDataHandler(handlerArg, buf, XmlEncode(n, (ICHAR *)buf)); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - } - break; - case XML_TOK_XML_DECL: - return XML_ERROR_MISPLACED_XML_PI; - case XML_TOK_DATA_NEWLINE: - if (characterDataHandler) { - XML_Char c = 0xA; - characterDataHandler(handlerArg, &c, 1); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - case XML_TOK_CDATA_SECT_OPEN: - { - enum XML_Error result; - if (startCdataSectionHandler) - startCdataSectionHandler(handlerArg); -#if 0 - /* Suppose you doing a transformation on a document that involves - changing only the character data. You set up a defaultHandler - and a characterDataHandler. The defaultHandler simply copies - characters through. The characterDataHandler does the - transformation and writes the characters out escaping them as - necessary. This case will fail to work if we leave out the - following two lines (because & and < inside CDATA sections will - be incorrectly escaped). - - However, now we have a start/endCdataSectionHandler, so it seems - easier to let the user deal with this. - */ - else if (characterDataHandler) - characterDataHandler(handlerArg, dataBuf, 0); -#endif - else if (defaultHandler) - reportDefault(parser, enc, s, next); - result = doCdataSection(parser, enc, &next, end, nextPtr, haveMore); - if (result != XML_ERROR_NONE) - return result; - else if (!next) { - processor = cdataSectionProcessor; - return result; - } - } - break; - case XML_TOK_TRAILING_RSQB: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - if (characterDataHandler) { - if (MUST_CONVERT(enc, s)) { - ICHAR *dataPtr = (ICHAR *)dataBuf; - XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd); - characterDataHandler(handlerArg, dataBuf, - (int)(dataPtr - (ICHAR *)dataBuf)); - } - else - characterDataHandler(handlerArg, - (XML_Char *)s, - (int)((XML_Char *)end - (XML_Char *)s)); - } - else if (defaultHandler) - reportDefault(parser, enc, s, end); - /* We are at the end of the final buffer, should we check for - XML_SUSPENDED, XML_FINISHED? - */ - if (startTagLevel == 0) { - *eventPP = end; - return XML_ERROR_NO_ELEMENTS; - } - if (tagLevel != startTagLevel) { - *eventPP = end; - return XML_ERROR_ASYNC_ENTITY; - } - *nextPtr = end; - return XML_ERROR_NONE; - case XML_TOK_DATA_CHARS: - { - XML_CharacterDataHandler charDataHandler = characterDataHandler; - if (charDataHandler) { - if (MUST_CONVERT(enc, s)) { - for (;;) { - ICHAR *dataPtr = (ICHAR *)dataBuf; - XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); - *eventEndPP = s; - charDataHandler(handlerArg, dataBuf, - (int)(dataPtr - (ICHAR *)dataBuf)); - if (s == next) - break; - *eventPP = s; - } - } - else - charDataHandler(handlerArg, - (XML_Char *)s, - (int)((XML_Char *)next - (XML_Char *)s)); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - } - break; - case XML_TOK_PI: - if (!reportProcessingInstruction(parser, enc, s, next)) - return XML_ERROR_NO_MEMORY; - break; - case XML_TOK_COMMENT: - if (!reportComment(parser, enc, s, next)) - return XML_ERROR_NO_MEMORY; - break; - default: - if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - } - *eventPP = s = next; - switch (ps_parsing) { - case XML_SUSPENDED: - *nextPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: ; - } - } - /* not reached */ -} - -/* Precondition: all arguments must be non-NULL; - Purpose: - - normalize attributes - - check attributes for well-formedness - - generate namespace aware attribute names (URI, prefix) - - build list of attributes for startElementHandler - - default attributes - - process namespace declarations (check and report them) - - generate namespace aware element name (URI, prefix) -*/ -static enum XML_Error -storeAtts(XML_Parser parser, const ENCODING *enc, - const char *attStr, TAG_NAME *tagNamePtr, - BINDING **bindingsPtr) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - ELEMENT_TYPE *elementType; - int nDefaultAtts; - const XML_Char **appAtts; /* the attribute list for the application */ - int attIndex = 0; - int prefixLen; - int i; - int n; - XML_Char *uri; - int nPrefixes = 0; - BINDING *binding; - const XML_Char *localPart; - - /* lookup the element type name */ - elementType = (ELEMENT_TYPE *)lookup(&dtd->elementTypes, tagNamePtr->str,0); - if (!elementType) { - const XML_Char *name = poolCopyString(&dtd->pool, tagNamePtr->str); - if (!name) - return XML_ERROR_NO_MEMORY; - elementType = (ELEMENT_TYPE *)lookup(&dtd->elementTypes, name, - sizeof(ELEMENT_TYPE)); - if (!elementType) - return XML_ERROR_NO_MEMORY; - if (ns && !setElementTypePrefix(parser, elementType)) - return XML_ERROR_NO_MEMORY; - } - nDefaultAtts = elementType->nDefaultAtts; - - /* get the attributes from the tokenizer */ - n = XmlGetAttributes(enc, attStr, attsSize, atts); - if (n + nDefaultAtts > attsSize) { - int oldAttsSize = attsSize; - ATTRIBUTE *temp; - attsSize = n + nDefaultAtts + INIT_ATTS_SIZE; - temp = (ATTRIBUTE *)REALLOC((void *)atts, attsSize * sizeof(ATTRIBUTE)); - if (temp == NULL) - return XML_ERROR_NO_MEMORY; - atts = temp; - if (n > oldAttsSize) - XmlGetAttributes(enc, attStr, n, atts); - } - - appAtts = (const XML_Char **)atts; - for (i = 0; i < n; i++) { - /* add the name and value to the attribute list */ - ATTRIBUTE_ID *attId = getAttributeId(parser, enc, atts[i].name, - atts[i].name - + XmlNameLength(enc, atts[i].name)); - if (!attId) - return XML_ERROR_NO_MEMORY; - /* Detect duplicate attributes by their QNames. This does not work when - namespace processing is turned on and different prefixes for the same - namespace are used. For this case we have a check further down. - */ - if ((attId->name)[-1]) { - if (enc == encoding) - eventPtr = atts[i].name; - return XML_ERROR_DUPLICATE_ATTRIBUTE; - } - (attId->name)[-1] = 1; - appAtts[attIndex++] = attId->name; - if (!atts[i].normalized) { - enum XML_Error result; - XML_Bool isCdata = XML_TRUE; - - /* figure out whether declared as other than CDATA */ - if (attId->maybeTokenized) { - int j; - for (j = 0; j < nDefaultAtts; j++) { - if (attId == elementType->defaultAtts[j].id) { - isCdata = elementType->defaultAtts[j].isCdata; - break; - } - } - } - - /* normalize the attribute value */ - result = storeAttributeValue(parser, enc, isCdata, - atts[i].valuePtr, atts[i].valueEnd, - &tempPool); - if (result) - return result; - appAtts[attIndex] = poolStart(&tempPool); - poolFinish(&tempPool); - } - else { - /* the value did not need normalizing */ - appAtts[attIndex] = poolStoreString(&tempPool, enc, atts[i].valuePtr, - atts[i].valueEnd); - if (appAtts[attIndex] == 0) - return XML_ERROR_NO_MEMORY; - poolFinish(&tempPool); - } - /* handle prefixed attribute names */ - if (attId->prefix) { - if (attId->xmlns) { - /* deal with namespace declarations here */ - enum XML_Error result = addBinding(parser, attId->prefix, attId, - appAtts[attIndex], bindingsPtr); - if (result) - return result; - --attIndex; - } - else { - /* deal with other prefixed names later */ - attIndex++; - nPrefixes++; - (attId->name)[-1] = 2; - } - } - else - attIndex++; - } - - /* set-up for XML_GetSpecifiedAttributeCount and XML_GetIdAttributeIndex */ - nSpecifiedAtts = attIndex; - if (elementType->idAtt && (elementType->idAtt->name)[-1]) { - for (i = 0; i < attIndex; i += 2) - if (appAtts[i] == elementType->idAtt->name) { - idAttIndex = i; - break; - } - } - else - idAttIndex = -1; - - /* do attribute defaulting */ - for (i = 0; i < nDefaultAtts; i++) { - const DEFAULT_ATTRIBUTE *da = elementType->defaultAtts + i; - if (!(da->id->name)[-1] && da->value) { - if (da->id->prefix) { - if (da->id->xmlns) { - enum XML_Error result = addBinding(parser, da->id->prefix, da->id, - da->value, bindingsPtr); - if (result) - return result; - } - else { - (da->id->name)[-1] = 2; - nPrefixes++; - appAtts[attIndex++] = da->id->name; - appAtts[attIndex++] = da->value; - } - } - else { - (da->id->name)[-1] = 1; - appAtts[attIndex++] = da->id->name; - appAtts[attIndex++] = da->value; - } - } - } - appAtts[attIndex] = 0; - - /* expand prefixed attribute names, check for duplicates, - and clear flags that say whether attributes were specified */ - i = 0; - if (nPrefixes) { - int j; /* hash table index */ - unsigned long version = nsAttsVersion; - int nsAttsSize = (int)1 << nsAttsPower; - /* size of hash table must be at least 2 * (# of prefixed attributes) */ - if ((nPrefixes << 1) >> nsAttsPower) { /* true for nsAttsPower = 0 */ - NS_ATT *temp; - /* hash table size must also be a power of 2 and >= 8 */ - while (nPrefixes >> nsAttsPower++); - if (nsAttsPower < 3) - nsAttsPower = 3; - nsAttsSize = (int)1 << nsAttsPower; - temp = (NS_ATT *)REALLOC(nsAtts, nsAttsSize * sizeof(NS_ATT)); - if (!temp) - return XML_ERROR_NO_MEMORY; - nsAtts = temp; - version = 0; /* force re-initialization of nsAtts hash table */ - } - /* using a version flag saves us from initializing nsAtts every time */ - if (!version) { /* initialize version flags when version wraps around */ - version = INIT_ATTS_VERSION; - for (j = nsAttsSize; j != 0; ) - nsAtts[--j].version = version; - } - nsAttsVersion = --version; - - /* expand prefixed names and check for duplicates */ - for (; i < attIndex; i += 2) { - const XML_Char *s = appAtts[i]; - if (s[-1] == 2) { /* prefixed */ - ATTRIBUTE_ID *id; - const BINDING *b; - unsigned long uriHash = 0; - ((XML_Char *)s)[-1] = 0; /* clear flag */ - id = (ATTRIBUTE_ID *)lookup(&dtd->attributeIds, s, 0); - b = id->prefix->binding; - if (!b) - return XML_ERROR_UNBOUND_PREFIX; - - /* as we expand the name we also calculate its hash value */ - for (j = 0; j < b->uriLen; j++) { - const XML_Char c = b->uri[j]; - if (!poolAppendChar(&tempPool, c)) - return XML_ERROR_NO_MEMORY; - uriHash = CHAR_HASH(uriHash, c); - } - while (*s++ != XML_T(ASCII_COLON)) - ; - do { /* copies null terminator */ - const XML_Char c = *s; - if (!poolAppendChar(&tempPool, *s)) - return XML_ERROR_NO_MEMORY; - uriHash = CHAR_HASH(uriHash, c); - } while (*s++); - - { /* Check hash table for duplicate of expanded name (uriName). - Derived from code in lookup(HASH_TABLE *table, ...). - */ - unsigned char step = 0; - unsigned long mask = nsAttsSize - 1; - j = uriHash & mask; /* index into hash table */ - while (nsAtts[j].version == version) { - /* for speed we compare stored hash values first */ - if (uriHash == nsAtts[j].hash) { - const XML_Char *s1 = poolStart(&tempPool); - const XML_Char *s2 = nsAtts[j].uriName; - /* s1 is null terminated, but not s2 */ - for (; *s1 == *s2 && *s1 != 0; s1++, s2++); - if (*s1 == 0) - return XML_ERROR_DUPLICATE_ATTRIBUTE; - } - if (!step) - step = PROBE_STEP(uriHash, mask, nsAttsPower); - j < step ? (j += nsAttsSize - step) : (j -= step); - } - } - - if (ns_triplets) { /* append namespace separator and prefix */ - tempPool.ptr[-1] = namespaceSeparator; - s = b->prefix->name; - do { - if (!poolAppendChar(&tempPool, *s)) - return XML_ERROR_NO_MEMORY; - } while (*s++); - } - - /* store expanded name in attribute list */ - s = poolStart(&tempPool); - poolFinish(&tempPool); - appAtts[i] = s; - - /* fill empty slot with new version, uriName and hash value */ - nsAtts[j].version = version; - nsAtts[j].hash = uriHash; - nsAtts[j].uriName = s; - - if (!--nPrefixes) { - i += 2; - break; - } - } - else /* not prefixed */ - ((XML_Char *)s)[-1] = 0; /* clear flag */ - } - } - /* clear flags for the remaining attributes */ - for (; i < attIndex; i += 2) - ((XML_Char *)(appAtts[i]))[-1] = 0; - for (binding = *bindingsPtr; binding; binding = binding->nextTagBinding) - binding->attId->name[-1] = 0; - - if (!ns) - return XML_ERROR_NONE; - - /* expand the element type name */ - if (elementType->prefix) { - binding = elementType->prefix->binding; - if (!binding) - return XML_ERROR_UNBOUND_PREFIX; - localPart = tagNamePtr->str; - while (*localPart++ != XML_T(ASCII_COLON)) - ; - } - else if (dtd->defaultPrefix.binding) { - binding = dtd->defaultPrefix.binding; - localPart = tagNamePtr->str; - } - else - return XML_ERROR_NONE; - prefixLen = 0; - if (ns_triplets && binding->prefix->name) { - for (; binding->prefix->name[prefixLen++];) - ; /* prefixLen includes null terminator */ - } - tagNamePtr->localPart = localPart; - tagNamePtr->uriLen = binding->uriLen; - tagNamePtr->prefix = binding->prefix->name; - tagNamePtr->prefixLen = prefixLen; - for (i = 0; localPart[i++];) - ; /* i includes null terminator */ - n = i + binding->uriLen + prefixLen; - if (n > binding->uriAlloc) { - TAG *p; - uri = (XML_Char *)MALLOC((n + EXPAND_SPARE) * sizeof(XML_Char)); - if (!uri) - return XML_ERROR_NO_MEMORY; - binding->uriAlloc = n + EXPAND_SPARE; - memcpy(uri, binding->uri, binding->uriLen * sizeof(XML_Char)); - for (p = tagStack; p; p = p->parent) - if (p->name.str == binding->uri) - p->name.str = uri; - FREE(binding->uri); - binding->uri = uri; - } - /* if namespaceSeparator != '\0' then uri includes it already */ - uri = binding->uri + binding->uriLen; - memcpy(uri, localPart, i * sizeof(XML_Char)); - /* we always have a namespace separator between localPart and prefix */ - if (prefixLen) { - uri += i - 1; - *uri = namespaceSeparator; /* replace null terminator */ - memcpy(uri + 1, binding->prefix->name, prefixLen * sizeof(XML_Char)); - } - tagNamePtr->str = binding->uri; - return XML_ERROR_NONE; -} - -/* addBinding() overwrites the value of prefix->binding without checking. - Therefore one must keep track of the old value outside of addBinding(). -*/ -static enum XML_Error -addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId, - const XML_Char *uri, BINDING **bindingsPtr) -{ - static const XML_Char xmlNamespace[] = { - ASCII_h, ASCII_t, ASCII_t, ASCII_p, ASCII_COLON, ASCII_SLASH, ASCII_SLASH, - ASCII_w, ASCII_w, ASCII_w, ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, - ASCII_o, ASCII_r, ASCII_g, ASCII_SLASH, ASCII_X, ASCII_M, ASCII_L, - ASCII_SLASH, ASCII_1, ASCII_9, ASCII_9, ASCII_8, ASCII_SLASH, - ASCII_n, ASCII_a, ASCII_m, ASCII_e, ASCII_s, ASCII_p, ASCII_a, ASCII_c, - ASCII_e, '\0' - }; - static const int xmlLen = - (int)sizeof(xmlNamespace)/sizeof(XML_Char) - 1; - static const XML_Char xmlnsNamespace[] = { - ASCII_h, ASCII_t, ASCII_t, ASCII_p, ASCII_COLON, ASCII_SLASH, ASCII_SLASH, - ASCII_w, ASCII_w, ASCII_w, ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, - ASCII_o, ASCII_r, ASCII_g, ASCII_SLASH, ASCII_2, ASCII_0, ASCII_0, - ASCII_0, ASCII_SLASH, ASCII_x, ASCII_m, ASCII_l, ASCII_n, ASCII_s, - ASCII_SLASH, '\0' - }; - static const int xmlnsLen = - (int)sizeof(xmlnsNamespace)/sizeof(XML_Char) - 1; - - XML_Bool mustBeXML = XML_FALSE; - XML_Bool isXML = XML_TRUE; - XML_Bool isXMLNS = XML_TRUE; - - BINDING *b; - int len; - - /* empty URI is only valid for default namespace per XML NS 1.0 (not 1.1) */ - if (*uri == XML_T('\0') && prefix->name) - return XML_ERROR_UNDECLARING_PREFIX; - - if (prefix->name - && prefix->name[0] == XML_T(ASCII_x) - && prefix->name[1] == XML_T(ASCII_m) - && prefix->name[2] == XML_T(ASCII_l)) { - - /* Not allowed to bind xmlns */ - if (prefix->name[3] == XML_T(ASCII_n) - && prefix->name[4] == XML_T(ASCII_s) - && prefix->name[5] == XML_T('\0')) - return XML_ERROR_RESERVED_PREFIX_XMLNS; - - if (prefix->name[3] == XML_T('\0')) - mustBeXML = XML_TRUE; - } - - for (len = 0; uri[len]; len++) { - if (isXML && (len > xmlLen || uri[len] != xmlNamespace[len])) - isXML = XML_FALSE; - - if (!mustBeXML && isXMLNS - && (len > xmlnsLen || uri[len] != xmlnsNamespace[len])) - isXMLNS = XML_FALSE; - } - isXML = isXML && len == xmlLen; - isXMLNS = isXMLNS && len == xmlnsLen; - - if (mustBeXML != isXML) - return mustBeXML ? XML_ERROR_RESERVED_PREFIX_XML - : XML_ERROR_RESERVED_NAMESPACE_URI; - - if (isXMLNS) - return XML_ERROR_RESERVED_NAMESPACE_URI; - - if (namespaceSeparator) - len++; - if (freeBindingList) { - b = freeBindingList; - if (len > b->uriAlloc) { - XML_Char *temp = (XML_Char *)REALLOC(b->uri, - sizeof(XML_Char) * (len + EXPAND_SPARE)); - if (temp == NULL) - return XML_ERROR_NO_MEMORY; - b->uri = temp; - b->uriAlloc = len + EXPAND_SPARE; - } - freeBindingList = b->nextTagBinding; - } - else { - b = (BINDING *)MALLOC(sizeof(BINDING)); - if (!b) - return XML_ERROR_NO_MEMORY; - b->uri = (XML_Char *)MALLOC(sizeof(XML_Char) * (len + EXPAND_SPARE)); - if (!b->uri) { - FREE(b); - return XML_ERROR_NO_MEMORY; - } - b->uriAlloc = len + EXPAND_SPARE; - } - b->uriLen = len; - memcpy(b->uri, uri, len * sizeof(XML_Char)); - if (namespaceSeparator) - b->uri[len - 1] = namespaceSeparator; - b->prefix = prefix; - b->attId = attId; - b->prevPrefixBinding = prefix->binding; - /* NULL binding when default namespace undeclared */ - if (*uri == XML_T('\0') && prefix == &_dtd->defaultPrefix) - prefix->binding = NULL; - else - prefix->binding = b; - b->nextTagBinding = *bindingsPtr; - *bindingsPtr = b; - /* if attId == NULL then we are not starting a namespace scope */ - if (attId && startNamespaceDeclHandler) - startNamespaceDeclHandler(handlerArg, prefix->name, - prefix->binding ? uri : 0); - return XML_ERROR_NONE; -} - -/* The idea here is to avoid using stack for each CDATA section when - the whole file is parsed with one call. -*/ -static enum XML_Error PTRCALL -cdataSectionProcessor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - enum XML_Error result = doCdataSection(parser, encoding, &start, end, - endPtr, (XML_Bool)!ps_finalBuffer); - if (result != XML_ERROR_NONE) - return result; - if (start) { - if (parentParser) { /* we are parsing an external entity */ - processor = externalEntityContentProcessor; - return externalEntityContentProcessor(parser, start, end, endPtr); - } - else { - processor = contentProcessor; - return contentProcessor(parser, start, end, endPtr); - } - } - return result; -} - -/* startPtr gets set to non-null if the section is closed, and to null if - the section is not yet closed. -*/ -static enum XML_Error -doCdataSection(XML_Parser parser, - const ENCODING *enc, - const char **startPtr, - const char *end, - const char **nextPtr, - XML_Bool haveMore) -{ - const char *s = *startPtr; - const char **eventPP; - const char **eventEndPP; - if (enc == encoding) { - eventPP = &eventPtr; - *eventPP = s; - eventEndPP = &eventEndPtr; - } - else { - eventPP = &(openInternalEntities->internalEventPtr); - eventEndPP = &(openInternalEntities->internalEventEndPtr); - } - *eventPP = s; - *startPtr = NULL; - - for (;;) { - const char *next; - int tok = XmlCdataSectionTok(enc, s, end, &next); - *eventEndPP = next; - switch (tok) { - case XML_TOK_CDATA_SECT_CLOSE: - if (endCdataSectionHandler) - endCdataSectionHandler(handlerArg); -#if 0 - /* see comment under XML_TOK_CDATA_SECT_OPEN */ - else if (characterDataHandler) - characterDataHandler(handlerArg, dataBuf, 0); -#endif - else if (defaultHandler) - reportDefault(parser, enc, s, next); - *startPtr = next; - *nextPtr = next; - if (ps_parsing == XML_FINISHED) - return XML_ERROR_ABORTED; - else - return XML_ERROR_NONE; - case XML_TOK_DATA_NEWLINE: - if (characterDataHandler) { - XML_Char c = 0xA; - characterDataHandler(handlerArg, &c, 1); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - break; - case XML_TOK_DATA_CHARS: - { - XML_CharacterDataHandler charDataHandler = characterDataHandler; - if (charDataHandler) { - if (MUST_CONVERT(enc, s)) { - for (;;) { - ICHAR *dataPtr = (ICHAR *)dataBuf; - XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); - *eventEndPP = next; - charDataHandler(handlerArg, dataBuf, - (int)(dataPtr - (ICHAR *)dataBuf)); - if (s == next) - break; - *eventPP = s; - } - } - else - charDataHandler(handlerArg, - (XML_Char *)s, - (int)((XML_Char *)next - (XML_Char *)s)); - } - else if (defaultHandler) - reportDefault(parser, enc, s, next); - } - break; - case XML_TOK_INVALID: - *eventPP = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_PARTIAL: - case XML_TOK_NONE: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_UNCLOSED_CDATA_SECTION; - default: - *eventPP = next; - return XML_ERROR_UNEXPECTED_STATE; - } - - *eventPP = s = next; - switch (ps_parsing) { - case XML_SUSPENDED: - *nextPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: ; - } - } - /* not reached */ -} - -#ifdef XML_DTD - -/* The idea here is to avoid using stack for each IGNORE section when - the whole file is parsed with one call. -*/ -static enum XML_Error PTRCALL -ignoreSectionProcessor(XML_Parser parser, - const char *start, - const char *end, - const char **endPtr) -{ - enum XML_Error result = doIgnoreSection(parser, encoding, &start, end, - endPtr, (XML_Bool)!ps_finalBuffer); - if (result != XML_ERROR_NONE) - return result; - if (start) { - processor = prologProcessor; - return prologProcessor(parser, start, end, endPtr); - } - return result; -} - -/* startPtr gets set to non-null is the section is closed, and to null - if the section is not yet closed. -*/ -static enum XML_Error -doIgnoreSection(XML_Parser parser, - const ENCODING *enc, - const char **startPtr, - const char *end, - const char **nextPtr, - XML_Bool haveMore) -{ - const char *next; - int tok; - const char *s = *startPtr; - const char **eventPP; - const char **eventEndPP; - if (enc == encoding) { - eventPP = &eventPtr; - *eventPP = s; - eventEndPP = &eventEndPtr; - } - else { - eventPP = &(openInternalEntities->internalEventPtr); - eventEndPP = &(openInternalEntities->internalEventEndPtr); - } - *eventPP = s; - *startPtr = NULL; - tok = XmlIgnoreSectionTok(enc, s, end, &next); - *eventEndPP = next; - switch (tok) { - case XML_TOK_IGNORE_SECT: - if (defaultHandler) - reportDefault(parser, enc, s, next); - *startPtr = next; - *nextPtr = next; - if (ps_parsing == XML_FINISHED) - return XML_ERROR_ABORTED; - else - return XML_ERROR_NONE; - case XML_TOK_INVALID: - *eventPP = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_PARTIAL: - case XML_TOK_NONE: - if (haveMore) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_SYNTAX; /* XML_ERROR_UNCLOSED_IGNORE_SECTION */ - default: - *eventPP = next; - return XML_ERROR_UNEXPECTED_STATE; - } - /* not reached */ -} - -#endif /* XML_DTD */ - -static enum XML_Error -initializeEncoding(XML_Parser parser) -{ - const char *s; -#ifdef XML_UNICODE - char encodingBuf[128]; - if (!protocolEncodingName) - s = NULL; - else { - int i; - for (i = 0; protocolEncodingName[i]; i++) { - if (i == sizeof(encodingBuf) - 1 - || (protocolEncodingName[i] & ~0x7f) != 0) { - encodingBuf[0] = '\0'; - break; - } - encodingBuf[i] = (char)protocolEncodingName[i]; - } - encodingBuf[i] = '\0'; - s = encodingBuf; - } -#else - s = protocolEncodingName; -#endif - if ((ns ? XmlInitEncodingNS : XmlInitEncoding)(&initEncoding, &encoding, s)) - return XML_ERROR_NONE; - return handleUnknownEncoding(parser, protocolEncodingName); -} - -static enum XML_Error -processXmlDecl(XML_Parser parser, int isGeneralTextEntity, - const char *s, const char *next) -{ - const char *encodingName = NULL; - const XML_Char *storedEncName = NULL; - const ENCODING *newEncoding = NULL; - const char *version = NULL; - const char *versionend; - const XML_Char *storedversion = NULL; - int standalone = -1; - if (!(ns - ? XmlParseXmlDeclNS - : XmlParseXmlDecl)(isGeneralTextEntity, - encoding, - s, - next, - &eventPtr, - &version, - &versionend, - &encodingName, - &newEncoding, - &standalone)) { - if (isGeneralTextEntity) - return XML_ERROR_TEXT_DECL; - else - return XML_ERROR_XML_DECL; - } - if (!isGeneralTextEntity && standalone == 1) { - _dtd->standalone = XML_TRUE; -#ifdef XML_DTD - if (paramEntityParsing == XML_PARAM_ENTITY_PARSING_UNLESS_STANDALONE) - paramEntityParsing = XML_PARAM_ENTITY_PARSING_NEVER; -#endif /* XML_DTD */ - } - if (xmlDeclHandler) { - if (encodingName != NULL) { - storedEncName = poolStoreString(&temp2Pool, - encoding, - encodingName, - encodingName - + XmlNameLength(encoding, encodingName)); - if (!storedEncName) - return XML_ERROR_NO_MEMORY; - poolFinish(&temp2Pool); - } - if (version) { - storedversion = poolStoreString(&temp2Pool, - encoding, - version, - versionend - encoding->minBytesPerChar); - if (!storedversion) - return XML_ERROR_NO_MEMORY; - } - xmlDeclHandler(handlerArg, storedversion, storedEncName, standalone); - } - else if (defaultHandler) - reportDefault(parser, encoding, s, next); - if (protocolEncodingName == NULL) { - if (newEncoding) { - if (newEncoding->minBytesPerChar != encoding->minBytesPerChar) { - eventPtr = encodingName; - return XML_ERROR_INCORRECT_ENCODING; - } - encoding = newEncoding; - } - else if (encodingName) { - enum XML_Error result; - if (!storedEncName) { - storedEncName = poolStoreString( - &temp2Pool, encoding, encodingName, - encodingName + XmlNameLength(encoding, encodingName)); - if (!storedEncName) - return XML_ERROR_NO_MEMORY; - } - result = handleUnknownEncoding(parser, storedEncName); - poolClear(&temp2Pool); - if (result == XML_ERROR_UNKNOWN_ENCODING) - eventPtr = encodingName; - return result; - } - } - - if (storedEncName || storedversion) - poolClear(&temp2Pool); - - return XML_ERROR_NONE; -} - -static enum XML_Error -handleUnknownEncoding(XML_Parser parser, const XML_Char *encodingName) -{ - if (unknownEncodingHandler) { - XML_Encoding info; - int i; - for (i = 0; i < 256; i++) - info.map[i] = -1; - info.convert = NULL; - info.data = NULL; - info.release = NULL; - if (unknownEncodingHandler(unknownEncodingHandlerData, encodingName, - &info)) { - ENCODING *enc; - unknownEncodingMem = MALLOC(XmlSizeOfUnknownEncoding()); - if (!unknownEncodingMem) { - if (info.release) - info.release(info.data); - return XML_ERROR_NO_MEMORY; - } - enc = (ns - ? XmlInitUnknownEncodingNS - : XmlInitUnknownEncoding)(unknownEncodingMem, - info.map, - info.convert, - info.data); - if (enc) { - unknownEncodingData = info.data; - unknownEncodingRelease = info.release; - encoding = enc; - return XML_ERROR_NONE; - } - } - if (info.release != NULL) - info.release(info.data); - } - return XML_ERROR_UNKNOWN_ENCODING; -} - -static enum XML_Error PTRCALL -prologInitProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - enum XML_Error result = initializeEncoding(parser); - if (result != XML_ERROR_NONE) - return result; - processor = prologProcessor; - return prologProcessor(parser, s, end, nextPtr); -} - -#ifdef XML_DTD - -static enum XML_Error PTRCALL -externalParEntInitProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - enum XML_Error result = initializeEncoding(parser); - if (result != XML_ERROR_NONE) - return result; - - /* we know now that XML_Parse(Buffer) has been called, - so we consider the external parameter entity read */ - _dtd->paramEntityRead = XML_TRUE; - - if (prologState.inEntityValue) { - processor = entityValueInitProcessor; - return entityValueInitProcessor(parser, s, end, nextPtr); - } - else { - processor = externalParEntProcessor; - return externalParEntProcessor(parser, s, end, nextPtr); - } -} - -static enum XML_Error PTRCALL -entityValueInitProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - int tok; - const char *start = s; - const char *next = start; - eventPtr = start; - - for (;;) { - tok = XmlPrologTok(encoding, start, end, &next); - eventEndPtr = next; - if (tok <= 0) { - if (!ps_finalBuffer && tok != XML_TOK_INVALID) { - *nextPtr = s; - return XML_ERROR_NONE; - } - switch (tok) { - case XML_TOK_INVALID: - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_NONE: /* start == end */ - default: - break; - } - /* found end of entity value - can store it now */ - return storeEntityValue(parser, encoding, s, end); - } - else if (tok == XML_TOK_XML_DECL) { - enum XML_Error result; - result = processXmlDecl(parser, 0, start, next); - if (result != XML_ERROR_NONE) - return result; - switch (ps_parsing) { - case XML_SUSPENDED: - *nextPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: - *nextPtr = next; - } - /* stop scanning for text declaration - we found one */ - processor = entityValueProcessor; - return entityValueProcessor(parser, next, end, nextPtr); - } - /* If we are at the end of the buffer, this would cause XmlPrologTok to - return XML_TOK_NONE on the next call, which would then cause the - function to exit with *nextPtr set to s - that is what we want for other - tokens, but not for the BOM - we would rather like to skip it; - then, when this routine is entered the next time, XmlPrologTok will - return XML_TOK_INVALID, since the BOM is still in the buffer - */ - else if (tok == XML_TOK_BOM && next == end && !ps_finalBuffer) { - *nextPtr = next; - return XML_ERROR_NONE; - } - start = next; - eventPtr = start; - } -} - -static enum XML_Error PTRCALL -externalParEntProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - const char *next = s; - int tok; - - tok = XmlPrologTok(encoding, s, end, &next); - if (tok <= 0) { - if (!ps_finalBuffer && tok != XML_TOK_INVALID) { - *nextPtr = s; - return XML_ERROR_NONE; - } - switch (tok) { - case XML_TOK_INVALID: - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_NONE: /* start == end */ - default: - break; - } - } - /* This would cause the next stage, i.e. doProlog to be passed XML_TOK_BOM. - However, when parsing an external subset, doProlog will not accept a BOM - as valid, and report a syntax error, so we have to skip the BOM - */ - else if (tok == XML_TOK_BOM) { - s = next; - tok = XmlPrologTok(encoding, s, end, &next); - } - - processor = prologProcessor; - return doProlog(parser, encoding, s, end, tok, next, - nextPtr, (XML_Bool)!ps_finalBuffer); -} - -static enum XML_Error PTRCALL -entityValueProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - const char *start = s; - const char *next = s; - const ENCODING *enc = encoding; - int tok; - - for (;;) { - tok = XmlPrologTok(enc, start, end, &next); - if (tok <= 0) { - if (!ps_finalBuffer && tok != XML_TOK_INVALID) { - *nextPtr = s; - return XML_ERROR_NONE; - } - switch (tok) { - case XML_TOK_INVALID: - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_NONE: /* start == end */ - default: - break; - } - /* found end of entity value - can store it now */ - return storeEntityValue(parser, enc, s, end); - } - start = next; - } -} - -#endif /* XML_DTD */ - -static enum XML_Error PTRCALL -prologProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - const char *next = s; - int tok = XmlPrologTok(encoding, s, end, &next); - return doProlog(parser, encoding, s, end, tok, next, - nextPtr, (XML_Bool)!ps_finalBuffer); -} - -static enum XML_Error -doProlog(XML_Parser parser, - const ENCODING *enc, - const char *s, - const char *end, - int tok, - const char *next, - const char **nextPtr, - XML_Bool haveMore) -{ -#ifdef XML_DTD - static const XML_Char externalSubsetName[] = { ASCII_HASH , '\0' }; -#endif /* XML_DTD */ - static const XML_Char atypeCDATA[] = - { ASCII_C, ASCII_D, ASCII_A, ASCII_T, ASCII_A, '\0' }; - static const XML_Char atypeID[] = { ASCII_I, ASCII_D, '\0' }; - static const XML_Char atypeIDREF[] = - { ASCII_I, ASCII_D, ASCII_R, ASCII_E, ASCII_F, '\0' }; - static const XML_Char atypeIDREFS[] = - { ASCII_I, ASCII_D, ASCII_R, ASCII_E, ASCII_F, ASCII_S, '\0' }; - static const XML_Char atypeENTITY[] = - { ASCII_E, ASCII_N, ASCII_T, ASCII_I, ASCII_T, ASCII_Y, '\0' }; - static const XML_Char atypeENTITIES[] = { ASCII_E, ASCII_N, - ASCII_T, ASCII_I, ASCII_T, ASCII_I, ASCII_E, ASCII_S, '\0' }; - static const XML_Char atypeNMTOKEN[] = { - ASCII_N, ASCII_M, ASCII_T, ASCII_O, ASCII_K, ASCII_E, ASCII_N, '\0' }; - static const XML_Char atypeNMTOKENS[] = { ASCII_N, ASCII_M, ASCII_T, - ASCII_O, ASCII_K, ASCII_E, ASCII_N, ASCII_S, '\0' }; - static const XML_Char notationPrefix[] = { ASCII_N, ASCII_O, ASCII_T, - ASCII_A, ASCII_T, ASCII_I, ASCII_O, ASCII_N, ASCII_LPAREN, '\0' }; - static const XML_Char enumValueSep[] = { ASCII_PIPE, '\0' }; - static const XML_Char enumValueStart[] = { ASCII_LPAREN, '\0' }; - - /* save one level of indirection */ - DTD * const dtd = _dtd; - - const char **eventPP; - const char **eventEndPP; - enum XML_Content_Quant quant; - - if (enc == encoding) { - eventPP = &eventPtr; - eventEndPP = &eventEndPtr; - } - else { - eventPP = &(openInternalEntities->internalEventPtr); - eventEndPP = &(openInternalEntities->internalEventEndPtr); - } - - for (;;) { - int role; - XML_Bool handleDefault = XML_TRUE; - *eventPP = s; - *eventEndPP = next; - if (tok <= 0) { - if (haveMore && tok != XML_TOK_INVALID) { - *nextPtr = s; - return XML_ERROR_NONE; - } - switch (tok) { - case XML_TOK_INVALID: - *eventPP = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - return XML_ERROR_PARTIAL_CHAR; - case XML_TOK_NONE: -#ifdef XML_DTD - /* for internal PE NOT referenced between declarations */ - if (enc != encoding && !openInternalEntities->betweenDecl) { - *nextPtr = s; - return XML_ERROR_NONE; - } - /* WFC: PE Between Declarations - must check that PE contains - complete markup, not only for external PEs, but also for - internal PEs if the reference occurs between declarations. - */ - if (isParamEntity || enc != encoding) { - if (XmlTokenRole(&prologState, XML_TOK_NONE, end, end, enc) - == XML_ROLE_ERROR) - return XML_ERROR_INCOMPLETE_PE; - *nextPtr = s; - return XML_ERROR_NONE; - } -#endif /* XML_DTD */ - return XML_ERROR_NO_ELEMENTS; - default: - tok = -tok; - next = end; - break; - } - } - role = XmlTokenRole(&prologState, tok, s, next, enc); - switch (role) { - case XML_ROLE_XML_DECL: - { - enum XML_Error result = processXmlDecl(parser, 0, s, next); - if (result != XML_ERROR_NONE) - return result; - enc = encoding; - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_DOCTYPE_NAME: - if (startDoctypeDeclHandler) { - doctypeName = poolStoreString(&tempPool, enc, s, next); - if (!doctypeName) - return XML_ERROR_NO_MEMORY; - poolFinish(&tempPool); - doctypePubid = NULL; - handleDefault = XML_FALSE; - } - doctypeSysid = NULL; /* always initialize to NULL */ - break; - case XML_ROLE_DOCTYPE_INTERNAL_SUBSET: - if (startDoctypeDeclHandler) { - startDoctypeDeclHandler(handlerArg, doctypeName, doctypeSysid, - doctypePubid, 1); - doctypeName = NULL; - poolClear(&tempPool); - handleDefault = XML_FALSE; - } - break; -#ifdef XML_DTD - case XML_ROLE_TEXT_DECL: - { - enum XML_Error result = processXmlDecl(parser, 1, s, next); - if (result != XML_ERROR_NONE) - return result; - enc = encoding; - handleDefault = XML_FALSE; - } - break; -#endif /* XML_DTD */ - case XML_ROLE_DOCTYPE_PUBLIC_ID: -#ifdef XML_DTD - useForeignDTD = XML_FALSE; - declEntity = (ENTITY *)lookup(&dtd->paramEntities, - externalSubsetName, - sizeof(ENTITY)); - if (!declEntity) - return XML_ERROR_NO_MEMORY; -#endif /* XML_DTD */ - dtd->hasParamEntityRefs = XML_TRUE; - if (startDoctypeDeclHandler) { - if (!XmlIsPublicId(enc, s, next, eventPP)) - return XML_ERROR_PUBLICID; - doctypePubid = poolStoreString(&tempPool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!doctypePubid) - return XML_ERROR_NO_MEMORY; - normalizePublicId((XML_Char *)doctypePubid); - poolFinish(&tempPool); - handleDefault = XML_FALSE; - goto alreadyChecked; - } - /* fall through */ - case XML_ROLE_ENTITY_PUBLIC_ID: - if (!XmlIsPublicId(enc, s, next, eventPP)) - return XML_ERROR_PUBLICID; - alreadyChecked: - if (dtd->keepProcessing && declEntity) { - XML_Char *tem = poolStoreString(&dtd->pool, - enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!tem) - return XML_ERROR_NO_MEMORY; - normalizePublicId(tem); - declEntity->publicId = tem; - poolFinish(&dtd->pool); - if (entityDeclHandler) - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_DOCTYPE_CLOSE: - if (doctypeName) { - startDoctypeDeclHandler(handlerArg, doctypeName, - doctypeSysid, doctypePubid, 0); - poolClear(&tempPool); - handleDefault = XML_FALSE; - } - /* doctypeSysid will be non-NULL in the case of a previous - XML_ROLE_DOCTYPE_SYSTEM_ID, even if startDoctypeDeclHandler - was not set, indicating an external subset - */ -#ifdef XML_DTD - if (doctypeSysid || useForeignDTD) { - XML_Bool hadParamEntityRefs = dtd->hasParamEntityRefs; - dtd->hasParamEntityRefs = XML_TRUE; - if (paramEntityParsing && externalEntityRefHandler) { - ENTITY *entity = (ENTITY *)lookup(&dtd->paramEntities, - externalSubsetName, - sizeof(ENTITY)); - if (!entity) - return XML_ERROR_NO_MEMORY; - if (useForeignDTD) - entity->base = curBase; - dtd->paramEntityRead = XML_FALSE; - if (!externalEntityRefHandler(externalEntityRefHandlerArg, - 0, - entity->base, - entity->systemId, - entity->publicId)) - return XML_ERROR_EXTERNAL_ENTITY_HANDLING; - if (dtd->paramEntityRead) { - if (!dtd->standalone && - notStandaloneHandler && - !notStandaloneHandler(handlerArg)) - return XML_ERROR_NOT_STANDALONE; - } - /* if we didn't read the foreign DTD then this means that there - is no external subset and we must reset dtd->hasParamEntityRefs - */ - else if (!doctypeSysid) - dtd->hasParamEntityRefs = hadParamEntityRefs; - /* end of DTD - no need to update dtd->keepProcessing */ - } - useForeignDTD = XML_FALSE; - } -#endif /* XML_DTD */ - if (endDoctypeDeclHandler) { - endDoctypeDeclHandler(handlerArg); - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_INSTANCE_START: -#ifdef XML_DTD - /* if there is no DOCTYPE declaration then now is the - last chance to read the foreign DTD - */ - if (useForeignDTD) { - XML_Bool hadParamEntityRefs = dtd->hasParamEntityRefs; - dtd->hasParamEntityRefs = XML_TRUE; - if (paramEntityParsing && externalEntityRefHandler) { - ENTITY *entity = (ENTITY *)lookup(&dtd->paramEntities, - externalSubsetName, - sizeof(ENTITY)); - if (!entity) - return XML_ERROR_NO_MEMORY; - entity->base = curBase; - dtd->paramEntityRead = XML_FALSE; - if (!externalEntityRefHandler(externalEntityRefHandlerArg, - 0, - entity->base, - entity->systemId, - entity->publicId)) - return XML_ERROR_EXTERNAL_ENTITY_HANDLING; - if (dtd->paramEntityRead) { - if (!dtd->standalone && - notStandaloneHandler && - !notStandaloneHandler(handlerArg)) - return XML_ERROR_NOT_STANDALONE; - } - /* if we didn't read the foreign DTD then this means that there - is no external subset and we must reset dtd->hasParamEntityRefs - */ - else - dtd->hasParamEntityRefs = hadParamEntityRefs; - /* end of DTD - no need to update dtd->keepProcessing */ - } - } -#endif /* XML_DTD */ - processor = contentProcessor; - return contentProcessor(parser, s, end, nextPtr); - case XML_ROLE_ATTLIST_ELEMENT_NAME: - declElementType = getElementType(parser, enc, s, next); - if (!declElementType) - return XML_ERROR_NO_MEMORY; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_NAME: - declAttributeId = getAttributeId(parser, enc, s, next); - if (!declAttributeId) - return XML_ERROR_NO_MEMORY; - declAttributeIsCdata = XML_FALSE; - declAttributeType = NULL; - declAttributeIsId = XML_FALSE; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_CDATA: - declAttributeIsCdata = XML_TRUE; - declAttributeType = atypeCDATA; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_ID: - declAttributeIsId = XML_TRUE; - declAttributeType = atypeID; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_IDREF: - declAttributeType = atypeIDREF; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_IDREFS: - declAttributeType = atypeIDREFS; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_ENTITY: - declAttributeType = atypeENTITY; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_ENTITIES: - declAttributeType = atypeENTITIES; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_NMTOKEN: - declAttributeType = atypeNMTOKEN; - goto checkAttListDeclHandler; - case XML_ROLE_ATTRIBUTE_TYPE_NMTOKENS: - declAttributeType = atypeNMTOKENS; - checkAttListDeclHandler: - if (dtd->keepProcessing && attlistDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_ATTRIBUTE_ENUM_VALUE: - case XML_ROLE_ATTRIBUTE_NOTATION_VALUE: - if (dtd->keepProcessing && attlistDeclHandler) { - const XML_Char *prefix; - if (declAttributeType) { - prefix = enumValueSep; - } - else { - prefix = (role == XML_ROLE_ATTRIBUTE_NOTATION_VALUE - ? notationPrefix - : enumValueStart); - } - if (!poolAppendString(&tempPool, prefix)) - return XML_ERROR_NO_MEMORY; - if (!poolAppend(&tempPool, enc, s, next)) - return XML_ERROR_NO_MEMORY; - declAttributeType = tempPool.start; - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_IMPLIED_ATTRIBUTE_VALUE: - case XML_ROLE_REQUIRED_ATTRIBUTE_VALUE: - if (dtd->keepProcessing) { - if (!defineAttribute(declElementType, declAttributeId, - declAttributeIsCdata, declAttributeIsId, - 0, parser)) - return XML_ERROR_NO_MEMORY; - if (attlistDeclHandler && declAttributeType) { - if (*declAttributeType == XML_T(ASCII_LPAREN) - || (*declAttributeType == XML_T(ASCII_N) - && declAttributeType[1] == XML_T(ASCII_O))) { - /* Enumerated or Notation type */ - if (!poolAppendChar(&tempPool, XML_T(ASCII_RPAREN)) - || !poolAppendChar(&tempPool, XML_T('\0'))) - return XML_ERROR_NO_MEMORY; - declAttributeType = tempPool.start; - poolFinish(&tempPool); - } - *eventEndPP = s; - attlistDeclHandler(handlerArg, declElementType->name, - declAttributeId->name, declAttributeType, - 0, role == XML_ROLE_REQUIRED_ATTRIBUTE_VALUE); - poolClear(&tempPool); - handleDefault = XML_FALSE; - } - } - break; - case XML_ROLE_DEFAULT_ATTRIBUTE_VALUE: - case XML_ROLE_FIXED_ATTRIBUTE_VALUE: - if (dtd->keepProcessing) { - const XML_Char *attVal; - enum XML_Error result = - storeAttributeValue(parser, enc, declAttributeIsCdata, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar, - &dtd->pool); - if (result) - return result; - attVal = poolStart(&dtd->pool); - poolFinish(&dtd->pool); - /* ID attributes aren't allowed to have a default */ - if (!defineAttribute(declElementType, declAttributeId, - declAttributeIsCdata, XML_FALSE, attVal, parser)) - return XML_ERROR_NO_MEMORY; - if (attlistDeclHandler && declAttributeType) { - if (*declAttributeType == XML_T(ASCII_LPAREN) - || (*declAttributeType == XML_T(ASCII_N) - && declAttributeType[1] == XML_T(ASCII_O))) { - /* Enumerated or Notation type */ - if (!poolAppendChar(&tempPool, XML_T(ASCII_RPAREN)) - || !poolAppendChar(&tempPool, XML_T('\0'))) - return XML_ERROR_NO_MEMORY; - declAttributeType = tempPool.start; - poolFinish(&tempPool); - } - *eventEndPP = s; - attlistDeclHandler(handlerArg, declElementType->name, - declAttributeId->name, declAttributeType, - attVal, - role == XML_ROLE_FIXED_ATTRIBUTE_VALUE); - poolClear(&tempPool); - handleDefault = XML_FALSE; - } - } - break; - case XML_ROLE_ENTITY_VALUE: - if (dtd->keepProcessing) { - enum XML_Error result = storeEntityValue(parser, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (declEntity) { - declEntity->textPtr = poolStart(&dtd->entityValuePool); - declEntity->textLen = (int)(poolLength(&dtd->entityValuePool)); - poolFinish(&dtd->entityValuePool); - if (entityDeclHandler) { - *eventEndPP = s; - entityDeclHandler(handlerArg, - declEntity->name, - declEntity->is_param, - declEntity->textPtr, - declEntity->textLen, - curBase, 0, 0, 0); - handleDefault = XML_FALSE; - } - } - else - poolDiscard(&dtd->entityValuePool); - if (result != XML_ERROR_NONE) - return result; - } - break; - case XML_ROLE_DOCTYPE_SYSTEM_ID: -#ifdef XML_DTD - useForeignDTD = XML_FALSE; -#endif /* XML_DTD */ - dtd->hasParamEntityRefs = XML_TRUE; - if (startDoctypeDeclHandler) { - doctypeSysid = poolStoreString(&tempPool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (doctypeSysid == NULL) - return XML_ERROR_NO_MEMORY; - poolFinish(&tempPool); - handleDefault = XML_FALSE; - } -#ifdef XML_DTD - else - /* use externalSubsetName to make doctypeSysid non-NULL - for the case where no startDoctypeDeclHandler is set */ - doctypeSysid = externalSubsetName; -#endif /* XML_DTD */ - if (!dtd->standalone -#ifdef XML_DTD - && !paramEntityParsing -#endif /* XML_DTD */ - && notStandaloneHandler - && !notStandaloneHandler(handlerArg)) - return XML_ERROR_NOT_STANDALONE; -#ifndef XML_DTD - break; -#else /* XML_DTD */ - if (!declEntity) { - declEntity = (ENTITY *)lookup(&dtd->paramEntities, - externalSubsetName, - sizeof(ENTITY)); - if (!declEntity) - return XML_ERROR_NO_MEMORY; - declEntity->publicId = NULL; - } - /* fall through */ -#endif /* XML_DTD */ - case XML_ROLE_ENTITY_SYSTEM_ID: - if (dtd->keepProcessing && declEntity) { - declEntity->systemId = poolStoreString(&dtd->pool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!declEntity->systemId) - return XML_ERROR_NO_MEMORY; - declEntity->base = curBase; - poolFinish(&dtd->pool); - if (entityDeclHandler) - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_ENTITY_COMPLETE: - if (dtd->keepProcessing && declEntity && entityDeclHandler) { - *eventEndPP = s; - entityDeclHandler(handlerArg, - declEntity->name, - declEntity->is_param, - 0,0, - declEntity->base, - declEntity->systemId, - declEntity->publicId, - 0); - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_ENTITY_NOTATION_NAME: - if (dtd->keepProcessing && declEntity) { - declEntity->notation = poolStoreString(&dtd->pool, enc, s, next); - if (!declEntity->notation) - return XML_ERROR_NO_MEMORY; - poolFinish(&dtd->pool); - if (unparsedEntityDeclHandler) { - *eventEndPP = s; - unparsedEntityDeclHandler(handlerArg, - declEntity->name, - declEntity->base, - declEntity->systemId, - declEntity->publicId, - declEntity->notation); - handleDefault = XML_FALSE; - } - else if (entityDeclHandler) { - *eventEndPP = s; - entityDeclHandler(handlerArg, - declEntity->name, - 0,0,0, - declEntity->base, - declEntity->systemId, - declEntity->publicId, - declEntity->notation); - handleDefault = XML_FALSE; - } - } - break; - case XML_ROLE_GENERAL_ENTITY_NAME: - { - if (XmlPredefinedEntityName(enc, s, next)) { - declEntity = NULL; - break; - } - if (dtd->keepProcessing) { - const XML_Char *name = poolStoreString(&dtd->pool, enc, s, next); - if (!name) - return XML_ERROR_NO_MEMORY; - declEntity = (ENTITY *)lookup(&dtd->generalEntities, name, - sizeof(ENTITY)); - if (!declEntity) - return XML_ERROR_NO_MEMORY; - if (declEntity->name != name) { - poolDiscard(&dtd->pool); - declEntity = NULL; - } - else { - poolFinish(&dtd->pool); - declEntity->publicId = NULL; - declEntity->is_param = XML_FALSE; - /* if we have a parent parser or are reading an internal parameter - entity, then the entity declaration is not considered "internal" - */ - declEntity->is_internal = !(parentParser || openInternalEntities); - if (entityDeclHandler) - handleDefault = XML_FALSE; - } - } - else { - poolDiscard(&dtd->pool); - declEntity = NULL; - } - } - break; - case XML_ROLE_PARAM_ENTITY_NAME: -#ifdef XML_DTD - if (dtd->keepProcessing) { - const XML_Char *name = poolStoreString(&dtd->pool, enc, s, next); - if (!name) - return XML_ERROR_NO_MEMORY; - declEntity = (ENTITY *)lookup(&dtd->paramEntities, - name, sizeof(ENTITY)); - if (!declEntity) - return XML_ERROR_NO_MEMORY; - if (declEntity->name != name) { - poolDiscard(&dtd->pool); - declEntity = NULL; - } - else { - poolFinish(&dtd->pool); - declEntity->publicId = NULL; - declEntity->is_param = XML_TRUE; - /* if we have a parent parser or are reading an internal parameter - entity, then the entity declaration is not considered "internal" - */ - declEntity->is_internal = !(parentParser || openInternalEntities); - if (entityDeclHandler) - handleDefault = XML_FALSE; - } - } - else { - poolDiscard(&dtd->pool); - declEntity = NULL; - } -#else /* not XML_DTD */ - declEntity = NULL; -#endif /* XML_DTD */ - break; - case XML_ROLE_NOTATION_NAME: - declNotationPublicId = NULL; - declNotationName = NULL; - if (notationDeclHandler) { - declNotationName = poolStoreString(&tempPool, enc, s, next); - if (!declNotationName) - return XML_ERROR_NO_MEMORY; - poolFinish(&tempPool); - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_NOTATION_PUBLIC_ID: - if (!XmlIsPublicId(enc, s, next, eventPP)) - return XML_ERROR_PUBLICID; - if (declNotationName) { /* means notationDeclHandler != NULL */ - XML_Char *tem = poolStoreString(&tempPool, - enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!tem) - return XML_ERROR_NO_MEMORY; - normalizePublicId(tem); - declNotationPublicId = tem; - poolFinish(&tempPool); - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_NOTATION_SYSTEM_ID: - if (declNotationName && notationDeclHandler) { - const XML_Char *systemId - = poolStoreString(&tempPool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!systemId) - return XML_ERROR_NO_MEMORY; - *eventEndPP = s; - notationDeclHandler(handlerArg, - declNotationName, - curBase, - systemId, - declNotationPublicId); - handleDefault = XML_FALSE; - } - poolClear(&tempPool); - break; - case XML_ROLE_NOTATION_NO_SYSTEM_ID: - if (declNotationPublicId && notationDeclHandler) { - *eventEndPP = s; - notationDeclHandler(handlerArg, - declNotationName, - curBase, - 0, - declNotationPublicId); - handleDefault = XML_FALSE; - } - poolClear(&tempPool); - break; - case XML_ROLE_ERROR: - switch (tok) { - case XML_TOK_PARAM_ENTITY_REF: - /* PE references in internal subset are - not allowed within declarations. */ - return XML_ERROR_PARAM_ENTITY_REF; - case XML_TOK_XML_DECL: - return XML_ERROR_MISPLACED_XML_PI; - default: - return XML_ERROR_SYNTAX; - } -#ifdef XML_DTD - case XML_ROLE_IGNORE_SECT: - { - enum XML_Error result; - if (defaultHandler) - reportDefault(parser, enc, s, next); - handleDefault = XML_FALSE; - result = doIgnoreSection(parser, enc, &next, end, nextPtr, haveMore); - if (result != XML_ERROR_NONE) - return result; - else if (!next) { - processor = ignoreSectionProcessor; - return result; - } - } - break; -#endif /* XML_DTD */ - case XML_ROLE_GROUP_OPEN: - if (prologState.level >= groupSize) { - if (groupSize) { - char *temp = (char *)REALLOC(groupConnector, groupSize *= 2); - if (temp == NULL) - return XML_ERROR_NO_MEMORY; - groupConnector = temp; - if (dtd->scaffIndex) { - int *temp = (int *)REALLOC(dtd->scaffIndex, - groupSize * sizeof(int)); - if (temp == NULL) - return XML_ERROR_NO_MEMORY; - dtd->scaffIndex = temp; - } - } - else { - groupConnector = (char *)MALLOC(groupSize = 32); - if (!groupConnector) - return XML_ERROR_NO_MEMORY; - } - } - groupConnector[prologState.level] = 0; - if (dtd->in_eldecl) { - int myindex = nextScaffoldPart(parser); - if (myindex < 0) - return XML_ERROR_NO_MEMORY; - dtd->scaffIndex[dtd->scaffLevel] = myindex; - dtd->scaffLevel++; - dtd->scaffold[myindex].type = XML_CTYPE_SEQ; - if (elementDeclHandler) - handleDefault = XML_FALSE; - } - break; - case XML_ROLE_GROUP_SEQUENCE: - if (groupConnector[prologState.level] == ASCII_PIPE) - return XML_ERROR_SYNTAX; - groupConnector[prologState.level] = ASCII_COMMA; - if (dtd->in_eldecl && elementDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_GROUP_CHOICE: - if (groupConnector[prologState.level] == ASCII_COMMA) - return XML_ERROR_SYNTAX; - if (dtd->in_eldecl - && !groupConnector[prologState.level] - && (dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type - != XML_CTYPE_MIXED) - ) { - dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type - = XML_CTYPE_CHOICE; - if (elementDeclHandler) - handleDefault = XML_FALSE; - } - groupConnector[prologState.level] = ASCII_PIPE; - break; - case XML_ROLE_PARAM_ENTITY_REF: -#ifdef XML_DTD - case XML_ROLE_INNER_PARAM_ENTITY_REF: - dtd->hasParamEntityRefs = XML_TRUE; - if (!paramEntityParsing) - dtd->keepProcessing = dtd->standalone; - else { - const XML_Char *name; - ENTITY *entity; - name = poolStoreString(&dtd->pool, enc, - s + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!name) - return XML_ERROR_NO_MEMORY; - entity = (ENTITY *)lookup(&dtd->paramEntities, name, 0); - poolDiscard(&dtd->pool); - /* first, determine if a check for an existing declaration is needed; - if yes, check that the entity exists, and that it is internal, - otherwise call the skipped entity handler - */ - if (prologState.documentEntity && - (dtd->standalone - ? !openInternalEntities - : !dtd->hasParamEntityRefs)) { - if (!entity) - return XML_ERROR_UNDEFINED_ENTITY; - else if (!entity->is_internal) - return XML_ERROR_ENTITY_DECLARED_IN_PE; - } - else if (!entity) { - dtd->keepProcessing = dtd->standalone; - /* cannot report skipped entities in declarations */ - if ((role == XML_ROLE_PARAM_ENTITY_REF) && skippedEntityHandler) { - skippedEntityHandler(handlerArg, name, 1); - handleDefault = XML_FALSE; - } - break; - } - if (entity->open) - return XML_ERROR_RECURSIVE_ENTITY_REF; - if (entity->textPtr) { - enum XML_Error result; - XML_Bool betweenDecl = - (role == XML_ROLE_PARAM_ENTITY_REF ? XML_TRUE : XML_FALSE); - result = processInternalEntity(parser, entity, betweenDecl); - if (result != XML_ERROR_NONE) - return result; - handleDefault = XML_FALSE; - break; - } - if (externalEntityRefHandler) { - dtd->paramEntityRead = XML_FALSE; - entity->open = XML_TRUE; - if (!externalEntityRefHandler(externalEntityRefHandlerArg, - 0, - entity->base, - entity->systemId, - entity->publicId)) { - entity->open = XML_FALSE; - return XML_ERROR_EXTERNAL_ENTITY_HANDLING; - } - entity->open = XML_FALSE; - handleDefault = XML_FALSE; - if (!dtd->paramEntityRead) { - dtd->keepProcessing = dtd->standalone; - break; - } - } - else { - dtd->keepProcessing = dtd->standalone; - break; - } - } -#endif /* XML_DTD */ - if (!dtd->standalone && - notStandaloneHandler && - !notStandaloneHandler(handlerArg)) - return XML_ERROR_NOT_STANDALONE; - break; - - /* Element declaration stuff */ - - case XML_ROLE_ELEMENT_NAME: - if (elementDeclHandler) { - declElementType = getElementType(parser, enc, s, next); - if (!declElementType) - return XML_ERROR_NO_MEMORY; - dtd->scaffLevel = 0; - dtd->scaffCount = 0; - dtd->in_eldecl = XML_TRUE; - handleDefault = XML_FALSE; - } - break; - - case XML_ROLE_CONTENT_ANY: - case XML_ROLE_CONTENT_EMPTY: - if (dtd->in_eldecl) { - if (elementDeclHandler) { - XML_Content * content = (XML_Content *) MALLOC(sizeof(XML_Content)); - if (!content) - return XML_ERROR_NO_MEMORY; - content->quant = XML_CQUANT_NONE; - content->name = NULL; - content->numchildren = 0; - content->children = NULL; - content->type = ((role == XML_ROLE_CONTENT_ANY) ? - XML_CTYPE_ANY : - XML_CTYPE_EMPTY); - *eventEndPP = s; - elementDeclHandler(handlerArg, declElementType->name, content); - handleDefault = XML_FALSE; - } - dtd->in_eldecl = XML_FALSE; - } - break; - - case XML_ROLE_CONTENT_PCDATA: - if (dtd->in_eldecl) { - dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type - = XML_CTYPE_MIXED; - if (elementDeclHandler) - handleDefault = XML_FALSE; - } - break; - - case XML_ROLE_CONTENT_ELEMENT: - quant = XML_CQUANT_NONE; - goto elementContent; - case XML_ROLE_CONTENT_ELEMENT_OPT: - quant = XML_CQUANT_OPT; - goto elementContent; - case XML_ROLE_CONTENT_ELEMENT_REP: - quant = XML_CQUANT_REP; - goto elementContent; - case XML_ROLE_CONTENT_ELEMENT_PLUS: - quant = XML_CQUANT_PLUS; - elementContent: - if (dtd->in_eldecl) { - ELEMENT_TYPE *el; - const XML_Char *name; - int nameLen; - const char *nxt = (quant == XML_CQUANT_NONE - ? next - : next - enc->minBytesPerChar); - int myindex = nextScaffoldPart(parser); - if (myindex < 0) - return XML_ERROR_NO_MEMORY; - dtd->scaffold[myindex].type = XML_CTYPE_NAME; - dtd->scaffold[myindex].quant = quant; - el = getElementType(parser, enc, s, nxt); - if (!el) - return XML_ERROR_NO_MEMORY; - name = el->name; - dtd->scaffold[myindex].name = name; - nameLen = 0; - for (; name[nameLen++]; ); - dtd->contentStringLen += nameLen; - if (elementDeclHandler) - handleDefault = XML_FALSE; - } - break; - - case XML_ROLE_GROUP_CLOSE: - quant = XML_CQUANT_NONE; - goto closeGroup; - case XML_ROLE_GROUP_CLOSE_OPT: - quant = XML_CQUANT_OPT; - goto closeGroup; - case XML_ROLE_GROUP_CLOSE_REP: - quant = XML_CQUANT_REP; - goto closeGroup; - case XML_ROLE_GROUP_CLOSE_PLUS: - quant = XML_CQUANT_PLUS; - closeGroup: - if (dtd->in_eldecl) { - if (elementDeclHandler) - handleDefault = XML_FALSE; - dtd->scaffLevel--; - dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel]].quant = quant; - if (dtd->scaffLevel == 0) { - if (!handleDefault) { - XML_Content *model = build_model(parser); - if (!model) - return XML_ERROR_NO_MEMORY; - *eventEndPP = s; - elementDeclHandler(handlerArg, declElementType->name, model); - } - dtd->in_eldecl = XML_FALSE; - dtd->contentStringLen = 0; - } - } - break; - /* End element declaration stuff */ - - case XML_ROLE_PI: - if (!reportProcessingInstruction(parser, enc, s, next)) - return XML_ERROR_NO_MEMORY; - handleDefault = XML_FALSE; - break; - case XML_ROLE_COMMENT: - if (!reportComment(parser, enc, s, next)) - return XML_ERROR_NO_MEMORY; - handleDefault = XML_FALSE; - break; - case XML_ROLE_NONE: - switch (tok) { - case XML_TOK_BOM: - handleDefault = XML_FALSE; - break; - } - break; - case XML_ROLE_DOCTYPE_NONE: - if (startDoctypeDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_ENTITY_NONE: - if (dtd->keepProcessing && entityDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_NOTATION_NONE: - if (notationDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_ATTLIST_NONE: - if (dtd->keepProcessing && attlistDeclHandler) - handleDefault = XML_FALSE; - break; - case XML_ROLE_ELEMENT_NONE: - if (elementDeclHandler) - handleDefault = XML_FALSE; - break; - } /* end of big switch */ - - if (handleDefault && defaultHandler) - reportDefault(parser, enc, s, next); - - switch (ps_parsing) { - case XML_SUSPENDED: - *nextPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: - s = next; - tok = XmlPrologTok(enc, s, end, &next); - } - } - /* not reached */ -} - -static enum XML_Error PTRCALL -epilogProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - processor = epilogProcessor; - eventPtr = s; - for (;;) { - const char *next = NULL; - int tok = XmlPrologTok(encoding, s, end, &next); - eventEndPtr = next; - switch (tok) { - /* report partial linebreak - it might be the last token */ - case -XML_TOK_PROLOG_S: - if (defaultHandler) { - reportDefault(parser, encoding, s, next); - if (ps_parsing == XML_FINISHED) - return XML_ERROR_ABORTED; - } - *nextPtr = next; - return XML_ERROR_NONE; - case XML_TOK_NONE: - *nextPtr = s; - return XML_ERROR_NONE; - case XML_TOK_PROLOG_S: - if (defaultHandler) - reportDefault(parser, encoding, s, next); - break; - case XML_TOK_PI: - if (!reportProcessingInstruction(parser, encoding, s, next)) - return XML_ERROR_NO_MEMORY; - break; - case XML_TOK_COMMENT: - if (!reportComment(parser, encoding, s, next)) - return XML_ERROR_NO_MEMORY; - break; - case XML_TOK_INVALID: - eventPtr = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - if (!ps_finalBuffer) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_UNCLOSED_TOKEN; - case XML_TOK_PARTIAL_CHAR: - if (!ps_finalBuffer) { - *nextPtr = s; - return XML_ERROR_NONE; - } - return XML_ERROR_PARTIAL_CHAR; - default: - return XML_ERROR_JUNK_AFTER_DOC_ELEMENT; - } - eventPtr = s = next; - switch (ps_parsing) { - case XML_SUSPENDED: - *nextPtr = next; - return XML_ERROR_NONE; - case XML_FINISHED: - return XML_ERROR_ABORTED; - default: ; - } - } -} - -static enum XML_Error -processInternalEntity(XML_Parser parser, ENTITY *entity, - XML_Bool betweenDecl) -{ - const char *textStart, *textEnd; - const char *next; - enum XML_Error result; - OPEN_INTERNAL_ENTITY *openEntity; - - if (freeInternalEntities) { - openEntity = freeInternalEntities; - freeInternalEntities = openEntity->next; - } - else { - openEntity = (OPEN_INTERNAL_ENTITY *)MALLOC(sizeof(OPEN_INTERNAL_ENTITY)); - if (!openEntity) - return XML_ERROR_NO_MEMORY; - } - entity->open = XML_TRUE; - entity->processed = 0; - openEntity->next = openInternalEntities; - openInternalEntities = openEntity; - openEntity->entity = entity; - openEntity->startTagLevel = tagLevel; - openEntity->betweenDecl = betweenDecl; - openEntity->internalEventPtr = NULL; - openEntity->internalEventEndPtr = NULL; - textStart = (char *)entity->textPtr; - textEnd = (char *)(entity->textPtr + entity->textLen); - -#ifdef XML_DTD - if (entity->is_param) { - int tok = XmlPrologTok(internalEncoding, textStart, textEnd, &next); - result = doProlog(parser, internalEncoding, textStart, textEnd, tok, - next, &next, XML_FALSE); - } - else -#endif /* XML_DTD */ - result = doContent(parser, tagLevel, internalEncoding, textStart, - textEnd, &next, XML_FALSE); - - if (result == XML_ERROR_NONE) { - if (textEnd != next && ps_parsing == XML_SUSPENDED) { - entity->processed = (int)(next - textStart); - processor = internalEntityProcessor; - } - else { - entity->open = XML_FALSE; - openInternalEntities = openEntity->next; - /* put openEntity back in list of free instances */ - openEntity->next = freeInternalEntities; - freeInternalEntities = openEntity; - } - } - return result; -} - -static enum XML_Error PTRCALL -internalEntityProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - ENTITY *entity; - const char *textStart, *textEnd; - const char *next; - enum XML_Error result; - OPEN_INTERNAL_ENTITY *openEntity = openInternalEntities; - if (!openEntity) - return XML_ERROR_UNEXPECTED_STATE; - - entity = openEntity->entity; - textStart = ((char *)entity->textPtr) + entity->processed; - textEnd = (char *)(entity->textPtr + entity->textLen); - -#ifdef XML_DTD - if (entity->is_param) { - int tok = XmlPrologTok(internalEncoding, textStart, textEnd, &next); - result = doProlog(parser, internalEncoding, textStart, textEnd, tok, - next, &next, XML_FALSE); - } - else -#endif /* XML_DTD */ - result = doContent(parser, openEntity->startTagLevel, internalEncoding, - textStart, textEnd, &next, XML_FALSE); - - if (result != XML_ERROR_NONE) - return result; - else if (textEnd != next && ps_parsing == XML_SUSPENDED) { - entity->processed = (int)(next - (char *)entity->textPtr); - return result; - } - else { - entity->open = XML_FALSE; - openInternalEntities = openEntity->next; - /* put openEntity back in list of free instances */ - openEntity->next = freeInternalEntities; - freeInternalEntities = openEntity; - } - -#ifdef XML_DTD - if (entity->is_param) { - int tok; - processor = prologProcessor; - tok = XmlPrologTok(encoding, s, end, &next); - return doProlog(parser, encoding, s, end, tok, next, nextPtr, - (XML_Bool)!ps_finalBuffer); - } - else -#endif /* XML_DTD */ - { - processor = contentProcessor; - /* see externalEntityContentProcessor vs contentProcessor */ - return doContent(parser, parentParser ? 1 : 0, encoding, s, end, - nextPtr, (XML_Bool)!ps_finalBuffer); - } -} - -static enum XML_Error PTRCALL -errorProcessor(XML_Parser parser, - const char *s, - const char *end, - const char **nextPtr) -{ - return errorCode; -} - -static enum XML_Error -storeAttributeValue(XML_Parser parser, const ENCODING *enc, XML_Bool isCdata, - const char *ptr, const char *end, - STRING_POOL *pool) -{ - enum XML_Error result = appendAttributeValue(parser, enc, isCdata, ptr, - end, pool); - if (result) - return result; - if (!isCdata && poolLength(pool) && poolLastChar(pool) == 0x20) - poolChop(pool); - if (!poolAppendChar(pool, XML_T('\0'))) - return XML_ERROR_NO_MEMORY; - return XML_ERROR_NONE; -} - -static enum XML_Error -appendAttributeValue(XML_Parser parser, const ENCODING *enc, XML_Bool isCdata, - const char *ptr, const char *end, - STRING_POOL *pool) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - for (;;) { - const char *next; - int tok = XmlAttributeValueTok(enc, ptr, end, &next); - switch (tok) { - case XML_TOK_NONE: - return XML_ERROR_NONE; - case XML_TOK_INVALID: - if (enc == encoding) - eventPtr = next; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_PARTIAL: - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_INVALID_TOKEN; - case XML_TOK_CHAR_REF: - { - XML_Char buf[XML_ENCODE_MAX]; - int i; - int n = XmlCharRefNumber(enc, ptr); - if (n < 0) { - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_BAD_CHAR_REF; - } - if (!isCdata - && n == 0x20 /* space */ - && (poolLength(pool) == 0 || poolLastChar(pool) == 0x20)) - break; - n = XmlEncode(n, (ICHAR *)buf); - if (!n) { - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_BAD_CHAR_REF; - } - for (i = 0; i < n; i++) { - if (!poolAppendChar(pool, buf[i])) - return XML_ERROR_NO_MEMORY; - } - } - break; - case XML_TOK_DATA_CHARS: - if (!poolAppend(pool, enc, ptr, next)) - return XML_ERROR_NO_MEMORY; - break; - case XML_TOK_TRAILING_CR: - next = ptr + enc->minBytesPerChar; - /* fall through */ - case XML_TOK_ATTRIBUTE_VALUE_S: - case XML_TOK_DATA_NEWLINE: - if (!isCdata && (poolLength(pool) == 0 || poolLastChar(pool) == 0x20)) - break; - if (!poolAppendChar(pool, 0x20)) - return XML_ERROR_NO_MEMORY; - break; - case XML_TOK_ENTITY_REF: - { - const XML_Char *name; - ENTITY *entity; - char checkEntityDecl; - XML_Char ch = (XML_Char) XmlPredefinedEntityName(enc, - ptr + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (ch) { - if (!poolAppendChar(pool, ch)) - return XML_ERROR_NO_MEMORY; - break; - } - name = poolStoreString(&temp2Pool, enc, - ptr + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!name) - return XML_ERROR_NO_MEMORY; - entity = (ENTITY *)lookup(&dtd->generalEntities, name, 0); - poolDiscard(&temp2Pool); - /* First, determine if a check for an existing declaration is needed; - if yes, check that the entity exists, and that it is internal. - */ - if (pool == &dtd->pool) /* are we called from prolog? */ - checkEntityDecl = -#ifdef XML_DTD - prologState.documentEntity && -#endif /* XML_DTD */ - (dtd->standalone - ? !openInternalEntities - : !dtd->hasParamEntityRefs); - else /* if (pool == &tempPool): we are called from content */ - checkEntityDecl = !dtd->hasParamEntityRefs || dtd->standalone; - if (checkEntityDecl) { - if (!entity) - return XML_ERROR_UNDEFINED_ENTITY; - else if (!entity->is_internal) - return XML_ERROR_ENTITY_DECLARED_IN_PE; - } - else if (!entity) { - /* Cannot report skipped entity here - see comments on - skippedEntityHandler. - if (skippedEntityHandler) - skippedEntityHandler(handlerArg, name, 0); - */ - /* Cannot call the default handler because this would be - out of sync with the call to the startElementHandler. - if ((pool == &tempPool) && defaultHandler) - reportDefault(parser, enc, ptr, next); - */ - break; - } - if (entity->open) { - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_RECURSIVE_ENTITY_REF; - } - if (entity->notation) { - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_BINARY_ENTITY_REF; - } - if (!entity->textPtr) { - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF; - } - else { - enum XML_Error result; - const XML_Char *textEnd = entity->textPtr + entity->textLen; - entity->open = XML_TRUE; - result = appendAttributeValue(parser, internalEncoding, isCdata, - (char *)entity->textPtr, - (char *)textEnd, pool); - entity->open = XML_FALSE; - if (result) - return result; - } - } - break; - default: - if (enc == encoding) - eventPtr = ptr; - return XML_ERROR_UNEXPECTED_STATE; - } - ptr = next; - } - /* not reached */ -} - -static enum XML_Error -storeEntityValue(XML_Parser parser, - const ENCODING *enc, - const char *entityTextPtr, - const char *entityTextEnd) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - STRING_POOL *pool = &(dtd->entityValuePool); - enum XML_Error result = XML_ERROR_NONE; -#ifdef XML_DTD - int oldInEntityValue = prologState.inEntityValue; - prologState.inEntityValue = 1; -#endif /* XML_DTD */ - /* never return Null for the value argument in EntityDeclHandler, - since this would indicate an external entity; therefore we - have to make sure that entityValuePool.start is not null */ - if (!pool->blocks) { - if (!poolGrow(pool)) - return XML_ERROR_NO_MEMORY; - } - - for (;;) { - const char *next; - int tok = XmlEntityValueTok(enc, entityTextPtr, entityTextEnd, &next); - switch (tok) { - case XML_TOK_PARAM_ENTITY_REF: -#ifdef XML_DTD - if (isParamEntity || enc != encoding) { - const XML_Char *name; - ENTITY *entity; - name = poolStoreString(&tempPool, enc, - entityTextPtr + enc->minBytesPerChar, - next - enc->minBytesPerChar); - if (!name) { - result = XML_ERROR_NO_MEMORY; - goto endEntityValue; - } - entity = (ENTITY *)lookup(&dtd->paramEntities, name, 0); - poolDiscard(&tempPool); - if (!entity) { - /* not a well-formedness error - see XML 1.0: WFC Entity Declared */ - /* cannot report skipped entity here - see comments on - skippedEntityHandler - if (skippedEntityHandler) - skippedEntityHandler(handlerArg, name, 0); - */ - dtd->keepProcessing = dtd->standalone; - goto endEntityValue; - } - if (entity->open) { - if (enc == encoding) - eventPtr = entityTextPtr; - result = XML_ERROR_RECURSIVE_ENTITY_REF; - goto endEntityValue; - } - if (entity->systemId) { - if (externalEntityRefHandler) { - dtd->paramEntityRead = XML_FALSE; - entity->open = XML_TRUE; - if (!externalEntityRefHandler(externalEntityRefHandlerArg, - 0, - entity->base, - entity->systemId, - entity->publicId)) { - entity->open = XML_FALSE; - result = XML_ERROR_EXTERNAL_ENTITY_HANDLING; - goto endEntityValue; - } - entity->open = XML_FALSE; - if (!dtd->paramEntityRead) - dtd->keepProcessing = dtd->standalone; - } - else - dtd->keepProcessing = dtd->standalone; - } - else { - entity->open = XML_TRUE; - result = storeEntityValue(parser, - internalEncoding, - (char *)entity->textPtr, - (char *)(entity->textPtr - + entity->textLen)); - entity->open = XML_FALSE; - if (result) - goto endEntityValue; - } - break; - } -#endif /* XML_DTD */ - /* In the internal subset, PE references are not legal - within markup declarations, e.g entity values in this case. */ - eventPtr = entityTextPtr; - result = XML_ERROR_PARAM_ENTITY_REF; - goto endEntityValue; - case XML_TOK_NONE: - result = XML_ERROR_NONE; - goto endEntityValue; - case XML_TOK_ENTITY_REF: - case XML_TOK_DATA_CHARS: - if (!poolAppend(pool, enc, entityTextPtr, next)) { - result = XML_ERROR_NO_MEMORY; - goto endEntityValue; - } - break; - case XML_TOK_TRAILING_CR: - next = entityTextPtr + enc->minBytesPerChar; - /* fall through */ - case XML_TOK_DATA_NEWLINE: - if (pool->end == pool->ptr && !poolGrow(pool)) { - result = XML_ERROR_NO_MEMORY; - goto endEntityValue; - } - *(pool->ptr)++ = 0xA; - break; - case XML_TOK_CHAR_REF: - { - XML_Char buf[XML_ENCODE_MAX]; - int i; - int n = XmlCharRefNumber(enc, entityTextPtr); - if (n < 0) { - if (enc == encoding) - eventPtr = entityTextPtr; - result = XML_ERROR_BAD_CHAR_REF; - goto endEntityValue; - } - n = XmlEncode(n, (ICHAR *)buf); - if (!n) { - if (enc == encoding) - eventPtr = entityTextPtr; - result = XML_ERROR_BAD_CHAR_REF; - goto endEntityValue; - } - for (i = 0; i < n; i++) { - if (pool->end == pool->ptr && !poolGrow(pool)) { - result = XML_ERROR_NO_MEMORY; - goto endEntityValue; - } - *(pool->ptr)++ = buf[i]; - } - } - break; - case XML_TOK_PARTIAL: - if (enc == encoding) - eventPtr = entityTextPtr; - result = XML_ERROR_INVALID_TOKEN; - goto endEntityValue; - case XML_TOK_INVALID: - if (enc == encoding) - eventPtr = next; - result = XML_ERROR_INVALID_TOKEN; - goto endEntityValue; - default: - if (enc == encoding) - eventPtr = entityTextPtr; - result = XML_ERROR_UNEXPECTED_STATE; - goto endEntityValue; - } - entityTextPtr = next; - } -endEntityValue: -#ifdef XML_DTD - prologState.inEntityValue = oldInEntityValue; -#endif /* XML_DTD */ - return result; -} - -static void FASTCALL -normalizeLines(XML_Char *s) -{ - XML_Char *p; - for (;; s++) { - if (*s == XML_T('\0')) - return; - if (*s == 0xD) - break; - } - p = s; - do { - if (*s == 0xD) { - *p++ = 0xA; - if (*++s == 0xA) - s++; - } - else - *p++ = *s++; - } while (*s); - *p = XML_T('\0'); -} - -static int -reportProcessingInstruction(XML_Parser parser, const ENCODING *enc, - const char *start, const char *end) -{ - const XML_Char *target; - XML_Char *data; - const char *tem; - if (!processingInstructionHandler) { - if (defaultHandler) - reportDefault(parser, enc, start, end); - return 1; - } - start += enc->minBytesPerChar * 2; - tem = start + XmlNameLength(enc, start); - target = poolStoreString(&tempPool, enc, start, tem); - if (!target) - return 0; - poolFinish(&tempPool); - data = poolStoreString(&tempPool, enc, - XmlSkipS(enc, tem), - end - enc->minBytesPerChar*2); - if (!data) - return 0; - normalizeLines(data); - processingInstructionHandler(handlerArg, target, data); - poolClear(&tempPool); - return 1; -} - -static int -reportComment(XML_Parser parser, const ENCODING *enc, - const char *start, const char *end) -{ - XML_Char *data; - if (!commentHandler) { - if (defaultHandler) - reportDefault(parser, enc, start, end); - return 1; - } - data = poolStoreString(&tempPool, - enc, - start + enc->minBytesPerChar * 4, - end - enc->minBytesPerChar * 3); - if (!data) - return 0; - normalizeLines(data); - commentHandler(handlerArg, data); - poolClear(&tempPool); - return 1; -} - -static void -reportDefault(XML_Parser parser, const ENCODING *enc, - const char *s, const char *end) -{ - if (MUST_CONVERT(enc, s)) { - const char **eventPP; - const char **eventEndPP; - if (enc == encoding) { - eventPP = &eventPtr; - eventEndPP = &eventEndPtr; - } - else { - eventPP = &(openInternalEntities->internalEventPtr); - eventEndPP = &(openInternalEntities->internalEventEndPtr); - } - do { - ICHAR *dataPtr = (ICHAR *)dataBuf; - XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd); - *eventEndPP = s; - defaultHandler(handlerArg, dataBuf, (int)(dataPtr - (ICHAR *)dataBuf)); - *eventPP = s; - } while (s != end); - } - else - defaultHandler(handlerArg, (XML_Char *)s, (int)((XML_Char *)end - (XML_Char *)s)); -} - - -static int -defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *attId, XML_Bool isCdata, - XML_Bool isId, const XML_Char *value, XML_Parser parser) -{ - DEFAULT_ATTRIBUTE *att; - if (value || isId) { - /* The handling of default attributes gets messed up if we have - a default which duplicates a non-default. */ - int i; - for (i = 0; i < type->nDefaultAtts; i++) - if (attId == type->defaultAtts[i].id) - return 1; - if (isId && !type->idAtt && !attId->xmlns) - type->idAtt = attId; - } - if (type->nDefaultAtts == type->allocDefaultAtts) { - if (type->allocDefaultAtts == 0) { - type->allocDefaultAtts = 8; - type->defaultAtts = (DEFAULT_ATTRIBUTE *)MALLOC(type->allocDefaultAtts - * sizeof(DEFAULT_ATTRIBUTE)); - if (!type->defaultAtts) - return 0; - } - else { - DEFAULT_ATTRIBUTE *temp; - int count = type->allocDefaultAtts * 2; - temp = (DEFAULT_ATTRIBUTE *) - REALLOC(type->defaultAtts, (count * sizeof(DEFAULT_ATTRIBUTE))); - if (temp == NULL) - return 0; - type->allocDefaultAtts = count; - type->defaultAtts = temp; - } - } - att = type->defaultAtts + type->nDefaultAtts; - att->id = attId; - att->value = value; - att->isCdata = isCdata; - if (!isCdata) - attId->maybeTokenized = XML_TRUE; - type->nDefaultAtts += 1; - return 1; -} - -static int -setElementTypePrefix(XML_Parser parser, ELEMENT_TYPE *elementType) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - const XML_Char *name; - for (name = elementType->name; *name; name++) { - if (*name == XML_T(ASCII_COLON)) { - PREFIX *prefix; - const XML_Char *s; - for (s = elementType->name; s != name; s++) { - if (!poolAppendChar(&dtd->pool, *s)) - return 0; - } - if (!poolAppendChar(&dtd->pool, XML_T('\0'))) - return 0; - prefix = (PREFIX *)lookup(&dtd->prefixes, poolStart(&dtd->pool), - sizeof(PREFIX)); - if (!prefix) - return 0; - if (prefix->name == poolStart(&dtd->pool)) - poolFinish(&dtd->pool); - else - poolDiscard(&dtd->pool); - elementType->prefix = prefix; - - } - } - return 1; -} - -static ATTRIBUTE_ID * -getAttributeId(XML_Parser parser, const ENCODING *enc, - const char *start, const char *end) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - ATTRIBUTE_ID *id; - const XML_Char *name; - if (!poolAppendChar(&dtd->pool, XML_T('\0'))) - return NULL; - name = poolStoreString(&dtd->pool, enc, start, end); - if (!name) - return NULL; - /* skip quotation mark - its storage will be re-used (like in name[-1]) */ - ++name; - id = (ATTRIBUTE_ID *)lookup(&dtd->attributeIds, name, sizeof(ATTRIBUTE_ID)); - if (!id) - return NULL; - if (id->name != name) - poolDiscard(&dtd->pool); - else { - poolFinish(&dtd->pool); - if (!ns) - ; - else if (name[0] == XML_T(ASCII_x) - && name[1] == XML_T(ASCII_m) - && name[2] == XML_T(ASCII_l) - && name[3] == XML_T(ASCII_n) - && name[4] == XML_T(ASCII_s) - && (name[5] == XML_T('\0') || name[5] == XML_T(ASCII_COLON))) { - if (name[5] == XML_T('\0')) - id->prefix = &dtd->defaultPrefix; - else - id->prefix = (PREFIX *)lookup(&dtd->prefixes, name + 6, sizeof(PREFIX)); - id->xmlns = XML_TRUE; - } - else { - int i; - for (i = 0; name[i]; i++) { - /* attributes without prefix are *not* in the default namespace */ - if (name[i] == XML_T(ASCII_COLON)) { - int j; - for (j = 0; j < i; j++) { - if (!poolAppendChar(&dtd->pool, name[j])) - return NULL; - } - if (!poolAppendChar(&dtd->pool, XML_T('\0'))) - return NULL; - id->prefix = (PREFIX *)lookup(&dtd->prefixes, poolStart(&dtd->pool), - sizeof(PREFIX)); - if (id->prefix->name == poolStart(&dtd->pool)) - poolFinish(&dtd->pool); - else - poolDiscard(&dtd->pool); - break; - } - } - } - } - return id; -} - -#define CONTEXT_SEP XML_T(ASCII_FF) - -static const XML_Char * -getContext(XML_Parser parser) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - HASH_TABLE_ITER iter; - XML_Bool needSep = XML_FALSE; - - if (dtd->defaultPrefix.binding) { - int i; - int len; - if (!poolAppendChar(&tempPool, XML_T(ASCII_EQUALS))) - return NULL; - len = dtd->defaultPrefix.binding->uriLen; - if (namespaceSeparator) - len--; - for (i = 0; i < len; i++) - if (!poolAppendChar(&tempPool, dtd->defaultPrefix.binding->uri[i])) - return NULL; - needSep = XML_TRUE; - } - - hashTableIterInit(&iter, &(dtd->prefixes)); - for (;;) { - int i; - int len; - const XML_Char *s; - PREFIX *prefix = (PREFIX *)hashTableIterNext(&iter); - if (!prefix) - break; - if (!prefix->binding) - continue; - if (needSep && !poolAppendChar(&tempPool, CONTEXT_SEP)) - return NULL; - for (s = prefix->name; *s; s++) - if (!poolAppendChar(&tempPool, *s)) - return NULL; - if (!poolAppendChar(&tempPool, XML_T(ASCII_EQUALS))) - return NULL; - len = prefix->binding->uriLen; - if (namespaceSeparator) - len--; - for (i = 0; i < len; i++) - if (!poolAppendChar(&tempPool, prefix->binding->uri[i])) - return NULL; - needSep = XML_TRUE; - } - - - hashTableIterInit(&iter, &(dtd->generalEntities)); - for (;;) { - const XML_Char *s; - ENTITY *e = (ENTITY *)hashTableIterNext(&iter); - if (!e) - break; - if (!e->open) - continue; - if (needSep && !poolAppendChar(&tempPool, CONTEXT_SEP)) - return NULL; - for (s = e->name; *s; s++) - if (!poolAppendChar(&tempPool, *s)) - return 0; - needSep = XML_TRUE; - } - - if (!poolAppendChar(&tempPool, XML_T('\0'))) - return NULL; - return tempPool.start; -} - -static XML_Bool -setContext(XML_Parser parser, const XML_Char *context) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - const XML_Char *s = context; - - while (*context != XML_T('\0')) { - if (*s == CONTEXT_SEP || *s == XML_T('\0')) { - ENTITY *e; - if (!poolAppendChar(&tempPool, XML_T('\0'))) - return XML_FALSE; - e = (ENTITY *)lookup(&dtd->generalEntities, poolStart(&tempPool), 0); - if (e) - e->open = XML_TRUE; - if (*s != XML_T('\0')) - s++; - context = s; - poolDiscard(&tempPool); - } - else if (*s == XML_T(ASCII_EQUALS)) { - PREFIX *prefix; - if (poolLength(&tempPool) == 0) - prefix = &dtd->defaultPrefix; - else { - if (!poolAppendChar(&tempPool, XML_T('\0'))) - return XML_FALSE; - prefix = (PREFIX *)lookup(&dtd->prefixes, poolStart(&tempPool), - sizeof(PREFIX)); - if (!prefix) - return XML_FALSE; - if (prefix->name == poolStart(&tempPool)) { - prefix->name = poolCopyString(&dtd->pool, prefix->name); - if (!prefix->name) - return XML_FALSE; - } - poolDiscard(&tempPool); - } - for (context = s + 1; - *context != CONTEXT_SEP && *context != XML_T('\0'); - context++) - if (!poolAppendChar(&tempPool, *context)) - return XML_FALSE; - if (!poolAppendChar(&tempPool, XML_T('\0'))) - return XML_FALSE; - if (addBinding(parser, prefix, NULL, poolStart(&tempPool), - &inheritedBindings) != XML_ERROR_NONE) - return XML_FALSE; - poolDiscard(&tempPool); - if (*context != XML_T('\0')) - ++context; - s = context; - } - else { - if (!poolAppendChar(&tempPool, *s)) - return XML_FALSE; - s++; - } - } - return XML_TRUE; -} - -static void FASTCALL -normalizePublicId(XML_Char *publicId) -{ - XML_Char *p = publicId; - XML_Char *s; - for (s = publicId; *s; s++) { - switch (*s) { - case 0x20: - case 0xD: - case 0xA: - if (p != publicId && p[-1] != 0x20) - *p++ = 0x20; - break; - default: - *p++ = *s; - } - } - if (p != publicId && p[-1] == 0x20) - --p; - *p = XML_T('\0'); -} - -static DTD * -dtdCreate(const XML_Memory_Handling_Suite *ms) -{ - DTD *p = (DTD *)ms->malloc_fcn(sizeof(DTD)); - if (p == NULL) - return p; - poolInit(&(p->pool), ms); - poolInit(&(p->entityValuePool), ms); - hashTableInit(&(p->generalEntities), ms); - hashTableInit(&(p->elementTypes), ms); - hashTableInit(&(p->attributeIds), ms); - hashTableInit(&(p->prefixes), ms); -#ifdef XML_DTD - p->paramEntityRead = XML_FALSE; - hashTableInit(&(p->paramEntities), ms); -#endif /* XML_DTD */ - p->defaultPrefix.name = NULL; - p->defaultPrefix.binding = NULL; - - p->in_eldecl = XML_FALSE; - p->scaffIndex = NULL; - p->scaffold = NULL; - p->scaffLevel = 0; - p->scaffSize = 0; - p->scaffCount = 0; - p->contentStringLen = 0; - - p->keepProcessing = XML_TRUE; - p->hasParamEntityRefs = XML_FALSE; - p->standalone = XML_FALSE; - return p; -} - -static void -dtdReset(DTD *p, const XML_Memory_Handling_Suite *ms) -{ - HASH_TABLE_ITER iter; - hashTableIterInit(&iter, &(p->elementTypes)); - for (;;) { - ELEMENT_TYPE *e = (ELEMENT_TYPE *)hashTableIterNext(&iter); - if (!e) - break; - if (e->allocDefaultAtts != 0) - ms->free_fcn(e->defaultAtts); - } - hashTableClear(&(p->generalEntities)); -#ifdef XML_DTD - p->paramEntityRead = XML_FALSE; - hashTableClear(&(p->paramEntities)); -#endif /* XML_DTD */ - hashTableClear(&(p->elementTypes)); - hashTableClear(&(p->attributeIds)); - hashTableClear(&(p->prefixes)); - poolClear(&(p->pool)); - poolClear(&(p->entityValuePool)); - p->defaultPrefix.name = NULL; - p->defaultPrefix.binding = NULL; - - p->in_eldecl = XML_FALSE; - - ms->free_fcn(p->scaffIndex); - p->scaffIndex = NULL; - ms->free_fcn(p->scaffold); - p->scaffold = NULL; - - p->scaffLevel = 0; - p->scaffSize = 0; - p->scaffCount = 0; - p->contentStringLen = 0; - - p->keepProcessing = XML_TRUE; - p->hasParamEntityRefs = XML_FALSE; - p->standalone = XML_FALSE; -} - -static void -dtdDestroy(DTD *p, XML_Bool isDocEntity, const XML_Memory_Handling_Suite *ms) -{ - HASH_TABLE_ITER iter; - hashTableIterInit(&iter, &(p->elementTypes)); - for (;;) { - ELEMENT_TYPE *e = (ELEMENT_TYPE *)hashTableIterNext(&iter); - if (!e) - break; - if (e->allocDefaultAtts != 0) - ms->free_fcn(e->defaultAtts); - } - hashTableDestroy(&(p->generalEntities)); -#ifdef XML_DTD - hashTableDestroy(&(p->paramEntities)); -#endif /* XML_DTD */ - hashTableDestroy(&(p->elementTypes)); - hashTableDestroy(&(p->attributeIds)); - hashTableDestroy(&(p->prefixes)); - poolDestroy(&(p->pool)); - poolDestroy(&(p->entityValuePool)); - if (isDocEntity) { - ms->free_fcn(p->scaffIndex); - ms->free_fcn(p->scaffold); - } - ms->free_fcn(p); -} - -/* Do a deep copy of the DTD. Return 0 for out of memory, non-zero otherwise. - The new DTD has already been initialized. -*/ -static int -dtdCopy(DTD *newDtd, const DTD *oldDtd, const XML_Memory_Handling_Suite *ms) -{ - HASH_TABLE_ITER iter; - - /* Copy the prefix table. */ - - hashTableIterInit(&iter, &(oldDtd->prefixes)); - for (;;) { - const XML_Char *name; - const PREFIX *oldP = (PREFIX *)hashTableIterNext(&iter); - if (!oldP) - break; - name = poolCopyString(&(newDtd->pool), oldP->name); - if (!name) - return 0; - if (!lookup(&(newDtd->prefixes), name, sizeof(PREFIX))) - return 0; - } - - hashTableIterInit(&iter, &(oldDtd->attributeIds)); - - /* Copy the attribute id table. */ - - for (;;) { - ATTRIBUTE_ID *newA; - const XML_Char *name; - const ATTRIBUTE_ID *oldA = (ATTRIBUTE_ID *)hashTableIterNext(&iter); - - if (!oldA) - break; - /* Remember to allocate the scratch byte before the name. */ - if (!poolAppendChar(&(newDtd->pool), XML_T('\0'))) - return 0; - name = poolCopyString(&(newDtd->pool), oldA->name); - if (!name) - return 0; - ++name; - newA = (ATTRIBUTE_ID *)lookup(&(newDtd->attributeIds), name, - sizeof(ATTRIBUTE_ID)); - if (!newA) - return 0; - newA->maybeTokenized = oldA->maybeTokenized; - if (oldA->prefix) { - newA->xmlns = oldA->xmlns; - if (oldA->prefix == &oldDtd->defaultPrefix) - newA->prefix = &newDtd->defaultPrefix; - else - newA->prefix = (PREFIX *)lookup(&(newDtd->prefixes), - oldA->prefix->name, 0); - } - } - - /* Copy the element type table. */ - - hashTableIterInit(&iter, &(oldDtd->elementTypes)); - - for (;;) { - int i; - ELEMENT_TYPE *newE; - const XML_Char *name; - const ELEMENT_TYPE *oldE = (ELEMENT_TYPE *)hashTableIterNext(&iter); - if (!oldE) - break; - name = poolCopyString(&(newDtd->pool), oldE->name); - if (!name) - return 0; - newE = (ELEMENT_TYPE *)lookup(&(newDtd->elementTypes), name, - sizeof(ELEMENT_TYPE)); - if (!newE) - return 0; - if (oldE->nDefaultAtts) { - newE->defaultAtts = (DEFAULT_ATTRIBUTE *) - ms->malloc_fcn(oldE->nDefaultAtts * sizeof(DEFAULT_ATTRIBUTE)); - if (!newE->defaultAtts) { - ms->free_fcn(newE); - return 0; - } - } - if (oldE->idAtt) - newE->idAtt = (ATTRIBUTE_ID *) - lookup(&(newDtd->attributeIds), oldE->idAtt->name, 0); - newE->allocDefaultAtts = newE->nDefaultAtts = oldE->nDefaultAtts; - if (oldE->prefix) - newE->prefix = (PREFIX *)lookup(&(newDtd->prefixes), - oldE->prefix->name, 0); - for (i = 0; i < newE->nDefaultAtts; i++) { - newE->defaultAtts[i].id = (ATTRIBUTE_ID *) - lookup(&(newDtd->attributeIds), oldE->defaultAtts[i].id->name, 0); - newE->defaultAtts[i].isCdata = oldE->defaultAtts[i].isCdata; - if (oldE->defaultAtts[i].value) { - newE->defaultAtts[i].value - = poolCopyString(&(newDtd->pool), oldE->defaultAtts[i].value); - if (!newE->defaultAtts[i].value) - return 0; - } - else - newE->defaultAtts[i].value = NULL; - } - } - - /* Copy the entity tables. */ - if (!copyEntityTable(&(newDtd->generalEntities), - &(newDtd->pool), - &(oldDtd->generalEntities))) - return 0; - -#ifdef XML_DTD - if (!copyEntityTable(&(newDtd->paramEntities), - &(newDtd->pool), - &(oldDtd->paramEntities))) - return 0; - newDtd->paramEntityRead = oldDtd->paramEntityRead; -#endif /* XML_DTD */ - - newDtd->keepProcessing = oldDtd->keepProcessing; - newDtd->hasParamEntityRefs = oldDtd->hasParamEntityRefs; - newDtd->standalone = oldDtd->standalone; - - /* Don't want deep copying for scaffolding */ - newDtd->in_eldecl = oldDtd->in_eldecl; - newDtd->scaffold = oldDtd->scaffold; - newDtd->contentStringLen = oldDtd->contentStringLen; - newDtd->scaffSize = oldDtd->scaffSize; - newDtd->scaffLevel = oldDtd->scaffLevel; - newDtd->scaffIndex = oldDtd->scaffIndex; - - return 1; -} /* End dtdCopy */ - -static int -copyEntityTable(HASH_TABLE *newTable, - STRING_POOL *newPool, - const HASH_TABLE *oldTable) -{ - HASH_TABLE_ITER iter; - const XML_Char *cachedOldBase = NULL; - const XML_Char *cachedNewBase = NULL; - - hashTableIterInit(&iter, oldTable); - - for (;;) { - ENTITY *newE; - const XML_Char *name; - const ENTITY *oldE = (ENTITY *)hashTableIterNext(&iter); - if (!oldE) - break; - name = poolCopyString(newPool, oldE->name); - if (!name) - return 0; - newE = (ENTITY *)lookup(newTable, name, sizeof(ENTITY)); - if (!newE) - return 0; - if (oldE->systemId) { - const XML_Char *tem = poolCopyString(newPool, oldE->systemId); - if (!tem) - return 0; - newE->systemId = tem; - if (oldE->base) { - if (oldE->base == cachedOldBase) - newE->base = cachedNewBase; - else { - cachedOldBase = oldE->base; - tem = poolCopyString(newPool, cachedOldBase); - if (!tem) - return 0; - cachedNewBase = newE->base = tem; - } - } - if (oldE->publicId) { - tem = poolCopyString(newPool, oldE->publicId); - if (!tem) - return 0; - newE->publicId = tem; - } - } - else { - const XML_Char *tem = poolCopyStringN(newPool, oldE->textPtr, - oldE->textLen); - if (!tem) - return 0; - newE->textPtr = tem; - newE->textLen = oldE->textLen; - } - if (oldE->notation) { - const XML_Char *tem = poolCopyString(newPool, oldE->notation); - if (!tem) - return 0; - newE->notation = tem; - } - newE->is_param = oldE->is_param; - newE->is_internal = oldE->is_internal; - } - return 1; -} - -#define INIT_POWER 6 - -static XML_Bool FASTCALL -keyeq(KEY s1, KEY s2) -{ - for (; *s1 == *s2; s1++, s2++) - if (*s1 == 0) - return XML_TRUE; - return XML_FALSE; -} - -static unsigned long FASTCALL -hash(KEY s) -{ - unsigned long h = 0; - while (*s) - h = CHAR_HASH(h, *s++); - return h; -} - -static NAMED * -lookup(HASH_TABLE *table, KEY name, size_t createSize) -{ - size_t i; - if (table->size == 0) { - size_t tsize; - if (!createSize) - return NULL; - table->power = INIT_POWER; - /* table->size is a power of 2 */ - table->size = (size_t)1 << INIT_POWER; - tsize = table->size * sizeof(NAMED *); - table->v = (NAMED **)table->mem->malloc_fcn(tsize); - if (!table->v) { - table->size = 0; - return NULL; - } - memset(table->v, 0, tsize); - i = hash(name) & ((unsigned long)table->size - 1); - } - else { - unsigned long h = hash(name); - unsigned long mask = (unsigned long)table->size - 1; - unsigned char step = 0; - i = h & mask; - while (table->v[i]) { - if (keyeq(name, table->v[i]->name)) - return table->v[i]; - if (!step) - step = PROBE_STEP(h, mask, table->power); - i < step ? (i += table->size - step) : (i -= step); - } - if (!createSize) - return NULL; - - /* check for overflow (table is half full) */ - if (table->used >> (table->power - 1)) { - unsigned char newPower = table->power + 1; - size_t newSize = (size_t)1 << newPower; - unsigned long newMask = (unsigned long)newSize - 1; - size_t tsize = newSize * sizeof(NAMED *); - NAMED **newV = (NAMED **)table->mem->malloc_fcn(tsize); - if (!newV) - return NULL; - memset(newV, 0, tsize); - for (i = 0; i < table->size; i++) - if (table->v[i]) { - unsigned long newHash = hash(table->v[i]->name); - size_t j = newHash & newMask; - step = 0; - while (newV[j]) { - if (!step) - step = PROBE_STEP(newHash, newMask, newPower); - j < step ? (j += newSize - step) : (j -= step); - } - newV[j] = table->v[i]; - } - table->mem->free_fcn(table->v); - table->v = newV; - table->power = newPower; - table->size = newSize; - i = h & newMask; - step = 0; - while (table->v[i]) { - if (!step) - step = PROBE_STEP(h, newMask, newPower); - i < step ? (i += newSize - step) : (i -= step); - } - } - } - table->v[i] = (NAMED *)table->mem->malloc_fcn(createSize); - if (!table->v[i]) - return NULL; - memset(table->v[i], 0, createSize); - table->v[i]->name = name; - (table->used)++; - return table->v[i]; -} - -static void FASTCALL -hashTableClear(HASH_TABLE *table) -{ - size_t i; - for (i = 0; i < table->size; i++) { - table->mem->free_fcn(table->v[i]); - table->v[i] = NULL; - } - table->used = 0; -} - -static void FASTCALL -hashTableDestroy(HASH_TABLE *table) -{ - size_t i; - for (i = 0; i < table->size; i++) - table->mem->free_fcn(table->v[i]); - table->mem->free_fcn(table->v); -} - -static void FASTCALL -hashTableInit(HASH_TABLE *p, const XML_Memory_Handling_Suite *ms) -{ - p->power = 0; - p->size = 0; - p->used = 0; - p->v = NULL; - p->mem = ms; -} - -static void FASTCALL -hashTableIterInit(HASH_TABLE_ITER *iter, const HASH_TABLE *table) -{ - iter->p = table->v; - iter->end = iter->p + table->size; -} - -static NAMED * FASTCALL -hashTableIterNext(HASH_TABLE_ITER *iter) -{ - while (iter->p != iter->end) { - NAMED *tem = *(iter->p)++; - if (tem) - return tem; - } - return NULL; -} - -static void FASTCALL -poolInit(STRING_POOL *pool, const XML_Memory_Handling_Suite *ms) -{ - pool->blocks = NULL; - pool->freeBlocks = NULL; - pool->start = NULL; - pool->ptr = NULL; - pool->end = NULL; - pool->mem = ms; -} - -static void FASTCALL -poolClear(STRING_POOL *pool) -{ - if (!pool->freeBlocks) - pool->freeBlocks = pool->blocks; - else { - BLOCK *p = pool->blocks; - while (p) { - BLOCK *tem = p->next; - p->next = pool->freeBlocks; - pool->freeBlocks = p; - p = tem; - } - } - pool->blocks = NULL; - pool->start = NULL; - pool->ptr = NULL; - pool->end = NULL; -} - -static void FASTCALL -poolDestroy(STRING_POOL *pool) -{ - BLOCK *p = pool->blocks; - while (p) { - BLOCK *tem = p->next; - pool->mem->free_fcn(p); - p = tem; - } - p = pool->freeBlocks; - while (p) { - BLOCK *tem = p->next; - pool->mem->free_fcn(p); - p = tem; - } -} - -static XML_Char * -poolAppend(STRING_POOL *pool, const ENCODING *enc, - const char *ptr, const char *end) -{ - if (!pool->ptr && !poolGrow(pool)) - return NULL; - for (;;) { - XmlConvert(enc, &ptr, end, (ICHAR **)&(pool->ptr), (ICHAR *)pool->end); - if (ptr == end) - break; - if (!poolGrow(pool)) - return NULL; - } - return pool->start; -} - -static const XML_Char * FASTCALL -poolCopyString(STRING_POOL *pool, const XML_Char *s) -{ - do { - if (!poolAppendChar(pool, *s)) - return NULL; - } while (*s++); - s = pool->start; - poolFinish(pool); - return s; -} - -static const XML_Char * -poolCopyStringN(STRING_POOL *pool, const XML_Char *s, int n) -{ - if (!pool->ptr && !poolGrow(pool)) - return NULL; - for (; n > 0; --n, s++) { - if (!poolAppendChar(pool, *s)) - return NULL; - } - s = pool->start; - poolFinish(pool); - return s; -} - -static const XML_Char * FASTCALL -poolAppendString(STRING_POOL *pool, const XML_Char *s) -{ - while (*s) { - if (!poolAppendChar(pool, *s)) - return NULL; - s++; - } - return pool->start; -} - -static XML_Char * -poolStoreString(STRING_POOL *pool, const ENCODING *enc, - const char *ptr, const char *end) -{ - if (!poolAppend(pool, enc, ptr, end)) - return NULL; - if (pool->ptr == pool->end && !poolGrow(pool)) - return NULL; - *(pool->ptr)++ = 0; - return pool->start; -} - -static XML_Bool FASTCALL -poolGrow(STRING_POOL *pool) -{ - if (pool->freeBlocks) { - if (pool->start == 0) { - pool->blocks = pool->freeBlocks; - pool->freeBlocks = pool->freeBlocks->next; - pool->blocks->next = NULL; - pool->start = pool->blocks->s; - pool->end = pool->start + pool->blocks->size; - pool->ptr = pool->start; - return XML_TRUE; - } - if (pool->end - pool->start < pool->freeBlocks->size) { - BLOCK *tem = pool->freeBlocks->next; - pool->freeBlocks->next = pool->blocks; - pool->blocks = pool->freeBlocks; - pool->freeBlocks = tem; - memcpy(pool->blocks->s, pool->start, - (pool->end - pool->start) * sizeof(XML_Char)); - pool->ptr = pool->blocks->s + (pool->ptr - pool->start); - pool->start = pool->blocks->s; - pool->end = pool->start + pool->blocks->size; - return XML_TRUE; - } - } - if (pool->blocks && pool->start == pool->blocks->s) { - int blockSize = (int)(pool->end - pool->start)*2; - pool->blocks = (BLOCK *) - pool->mem->realloc_fcn(pool->blocks, - (offsetof(BLOCK, s) - + blockSize * sizeof(XML_Char))); - if (pool->blocks == NULL) - return XML_FALSE; - pool->blocks->size = blockSize; - pool->ptr = pool->blocks->s + (pool->ptr - pool->start); - pool->start = pool->blocks->s; - pool->end = pool->start + blockSize; - } - else { - BLOCK *tem; - int blockSize = (int)(pool->end - pool->start); - if (blockSize < INIT_BLOCK_SIZE) - blockSize = INIT_BLOCK_SIZE; - else - blockSize *= 2; - tem = (BLOCK *)pool->mem->malloc_fcn(offsetof(BLOCK, s) - + blockSize * sizeof(XML_Char)); - if (!tem) - return XML_FALSE; - tem->size = blockSize; - tem->next = pool->blocks; - pool->blocks = tem; - if (pool->ptr != pool->start) - memcpy(tem->s, pool->start, - (pool->ptr - pool->start) * sizeof(XML_Char)); - pool->ptr = tem->s + (pool->ptr - pool->start); - pool->start = tem->s; - pool->end = tem->s + blockSize; - } - return XML_TRUE; -} - -static int FASTCALL -nextScaffoldPart(XML_Parser parser) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - CONTENT_SCAFFOLD * me; - int next; - - if (!dtd->scaffIndex) { - dtd->scaffIndex = (int *)MALLOC(groupSize * sizeof(int)); - if (!dtd->scaffIndex) - return -1; - dtd->scaffIndex[0] = 0; - } - - if (dtd->scaffCount >= dtd->scaffSize) { - CONTENT_SCAFFOLD *temp; - if (dtd->scaffold) { - temp = (CONTENT_SCAFFOLD *) - REALLOC(dtd->scaffold, dtd->scaffSize * 2 * sizeof(CONTENT_SCAFFOLD)); - if (temp == NULL) - return -1; - dtd->scaffSize *= 2; - } - else { - temp = (CONTENT_SCAFFOLD *)MALLOC(INIT_SCAFFOLD_ELEMENTS - * sizeof(CONTENT_SCAFFOLD)); - if (temp == NULL) - return -1; - dtd->scaffSize = INIT_SCAFFOLD_ELEMENTS; - } - dtd->scaffold = temp; - } - next = dtd->scaffCount++; - me = &dtd->scaffold[next]; - if (dtd->scaffLevel) { - CONTENT_SCAFFOLD *parent = &dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel-1]]; - if (parent->lastchild) { - dtd->scaffold[parent->lastchild].nextsib = next; - } - if (!parent->childcnt) - parent->firstchild = next; - parent->lastchild = next; - parent->childcnt++; - } - me->firstchild = me->lastchild = me->childcnt = me->nextsib = 0; - return next; -} - -static void -build_node(XML_Parser parser, - int src_node, - XML_Content *dest, - XML_Content **contpos, - XML_Char **strpos) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - dest->type = dtd->scaffold[src_node].type; - dest->quant = dtd->scaffold[src_node].quant; - if (dest->type == XML_CTYPE_NAME) { - const XML_Char *src; - dest->name = *strpos; - src = dtd->scaffold[src_node].name; - for (;;) { - *(*strpos)++ = *src; - if (!*src) - break; - src++; - } - dest->numchildren = 0; - dest->children = NULL; - } - else { - unsigned int i; - int cn; - dest->numchildren = dtd->scaffold[src_node].childcnt; - dest->children = *contpos; - *contpos += dest->numchildren; - for (i = 0, cn = dtd->scaffold[src_node].firstchild; - i < dest->numchildren; - i++, cn = dtd->scaffold[cn].nextsib) { - build_node(parser, cn, &(dest->children[i]), contpos, strpos); - } - dest->name = NULL; - } -} - -static XML_Content * -build_model (XML_Parser parser) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - XML_Content *ret; - XML_Content *cpos; - XML_Char * str; - int allocsize = (dtd->scaffCount * sizeof(XML_Content) - + (dtd->contentStringLen * sizeof(XML_Char))); - - ret = (XML_Content *)MALLOC(allocsize); - if (!ret) - return NULL; - - str = (XML_Char *) (&ret[dtd->scaffCount]); - cpos = &ret[1]; - - build_node(parser, 0, ret, &cpos, &str); - return ret; -} - -static ELEMENT_TYPE * -getElementType(XML_Parser parser, - const ENCODING *enc, - const char *ptr, - const char *end) -{ - DTD * const dtd = _dtd; /* save one level of indirection */ - const XML_Char *name = poolStoreString(&dtd->pool, enc, ptr, end); - ELEMENT_TYPE *ret; - - if (!name) - return NULL; - ret = (ELEMENT_TYPE *) lookup(&dtd->elementTypes, name, sizeof(ELEMENT_TYPE)); + + 77fea421d361dca90041d0040ecf1dca651167fadf2af79e990e35168d70d933 (2.2.1+) +*/ + +#define _GNU_SOURCE /* syscall prototype */ + +#include +#include /* memset(), memcpy() */ +#include +#include /* UINT_MAX */ +#include /* fprintf */ +#include /* getenv */ + +#ifdef _WIN32 +#define getpid GetCurrentProcessId +#else +#include /* gettimeofday() */ +#include /* getpid() */ +#include /* getpid() */ +#endif + +#define XML_BUILDING_EXPAT 1 + +#ifdef _WIN32 +#include "winconfig.h" +#elif defined(HAVE_EXPAT_CONFIG_H) +#include +#endif /* ndef _WIN32 */ + +#include "ascii.h" +#include "expat.h" +#include "siphash.h" + +#ifdef XML_UNICODE +#define XML_ENCODE_MAX XML_UTF16_ENCODE_MAX +#define XmlConvert XmlUtf16Convert +#define XmlGetInternalEncoding XmlGetUtf16InternalEncoding +#define XmlGetInternalEncodingNS XmlGetUtf16InternalEncodingNS +#define XmlEncode XmlUtf16Encode +/* Using pointer subtraction to convert to integer type. */ +#define MUST_CONVERT(enc, s) (!(enc)->isUtf16 || (((char *)(s) - (char *)NULL) & 1)) +typedef unsigned short ICHAR; +#else +#define XML_ENCODE_MAX XML_UTF8_ENCODE_MAX +#define XmlConvert XmlUtf8Convert +#define XmlGetInternalEncoding XmlGetUtf8InternalEncoding +#define XmlGetInternalEncodingNS XmlGetUtf8InternalEncodingNS +#define XmlEncode XmlUtf8Encode +#define MUST_CONVERT(enc, s) (!(enc)->isUtf8) +typedef char ICHAR; +#endif + + +#ifndef XML_NS + +#define XmlInitEncodingNS XmlInitEncoding +#define XmlInitUnknownEncodingNS XmlInitUnknownEncoding +#undef XmlGetInternalEncodingNS +#define XmlGetInternalEncodingNS XmlGetInternalEncoding +#define XmlParseXmlDeclNS XmlParseXmlDecl + +#endif + +#ifdef XML_UNICODE + +#ifdef XML_UNICODE_WCHAR_T +#define XML_T(x) (const wchar_t)x +#define XML_L(x) L ## x +#else +#define XML_T(x) (const unsigned short)x +#define XML_L(x) x +#endif + +#else + +#define XML_T(x) x +#define XML_L(x) x + +#endif + +/* Round up n to be a multiple of sz, where sz is a power of 2. */ +#define ROUND_UP(n, sz) (((n) + ((sz) - 1)) & ~((sz) - 1)) + +/* Handle the case where memmove() doesn't exist. */ +#ifndef HAVE_MEMMOVE +#ifdef HAVE_BCOPY +#define memmove(d,s,l) bcopy((s),(d),(l)) +#else +#error memmove does not exist on this platform, nor is a substitute available +#endif /* HAVE_BCOPY */ +#endif /* HAVE_MEMMOVE */ + +#include "internal.h" +#include "xmltok.h" +#include "xmlrole.h" + +typedef const XML_Char *KEY; + +typedef struct { + KEY name; +} NAMED; + +typedef struct { + NAMED **v; + unsigned char power; + size_t size; + size_t used; + const XML_Memory_Handling_Suite *mem; +} HASH_TABLE; + +static size_t +keylen(KEY s); + +static void +copy_salt_to_sipkey(XML_Parser parser, struct sipkey * key); + +/* For probing (after a collision) we need a step size relative prime + to the hash table size, which is a power of 2. We use double-hashing, + since we can calculate a second hash value cheaply by taking those bits + of the first hash value that were discarded (masked out) when the table + index was calculated: index = hash & mask, where mask = table->size - 1. + We limit the maximum step size to table->size / 4 (mask >> 2) and make + it odd, since odd numbers are always relative prime to a power of 2. +*/ +#define SECOND_HASH(hash, mask, power) \ + ((((hash) & ~(mask)) >> ((power) - 1)) & ((mask) >> 2)) +#define PROBE_STEP(hash, mask, power) \ + ((unsigned char)((SECOND_HASH(hash, mask, power)) | 1)) + +typedef struct { + NAMED **p; + NAMED **end; +} HASH_TABLE_ITER; + +#define INIT_TAG_BUF_SIZE 32 /* must be a multiple of sizeof(XML_Char) */ +#define INIT_DATA_BUF_SIZE 1024 +#define INIT_ATTS_SIZE 16 +#define INIT_ATTS_VERSION 0xFFFFFFFF +#define INIT_BLOCK_SIZE 1024 +#define INIT_BUFFER_SIZE 1024 + +#define EXPAND_SPARE 24 + +typedef struct binding { + struct prefix *prefix; + struct binding *nextTagBinding; + struct binding *prevPrefixBinding; + const struct attribute_id *attId; + XML_Char *uri; + int uriLen; + int uriAlloc; +} BINDING; + +typedef struct prefix { + const XML_Char *name; + BINDING *binding; +} PREFIX; + +typedef struct { + const XML_Char *str; + const XML_Char *localPart; + const XML_Char *prefix; + int strLen; + int uriLen; + int prefixLen; +} TAG_NAME; + +/* TAG represents an open element. + The name of the element is stored in both the document and API + encodings. The memory buffer 'buf' is a separately-allocated + memory area which stores the name. During the XML_Parse()/ + XMLParseBuffer() when the element is open, the memory for the 'raw' + version of the name (in the document encoding) is shared with the + document buffer. If the element is open across calls to + XML_Parse()/XML_ParseBuffer(), the buffer is re-allocated to + contain the 'raw' name as well. + + A parser re-uses these structures, maintaining a list of allocated + TAG objects in a free list. +*/ +typedef struct tag { + struct tag *parent; /* parent of this element */ + const char *rawName; /* tagName in the original encoding */ + int rawNameLength; + TAG_NAME name; /* tagName in the API encoding */ + char *buf; /* buffer for name components */ + char *bufEnd; /* end of the buffer */ + BINDING *bindings; +} TAG; + +typedef struct { + const XML_Char *name; + const XML_Char *textPtr; + int textLen; /* length in XML_Chars */ + int processed; /* # of processed bytes - when suspended */ + const XML_Char *systemId; + const XML_Char *base; + const XML_Char *publicId; + const XML_Char *notation; + XML_Bool open; + XML_Bool is_param; + XML_Bool is_internal; /* true if declared in internal subset outside PE */ +} ENTITY; + +typedef struct { + enum XML_Content_Type type; + enum XML_Content_Quant quant; + const XML_Char * name; + int firstchild; + int lastchild; + int childcnt; + int nextsib; +} CONTENT_SCAFFOLD; + +#define INIT_SCAFFOLD_ELEMENTS 32 + +typedef struct block { + struct block *next; + int size; + XML_Char s[1]; +} BLOCK; + +typedef struct { + BLOCK *blocks; + BLOCK *freeBlocks; + const XML_Char *end; + XML_Char *ptr; + XML_Char *start; + const XML_Memory_Handling_Suite *mem; +} STRING_POOL; + +/* The XML_Char before the name is used to determine whether + an attribute has been specified. */ +typedef struct attribute_id { + XML_Char *name; + PREFIX *prefix; + XML_Bool maybeTokenized; + XML_Bool xmlns; +} ATTRIBUTE_ID; + +typedef struct { + const ATTRIBUTE_ID *id; + XML_Bool isCdata; + const XML_Char *value; +} DEFAULT_ATTRIBUTE; + +typedef struct { + unsigned long version; + unsigned long hash; + const XML_Char *uriName; +} NS_ATT; + +typedef struct { + const XML_Char *name; + PREFIX *prefix; + const ATTRIBUTE_ID *idAtt; + int nDefaultAtts; + int allocDefaultAtts; + DEFAULT_ATTRIBUTE *defaultAtts; +} ELEMENT_TYPE; + +typedef struct { + HASH_TABLE generalEntities; + HASH_TABLE elementTypes; + HASH_TABLE attributeIds; + HASH_TABLE prefixes; + STRING_POOL pool; + STRING_POOL entityValuePool; + /* false once a parameter entity reference has been skipped */ + XML_Bool keepProcessing; + /* true once an internal or external PE reference has been encountered; + this includes the reference to an external subset */ + XML_Bool hasParamEntityRefs; + XML_Bool standalone; +#ifdef XML_DTD + /* indicates if external PE has been read */ + XML_Bool paramEntityRead; + HASH_TABLE paramEntities; +#endif /* XML_DTD */ + PREFIX defaultPrefix; + /* === scaffolding for building content model === */ + XML_Bool in_eldecl; + CONTENT_SCAFFOLD *scaffold; + unsigned contentStringLen; + unsigned scaffSize; + unsigned scaffCount; + int scaffLevel; + int *scaffIndex; +} DTD; + +typedef struct open_internal_entity { + const char *internalEventPtr; + const char *internalEventEndPtr; + struct open_internal_entity *next; + ENTITY *entity; + int startTagLevel; + XML_Bool betweenDecl; /* WFC: PE Between Declarations */ +} OPEN_INTERNAL_ENTITY; + +typedef enum XML_Error PTRCALL Processor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr); + +static Processor prologProcessor; +static Processor prologInitProcessor; +static Processor contentProcessor; +static Processor cdataSectionProcessor; +#ifdef XML_DTD +static Processor ignoreSectionProcessor; +static Processor externalParEntProcessor; +static Processor externalParEntInitProcessor; +static Processor entityValueProcessor; +static Processor entityValueInitProcessor; +#endif /* XML_DTD */ +static Processor epilogProcessor; +static Processor errorProcessor; +static Processor externalEntityInitProcessor; +static Processor externalEntityInitProcessor2; +static Processor externalEntityInitProcessor3; +static Processor externalEntityContentProcessor; +static Processor internalEntityProcessor; + +static enum XML_Error +handleUnknownEncoding(XML_Parser parser, const XML_Char *encodingName); +static enum XML_Error +processXmlDecl(XML_Parser parser, int isGeneralTextEntity, + const char *s, const char *next); +static enum XML_Error +initializeEncoding(XML_Parser parser); +static enum XML_Error +doProlog(XML_Parser parser, const ENCODING *enc, const char *s, + const char *end, int tok, const char *next, const char **nextPtr, + XML_Bool haveMore); +static enum XML_Error +processInternalEntity(XML_Parser parser, ENTITY *entity, + XML_Bool betweenDecl); +static enum XML_Error +doContent(XML_Parser parser, int startTagLevel, const ENCODING *enc, + const char *start, const char *end, const char **endPtr, + XML_Bool haveMore); +static enum XML_Error +doCdataSection(XML_Parser parser, const ENCODING *, const char **startPtr, + const char *end, const char **nextPtr, XML_Bool haveMore); +#ifdef XML_DTD +static enum XML_Error +doIgnoreSection(XML_Parser parser, const ENCODING *, const char **startPtr, + const char *end, const char **nextPtr, XML_Bool haveMore); +#endif /* XML_DTD */ + +static void +freeBindings(XML_Parser parser, BINDING *bindings); +static enum XML_Error +storeAtts(XML_Parser parser, const ENCODING *, const char *s, + TAG_NAME *tagNamePtr, BINDING **bindingsPtr); +static enum XML_Error +addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId, + const XML_Char *uri, BINDING **bindingsPtr); +static int +defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *, XML_Bool isCdata, + XML_Bool isId, const XML_Char *dfltValue, XML_Parser parser); +static enum XML_Error +storeAttributeValue(XML_Parser parser, const ENCODING *, XML_Bool isCdata, + const char *, const char *, STRING_POOL *); +static enum XML_Error +appendAttributeValue(XML_Parser parser, const ENCODING *, XML_Bool isCdata, + const char *, const char *, STRING_POOL *); +static ATTRIBUTE_ID * +getAttributeId(XML_Parser parser, const ENCODING *enc, const char *start, + const char *end); +static int +setElementTypePrefix(XML_Parser parser, ELEMENT_TYPE *); +static enum XML_Error +storeEntityValue(XML_Parser parser, const ENCODING *enc, const char *start, + const char *end); +static int +reportProcessingInstruction(XML_Parser parser, const ENCODING *enc, + const char *start, const char *end); +static int +reportComment(XML_Parser parser, const ENCODING *enc, const char *start, + const char *end); +static void +reportDefault(XML_Parser parser, const ENCODING *enc, const char *start, + const char *end); + +static const XML_Char * getContext(XML_Parser parser); +static XML_Bool +setContext(XML_Parser parser, const XML_Char *context); + +static void FASTCALL normalizePublicId(XML_Char *s); + +static DTD * dtdCreate(const XML_Memory_Handling_Suite *ms); +/* do not call if parentParser != NULL */ +static void dtdReset(DTD *p, const XML_Memory_Handling_Suite *ms); +static void +dtdDestroy(DTD *p, XML_Bool isDocEntity, const XML_Memory_Handling_Suite *ms); +static int +dtdCopy(XML_Parser oldParser, + DTD *newDtd, const DTD *oldDtd, const XML_Memory_Handling_Suite *ms); +static int +copyEntityTable(XML_Parser oldParser, + HASH_TABLE *, STRING_POOL *, const HASH_TABLE *); +static NAMED * +lookup(XML_Parser parser, HASH_TABLE *table, KEY name, size_t createSize); +static void FASTCALL +hashTableInit(HASH_TABLE *, const XML_Memory_Handling_Suite *ms); +static void FASTCALL hashTableClear(HASH_TABLE *); +static void FASTCALL hashTableDestroy(HASH_TABLE *); +static void FASTCALL +hashTableIterInit(HASH_TABLE_ITER *, const HASH_TABLE *); +static NAMED * FASTCALL hashTableIterNext(HASH_TABLE_ITER *); + +static void FASTCALL +poolInit(STRING_POOL *, const XML_Memory_Handling_Suite *ms); +static void FASTCALL poolClear(STRING_POOL *); +static void FASTCALL poolDestroy(STRING_POOL *); +static XML_Char * +poolAppend(STRING_POOL *pool, const ENCODING *enc, + const char *ptr, const char *end); +static XML_Char * +poolStoreString(STRING_POOL *pool, const ENCODING *enc, + const char *ptr, const char *end); +static XML_Bool FASTCALL poolGrow(STRING_POOL *pool); +static const XML_Char * FASTCALL +poolCopyString(STRING_POOL *pool, const XML_Char *s); +static const XML_Char * +poolCopyStringN(STRING_POOL *pool, const XML_Char *s, int n); +static const XML_Char * FASTCALL +poolAppendString(STRING_POOL *pool, const XML_Char *s); + +static int FASTCALL nextScaffoldPart(XML_Parser parser); +static XML_Content * build_model(XML_Parser parser); +static ELEMENT_TYPE * +getElementType(XML_Parser parser, const ENCODING *enc, + const char *ptr, const char *end); + +static unsigned long generate_hash_secret_salt(XML_Parser parser); +static XML_Bool startParsing(XML_Parser parser); + +static XML_Parser +parserCreate(const XML_Char *encodingName, + const XML_Memory_Handling_Suite *memsuite, + const XML_Char *nameSep, + DTD *dtd); + +static void +parserInit(XML_Parser parser, const XML_Char *encodingName); + +#define poolStart(pool) ((pool)->start) +#define poolEnd(pool) ((pool)->ptr) +#define poolLength(pool) ((pool)->ptr - (pool)->start) +#define poolChop(pool) ((void)--(pool->ptr)) +#define poolLastChar(pool) (((pool)->ptr)[-1]) +#define poolDiscard(pool) ((pool)->ptr = (pool)->start) +#define poolFinish(pool) ((pool)->start = (pool)->ptr) +#define poolAppendChar(pool, c) \ + (((pool)->ptr == (pool)->end && !poolGrow(pool)) \ + ? 0 \ + : ((*((pool)->ptr)++ = c), 1)) + +struct XML_ParserStruct { + /* The first member must be userData so that the XML_GetUserData + macro works. */ + void *m_userData; + void *m_handlerArg; + char *m_buffer; + const XML_Memory_Handling_Suite m_mem; + /* first character to be parsed */ + const char *m_bufferPtr; + /* past last character to be parsed */ + char *m_bufferEnd; + /* allocated end of buffer */ + const char *m_bufferLim; + XML_Index m_parseEndByteIndex; + const char *m_parseEndPtr; + XML_Char *m_dataBuf; + XML_Char *m_dataBufEnd; + XML_StartElementHandler m_startElementHandler; + XML_EndElementHandler m_endElementHandler; + XML_CharacterDataHandler m_characterDataHandler; + XML_ProcessingInstructionHandler m_processingInstructionHandler; + XML_CommentHandler m_commentHandler; + XML_StartCdataSectionHandler m_startCdataSectionHandler; + XML_EndCdataSectionHandler m_endCdataSectionHandler; + XML_DefaultHandler m_defaultHandler; + XML_StartDoctypeDeclHandler m_startDoctypeDeclHandler; + XML_EndDoctypeDeclHandler m_endDoctypeDeclHandler; + XML_UnparsedEntityDeclHandler m_unparsedEntityDeclHandler; + XML_NotationDeclHandler m_notationDeclHandler; + XML_StartNamespaceDeclHandler m_startNamespaceDeclHandler; + XML_EndNamespaceDeclHandler m_endNamespaceDeclHandler; + XML_NotStandaloneHandler m_notStandaloneHandler; + XML_ExternalEntityRefHandler m_externalEntityRefHandler; + XML_Parser m_externalEntityRefHandlerArg; + XML_SkippedEntityHandler m_skippedEntityHandler; + XML_UnknownEncodingHandler m_unknownEncodingHandler; + XML_ElementDeclHandler m_elementDeclHandler; + XML_AttlistDeclHandler m_attlistDeclHandler; + XML_EntityDeclHandler m_entityDeclHandler; + XML_XmlDeclHandler m_xmlDeclHandler; + const ENCODING *m_encoding; + INIT_ENCODING m_initEncoding; + const ENCODING *m_internalEncoding; + const XML_Char *m_protocolEncodingName; + XML_Bool m_ns; + XML_Bool m_ns_triplets; + void *m_unknownEncodingMem; + void *m_unknownEncodingData; + void *m_unknownEncodingHandlerData; + void (XMLCALL *m_unknownEncodingRelease)(void *); + PROLOG_STATE m_prologState; + Processor *m_processor; + enum XML_Error m_errorCode; + const char *m_eventPtr; + const char *m_eventEndPtr; + const char *m_positionPtr; + OPEN_INTERNAL_ENTITY *m_openInternalEntities; + OPEN_INTERNAL_ENTITY *m_freeInternalEntities; + XML_Bool m_defaultExpandInternalEntities; + int m_tagLevel; + ENTITY *m_declEntity; + const XML_Char *m_doctypeName; + const XML_Char *m_doctypeSysid; + const XML_Char *m_doctypePubid; + const XML_Char *m_declAttributeType; + const XML_Char *m_declNotationName; + const XML_Char *m_declNotationPublicId; + ELEMENT_TYPE *m_declElementType; + ATTRIBUTE_ID *m_declAttributeId; + XML_Bool m_declAttributeIsCdata; + XML_Bool m_declAttributeIsId; + DTD *m_dtd; + const XML_Char *m_curBase; + TAG *m_tagStack; + TAG *m_freeTagList; + BINDING *m_inheritedBindings; + BINDING *m_freeBindingList; + int m_attsSize; + int m_nSpecifiedAtts; + int m_idAttIndex; + ATTRIBUTE *m_atts; + NS_ATT *m_nsAtts; + unsigned long m_nsAttsVersion; + unsigned char m_nsAttsPower; +#ifdef XML_ATTR_INFO + XML_AttrInfo *m_attInfo; +#endif + POSITION m_position; + STRING_POOL m_tempPool; + STRING_POOL m_temp2Pool; + char *m_groupConnector; + unsigned int m_groupSize; + XML_Char m_namespaceSeparator; + XML_Parser m_parentParser; + XML_ParsingStatus m_parsingStatus; +#ifdef XML_DTD + XML_Bool m_isParamEntity; + XML_Bool m_useForeignDTD; + enum XML_ParamEntityParsing m_paramEntityParsing; +#endif + unsigned long m_hash_secret_salt; +}; + +#define MALLOC(s) (parser->m_mem.malloc_fcn((s))) +#define REALLOC(p,s) (parser->m_mem.realloc_fcn((p),(s))) +#define FREE(p) (parser->m_mem.free_fcn((p))) + +#define userData (parser->m_userData) +#define handlerArg (parser->m_handlerArg) +#define startElementHandler (parser->m_startElementHandler) +#define endElementHandler (parser->m_endElementHandler) +#define characterDataHandler (parser->m_characterDataHandler) +#define processingInstructionHandler \ + (parser->m_processingInstructionHandler) +#define commentHandler (parser->m_commentHandler) +#define startCdataSectionHandler \ + (parser->m_startCdataSectionHandler) +#define endCdataSectionHandler (parser->m_endCdataSectionHandler) +#define defaultHandler (parser->m_defaultHandler) +#define startDoctypeDeclHandler (parser->m_startDoctypeDeclHandler) +#define endDoctypeDeclHandler (parser->m_endDoctypeDeclHandler) +#define unparsedEntityDeclHandler \ + (parser->m_unparsedEntityDeclHandler) +#define notationDeclHandler (parser->m_notationDeclHandler) +#define startNamespaceDeclHandler \ + (parser->m_startNamespaceDeclHandler) +#define endNamespaceDeclHandler (parser->m_endNamespaceDeclHandler) +#define notStandaloneHandler (parser->m_notStandaloneHandler) +#define externalEntityRefHandler \ + (parser->m_externalEntityRefHandler) +#define externalEntityRefHandlerArg \ + (parser->m_externalEntityRefHandlerArg) +#define internalEntityRefHandler \ + (parser->m_internalEntityRefHandler) +#define skippedEntityHandler (parser->m_skippedEntityHandler) +#define unknownEncodingHandler (parser->m_unknownEncodingHandler) +#define elementDeclHandler (parser->m_elementDeclHandler) +#define attlistDeclHandler (parser->m_attlistDeclHandler) +#define entityDeclHandler (parser->m_entityDeclHandler) +#define xmlDeclHandler (parser->m_xmlDeclHandler) +#define encoding (parser->m_encoding) +#define initEncoding (parser->m_initEncoding) +#define internalEncoding (parser->m_internalEncoding) +#define unknownEncodingMem (parser->m_unknownEncodingMem) +#define unknownEncodingData (parser->m_unknownEncodingData) +#define unknownEncodingHandlerData \ + (parser->m_unknownEncodingHandlerData) +#define unknownEncodingRelease (parser->m_unknownEncodingRelease) +#define protocolEncodingName (parser->m_protocolEncodingName) +#define ns (parser->m_ns) +#define ns_triplets (parser->m_ns_triplets) +#define prologState (parser->m_prologState) +#define processor (parser->m_processor) +#define errorCode (parser->m_errorCode) +#define eventPtr (parser->m_eventPtr) +#define eventEndPtr (parser->m_eventEndPtr) +#define positionPtr (parser->m_positionPtr) +#define position (parser->m_position) +#define openInternalEntities (parser->m_openInternalEntities) +#define freeInternalEntities (parser->m_freeInternalEntities) +#define defaultExpandInternalEntities \ + (parser->m_defaultExpandInternalEntities) +#define tagLevel (parser->m_tagLevel) +#define buffer (parser->m_buffer) +#define bufferPtr (parser->m_bufferPtr) +#define bufferEnd (parser->m_bufferEnd) +#define parseEndByteIndex (parser->m_parseEndByteIndex) +#define parseEndPtr (parser->m_parseEndPtr) +#define bufferLim (parser->m_bufferLim) +#define dataBuf (parser->m_dataBuf) +#define dataBufEnd (parser->m_dataBufEnd) +#define _dtd (parser->m_dtd) +#define curBase (parser->m_curBase) +#define declEntity (parser->m_declEntity) +#define doctypeName (parser->m_doctypeName) +#define doctypeSysid (parser->m_doctypeSysid) +#define doctypePubid (parser->m_doctypePubid) +#define declAttributeType (parser->m_declAttributeType) +#define declNotationName (parser->m_declNotationName) +#define declNotationPublicId (parser->m_declNotationPublicId) +#define declElementType (parser->m_declElementType) +#define declAttributeId (parser->m_declAttributeId) +#define declAttributeIsCdata (parser->m_declAttributeIsCdata) +#define declAttributeIsId (parser->m_declAttributeIsId) +#define freeTagList (parser->m_freeTagList) +#define freeBindingList (parser->m_freeBindingList) +#define inheritedBindings (parser->m_inheritedBindings) +#define tagStack (parser->m_tagStack) +#define atts (parser->m_atts) +#define attsSize (parser->m_attsSize) +#define nSpecifiedAtts (parser->m_nSpecifiedAtts) +#define idAttIndex (parser->m_idAttIndex) +#define nsAtts (parser->m_nsAtts) +#define nsAttsVersion (parser->m_nsAttsVersion) +#define nsAttsPower (parser->m_nsAttsPower) +#define attInfo (parser->m_attInfo) +#define tempPool (parser->m_tempPool) +#define temp2Pool (parser->m_temp2Pool) +#define groupConnector (parser->m_groupConnector) +#define groupSize (parser->m_groupSize) +#define namespaceSeparator (parser->m_namespaceSeparator) +#define parentParser (parser->m_parentParser) +#define ps_parsing (parser->m_parsingStatus.parsing) +#define ps_finalBuffer (parser->m_parsingStatus.finalBuffer) +#ifdef XML_DTD +#define isParamEntity (parser->m_isParamEntity) +#define useForeignDTD (parser->m_useForeignDTD) +#define paramEntityParsing (parser->m_paramEntityParsing) +#endif /* XML_DTD */ +#define hash_secret_salt (parser->m_hash_secret_salt) + +XML_Parser XMLCALL +XML_ParserCreate(const XML_Char *encodingName) +{ + return XML_ParserCreate_MM(encodingName, NULL, NULL); +} + +XML_Parser XMLCALL +XML_ParserCreateNS(const XML_Char *encodingName, XML_Char nsSep) +{ + XML_Char tmp[2]; + *tmp = nsSep; + return XML_ParserCreate_MM(encodingName, NULL, tmp); +} + +static const XML_Char implicitContext[] = { + ASCII_x, ASCII_m, ASCII_l, ASCII_EQUALS, ASCII_h, ASCII_t, ASCII_t, ASCII_p, + ASCII_COLON, ASCII_SLASH, ASCII_SLASH, ASCII_w, ASCII_w, ASCII_w, + ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, ASCII_o, ASCII_r, ASCII_g, + ASCII_SLASH, ASCII_X, ASCII_M, ASCII_L, ASCII_SLASH, ASCII_1, ASCII_9, + ASCII_9, ASCII_8, ASCII_SLASH, ASCII_n, ASCII_a, ASCII_m, ASCII_e, + ASCII_s, ASCII_p, ASCII_a, ASCII_c, ASCII_e, '\0' +}; + + +#if defined(HAVE_GETRANDOM) || defined(HAVE_SYSCALL_GETRANDOM) +# include + +# if defined(HAVE_GETRANDOM) +# include /* getrandom */ +# else +# include /* syscall */ +# include /* SYS_getrandom */ +# endif + +/* Obtain entropy on Linux 3.17+ */ +static int +writeRandomBytes_getrandom(void * target, size_t count) { + int success = 0; /* full count bytes written? */ + size_t bytesWrittenTotal = 0; + const unsigned int getrandomFlags = 0; + + do { + void * const currentTarget = (void*)((char*)target + bytesWrittenTotal); + const size_t bytesToWrite = count - bytesWrittenTotal; + + const int bytesWrittenMore = +#if defined(HAVE_GETRANDOM) + getrandom(currentTarget, bytesToWrite, getrandomFlags); +#else + syscall(SYS_getrandom, currentTarget, bytesToWrite, getrandomFlags); +#endif + + if (bytesWrittenMore > 0) { + bytesWrittenTotal += bytesWrittenMore; + if (bytesWrittenTotal >= count) + success = 1; + } + } while (! success && (errno == EINTR || errno == EAGAIN)); + + return success; +} + +#endif /* defined(HAVE_GETRANDOM) || defined(HAVE_SYSCALL_GETRANDOM) */ + + +#ifdef _WIN32 + +typedef BOOLEAN (APIENTRY *RTLGENRANDOM_FUNC)(PVOID, ULONG); + +/* Obtain entropy on Windows XP / Windows Server 2003 and later. + * Hint on RtlGenRandom and the following article from libsodioum. + * + * Michael Howard: Cryptographically Secure Random number on Windows without using CryptoAPI + * https://blogs.msdn.microsoft.com/michael_howard/2005/01/14/cryptographically-secure-random-number-on-windows-without-using-cryptoapi/ + */ +static int +writeRandomBytes_RtlGenRandom(void * target, size_t count) { + int success = 0; /* full count bytes written? */ + const HMODULE advapi32 = LoadLibrary("ADVAPI32.DLL"); + + if (advapi32) { + const RTLGENRANDOM_FUNC RtlGenRandom + = (RTLGENRANDOM_FUNC)GetProcAddress(advapi32, "SystemFunction036"); + if (RtlGenRandom) { + if (RtlGenRandom((PVOID)target, (ULONG)count) == TRUE) { + success = 1; + } + } + FreeLibrary(advapi32); + } + + return success; +} + +#endif /* _WIN32 */ + + +static unsigned long +gather_time_entropy(void) +{ +#ifdef _WIN32 + FILETIME ft; + GetSystemTimeAsFileTime(&ft); /* never fails */ + return ft.dwHighDateTime ^ ft.dwLowDateTime; +#else + struct timeval tv; + int gettimeofday_res; + + gettimeofday_res = gettimeofday(&tv, NULL); + assert (gettimeofday_res == 0); + + /* Microseconds time is <20 bits entropy */ + return tv.tv_usec; +#endif +} + +#if defined(HAVE_ARC4RANDOM_BUF) && defined(HAVE_LIBBSD) +# include +#endif + +static unsigned long +ENTROPY_DEBUG(const char * label, unsigned long entropy) { + const char * const EXPAT_ENTROPY_DEBUG = getenv("EXPAT_ENTROPY_DEBUG"); + if (EXPAT_ENTROPY_DEBUG && ! strcmp(EXPAT_ENTROPY_DEBUG, "1")) { + fprintf(stderr, "Entropy: %s --> 0x%0*lx (%lu bytes)\n", + label, + (int)sizeof(entropy) * 2, entropy, + (unsigned long)sizeof(entropy)); + } + return entropy; +} + +static unsigned long +generate_hash_secret_salt(XML_Parser parser) +{ + unsigned long entropy; + (void)parser; +#if defined(HAVE_ARC4RANDOM_BUF) || defined(__CloudABI__) + (void)gather_time_entropy; + arc4random_buf(&entropy, sizeof(entropy)); + return ENTROPY_DEBUG("arc4random_buf", entropy); +#else + /* Try high quality providers first .. */ +#ifdef _WIN32 + if (writeRandomBytes_RtlGenRandom((void *)&entropy, sizeof(entropy))) { + return ENTROPY_DEBUG("RtlGenRandom", entropy); + } +#elif defined(HAVE_GETRANDOM) || defined(HAVE_SYSCALL_GETRANDOM) + if (writeRandomBytes_getrandom((void *)&entropy, sizeof(entropy))) { + return ENTROPY_DEBUG("getrandom", entropy); + } +#endif + /* .. and self-made low quality for backup: */ + + /* Process ID is 0 bits entropy if attacker has local access */ + entropy = gather_time_entropy() ^ getpid(); + + /* Factors are 2^31-1 and 2^61-1 (Mersenne primes M31 and M61) */ + if (sizeof(unsigned long) == 4) { + return ENTROPY_DEBUG("fallback(4)", entropy * 2147483647); + } else { + return ENTROPY_DEBUG("fallback(8)", + entropy * (unsigned long)2305843009213693951); + } +#endif +} + +static unsigned long +get_hash_secret_salt(XML_Parser parser) { + if (parser->m_parentParser != NULL) + return get_hash_secret_salt(parser->m_parentParser); + return parser->m_hash_secret_salt; +} + +static XML_Bool /* only valid for root parser */ +startParsing(XML_Parser parser) +{ + /* hash functions must be initialized before setContext() is called */ + if (hash_secret_salt == 0) + hash_secret_salt = generate_hash_secret_salt(parser); + if (ns) { + /* implicit context only set for root parser, since child + parsers (i.e. external entity parsers) will inherit it + */ + return setContext(parser, implicitContext); + } + return XML_TRUE; +} + +XML_Parser XMLCALL +XML_ParserCreate_MM(const XML_Char *encodingName, + const XML_Memory_Handling_Suite *memsuite, + const XML_Char *nameSep) +{ + return parserCreate(encodingName, memsuite, nameSep, NULL); +} + +static XML_Parser +parserCreate(const XML_Char *encodingName, + const XML_Memory_Handling_Suite *memsuite, + const XML_Char *nameSep, + DTD *dtd) +{ + XML_Parser parser; + + if (memsuite) { + XML_Memory_Handling_Suite *mtemp; + parser = (XML_Parser) + memsuite->malloc_fcn(sizeof(struct XML_ParserStruct)); + if (parser != NULL) { + mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem); + mtemp->malloc_fcn = memsuite->malloc_fcn; + mtemp->realloc_fcn = memsuite->realloc_fcn; + mtemp->free_fcn = memsuite->free_fcn; + } + } + else { + XML_Memory_Handling_Suite *mtemp; + parser = (XML_Parser)malloc(sizeof(struct XML_ParserStruct)); + if (parser != NULL) { + mtemp = (XML_Memory_Handling_Suite *)&(parser->m_mem); + mtemp->malloc_fcn = malloc; + mtemp->realloc_fcn = realloc; + mtemp->free_fcn = free; + } + } + + if (!parser) + return parser; + + buffer = NULL; + bufferLim = NULL; + + attsSize = INIT_ATTS_SIZE; + atts = (ATTRIBUTE *)MALLOC(attsSize * sizeof(ATTRIBUTE)); + if (atts == NULL) { + FREE(parser); + return NULL; + } +#ifdef XML_ATTR_INFO + attInfo = (XML_AttrInfo*)MALLOC(attsSize * sizeof(XML_AttrInfo)); + if (attInfo == NULL) { + FREE(atts); + FREE(parser); + return NULL; + } +#endif + dataBuf = (XML_Char *)MALLOC(INIT_DATA_BUF_SIZE * sizeof(XML_Char)); + if (dataBuf == NULL) { + FREE(atts); +#ifdef XML_ATTR_INFO + FREE(attInfo); +#endif + FREE(parser); + return NULL; + } + dataBufEnd = dataBuf + INIT_DATA_BUF_SIZE; + + if (dtd) + _dtd = dtd; + else { + _dtd = dtdCreate(&parser->m_mem); + if (_dtd == NULL) { + FREE(dataBuf); + FREE(atts); +#ifdef XML_ATTR_INFO + FREE(attInfo); +#endif + FREE(parser); + return NULL; + } + } + + freeBindingList = NULL; + freeTagList = NULL; + freeInternalEntities = NULL; + + groupSize = 0; + groupConnector = NULL; + + unknownEncodingHandler = NULL; + unknownEncodingHandlerData = NULL; + + namespaceSeparator = ASCII_EXCL; + ns = XML_FALSE; + ns_triplets = XML_FALSE; + + nsAtts = NULL; + nsAttsVersion = 0; + nsAttsPower = 0; + + poolInit(&tempPool, &(parser->m_mem)); + poolInit(&temp2Pool, &(parser->m_mem)); + parserInit(parser, encodingName); + + if (encodingName && !protocolEncodingName) { + XML_ParserFree(parser); + return NULL; + } + + if (nameSep) { + ns = XML_TRUE; + internalEncoding = XmlGetInternalEncodingNS(); + namespaceSeparator = *nameSep; + } + else { + internalEncoding = XmlGetInternalEncoding(); + } + + return parser; +} + +static void +parserInit(XML_Parser parser, const XML_Char *encodingName) +{ + processor = prologInitProcessor; + XmlPrologStateInit(&prologState); + protocolEncodingName = (encodingName != NULL + ? poolCopyString(&tempPool, encodingName) + : NULL); + curBase = NULL; + XmlInitEncoding(&initEncoding, &encoding, 0); + userData = NULL; + handlerArg = NULL; + startElementHandler = NULL; + endElementHandler = NULL; + characterDataHandler = NULL; + processingInstructionHandler = NULL; + commentHandler = NULL; + startCdataSectionHandler = NULL; + endCdataSectionHandler = NULL; + defaultHandler = NULL; + startDoctypeDeclHandler = NULL; + endDoctypeDeclHandler = NULL; + unparsedEntityDeclHandler = NULL; + notationDeclHandler = NULL; + startNamespaceDeclHandler = NULL; + endNamespaceDeclHandler = NULL; + notStandaloneHandler = NULL; + externalEntityRefHandler = NULL; + externalEntityRefHandlerArg = parser; + skippedEntityHandler = NULL; + elementDeclHandler = NULL; + attlistDeclHandler = NULL; + entityDeclHandler = NULL; + xmlDeclHandler = NULL; + bufferPtr = buffer; + bufferEnd = buffer; + parseEndByteIndex = 0; + parseEndPtr = NULL; + declElementType = NULL; + declAttributeId = NULL; + declEntity = NULL; + doctypeName = NULL; + doctypeSysid = NULL; + doctypePubid = NULL; + declAttributeType = NULL; + declNotationName = NULL; + declNotationPublicId = NULL; + declAttributeIsCdata = XML_FALSE; + declAttributeIsId = XML_FALSE; + memset(&position, 0, sizeof(POSITION)); + errorCode = XML_ERROR_NONE; + eventPtr = NULL; + eventEndPtr = NULL; + positionPtr = NULL; + openInternalEntities = NULL; + defaultExpandInternalEntities = XML_TRUE; + tagLevel = 0; + tagStack = NULL; + inheritedBindings = NULL; + nSpecifiedAtts = 0; + unknownEncodingMem = NULL; + unknownEncodingRelease = NULL; + unknownEncodingData = NULL; + parentParser = NULL; + ps_parsing = XML_INITIALIZED; +#ifdef XML_DTD + isParamEntity = XML_FALSE; + useForeignDTD = XML_FALSE; + paramEntityParsing = XML_PARAM_ENTITY_PARSING_NEVER; +#endif + hash_secret_salt = 0; +} + +/* moves list of bindings to freeBindingList */ +static void FASTCALL +moveToFreeBindingList(XML_Parser parser, BINDING *bindings) +{ + while (bindings) { + BINDING *b = bindings; + bindings = bindings->nextTagBinding; + b->nextTagBinding = freeBindingList; + freeBindingList = b; + } +} + +XML_Bool XMLCALL +XML_ParserReset(XML_Parser parser, const XML_Char *encodingName) +{ + TAG *tStk; + OPEN_INTERNAL_ENTITY *openEntityList; + + if (parser == NULL) + return XML_FALSE; + + if (parentParser) + return XML_FALSE; + /* move tagStack to freeTagList */ + tStk = tagStack; + while (tStk) { + TAG *tag = tStk; + tStk = tStk->parent; + tag->parent = freeTagList; + moveToFreeBindingList(parser, tag->bindings); + tag->bindings = NULL; + freeTagList = tag; + } + /* move openInternalEntities to freeInternalEntities */ + openEntityList = openInternalEntities; + while (openEntityList) { + OPEN_INTERNAL_ENTITY *openEntity = openEntityList; + openEntityList = openEntity->next; + openEntity->next = freeInternalEntities; + freeInternalEntities = openEntity; + } + moveToFreeBindingList(parser, inheritedBindings); + FREE(unknownEncodingMem); + if (unknownEncodingRelease) + unknownEncodingRelease(unknownEncodingData); + poolClear(&tempPool); + poolClear(&temp2Pool); + parserInit(parser, encodingName); + dtdReset(_dtd, &parser->m_mem); + return XML_TRUE; +} + +enum XML_Status XMLCALL +XML_SetEncoding(XML_Parser parser, const XML_Char *encodingName) +{ + if (parser == NULL) + return XML_STATUS_ERROR; + /* Block after XML_Parse()/XML_ParseBuffer() has been called. + XXX There's no way for the caller to determine which of the + XXX possible error cases caused the XML_STATUS_ERROR return. + */ + if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) + return XML_STATUS_ERROR; + if (encodingName == NULL) + protocolEncodingName = NULL; + else { + protocolEncodingName = poolCopyString(&tempPool, encodingName); + if (!protocolEncodingName) + return XML_STATUS_ERROR; + } + return XML_STATUS_OK; +} + +XML_Parser XMLCALL +XML_ExternalEntityParserCreate(XML_Parser oldParser, + const XML_Char *context, + const XML_Char *encodingName) +{ + XML_Parser parser = oldParser; + DTD *newDtd = NULL; + DTD *oldDtd; + XML_StartElementHandler oldStartElementHandler; + XML_EndElementHandler oldEndElementHandler; + XML_CharacterDataHandler oldCharacterDataHandler; + XML_ProcessingInstructionHandler oldProcessingInstructionHandler; + XML_CommentHandler oldCommentHandler; + XML_StartCdataSectionHandler oldStartCdataSectionHandler; + XML_EndCdataSectionHandler oldEndCdataSectionHandler; + XML_DefaultHandler oldDefaultHandler; + XML_UnparsedEntityDeclHandler oldUnparsedEntityDeclHandler; + XML_NotationDeclHandler oldNotationDeclHandler; + XML_StartNamespaceDeclHandler oldStartNamespaceDeclHandler; + XML_EndNamespaceDeclHandler oldEndNamespaceDeclHandler; + XML_NotStandaloneHandler oldNotStandaloneHandler; + XML_ExternalEntityRefHandler oldExternalEntityRefHandler; + XML_SkippedEntityHandler oldSkippedEntityHandler; + XML_UnknownEncodingHandler oldUnknownEncodingHandler; + XML_ElementDeclHandler oldElementDeclHandler; + XML_AttlistDeclHandler oldAttlistDeclHandler; + XML_EntityDeclHandler oldEntityDeclHandler; + XML_XmlDeclHandler oldXmlDeclHandler; + ELEMENT_TYPE * oldDeclElementType; + + void *oldUserData; + void *oldHandlerArg; + XML_Bool oldDefaultExpandInternalEntities; + XML_Parser oldExternalEntityRefHandlerArg; +#ifdef XML_DTD + enum XML_ParamEntityParsing oldParamEntityParsing; + int oldInEntityValue; +#endif + XML_Bool oldns_triplets; + /* Note that the new parser shares the same hash secret as the old + parser, so that dtdCopy and copyEntityTable can lookup values + from hash tables associated with either parser without us having + to worry which hash secrets each table has. + */ + unsigned long oldhash_secret_salt; + + /* Validate the oldParser parameter before we pull everything out of it */ + if (oldParser == NULL) + return NULL; + + /* Stash the original parser contents on the stack */ + oldDtd = _dtd; + oldStartElementHandler = startElementHandler; + oldEndElementHandler = endElementHandler; + oldCharacterDataHandler = characterDataHandler; + oldProcessingInstructionHandler = processingInstructionHandler; + oldCommentHandler = commentHandler; + oldStartCdataSectionHandler = startCdataSectionHandler; + oldEndCdataSectionHandler = endCdataSectionHandler; + oldDefaultHandler = defaultHandler; + oldUnparsedEntityDeclHandler = unparsedEntityDeclHandler; + oldNotationDeclHandler = notationDeclHandler; + oldStartNamespaceDeclHandler = startNamespaceDeclHandler; + oldEndNamespaceDeclHandler = endNamespaceDeclHandler; + oldNotStandaloneHandler = notStandaloneHandler; + oldExternalEntityRefHandler = externalEntityRefHandler; + oldSkippedEntityHandler = skippedEntityHandler; + oldUnknownEncodingHandler = unknownEncodingHandler; + oldElementDeclHandler = elementDeclHandler; + oldAttlistDeclHandler = attlistDeclHandler; + oldEntityDeclHandler = entityDeclHandler; + oldXmlDeclHandler = xmlDeclHandler; + oldDeclElementType = declElementType; + + oldUserData = userData; + oldHandlerArg = handlerArg; + oldDefaultExpandInternalEntities = defaultExpandInternalEntities; + oldExternalEntityRefHandlerArg = externalEntityRefHandlerArg; +#ifdef XML_DTD + oldParamEntityParsing = paramEntityParsing; + oldInEntityValue = prologState.inEntityValue; +#endif + oldns_triplets = ns_triplets; + /* Note that the new parser shares the same hash secret as the old + parser, so that dtdCopy and copyEntityTable can lookup values + from hash tables associated with either parser without us having + to worry which hash secrets each table has. + */ + oldhash_secret_salt = hash_secret_salt; + +#ifdef XML_DTD + if (!context) + newDtd = oldDtd; +#endif /* XML_DTD */ + + /* Note that the magical uses of the pre-processor to make field + access look more like C++ require that `parser' be overwritten + here. This makes this function more painful to follow than it + would be otherwise. + */ + if (ns) { + XML_Char tmp[2]; + *tmp = namespaceSeparator; + parser = parserCreate(encodingName, &parser->m_mem, tmp, newDtd); + } + else { + parser = parserCreate(encodingName, &parser->m_mem, NULL, newDtd); + } + + if (!parser) + return NULL; + + startElementHandler = oldStartElementHandler; + endElementHandler = oldEndElementHandler; + characterDataHandler = oldCharacterDataHandler; + processingInstructionHandler = oldProcessingInstructionHandler; + commentHandler = oldCommentHandler; + startCdataSectionHandler = oldStartCdataSectionHandler; + endCdataSectionHandler = oldEndCdataSectionHandler; + defaultHandler = oldDefaultHandler; + unparsedEntityDeclHandler = oldUnparsedEntityDeclHandler; + notationDeclHandler = oldNotationDeclHandler; + startNamespaceDeclHandler = oldStartNamespaceDeclHandler; + endNamespaceDeclHandler = oldEndNamespaceDeclHandler; + notStandaloneHandler = oldNotStandaloneHandler; + externalEntityRefHandler = oldExternalEntityRefHandler; + skippedEntityHandler = oldSkippedEntityHandler; + unknownEncodingHandler = oldUnknownEncodingHandler; + elementDeclHandler = oldElementDeclHandler; + attlistDeclHandler = oldAttlistDeclHandler; + entityDeclHandler = oldEntityDeclHandler; + xmlDeclHandler = oldXmlDeclHandler; + declElementType = oldDeclElementType; + userData = oldUserData; + if (oldUserData == oldHandlerArg) + handlerArg = userData; + else + handlerArg = parser; + if (oldExternalEntityRefHandlerArg != oldParser) + externalEntityRefHandlerArg = oldExternalEntityRefHandlerArg; + defaultExpandInternalEntities = oldDefaultExpandInternalEntities; + ns_triplets = oldns_triplets; + hash_secret_salt = oldhash_secret_salt; + parentParser = oldParser; +#ifdef XML_DTD + paramEntityParsing = oldParamEntityParsing; + prologState.inEntityValue = oldInEntityValue; + if (context) { +#endif /* XML_DTD */ + if (!dtdCopy(oldParser, _dtd, oldDtd, &parser->m_mem) + || !setContext(parser, context)) { + XML_ParserFree(parser); + return NULL; + } + processor = externalEntityInitProcessor; +#ifdef XML_DTD + } + else { + /* The DTD instance referenced by _dtd is shared between the document's + root parser and external PE parsers, therefore one does not need to + call setContext. In addition, one also *must* not call setContext, + because this would overwrite existing prefix->binding pointers in + _dtd with ones that get destroyed with the external PE parser. + This would leave those prefixes with dangling pointers. + */ + isParamEntity = XML_TRUE; + XmlPrologStateInitExternalEntity(&prologState); + processor = externalParEntInitProcessor; + } +#endif /* XML_DTD */ + return parser; +} + +static void FASTCALL +destroyBindings(BINDING *bindings, XML_Parser parser) +{ + for (;;) { + BINDING *b = bindings; + if (!b) + break; + bindings = b->nextTagBinding; + FREE(b->uri); + FREE(b); + } +} + +void XMLCALL +XML_ParserFree(XML_Parser parser) +{ + TAG *tagList; + OPEN_INTERNAL_ENTITY *entityList; + if (parser == NULL) + return; + /* free tagStack and freeTagList */ + tagList = tagStack; + for (;;) { + TAG *p; + if (tagList == NULL) { + if (freeTagList == NULL) + break; + tagList = freeTagList; + freeTagList = NULL; + } + p = tagList; + tagList = tagList->parent; + FREE(p->buf); + destroyBindings(p->bindings, parser); + FREE(p); + } + /* free openInternalEntities and freeInternalEntities */ + entityList = openInternalEntities; + for (;;) { + OPEN_INTERNAL_ENTITY *openEntity; + if (entityList == NULL) { + if (freeInternalEntities == NULL) + break; + entityList = freeInternalEntities; + freeInternalEntities = NULL; + } + openEntity = entityList; + entityList = entityList->next; + FREE(openEntity); + } + + destroyBindings(freeBindingList, parser); + destroyBindings(inheritedBindings, parser); + poolDestroy(&tempPool); + poolDestroy(&temp2Pool); +#ifdef XML_DTD + /* external parameter entity parsers share the DTD structure + parser->m_dtd with the root parser, so we must not destroy it + */ + if (!isParamEntity && _dtd) +#else + if (_dtd) +#endif /* XML_DTD */ + dtdDestroy(_dtd, (XML_Bool)!parentParser, &parser->m_mem); + FREE((void *)atts); +#ifdef XML_ATTR_INFO + FREE((void *)attInfo); +#endif + FREE(groupConnector); + FREE(buffer); + FREE(dataBuf); + FREE(nsAtts); + FREE(unknownEncodingMem); + if (unknownEncodingRelease) + unknownEncodingRelease(unknownEncodingData); + FREE(parser); +} + +void XMLCALL +XML_UseParserAsHandlerArg(XML_Parser parser) +{ + if (parser != NULL) + handlerArg = parser; +} + +enum XML_Error XMLCALL +XML_UseForeignDTD(XML_Parser parser, XML_Bool useDTD) +{ + if (parser == NULL) + return XML_ERROR_INVALID_ARGUMENT; +#ifdef XML_DTD + /* block after XML_Parse()/XML_ParseBuffer() has been called */ + if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) + return XML_ERROR_CANT_CHANGE_FEATURE_ONCE_PARSING; + useForeignDTD = useDTD; + return XML_ERROR_NONE; +#else + return XML_ERROR_FEATURE_REQUIRES_XML_DTD; +#endif +} + +void XMLCALL +XML_SetReturnNSTriplet(XML_Parser parser, int do_nst) +{ + if (parser == NULL) + return; + /* block after XML_Parse()/XML_ParseBuffer() has been called */ + if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) + return; + ns_triplets = do_nst ? XML_TRUE : XML_FALSE; +} + +void XMLCALL +XML_SetUserData(XML_Parser parser, void *p) +{ + if (parser == NULL) + return; + if (handlerArg == userData) + handlerArg = userData = p; + else + userData = p; +} + +enum XML_Status XMLCALL +XML_SetBase(XML_Parser parser, const XML_Char *p) +{ + if (parser == NULL) + return XML_STATUS_ERROR; + if (p) { + p = poolCopyString(&_dtd->pool, p); + if (!p) + return XML_STATUS_ERROR; + curBase = p; + } + else + curBase = NULL; + return XML_STATUS_OK; +} + +const XML_Char * XMLCALL +XML_GetBase(XML_Parser parser) +{ + if (parser == NULL) + return NULL; + return curBase; +} + +int XMLCALL +XML_GetSpecifiedAttributeCount(XML_Parser parser) +{ + if (parser == NULL) + return -1; + return nSpecifiedAtts; +} + +int XMLCALL +XML_GetIdAttributeIndex(XML_Parser parser) +{ + if (parser == NULL) + return -1; + return idAttIndex; +} + +#ifdef XML_ATTR_INFO +const XML_AttrInfo * XMLCALL +XML_GetAttributeInfo(XML_Parser parser) +{ + if (parser == NULL) + return NULL; + return attInfo; +} +#endif + +void XMLCALL +XML_SetElementHandler(XML_Parser parser, + XML_StartElementHandler start, + XML_EndElementHandler end) +{ + if (parser == NULL) + return; + startElementHandler = start; + endElementHandler = end; +} + +void XMLCALL +XML_SetStartElementHandler(XML_Parser parser, + XML_StartElementHandler start) { + if (parser != NULL) + startElementHandler = start; +} + +void XMLCALL +XML_SetEndElementHandler(XML_Parser parser, + XML_EndElementHandler end) { + if (parser != NULL) + endElementHandler = end; +} + +void XMLCALL +XML_SetCharacterDataHandler(XML_Parser parser, + XML_CharacterDataHandler handler) +{ + if (parser != NULL) + characterDataHandler = handler; +} + +void XMLCALL +XML_SetProcessingInstructionHandler(XML_Parser parser, + XML_ProcessingInstructionHandler handler) +{ + if (parser != NULL) + processingInstructionHandler = handler; +} + +void XMLCALL +XML_SetCommentHandler(XML_Parser parser, + XML_CommentHandler handler) +{ + if (parser != NULL) + commentHandler = handler; +} + +void XMLCALL +XML_SetCdataSectionHandler(XML_Parser parser, + XML_StartCdataSectionHandler start, + XML_EndCdataSectionHandler end) +{ + if (parser == NULL) + return; + startCdataSectionHandler = start; + endCdataSectionHandler = end; +} + +void XMLCALL +XML_SetStartCdataSectionHandler(XML_Parser parser, + XML_StartCdataSectionHandler start) { + if (parser != NULL) + startCdataSectionHandler = start; +} + +void XMLCALL +XML_SetEndCdataSectionHandler(XML_Parser parser, + XML_EndCdataSectionHandler end) { + if (parser != NULL) + endCdataSectionHandler = end; +} + +void XMLCALL +XML_SetDefaultHandler(XML_Parser parser, + XML_DefaultHandler handler) +{ + if (parser == NULL) + return; + defaultHandler = handler; + defaultExpandInternalEntities = XML_FALSE; +} + +void XMLCALL +XML_SetDefaultHandlerExpand(XML_Parser parser, + XML_DefaultHandler handler) +{ + if (parser == NULL) + return; + defaultHandler = handler; + defaultExpandInternalEntities = XML_TRUE; +} + +void XMLCALL +XML_SetDoctypeDeclHandler(XML_Parser parser, + XML_StartDoctypeDeclHandler start, + XML_EndDoctypeDeclHandler end) +{ + if (parser == NULL) + return; + startDoctypeDeclHandler = start; + endDoctypeDeclHandler = end; +} + +void XMLCALL +XML_SetStartDoctypeDeclHandler(XML_Parser parser, + XML_StartDoctypeDeclHandler start) { + if (parser != NULL) + startDoctypeDeclHandler = start; +} + +void XMLCALL +XML_SetEndDoctypeDeclHandler(XML_Parser parser, + XML_EndDoctypeDeclHandler end) { + if (parser != NULL) + endDoctypeDeclHandler = end; +} + +void XMLCALL +XML_SetUnparsedEntityDeclHandler(XML_Parser parser, + XML_UnparsedEntityDeclHandler handler) +{ + if (parser != NULL) + unparsedEntityDeclHandler = handler; +} + +void XMLCALL +XML_SetNotationDeclHandler(XML_Parser parser, + XML_NotationDeclHandler handler) +{ + if (parser != NULL) + notationDeclHandler = handler; +} + +void XMLCALL +XML_SetNamespaceDeclHandler(XML_Parser parser, + XML_StartNamespaceDeclHandler start, + XML_EndNamespaceDeclHandler end) +{ + if (parser == NULL) + return; + startNamespaceDeclHandler = start; + endNamespaceDeclHandler = end; +} + +void XMLCALL +XML_SetStartNamespaceDeclHandler(XML_Parser parser, + XML_StartNamespaceDeclHandler start) { + if (parser != NULL) + startNamespaceDeclHandler = start; +} + +void XMLCALL +XML_SetEndNamespaceDeclHandler(XML_Parser parser, + XML_EndNamespaceDeclHandler end) { + if (parser != NULL) + endNamespaceDeclHandler = end; +} + +void XMLCALL +XML_SetNotStandaloneHandler(XML_Parser parser, + XML_NotStandaloneHandler handler) +{ + if (parser != NULL) + notStandaloneHandler = handler; +} + +void XMLCALL +XML_SetExternalEntityRefHandler(XML_Parser parser, + XML_ExternalEntityRefHandler handler) +{ + if (parser != NULL) + externalEntityRefHandler = handler; +} + +void XMLCALL +XML_SetExternalEntityRefHandlerArg(XML_Parser parser, void *arg) +{ + if (parser == NULL) + return; + if (arg) + externalEntityRefHandlerArg = (XML_Parser)arg; + else + externalEntityRefHandlerArg = parser; +} + +void XMLCALL +XML_SetSkippedEntityHandler(XML_Parser parser, + XML_SkippedEntityHandler handler) +{ + if (parser != NULL) + skippedEntityHandler = handler; +} + +void XMLCALL +XML_SetUnknownEncodingHandler(XML_Parser parser, + XML_UnknownEncodingHandler handler, + void *data) +{ + if (parser == NULL) + return; + unknownEncodingHandler = handler; + unknownEncodingHandlerData = data; +} + +void XMLCALL +XML_SetElementDeclHandler(XML_Parser parser, + XML_ElementDeclHandler eldecl) +{ + if (parser != NULL) + elementDeclHandler = eldecl; +} + +void XMLCALL +XML_SetAttlistDeclHandler(XML_Parser parser, + XML_AttlistDeclHandler attdecl) +{ + if (parser != NULL) + attlistDeclHandler = attdecl; +} + +void XMLCALL +XML_SetEntityDeclHandler(XML_Parser parser, + XML_EntityDeclHandler handler) +{ + if (parser != NULL) + entityDeclHandler = handler; +} + +void XMLCALL +XML_SetXmlDeclHandler(XML_Parser parser, + XML_XmlDeclHandler handler) { + if (parser != NULL) + xmlDeclHandler = handler; +} + +int XMLCALL +XML_SetParamEntityParsing(XML_Parser parser, + enum XML_ParamEntityParsing peParsing) +{ + if (parser == NULL) + return 0; + /* block after XML_Parse()/XML_ParseBuffer() has been called */ + if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) + return 0; +#ifdef XML_DTD + paramEntityParsing = peParsing; + return 1; +#else + return peParsing == XML_PARAM_ENTITY_PARSING_NEVER; +#endif +} + +int XMLCALL +XML_SetHashSalt(XML_Parser parser, + unsigned long hash_salt) +{ + if (parser == NULL) + return 0; + if (parser->m_parentParser) + return XML_SetHashSalt(parser->m_parentParser, hash_salt); + /* block after XML_Parse()/XML_ParseBuffer() has been called */ + if (ps_parsing == XML_PARSING || ps_parsing == XML_SUSPENDED) + return 0; + hash_secret_salt = hash_salt; + return 1; +} + +enum XML_Status XMLCALL +XML_Parse(XML_Parser parser, const char *s, int len, int isFinal) +{ + if ((parser == NULL) || (len < 0) || ((s == NULL) && (len != 0))) { + errorCode = XML_ERROR_INVALID_ARGUMENT; + return XML_STATUS_ERROR; + } + switch (ps_parsing) { + case XML_SUSPENDED: + errorCode = XML_ERROR_SUSPENDED; + return XML_STATUS_ERROR; + case XML_FINISHED: + errorCode = XML_ERROR_FINISHED; + return XML_STATUS_ERROR; + case XML_INITIALIZED: + if (parentParser == NULL && !startParsing(parser)) { + errorCode = XML_ERROR_NO_MEMORY; + return XML_STATUS_ERROR; + } + default: + ps_parsing = XML_PARSING; + } + + if (len == 0) { + ps_finalBuffer = (XML_Bool)isFinal; + if (!isFinal) + return XML_STATUS_OK; + positionPtr = bufferPtr; + parseEndPtr = bufferEnd; + + /* If data are left over from last buffer, and we now know that these + data are the final chunk of input, then we have to check them again + to detect errors based on that fact. + */ + errorCode = processor(parser, bufferPtr, parseEndPtr, &bufferPtr); + + if (errorCode == XML_ERROR_NONE) { + switch (ps_parsing) { + case XML_SUSPENDED: + XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); + positionPtr = bufferPtr; + return XML_STATUS_SUSPENDED; + case XML_INITIALIZED: + case XML_PARSING: + ps_parsing = XML_FINISHED; + /* fall through */ + default: + return XML_STATUS_OK; + } + } + eventEndPtr = eventPtr; + processor = errorProcessor; + return XML_STATUS_ERROR; + } +#ifndef XML_CONTEXT_BYTES + else if (bufferPtr == bufferEnd) { + const char *end; + int nLeftOver; + enum XML_Status result; + /* Detect overflow (a+b > MAX <==> b > MAX-a) */ + if (len > ((XML_Size)-1) / 2 - parseEndByteIndex) { + errorCode = XML_ERROR_NO_MEMORY; + eventPtr = eventEndPtr = NULL; + processor = errorProcessor; + return XML_STATUS_ERROR; + } + parseEndByteIndex += len; + positionPtr = s; + ps_finalBuffer = (XML_Bool)isFinal; + + errorCode = processor(parser, s, parseEndPtr = s + len, &end); + + if (errorCode != XML_ERROR_NONE) { + eventEndPtr = eventPtr; + processor = errorProcessor; + return XML_STATUS_ERROR; + } + else { + switch (ps_parsing) { + case XML_SUSPENDED: + result = XML_STATUS_SUSPENDED; + break; + case XML_INITIALIZED: + case XML_PARSING: + if (isFinal) { + ps_parsing = XML_FINISHED; + return XML_STATUS_OK; + } + /* fall through */ + default: + result = XML_STATUS_OK; + } + } + + XmlUpdatePosition(encoding, positionPtr, end, &position); + nLeftOver = s + len - end; + if (nLeftOver) { + if (buffer == NULL || nLeftOver > bufferLim - buffer) { + /* avoid _signed_ integer overflow */ + char *temp = NULL; + const int bytesToAllocate = (int)((unsigned)len * 2U); + if (bytesToAllocate > 0) { + temp = (buffer == NULL + ? (char *)MALLOC(bytesToAllocate) + : (char *)REALLOC(buffer, bytesToAllocate)); + } + if (temp == NULL) { + errorCode = XML_ERROR_NO_MEMORY; + eventPtr = eventEndPtr = NULL; + processor = errorProcessor; + return XML_STATUS_ERROR; + } + buffer = temp; + bufferLim = buffer + bytesToAllocate; + } + memcpy(buffer, end, nLeftOver); + } + bufferPtr = buffer; + bufferEnd = buffer + nLeftOver; + positionPtr = bufferPtr; + parseEndPtr = bufferEnd; + eventPtr = bufferPtr; + eventEndPtr = bufferPtr; + return result; + } +#endif /* not defined XML_CONTEXT_BYTES */ + else { + void *buff = XML_GetBuffer(parser, len); + if (buff == NULL) + return XML_STATUS_ERROR; + else { + memcpy(buff, s, len); + return XML_ParseBuffer(parser, len, isFinal); + } + } +} + +enum XML_Status XMLCALL +XML_ParseBuffer(XML_Parser parser, int len, int isFinal) +{ + const char *start; + enum XML_Status result = XML_STATUS_OK; + + if (parser == NULL) + return XML_STATUS_ERROR; + switch (ps_parsing) { + case XML_SUSPENDED: + errorCode = XML_ERROR_SUSPENDED; + return XML_STATUS_ERROR; + case XML_FINISHED: + errorCode = XML_ERROR_FINISHED; + return XML_STATUS_ERROR; + case XML_INITIALIZED: + if (parentParser == NULL && !startParsing(parser)) { + errorCode = XML_ERROR_NO_MEMORY; + return XML_STATUS_ERROR; + } + default: + ps_parsing = XML_PARSING; + } + + start = bufferPtr; + positionPtr = start; + bufferEnd += len; + parseEndPtr = bufferEnd; + parseEndByteIndex += len; + ps_finalBuffer = (XML_Bool)isFinal; + + errorCode = processor(parser, start, parseEndPtr, &bufferPtr); + + if (errorCode != XML_ERROR_NONE) { + eventEndPtr = eventPtr; + processor = errorProcessor; + return XML_STATUS_ERROR; + } + else { + switch (ps_parsing) { + case XML_SUSPENDED: + result = XML_STATUS_SUSPENDED; + break; + case XML_INITIALIZED: + case XML_PARSING: + if (isFinal) { + ps_parsing = XML_FINISHED; + return result; + } + default: ; /* should not happen */ + } + } + + XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); + positionPtr = bufferPtr; + return result; +} + +void * XMLCALL +XML_GetBuffer(XML_Parser parser, int len) +{ + if (parser == NULL) + return NULL; + if (len < 0) { + errorCode = XML_ERROR_NO_MEMORY; + return NULL; + } + switch (ps_parsing) { + case XML_SUSPENDED: + errorCode = XML_ERROR_SUSPENDED; + return NULL; + case XML_FINISHED: + errorCode = XML_ERROR_FINISHED; + return NULL; + default: ; + } + + if (len > bufferLim - bufferEnd) { +#ifdef XML_CONTEXT_BYTES + int keep; +#endif /* defined XML_CONTEXT_BYTES */ + /* Do not invoke signed arithmetic overflow: */ + int neededSize = (int) ((unsigned)len + (unsigned)(bufferEnd - bufferPtr)); + if (neededSize < 0) { + errorCode = XML_ERROR_NO_MEMORY; + return NULL; + } +#ifdef XML_CONTEXT_BYTES + keep = (int)(bufferPtr - buffer); + if (keep > XML_CONTEXT_BYTES) + keep = XML_CONTEXT_BYTES; + neededSize += keep; +#endif /* defined XML_CONTEXT_BYTES */ + if (neededSize <= bufferLim - buffer) { +#ifdef XML_CONTEXT_BYTES + if (keep < bufferPtr - buffer) { + int offset = (int)(bufferPtr - buffer) - keep; + memmove(buffer, &buffer[offset], bufferEnd - bufferPtr + keep); + bufferEnd -= offset; + bufferPtr -= offset; + } +#else + memmove(buffer, bufferPtr, bufferEnd - bufferPtr); + bufferEnd = buffer + (bufferEnd - bufferPtr); + bufferPtr = buffer; +#endif /* not defined XML_CONTEXT_BYTES */ + } + else { + char *newBuf; + int bufferSize = (int)(bufferLim - bufferPtr); + if (bufferSize == 0) + bufferSize = INIT_BUFFER_SIZE; + do { + /* Do not invoke signed arithmetic overflow: */ + bufferSize = (int) (2U * (unsigned) bufferSize); + } while (bufferSize < neededSize && bufferSize > 0); + if (bufferSize <= 0) { + errorCode = XML_ERROR_NO_MEMORY; + return NULL; + } + newBuf = (char *)MALLOC(bufferSize); + if (newBuf == 0) { + errorCode = XML_ERROR_NO_MEMORY; + return NULL; + } + bufferLim = newBuf + bufferSize; +#ifdef XML_CONTEXT_BYTES + if (bufferPtr) { + int keep = (int)(bufferPtr - buffer); + if (keep > XML_CONTEXT_BYTES) + keep = XML_CONTEXT_BYTES; + memcpy(newBuf, &bufferPtr[-keep], bufferEnd - bufferPtr + keep); + FREE(buffer); + buffer = newBuf; + bufferEnd = buffer + (bufferEnd - bufferPtr) + keep; + bufferPtr = buffer + keep; + } + else { + bufferEnd = newBuf + (bufferEnd - bufferPtr); + bufferPtr = buffer = newBuf; + } +#else + if (bufferPtr) { + memcpy(newBuf, bufferPtr, bufferEnd - bufferPtr); + FREE(buffer); + } + bufferEnd = newBuf + (bufferEnd - bufferPtr); + bufferPtr = buffer = newBuf; +#endif /* not defined XML_CONTEXT_BYTES */ + } + eventPtr = eventEndPtr = NULL; + positionPtr = NULL; + } + return bufferEnd; +} + +enum XML_Status XMLCALL +XML_StopParser(XML_Parser parser, XML_Bool resumable) +{ + if (parser == NULL) + return XML_STATUS_ERROR; + switch (ps_parsing) { + case XML_SUSPENDED: + if (resumable) { + errorCode = XML_ERROR_SUSPENDED; + return XML_STATUS_ERROR; + } + ps_parsing = XML_FINISHED; + break; + case XML_FINISHED: + errorCode = XML_ERROR_FINISHED; + return XML_STATUS_ERROR; + default: + if (resumable) { +#ifdef XML_DTD + if (isParamEntity) { + errorCode = XML_ERROR_SUSPEND_PE; + return XML_STATUS_ERROR; + } +#endif + ps_parsing = XML_SUSPENDED; + } + else + ps_parsing = XML_FINISHED; + } + return XML_STATUS_OK; +} + +enum XML_Status XMLCALL +XML_ResumeParser(XML_Parser parser) +{ + enum XML_Status result = XML_STATUS_OK; + + if (parser == NULL) + return XML_STATUS_ERROR; + if (ps_parsing != XML_SUSPENDED) { + errorCode = XML_ERROR_NOT_SUSPENDED; + return XML_STATUS_ERROR; + } + ps_parsing = XML_PARSING; + + errorCode = processor(parser, bufferPtr, parseEndPtr, &bufferPtr); + + if (errorCode != XML_ERROR_NONE) { + eventEndPtr = eventPtr; + processor = errorProcessor; + return XML_STATUS_ERROR; + } + else { + switch (ps_parsing) { + case XML_SUSPENDED: + result = XML_STATUS_SUSPENDED; + break; + case XML_INITIALIZED: + case XML_PARSING: + if (ps_finalBuffer) { + ps_parsing = XML_FINISHED; + return result; + } + default: ; + } + } + + XmlUpdatePosition(encoding, positionPtr, bufferPtr, &position); + positionPtr = bufferPtr; + return result; +} + +void XMLCALL +XML_GetParsingStatus(XML_Parser parser, XML_ParsingStatus *status) +{ + if (parser == NULL) + return; + assert(status != NULL); + *status = parser->m_parsingStatus; +} + +enum XML_Error XMLCALL +XML_GetErrorCode(XML_Parser parser) +{ + if (parser == NULL) + return XML_ERROR_INVALID_ARGUMENT; + return errorCode; +} + +XML_Index XMLCALL +XML_GetCurrentByteIndex(XML_Parser parser) +{ + if (parser == NULL) + return -1; + if (eventPtr) + return (XML_Index)(parseEndByteIndex - (parseEndPtr - eventPtr)); + return -1; +} + +int XMLCALL +XML_GetCurrentByteCount(XML_Parser parser) +{ + if (parser == NULL) + return 0; + if (eventEndPtr && eventPtr) + return (int)(eventEndPtr - eventPtr); + return 0; +} + +const char * XMLCALL +XML_GetInputContext(XML_Parser parser, int *offset, int *size) +{ +#ifdef XML_CONTEXT_BYTES + if (parser == NULL) + return NULL; + if (eventPtr && buffer) { + if (offset != NULL) + *offset = (int)(eventPtr - buffer); + if (size != NULL) + *size = (int)(bufferEnd - buffer); + return buffer; + } +#else + (void)parser; + (void)offset; + (void)size; +#endif /* defined XML_CONTEXT_BYTES */ + return (char *) 0; +} + +XML_Size XMLCALL +XML_GetCurrentLineNumber(XML_Parser parser) +{ + if (parser == NULL) + return 0; + if (eventPtr && eventPtr >= positionPtr) { + XmlUpdatePosition(encoding, positionPtr, eventPtr, &position); + positionPtr = eventPtr; + } + return position.lineNumber + 1; +} + +XML_Size XMLCALL +XML_GetCurrentColumnNumber(XML_Parser parser) +{ + if (parser == NULL) + return 0; + if (eventPtr && eventPtr >= positionPtr) { + XmlUpdatePosition(encoding, positionPtr, eventPtr, &position); + positionPtr = eventPtr; + } + return position.columnNumber; +} + +void XMLCALL +XML_FreeContentModel(XML_Parser parser, XML_Content *model) +{ + if (parser != NULL) + FREE(model); +} + +void * XMLCALL +XML_MemMalloc(XML_Parser parser, size_t size) +{ + if (parser == NULL) + return NULL; + return MALLOC(size); +} + +void * XMLCALL +XML_MemRealloc(XML_Parser parser, void *ptr, size_t size) +{ + if (parser == NULL) + return NULL; + return REALLOC(ptr, size); +} + +void XMLCALL +XML_MemFree(XML_Parser parser, void *ptr) +{ + if (parser != NULL) + FREE(ptr); +} + +void XMLCALL +XML_DefaultCurrent(XML_Parser parser) +{ + if (parser == NULL) + return; + if (defaultHandler) { + if (openInternalEntities) + reportDefault(parser, + internalEncoding, + openInternalEntities->internalEventPtr, + openInternalEntities->internalEventEndPtr); + else + reportDefault(parser, encoding, eventPtr, eventEndPtr); + } +} + +const XML_LChar * XMLCALL +XML_ErrorString(enum XML_Error code) +{ + static const XML_LChar* const message[] = { + 0, + XML_L("out of memory"), + XML_L("syntax error"), + XML_L("no element found"), + XML_L("not well-formed (invalid token)"), + XML_L("unclosed token"), + XML_L("partial character"), + XML_L("mismatched tag"), + XML_L("duplicate attribute"), + XML_L("junk after document element"), + XML_L("illegal parameter entity reference"), + XML_L("undefined entity"), + XML_L("recursive entity reference"), + XML_L("asynchronous entity"), + XML_L("reference to invalid character number"), + XML_L("reference to binary entity"), + XML_L("reference to external entity in attribute"), + XML_L("XML or text declaration not at start of entity"), + XML_L("unknown encoding"), + XML_L("encoding specified in XML declaration is incorrect"), + XML_L("unclosed CDATA section"), + XML_L("error in processing external entity reference"), + XML_L("document is not standalone"), + XML_L("unexpected parser state - please send a bug report"), + XML_L("entity declared in parameter entity"), + XML_L("requested feature requires XML_DTD support in Expat"), + XML_L("cannot change setting once parsing has begun"), + XML_L("unbound prefix"), + XML_L("must not undeclare prefix"), + XML_L("incomplete markup in parameter entity"), + XML_L("XML declaration not well-formed"), + XML_L("text declaration not well-formed"), + XML_L("illegal character(s) in public id"), + XML_L("parser suspended"), + XML_L("parser not suspended"), + XML_L("parsing aborted"), + XML_L("parsing finished"), + XML_L("cannot suspend in external parameter entity"), + XML_L("reserved prefix (xml) must not be undeclared or bound to another namespace name"), + XML_L("reserved prefix (xmlns) must not be declared or undeclared"), + XML_L("prefix must not be bound to one of the reserved namespace names") + }; + if (code > 0 && code < sizeof(message)/sizeof(message[0])) + return message[code]; + return NULL; +} + +const XML_LChar * XMLCALL +XML_ExpatVersion(void) { + + /* V1 is used to string-ize the version number. However, it would + string-ize the actual version macro *names* unless we get them + substituted before being passed to V1. CPP is defined to expand + a macro, then rescan for more expansions. Thus, we use V2 to expand + the version macros, then CPP will expand the resulting V1() macro + with the correct numerals. */ + /* ### I'm assuming cpp is portable in this respect... */ + +#define V1(a,b,c) XML_L(#a)XML_L(".")XML_L(#b)XML_L(".")XML_L(#c) +#define V2(a,b,c) XML_L("expat_")V1(a,b,c) + + return V2(XML_MAJOR_VERSION, XML_MINOR_VERSION, XML_MICRO_VERSION); + +#undef V1 +#undef V2 +} + +XML_Expat_Version XMLCALL +XML_ExpatVersionInfo(void) +{ + XML_Expat_Version version; + + version.major = XML_MAJOR_VERSION; + version.minor = XML_MINOR_VERSION; + version.micro = XML_MICRO_VERSION; + + return version; +} + +const XML_Feature * XMLCALL +XML_GetFeatureList(void) +{ + static const XML_Feature features[] = { + {XML_FEATURE_SIZEOF_XML_CHAR, XML_L("sizeof(XML_Char)"), + sizeof(XML_Char)}, + {XML_FEATURE_SIZEOF_XML_LCHAR, XML_L("sizeof(XML_LChar)"), + sizeof(XML_LChar)}, +#ifdef XML_UNICODE + {XML_FEATURE_UNICODE, XML_L("XML_UNICODE"), 0}, +#endif +#ifdef XML_UNICODE_WCHAR_T + {XML_FEATURE_UNICODE_WCHAR_T, XML_L("XML_UNICODE_WCHAR_T"), 0}, +#endif +#ifdef XML_DTD + {XML_FEATURE_DTD, XML_L("XML_DTD"), 0}, +#endif +#ifdef XML_CONTEXT_BYTES + {XML_FEATURE_CONTEXT_BYTES, XML_L("XML_CONTEXT_BYTES"), + XML_CONTEXT_BYTES}, +#endif +#ifdef XML_MIN_SIZE + {XML_FEATURE_MIN_SIZE, XML_L("XML_MIN_SIZE"), 0}, +#endif +#ifdef XML_NS + {XML_FEATURE_NS, XML_L("XML_NS"), 0}, +#endif +#ifdef XML_LARGE_SIZE + {XML_FEATURE_LARGE_SIZE, XML_L("XML_LARGE_SIZE"), 0}, +#endif +#ifdef XML_ATTR_INFO + {XML_FEATURE_ATTR_INFO, XML_L("XML_ATTR_INFO"), 0}, +#endif + {XML_FEATURE_END, NULL, 0} + }; + + return features; +} + +/* Initially tag->rawName always points into the parse buffer; + for those TAG instances opened while the current parse buffer was + processed, and not yet closed, we need to store tag->rawName in a more + permanent location, since the parse buffer is about to be discarded. +*/ +static XML_Bool +storeRawNames(XML_Parser parser) +{ + TAG *tag = tagStack; + while (tag) { + int bufSize; + int nameLen = sizeof(XML_Char) * (tag->name.strLen + 1); + char *rawNameBuf = tag->buf + nameLen; + /* Stop if already stored. Since tagStack is a stack, we can stop + at the first entry that has already been copied; everything + below it in the stack is already been accounted for in a + previous call to this function. + */ + if (tag->rawName == rawNameBuf) + break; + /* For re-use purposes we need to ensure that the + size of tag->buf is a multiple of sizeof(XML_Char). + */ + bufSize = nameLen + ROUND_UP(tag->rawNameLength, sizeof(XML_Char)); + if (bufSize > tag->bufEnd - tag->buf) { + char *temp = (char *)REALLOC(tag->buf, bufSize); + if (temp == NULL) + return XML_FALSE; + /* if tag->name.str points to tag->buf (only when namespace + processing is off) then we have to update it + */ + if (tag->name.str == (XML_Char *)tag->buf) + tag->name.str = (XML_Char *)temp; + /* if tag->name.localPart is set (when namespace processing is on) + then update it as well, since it will always point into tag->buf + */ + if (tag->name.localPart) + tag->name.localPart = (XML_Char *)temp + (tag->name.localPart - + (XML_Char *)tag->buf); + tag->buf = temp; + tag->bufEnd = temp + bufSize; + rawNameBuf = temp + nameLen; + } + memcpy(rawNameBuf, tag->rawName, tag->rawNameLength); + tag->rawName = rawNameBuf; + tag = tag->parent; + } + return XML_TRUE; +} + +static enum XML_Error PTRCALL +contentProcessor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + enum XML_Error result = doContent(parser, 0, encoding, start, end, + endPtr, (XML_Bool)!ps_finalBuffer); + if (result == XML_ERROR_NONE) { + if (!storeRawNames(parser)) + return XML_ERROR_NO_MEMORY; + } + return result; +} + +static enum XML_Error PTRCALL +externalEntityInitProcessor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + enum XML_Error result = initializeEncoding(parser); + if (result != XML_ERROR_NONE) + return result; + processor = externalEntityInitProcessor2; + return externalEntityInitProcessor2(parser, start, end, endPtr); +} + +static enum XML_Error PTRCALL +externalEntityInitProcessor2(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + const char *next = start; /* XmlContentTok doesn't always set the last arg */ + int tok = XmlContentTok(encoding, start, end, &next); + switch (tok) { + case XML_TOK_BOM: + /* If we are at the end of the buffer, this would cause the next stage, + i.e. externalEntityInitProcessor3, to pass control directly to + doContent (by detecting XML_TOK_NONE) without processing any xml text + declaration - causing the error XML_ERROR_MISPLACED_XML_PI in doContent. + */ + if (next == end && !ps_finalBuffer) { + *endPtr = next; + return XML_ERROR_NONE; + } + start = next; + break; + case XML_TOK_PARTIAL: + if (!ps_finalBuffer) { + *endPtr = start; + return XML_ERROR_NONE; + } + eventPtr = start; + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (!ps_finalBuffer) { + *endPtr = start; + return XML_ERROR_NONE; + } + eventPtr = start; + return XML_ERROR_PARTIAL_CHAR; + } + processor = externalEntityInitProcessor3; + return externalEntityInitProcessor3(parser, start, end, endPtr); +} + +static enum XML_Error PTRCALL +externalEntityInitProcessor3(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + int tok; + const char *next = start; /* XmlContentTok doesn't always set the last arg */ + eventPtr = start; + tok = XmlContentTok(encoding, start, end, &next); + eventEndPtr = next; + + switch (tok) { + case XML_TOK_XML_DECL: + { + enum XML_Error result; + result = processXmlDecl(parser, 1, start, next); + if (result != XML_ERROR_NONE) + return result; + switch (ps_parsing) { + case XML_SUSPENDED: + *endPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: + start = next; + } + } + break; + case XML_TOK_PARTIAL: + if (!ps_finalBuffer) { + *endPtr = start; + return XML_ERROR_NONE; + } + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (!ps_finalBuffer) { + *endPtr = start; + return XML_ERROR_NONE; + } + return XML_ERROR_PARTIAL_CHAR; + } + processor = externalEntityContentProcessor; + tagLevel = 1; + return externalEntityContentProcessor(parser, start, end, endPtr); +} + +static enum XML_Error PTRCALL +externalEntityContentProcessor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + enum XML_Error result = doContent(parser, 1, encoding, start, end, + endPtr, (XML_Bool)!ps_finalBuffer); + if (result == XML_ERROR_NONE) { + if (!storeRawNames(parser)) + return XML_ERROR_NO_MEMORY; + } + return result; +} + +static enum XML_Error +doContent(XML_Parser parser, + int startTagLevel, + const ENCODING *enc, + const char *s, + const char *end, + const char **nextPtr, + XML_Bool haveMore) +{ + /* save one level of indirection */ + DTD * const dtd = _dtd; + + const char **eventPP; + const char **eventEndPP; + if (enc == encoding) { + eventPP = &eventPtr; + eventEndPP = &eventEndPtr; + } + else { + eventPP = &(openInternalEntities->internalEventPtr); + eventEndPP = &(openInternalEntities->internalEventEndPtr); + } + *eventPP = s; + + for (;;) { + const char *next = s; /* XmlContentTok doesn't always set the last arg */ + int tok = XmlContentTok(enc, s, end, &next); + *eventEndPP = next; + switch (tok) { + case XML_TOK_TRAILING_CR: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + *eventEndPP = end; + if (characterDataHandler) { + XML_Char c = 0xA; + characterDataHandler(handlerArg, &c, 1); + } + else if (defaultHandler) + reportDefault(parser, enc, s, end); + /* We are at the end of the final buffer, should we check for + XML_SUSPENDED, XML_FINISHED? + */ + if (startTagLevel == 0) + return XML_ERROR_NO_ELEMENTS; + if (tagLevel != startTagLevel) + return XML_ERROR_ASYNC_ENTITY; + *nextPtr = end; + return XML_ERROR_NONE; + case XML_TOK_NONE: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + if (startTagLevel > 0) { + if (tagLevel != startTagLevel) + return XML_ERROR_ASYNC_ENTITY; + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_NO_ELEMENTS; + case XML_TOK_INVALID: + *eventPP = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_PARTIAL_CHAR; + case XML_TOK_ENTITY_REF: + { + const XML_Char *name; + ENTITY *entity; + XML_Char ch = (XML_Char) XmlPredefinedEntityName(enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (ch) { + if (characterDataHandler) + characterDataHandler(handlerArg, &ch, 1); + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + } + name = poolStoreString(&dtd->pool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!name) + return XML_ERROR_NO_MEMORY; + entity = (ENTITY *)lookup(parser, &dtd->generalEntities, name, 0); + poolDiscard(&dtd->pool); + /* First, determine if a check for an existing declaration is needed; + if yes, check that the entity exists, and that it is internal, + otherwise call the skipped entity or default handler. + */ + if (!dtd->hasParamEntityRefs || dtd->standalone) { + if (!entity) + return XML_ERROR_UNDEFINED_ENTITY; + else if (!entity->is_internal) + return XML_ERROR_ENTITY_DECLARED_IN_PE; + } + else if (!entity) { + if (skippedEntityHandler) + skippedEntityHandler(handlerArg, name, 0); + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + } + if (entity->open) + return XML_ERROR_RECURSIVE_ENTITY_REF; + if (entity->notation) + return XML_ERROR_BINARY_ENTITY_REF; + if (entity->textPtr) { + enum XML_Error result; + if (!defaultExpandInternalEntities) { + if (skippedEntityHandler) + skippedEntityHandler(handlerArg, entity->name, 0); + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + } + result = processInternalEntity(parser, entity, XML_FALSE); + if (result != XML_ERROR_NONE) + return result; + } + else if (externalEntityRefHandler) { + const XML_Char *context; + entity->open = XML_TRUE; + context = getContext(parser); + entity->open = XML_FALSE; + if (!context) + return XML_ERROR_NO_MEMORY; + if (!externalEntityRefHandler(externalEntityRefHandlerArg, + context, + entity->base, + entity->systemId, + entity->publicId)) + return XML_ERROR_EXTERNAL_ENTITY_HANDLING; + poolDiscard(&tempPool); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + } + case XML_TOK_START_TAG_NO_ATTS: + /* fall through */ + case XML_TOK_START_TAG_WITH_ATTS: + { + TAG *tag; + enum XML_Error result; + XML_Char *toPtr; + if (freeTagList) { + tag = freeTagList; + freeTagList = freeTagList->parent; + } + else { + tag = (TAG *)MALLOC(sizeof(TAG)); + if (!tag) + return XML_ERROR_NO_MEMORY; + tag->buf = (char *)MALLOC(INIT_TAG_BUF_SIZE); + if (!tag->buf) { + FREE(tag); + return XML_ERROR_NO_MEMORY; + } + tag->bufEnd = tag->buf + INIT_TAG_BUF_SIZE; + } + tag->bindings = NULL; + tag->parent = tagStack; + tagStack = tag; + tag->name.localPart = NULL; + tag->name.prefix = NULL; + tag->rawName = s + enc->minBytesPerChar; + tag->rawNameLength = XmlNameLength(enc, tag->rawName); + ++tagLevel; + { + const char *rawNameEnd = tag->rawName + tag->rawNameLength; + const char *fromPtr = tag->rawName; + toPtr = (XML_Char *)tag->buf; + for (;;) { + int bufSize; + int convLen; + const enum XML_Convert_Result convert_res = XmlConvert(enc, + &fromPtr, rawNameEnd, + (ICHAR **)&toPtr, (ICHAR *)tag->bufEnd - 1); + convLen = (int)(toPtr - (XML_Char *)tag->buf); + if ((fromPtr >= rawNameEnd) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE)) { + tag->name.strLen = convLen; + break; + } + bufSize = (int)(tag->bufEnd - tag->buf) << 1; + { + char *temp = (char *)REALLOC(tag->buf, bufSize); + if (temp == NULL) + return XML_ERROR_NO_MEMORY; + tag->buf = temp; + tag->bufEnd = temp + bufSize; + toPtr = (XML_Char *)temp + convLen; + } + } + } + tag->name.str = (XML_Char *)tag->buf; + *toPtr = XML_T('\0'); + result = storeAtts(parser, enc, s, &(tag->name), &(tag->bindings)); + if (result) + return result; + if (startElementHandler) + startElementHandler(handlerArg, tag->name.str, + (const XML_Char **)atts); + else if (defaultHandler) + reportDefault(parser, enc, s, next); + poolClear(&tempPool); + break; + } + case XML_TOK_EMPTY_ELEMENT_NO_ATTS: + /* fall through */ + case XML_TOK_EMPTY_ELEMENT_WITH_ATTS: + { + const char *rawName = s + enc->minBytesPerChar; + enum XML_Error result; + BINDING *bindings = NULL; + XML_Bool noElmHandlers = XML_TRUE; + TAG_NAME name; + name.str = poolStoreString(&tempPool, enc, rawName, + rawName + XmlNameLength(enc, rawName)); + if (!name.str) + return XML_ERROR_NO_MEMORY; + poolFinish(&tempPool); + result = storeAtts(parser, enc, s, &name, &bindings); + if (result != XML_ERROR_NONE) { + freeBindings(parser, bindings); + return result; + } + poolFinish(&tempPool); + if (startElementHandler) { + startElementHandler(handlerArg, name.str, (const XML_Char **)atts); + noElmHandlers = XML_FALSE; + } + if (endElementHandler) { + if (startElementHandler) + *eventPP = *eventEndPP; + endElementHandler(handlerArg, name.str); + noElmHandlers = XML_FALSE; + } + if (noElmHandlers && defaultHandler) + reportDefault(parser, enc, s, next); + poolClear(&tempPool); + freeBindings(parser, bindings); + } + if (tagLevel == 0) + return epilogProcessor(parser, next, end, nextPtr); + break; + case XML_TOK_END_TAG: + if (tagLevel == startTagLevel) + return XML_ERROR_ASYNC_ENTITY; + else { + int len; + const char *rawName; + TAG *tag = tagStack; + tagStack = tag->parent; + tag->parent = freeTagList; + freeTagList = tag; + rawName = s + enc->minBytesPerChar*2; + len = XmlNameLength(enc, rawName); + if (len != tag->rawNameLength + || memcmp(tag->rawName, rawName, len) != 0) { + *eventPP = rawName; + return XML_ERROR_TAG_MISMATCH; + } + --tagLevel; + if (endElementHandler) { + const XML_Char *localPart; + const XML_Char *prefix; + XML_Char *uri; + localPart = tag->name.localPart; + if (ns && localPart) { + /* localPart and prefix may have been overwritten in + tag->name.str, since this points to the binding->uri + buffer which gets re-used; so we have to add them again + */ + uri = (XML_Char *)tag->name.str + tag->name.uriLen; + /* don't need to check for space - already done in storeAtts() */ + while (*localPart) *uri++ = *localPart++; + prefix = (XML_Char *)tag->name.prefix; + if (ns_triplets && prefix) { + *uri++ = namespaceSeparator; + while (*prefix) *uri++ = *prefix++; + } + *uri = XML_T('\0'); + } + endElementHandler(handlerArg, tag->name.str); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + while (tag->bindings) { + BINDING *b = tag->bindings; + if (endNamespaceDeclHandler) + endNamespaceDeclHandler(handlerArg, b->prefix->name); + tag->bindings = tag->bindings->nextTagBinding; + b->nextTagBinding = freeBindingList; + freeBindingList = b; + b->prefix->binding = b->prevPrefixBinding; + } + if (tagLevel == 0) + return epilogProcessor(parser, next, end, nextPtr); + } + break; + case XML_TOK_CHAR_REF: + { + int n = XmlCharRefNumber(enc, s); + if (n < 0) + return XML_ERROR_BAD_CHAR_REF; + if (characterDataHandler) { + XML_Char buf[XML_ENCODE_MAX]; + characterDataHandler(handlerArg, buf, XmlEncode(n, (ICHAR *)buf)); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + } + break; + case XML_TOK_XML_DECL: + return XML_ERROR_MISPLACED_XML_PI; + case XML_TOK_DATA_NEWLINE: + if (characterDataHandler) { + XML_Char c = 0xA; + characterDataHandler(handlerArg, &c, 1); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + case XML_TOK_CDATA_SECT_OPEN: + { + enum XML_Error result; + if (startCdataSectionHandler) + startCdataSectionHandler(handlerArg); +#if 0 + /* Suppose you doing a transformation on a document that involves + changing only the character data. You set up a defaultHandler + and a characterDataHandler. The defaultHandler simply copies + characters through. The characterDataHandler does the + transformation and writes the characters out escaping them as + necessary. This case will fail to work if we leave out the + following two lines (because & and < inside CDATA sections will + be incorrectly escaped). + + However, now we have a start/endCdataSectionHandler, so it seems + easier to let the user deal with this. + */ + else if (characterDataHandler) + characterDataHandler(handlerArg, dataBuf, 0); +#endif + else if (defaultHandler) + reportDefault(parser, enc, s, next); + result = doCdataSection(parser, enc, &next, end, nextPtr, haveMore); + if (result != XML_ERROR_NONE) + return result; + else if (!next) { + processor = cdataSectionProcessor; + return result; + } + } + break; + case XML_TOK_TRAILING_RSQB: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + if (characterDataHandler) { + if (MUST_CONVERT(enc, s)) { + ICHAR *dataPtr = (ICHAR *)dataBuf; + XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd); + characterDataHandler(handlerArg, dataBuf, + (int)(dataPtr - (ICHAR *)dataBuf)); + } + else + characterDataHandler(handlerArg, + (XML_Char *)s, + (int)((XML_Char *)end - (XML_Char *)s)); + } + else if (defaultHandler) + reportDefault(parser, enc, s, end); + /* We are at the end of the final buffer, should we check for + XML_SUSPENDED, XML_FINISHED? + */ + if (startTagLevel == 0) { + *eventPP = end; + return XML_ERROR_NO_ELEMENTS; + } + if (tagLevel != startTagLevel) { + *eventPP = end; + return XML_ERROR_ASYNC_ENTITY; + } + *nextPtr = end; + return XML_ERROR_NONE; + case XML_TOK_DATA_CHARS: + { + XML_CharacterDataHandler charDataHandler = characterDataHandler; + if (charDataHandler) { + if (MUST_CONVERT(enc, s)) { + for (;;) { + ICHAR *dataPtr = (ICHAR *)dataBuf; + const enum XML_Convert_Result convert_res = XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); + *eventEndPP = s; + charDataHandler(handlerArg, dataBuf, + (int)(dataPtr - (ICHAR *)dataBuf)); + if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE)) + break; + *eventPP = s; + } + } + else + charDataHandler(handlerArg, + (XML_Char *)s, + (int)((XML_Char *)next - (XML_Char *)s)); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + } + break; + case XML_TOK_PI: + if (!reportProcessingInstruction(parser, enc, s, next)) + return XML_ERROR_NO_MEMORY; + break; + case XML_TOK_COMMENT: + if (!reportComment(parser, enc, s, next)) + return XML_ERROR_NO_MEMORY; + break; + default: + if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + } + *eventPP = s = next; + switch (ps_parsing) { + case XML_SUSPENDED: + *nextPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: ; + } + } + /* not reached */ +} + +/* This function does not call free() on the allocated memory, merely + * moving it to the parser's freeBindingList where it can be freed or + * reused as appropriate. + */ +static void +freeBindings(XML_Parser parser, BINDING *bindings) +{ + while (bindings) { + BINDING *b = bindings; + + /* startNamespaceDeclHandler will have been called for this + * binding in addBindings(), so call the end handler now. + */ + if (endNamespaceDeclHandler) + endNamespaceDeclHandler(handlerArg, b->prefix->name); + + bindings = bindings->nextTagBinding; + b->nextTagBinding = freeBindingList; + freeBindingList = b; + b->prefix->binding = b->prevPrefixBinding; + } +} + +/* Precondition: all arguments must be non-NULL; + Purpose: + - normalize attributes + - check attributes for well-formedness + - generate namespace aware attribute names (URI, prefix) + - build list of attributes for startElementHandler + - default attributes + - process namespace declarations (check and report them) + - generate namespace aware element name (URI, prefix) +*/ +static enum XML_Error +storeAtts(XML_Parser parser, const ENCODING *enc, + const char *attStr, TAG_NAME *tagNamePtr, + BINDING **bindingsPtr) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + ELEMENT_TYPE *elementType; + int nDefaultAtts; + const XML_Char **appAtts; /* the attribute list for the application */ + int attIndex = 0; + int prefixLen; + int i; + int n; + XML_Char *uri; + int nPrefixes = 0; + BINDING *binding; + const XML_Char *localPart; + + /* lookup the element type name */ + elementType = (ELEMENT_TYPE *)lookup(parser, &dtd->elementTypes, tagNamePtr->str,0); + if (!elementType) { + const XML_Char *name = poolCopyString(&dtd->pool, tagNamePtr->str); + if (!name) + return XML_ERROR_NO_MEMORY; + elementType = (ELEMENT_TYPE *)lookup(parser, &dtd->elementTypes, name, + sizeof(ELEMENT_TYPE)); + if (!elementType) + return XML_ERROR_NO_MEMORY; + if (ns && !setElementTypePrefix(parser, elementType)) + return XML_ERROR_NO_MEMORY; + } + nDefaultAtts = elementType->nDefaultAtts; + + /* get the attributes from the tokenizer */ + n = XmlGetAttributes(enc, attStr, attsSize, atts); + if (n + nDefaultAtts > attsSize) { + int oldAttsSize = attsSize; + ATTRIBUTE *temp; +#ifdef XML_ATTR_INFO + XML_AttrInfo *temp2; +#endif + attsSize = n + nDefaultAtts + INIT_ATTS_SIZE; + temp = (ATTRIBUTE *)REALLOC((void *)atts, attsSize * sizeof(ATTRIBUTE)); + if (temp == NULL) + return XML_ERROR_NO_MEMORY; + atts = temp; +#ifdef XML_ATTR_INFO + temp2 = (XML_AttrInfo *)REALLOC((void *)attInfo, attsSize * sizeof(XML_AttrInfo)); + if (temp2 == NULL) + return XML_ERROR_NO_MEMORY; + attInfo = temp2; +#endif + if (n > oldAttsSize) + XmlGetAttributes(enc, attStr, n, atts); + } + + appAtts = (const XML_Char **)atts; + for (i = 0; i < n; i++) { + ATTRIBUTE *currAtt = &atts[i]; +#ifdef XML_ATTR_INFO + XML_AttrInfo *currAttInfo = &attInfo[i]; +#endif + /* add the name and value to the attribute list */ + ATTRIBUTE_ID *attId = getAttributeId(parser, enc, currAtt->name, + currAtt->name + + XmlNameLength(enc, currAtt->name)); + if (!attId) + return XML_ERROR_NO_MEMORY; +#ifdef XML_ATTR_INFO + currAttInfo->nameStart = parseEndByteIndex - (parseEndPtr - currAtt->name); + currAttInfo->nameEnd = currAttInfo->nameStart + + XmlNameLength(enc, currAtt->name); + currAttInfo->valueStart = parseEndByteIndex - + (parseEndPtr - currAtt->valuePtr); + currAttInfo->valueEnd = parseEndByteIndex - (parseEndPtr - currAtt->valueEnd); +#endif + /* Detect duplicate attributes by their QNames. This does not work when + namespace processing is turned on and different prefixes for the same + namespace are used. For this case we have a check further down. + */ + if ((attId->name)[-1]) { + if (enc == encoding) + eventPtr = atts[i].name; + return XML_ERROR_DUPLICATE_ATTRIBUTE; + } + (attId->name)[-1] = 1; + appAtts[attIndex++] = attId->name; + if (!atts[i].normalized) { + enum XML_Error result; + XML_Bool isCdata = XML_TRUE; + + /* figure out whether declared as other than CDATA */ + if (attId->maybeTokenized) { + int j; + for (j = 0; j < nDefaultAtts; j++) { + if (attId == elementType->defaultAtts[j].id) { + isCdata = elementType->defaultAtts[j].isCdata; + break; + } + } + } + + /* normalize the attribute value */ + result = storeAttributeValue(parser, enc, isCdata, + atts[i].valuePtr, atts[i].valueEnd, + &tempPool); + if (result) + return result; + appAtts[attIndex] = poolStart(&tempPool); + poolFinish(&tempPool); + } + else { + /* the value did not need normalizing */ + appAtts[attIndex] = poolStoreString(&tempPool, enc, atts[i].valuePtr, + atts[i].valueEnd); + if (appAtts[attIndex] == 0) + return XML_ERROR_NO_MEMORY; + poolFinish(&tempPool); + } + /* handle prefixed attribute names */ + if (attId->prefix) { + if (attId->xmlns) { + /* deal with namespace declarations here */ + enum XML_Error result = addBinding(parser, attId->prefix, attId, + appAtts[attIndex], bindingsPtr); + if (result) + return result; + --attIndex; + } + else { + /* deal with other prefixed names later */ + attIndex++; + nPrefixes++; + (attId->name)[-1] = 2; + } + } + else + attIndex++; + } + + /* set-up for XML_GetSpecifiedAttributeCount and XML_GetIdAttributeIndex */ + nSpecifiedAtts = attIndex; + if (elementType->idAtt && (elementType->idAtt->name)[-1]) { + for (i = 0; i < attIndex; i += 2) + if (appAtts[i] == elementType->idAtt->name) { + idAttIndex = i; + break; + } + } + else + idAttIndex = -1; + + /* do attribute defaulting */ + for (i = 0; i < nDefaultAtts; i++) { + const DEFAULT_ATTRIBUTE *da = elementType->defaultAtts + i; + if (!(da->id->name)[-1] && da->value) { + if (da->id->prefix) { + if (da->id->xmlns) { + enum XML_Error result = addBinding(parser, da->id->prefix, da->id, + da->value, bindingsPtr); + if (result) + return result; + } + else { + (da->id->name)[-1] = 2; + nPrefixes++; + appAtts[attIndex++] = da->id->name; + appAtts[attIndex++] = da->value; + } + } + else { + (da->id->name)[-1] = 1; + appAtts[attIndex++] = da->id->name; + appAtts[attIndex++] = da->value; + } + } + } + appAtts[attIndex] = 0; + + /* expand prefixed attribute names, check for duplicates, + and clear flags that say whether attributes were specified */ + i = 0; + if (nPrefixes) { + int j; /* hash table index */ + unsigned long version = nsAttsVersion; + int nsAttsSize = (int)1 << nsAttsPower; + /* size of hash table must be at least 2 * (# of prefixed attributes) */ + if ((nPrefixes << 1) >> nsAttsPower) { /* true for nsAttsPower = 0 */ + NS_ATT *temp; + /* hash table size must also be a power of 2 and >= 8 */ + while (nPrefixes >> nsAttsPower++); + if (nsAttsPower < 3) + nsAttsPower = 3; + nsAttsSize = (int)1 << nsAttsPower; + temp = (NS_ATT *)REALLOC(nsAtts, nsAttsSize * sizeof(NS_ATT)); + if (!temp) + return XML_ERROR_NO_MEMORY; + nsAtts = temp; + version = 0; /* force re-initialization of nsAtts hash table */ + } + /* using a version flag saves us from initializing nsAtts every time */ + if (!version) { /* initialize version flags when version wraps around */ + version = INIT_ATTS_VERSION; + for (j = nsAttsSize; j != 0; ) + nsAtts[--j].version = version; + } + nsAttsVersion = --version; + + /* expand prefixed names and check for duplicates */ + for (; i < attIndex; i += 2) { + const XML_Char *s = appAtts[i]; + if (s[-1] == 2) { /* prefixed */ + ATTRIBUTE_ID *id; + const BINDING *b; + unsigned long uriHash; + struct siphash sip_state; + struct sipkey sip_key; + + copy_salt_to_sipkey(parser, &sip_key); + sip24_init(&sip_state, &sip_key); + + ((XML_Char *)s)[-1] = 0; /* clear flag */ + id = (ATTRIBUTE_ID *)lookup(parser, &dtd->attributeIds, s, 0); + if (!id || !id->prefix) + return XML_ERROR_NO_MEMORY; + b = id->prefix->binding; + if (!b) + return XML_ERROR_UNBOUND_PREFIX; + + for (j = 0; j < b->uriLen; j++) { + const XML_Char c = b->uri[j]; + if (!poolAppendChar(&tempPool, c)) + return XML_ERROR_NO_MEMORY; + } + + sip24_update(&sip_state, b->uri, b->uriLen * sizeof(XML_Char)); + + while (*s++ != XML_T(ASCII_COLON)) + ; + + sip24_update(&sip_state, s, keylen(s) * sizeof(XML_Char)); + + do { /* copies null terminator */ + if (!poolAppendChar(&tempPool, *s)) + return XML_ERROR_NO_MEMORY; + } while (*s++); + + uriHash = (unsigned long)sip24_final(&sip_state); + + { /* Check hash table for duplicate of expanded name (uriName). + Derived from code in lookup(parser, HASH_TABLE *table, ...). + */ + unsigned char step = 0; + unsigned long mask = nsAttsSize - 1; + j = uriHash & mask; /* index into hash table */ + while (nsAtts[j].version == version) { + /* for speed we compare stored hash values first */ + if (uriHash == nsAtts[j].hash) { + const XML_Char *s1 = poolStart(&tempPool); + const XML_Char *s2 = nsAtts[j].uriName; + /* s1 is null terminated, but not s2 */ + for (; *s1 == *s2 && *s1 != 0; s1++, s2++); + if (*s1 == 0) + return XML_ERROR_DUPLICATE_ATTRIBUTE; + } + if (!step) + step = PROBE_STEP(uriHash, mask, nsAttsPower); + j < step ? (j += nsAttsSize - step) : (j -= step); + } + } + + if (ns_triplets) { /* append namespace separator and prefix */ + tempPool.ptr[-1] = namespaceSeparator; + s = b->prefix->name; + do { + if (!poolAppendChar(&tempPool, *s)) + return XML_ERROR_NO_MEMORY; + } while (*s++); + } + + /* store expanded name in attribute list */ + s = poolStart(&tempPool); + poolFinish(&tempPool); + appAtts[i] = s; + + /* fill empty slot with new version, uriName and hash value */ + nsAtts[j].version = version; + nsAtts[j].hash = uriHash; + nsAtts[j].uriName = s; + + if (!--nPrefixes) { + i += 2; + break; + } + } + else /* not prefixed */ + ((XML_Char *)s)[-1] = 0; /* clear flag */ + } + } + /* clear flags for the remaining attributes */ + for (; i < attIndex; i += 2) + ((XML_Char *)(appAtts[i]))[-1] = 0; + for (binding = *bindingsPtr; binding; binding = binding->nextTagBinding) + binding->attId->name[-1] = 0; + + if (!ns) + return XML_ERROR_NONE; + + /* expand the element type name */ + if (elementType->prefix) { + binding = elementType->prefix->binding; + if (!binding) + return XML_ERROR_UNBOUND_PREFIX; + localPart = tagNamePtr->str; + while (*localPart++ != XML_T(ASCII_COLON)) + ; + } + else if (dtd->defaultPrefix.binding) { + binding = dtd->defaultPrefix.binding; + localPart = tagNamePtr->str; + } + else + return XML_ERROR_NONE; + prefixLen = 0; + if (ns_triplets && binding->prefix->name) { + for (; binding->prefix->name[prefixLen++];) + ; /* prefixLen includes null terminator */ + } + tagNamePtr->localPart = localPart; + tagNamePtr->uriLen = binding->uriLen; + tagNamePtr->prefix = binding->prefix->name; + tagNamePtr->prefixLen = prefixLen; + for (i = 0; localPart[i++];) + ; /* i includes null terminator */ + n = i + binding->uriLen + prefixLen; + if (n > binding->uriAlloc) { + TAG *p; + uri = (XML_Char *)MALLOC((n + EXPAND_SPARE) * sizeof(XML_Char)); + if (!uri) + return XML_ERROR_NO_MEMORY; + binding->uriAlloc = n + EXPAND_SPARE; + memcpy(uri, binding->uri, binding->uriLen * sizeof(XML_Char)); + for (p = tagStack; p; p = p->parent) + if (p->name.str == binding->uri) + p->name.str = uri; + FREE(binding->uri); + binding->uri = uri; + } + /* if namespaceSeparator != '\0' then uri includes it already */ + uri = binding->uri + binding->uriLen; + memcpy(uri, localPart, i * sizeof(XML_Char)); + /* we always have a namespace separator between localPart and prefix */ + if (prefixLen) { + uri += i - 1; + *uri = namespaceSeparator; /* replace null terminator */ + memcpy(uri + 1, binding->prefix->name, prefixLen * sizeof(XML_Char)); + } + tagNamePtr->str = binding->uri; + return XML_ERROR_NONE; +} + +/* addBinding() overwrites the value of prefix->binding without checking. + Therefore one must keep track of the old value outside of addBinding(). +*/ +static enum XML_Error +addBinding(XML_Parser parser, PREFIX *prefix, const ATTRIBUTE_ID *attId, + const XML_Char *uri, BINDING **bindingsPtr) +{ + static const XML_Char xmlNamespace[] = { + ASCII_h, ASCII_t, ASCII_t, ASCII_p, ASCII_COLON, ASCII_SLASH, ASCII_SLASH, + ASCII_w, ASCII_w, ASCII_w, ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, + ASCII_o, ASCII_r, ASCII_g, ASCII_SLASH, ASCII_X, ASCII_M, ASCII_L, + ASCII_SLASH, ASCII_1, ASCII_9, ASCII_9, ASCII_8, ASCII_SLASH, + ASCII_n, ASCII_a, ASCII_m, ASCII_e, ASCII_s, ASCII_p, ASCII_a, ASCII_c, + ASCII_e, '\0' + }; + static const int xmlLen = + (int)sizeof(xmlNamespace)/sizeof(XML_Char) - 1; + static const XML_Char xmlnsNamespace[] = { + ASCII_h, ASCII_t, ASCII_t, ASCII_p, ASCII_COLON, ASCII_SLASH, ASCII_SLASH, + ASCII_w, ASCII_w, ASCII_w, ASCII_PERIOD, ASCII_w, ASCII_3, ASCII_PERIOD, + ASCII_o, ASCII_r, ASCII_g, ASCII_SLASH, ASCII_2, ASCII_0, ASCII_0, + ASCII_0, ASCII_SLASH, ASCII_x, ASCII_m, ASCII_l, ASCII_n, ASCII_s, + ASCII_SLASH, '\0' + }; + static const int xmlnsLen = + (int)sizeof(xmlnsNamespace)/sizeof(XML_Char) - 1; + + XML_Bool mustBeXML = XML_FALSE; + XML_Bool isXML = XML_TRUE; + XML_Bool isXMLNS = XML_TRUE; + + BINDING *b; + int len; + + /* empty URI is only valid for default namespace per XML NS 1.0 (not 1.1) */ + if (*uri == XML_T('\0') && prefix->name) + return XML_ERROR_UNDECLARING_PREFIX; + + if (prefix->name + && prefix->name[0] == XML_T(ASCII_x) + && prefix->name[1] == XML_T(ASCII_m) + && prefix->name[2] == XML_T(ASCII_l)) { + + /* Not allowed to bind xmlns */ + if (prefix->name[3] == XML_T(ASCII_n) + && prefix->name[4] == XML_T(ASCII_s) + && prefix->name[5] == XML_T('\0')) + return XML_ERROR_RESERVED_PREFIX_XMLNS; + + if (prefix->name[3] == XML_T('\0')) + mustBeXML = XML_TRUE; + } + + for (len = 0; uri[len]; len++) { + if (isXML && (len > xmlLen || uri[len] != xmlNamespace[len])) + isXML = XML_FALSE; + + if (!mustBeXML && isXMLNS + && (len > xmlnsLen || uri[len] != xmlnsNamespace[len])) + isXMLNS = XML_FALSE; + } + isXML = isXML && len == xmlLen; + isXMLNS = isXMLNS && len == xmlnsLen; + + if (mustBeXML != isXML) + return mustBeXML ? XML_ERROR_RESERVED_PREFIX_XML + : XML_ERROR_RESERVED_NAMESPACE_URI; + + if (isXMLNS) + return XML_ERROR_RESERVED_NAMESPACE_URI; + + if (namespaceSeparator) + len++; + if (freeBindingList) { + b = freeBindingList; + if (len > b->uriAlloc) { + XML_Char *temp = (XML_Char *)REALLOC(b->uri, + sizeof(XML_Char) * (len + EXPAND_SPARE)); + if (temp == NULL) + return XML_ERROR_NO_MEMORY; + b->uri = temp; + b->uriAlloc = len + EXPAND_SPARE; + } + freeBindingList = b->nextTagBinding; + } + else { + b = (BINDING *)MALLOC(sizeof(BINDING)); + if (!b) + return XML_ERROR_NO_MEMORY; + b->uri = (XML_Char *)MALLOC(sizeof(XML_Char) * (len + EXPAND_SPARE)); + if (!b->uri) { + FREE(b); + return XML_ERROR_NO_MEMORY; + } + b->uriAlloc = len + EXPAND_SPARE; + } + b->uriLen = len; + memcpy(b->uri, uri, len * sizeof(XML_Char)); + if (namespaceSeparator) + b->uri[len - 1] = namespaceSeparator; + b->prefix = prefix; + b->attId = attId; + b->prevPrefixBinding = prefix->binding; + /* NULL binding when default namespace undeclared */ + if (*uri == XML_T('\0') && prefix == &_dtd->defaultPrefix) + prefix->binding = NULL; + else + prefix->binding = b; + b->nextTagBinding = *bindingsPtr; + *bindingsPtr = b; + /* if attId == NULL then we are not starting a namespace scope */ + if (attId && startNamespaceDeclHandler) + startNamespaceDeclHandler(handlerArg, prefix->name, + prefix->binding ? uri : 0); + return XML_ERROR_NONE; +} + +/* The idea here is to avoid using stack for each CDATA section when + the whole file is parsed with one call. +*/ +static enum XML_Error PTRCALL +cdataSectionProcessor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + enum XML_Error result = doCdataSection(parser, encoding, &start, end, + endPtr, (XML_Bool)!ps_finalBuffer); + if (result != XML_ERROR_NONE) + return result; + if (start) { + if (parentParser) { /* we are parsing an external entity */ + processor = externalEntityContentProcessor; + return externalEntityContentProcessor(parser, start, end, endPtr); + } + else { + processor = contentProcessor; + return contentProcessor(parser, start, end, endPtr); + } + } + return result; +} + +/* startPtr gets set to non-null if the section is closed, and to null if + the section is not yet closed. +*/ +static enum XML_Error +doCdataSection(XML_Parser parser, + const ENCODING *enc, + const char **startPtr, + const char *end, + const char **nextPtr, + XML_Bool haveMore) +{ + const char *s = *startPtr; + const char **eventPP; + const char **eventEndPP; + if (enc == encoding) { + eventPP = &eventPtr; + *eventPP = s; + eventEndPP = &eventEndPtr; + } + else { + eventPP = &(openInternalEntities->internalEventPtr); + eventEndPP = &(openInternalEntities->internalEventEndPtr); + } + *eventPP = s; + *startPtr = NULL; + + for (;;) { + const char *next; + int tok = XmlCdataSectionTok(enc, s, end, &next); + *eventEndPP = next; + switch (tok) { + case XML_TOK_CDATA_SECT_CLOSE: + if (endCdataSectionHandler) + endCdataSectionHandler(handlerArg); +#if 0 + /* see comment under XML_TOK_CDATA_SECT_OPEN */ + else if (characterDataHandler) + characterDataHandler(handlerArg, dataBuf, 0); +#endif + else if (defaultHandler) + reportDefault(parser, enc, s, next); + *startPtr = next; + *nextPtr = next; + if (ps_parsing == XML_FINISHED) + return XML_ERROR_ABORTED; + else + return XML_ERROR_NONE; + case XML_TOK_DATA_NEWLINE: + if (characterDataHandler) { + XML_Char c = 0xA; + characterDataHandler(handlerArg, &c, 1); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + break; + case XML_TOK_DATA_CHARS: + { + XML_CharacterDataHandler charDataHandler = characterDataHandler; + if (charDataHandler) { + if (MUST_CONVERT(enc, s)) { + for (;;) { + ICHAR *dataPtr = (ICHAR *)dataBuf; + const enum XML_Convert_Result convert_res = XmlConvert(enc, &s, next, &dataPtr, (ICHAR *)dataBufEnd); + *eventEndPP = next; + charDataHandler(handlerArg, dataBuf, + (int)(dataPtr - (ICHAR *)dataBuf)); + if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE)) + break; + *eventPP = s; + } + } + else + charDataHandler(handlerArg, + (XML_Char *)s, + (int)((XML_Char *)next - (XML_Char *)s)); + } + else if (defaultHandler) + reportDefault(parser, enc, s, next); + } + break; + case XML_TOK_INVALID: + *eventPP = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_PARTIAL_CHAR; + case XML_TOK_PARTIAL: + case XML_TOK_NONE: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_UNCLOSED_CDATA_SECTION; + default: + *eventPP = next; + return XML_ERROR_UNEXPECTED_STATE; + } + + *eventPP = s = next; + switch (ps_parsing) { + case XML_SUSPENDED: + *nextPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: ; + } + } + /* not reached */ +} + +#ifdef XML_DTD + +/* The idea here is to avoid using stack for each IGNORE section when + the whole file is parsed with one call. +*/ +static enum XML_Error PTRCALL +ignoreSectionProcessor(XML_Parser parser, + const char *start, + const char *end, + const char **endPtr) +{ + enum XML_Error result = doIgnoreSection(parser, encoding, &start, end, + endPtr, (XML_Bool)!ps_finalBuffer); + if (result != XML_ERROR_NONE) + return result; + if (start) { + processor = prologProcessor; + return prologProcessor(parser, start, end, endPtr); + } + return result; +} + +/* startPtr gets set to non-null is the section is closed, and to null + if the section is not yet closed. +*/ +static enum XML_Error +doIgnoreSection(XML_Parser parser, + const ENCODING *enc, + const char **startPtr, + const char *end, + const char **nextPtr, + XML_Bool haveMore) +{ + const char *next; + int tok; + const char *s = *startPtr; + const char **eventPP; + const char **eventEndPP; + if (enc == encoding) { + eventPP = &eventPtr; + *eventPP = s; + eventEndPP = &eventEndPtr; + } + else { + eventPP = &(openInternalEntities->internalEventPtr); + eventEndPP = &(openInternalEntities->internalEventEndPtr); + } + *eventPP = s; + *startPtr = NULL; + tok = XmlIgnoreSectionTok(enc, s, end, &next); + *eventEndPP = next; + switch (tok) { + case XML_TOK_IGNORE_SECT: + if (defaultHandler) + reportDefault(parser, enc, s, next); + *startPtr = next; + *nextPtr = next; + if (ps_parsing == XML_FINISHED) + return XML_ERROR_ABORTED; + else + return XML_ERROR_NONE; + case XML_TOK_INVALID: + *eventPP = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_PARTIAL_CHAR; + case XML_TOK_PARTIAL: + case XML_TOK_NONE: + if (haveMore) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_SYNTAX; /* XML_ERROR_UNCLOSED_IGNORE_SECTION */ + default: + *eventPP = next; + return XML_ERROR_UNEXPECTED_STATE; + } + /* not reached */ +} + +#endif /* XML_DTD */ + +static enum XML_Error +initializeEncoding(XML_Parser parser) +{ + const char *s; +#ifdef XML_UNICODE + char encodingBuf[128]; + if (!protocolEncodingName) + s = NULL; + else { + int i; + for (i = 0; protocolEncodingName[i]; i++) { + if (i == sizeof(encodingBuf) - 1 + || (protocolEncodingName[i] & ~0x7f) != 0) { + encodingBuf[0] = '\0'; + break; + } + encodingBuf[i] = (char)protocolEncodingName[i]; + } + encodingBuf[i] = '\0'; + s = encodingBuf; + } +#else + s = protocolEncodingName; +#endif + if ((ns ? XmlInitEncodingNS : XmlInitEncoding)(&initEncoding, &encoding, s)) + return XML_ERROR_NONE; + return handleUnknownEncoding(parser, protocolEncodingName); +} + +static enum XML_Error +processXmlDecl(XML_Parser parser, int isGeneralTextEntity, + const char *s, const char *next) +{ + const char *encodingName = NULL; + const XML_Char *storedEncName = NULL; + const ENCODING *newEncoding = NULL; + const char *version = NULL; + const char *versionend; + const XML_Char *storedversion = NULL; + int standalone = -1; + if (!(ns + ? XmlParseXmlDeclNS + : XmlParseXmlDecl)(isGeneralTextEntity, + encoding, + s, + next, + &eventPtr, + &version, + &versionend, + &encodingName, + &newEncoding, + &standalone)) { + if (isGeneralTextEntity) + return XML_ERROR_TEXT_DECL; + else + return XML_ERROR_XML_DECL; + } + if (!isGeneralTextEntity && standalone == 1) { + _dtd->standalone = XML_TRUE; +#ifdef XML_DTD + if (paramEntityParsing == XML_PARAM_ENTITY_PARSING_UNLESS_STANDALONE) + paramEntityParsing = XML_PARAM_ENTITY_PARSING_NEVER; +#endif /* XML_DTD */ + } + if (xmlDeclHandler) { + if (encodingName != NULL) { + storedEncName = poolStoreString(&temp2Pool, + encoding, + encodingName, + encodingName + + XmlNameLength(encoding, encodingName)); + if (!storedEncName) + return XML_ERROR_NO_MEMORY; + poolFinish(&temp2Pool); + } + if (version) { + storedversion = poolStoreString(&temp2Pool, + encoding, + version, + versionend - encoding->minBytesPerChar); + if (!storedversion) + return XML_ERROR_NO_MEMORY; + } + xmlDeclHandler(handlerArg, storedversion, storedEncName, standalone); + } + else if (defaultHandler) + reportDefault(parser, encoding, s, next); + if (protocolEncodingName == NULL) { + if (newEncoding) { + if (newEncoding->minBytesPerChar != encoding->minBytesPerChar) { + eventPtr = encodingName; + return XML_ERROR_INCORRECT_ENCODING; + } + encoding = newEncoding; + } + else if (encodingName) { + enum XML_Error result; + if (!storedEncName) { + storedEncName = poolStoreString( + &temp2Pool, encoding, encodingName, + encodingName + XmlNameLength(encoding, encodingName)); + if (!storedEncName) + return XML_ERROR_NO_MEMORY; + } + result = handleUnknownEncoding(parser, storedEncName); + poolClear(&temp2Pool); + if (result == XML_ERROR_UNKNOWN_ENCODING) + eventPtr = encodingName; + return result; + } + } + + if (storedEncName || storedversion) + poolClear(&temp2Pool); + + return XML_ERROR_NONE; +} + +static enum XML_Error +handleUnknownEncoding(XML_Parser parser, const XML_Char *encodingName) +{ + if (unknownEncodingHandler) { + XML_Encoding info; + int i; + for (i = 0; i < 256; i++) + info.map[i] = -1; + info.convert = NULL; + info.data = NULL; + info.release = NULL; + if (unknownEncodingHandler(unknownEncodingHandlerData, encodingName, + &info)) { + ENCODING *enc; + unknownEncodingMem = MALLOC(XmlSizeOfUnknownEncoding()); + if (!unknownEncodingMem) { + if (info.release) + info.release(info.data); + return XML_ERROR_NO_MEMORY; + } + enc = (ns + ? XmlInitUnknownEncodingNS + : XmlInitUnknownEncoding)(unknownEncodingMem, + info.map, + info.convert, + info.data); + if (enc) { + unknownEncodingData = info.data; + unknownEncodingRelease = info.release; + encoding = enc; + return XML_ERROR_NONE; + } + } + if (info.release != NULL) + info.release(info.data); + } + return XML_ERROR_UNKNOWN_ENCODING; +} + +static enum XML_Error PTRCALL +prologInitProcessor(XML_Parser parser, + const char *s, + const char *end, + const char **nextPtr) +{ + enum XML_Error result = initializeEncoding(parser); + if (result != XML_ERROR_NONE) + return result; + processor = prologProcessor; + return prologProcessor(parser, s, end, nextPtr); +} + +#ifdef XML_DTD + +static enum XML_Error PTRCALL +externalParEntInitProcessor(XML_Parser parser, + const char *s, + const char *end, + const char **nextPtr) +{ + enum XML_Error result = initializeEncoding(parser); + if (result != XML_ERROR_NONE) + return result; + + /* we know now that XML_Parse(Buffer) has been called, + so we consider the external parameter entity read */ + _dtd->paramEntityRead = XML_TRUE; + + if (prologState.inEntityValue) { + processor = entityValueInitProcessor; + return entityValueInitProcessor(parser, s, end, nextPtr); + } + else { + processor = externalParEntProcessor; + return externalParEntProcessor(parser, s, end, nextPtr); + } +} + +static enum XML_Error PTRCALL +entityValueInitProcessor(XML_Parser parser, + const char *s, + const char *end, + const char **nextPtr) +{ + int tok; + const char *start = s; + const char *next = start; + eventPtr = start; + + for (;;) { + tok = XmlPrologTok(encoding, start, end, &next); + eventEndPtr = next; + if (tok <= 0) { + if (!ps_finalBuffer && tok != XML_TOK_INVALID) { + *nextPtr = s; + return XML_ERROR_NONE; + } + switch (tok) { + case XML_TOK_INVALID: + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL: + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + return XML_ERROR_PARTIAL_CHAR; + case XML_TOK_NONE: /* start == end */ + default: + break; + } + /* found end of entity value - can store it now */ + return storeEntityValue(parser, encoding, s, end); + } + else if (tok == XML_TOK_XML_DECL) { + enum XML_Error result; + result = processXmlDecl(parser, 0, start, next); + if (result != XML_ERROR_NONE) + return result; + switch (ps_parsing) { + case XML_SUSPENDED: + *nextPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: + *nextPtr = next; + } + /* stop scanning for text declaration - we found one */ + processor = entityValueProcessor; + return entityValueProcessor(parser, next, end, nextPtr); + } + /* If we are at the end of the buffer, this would cause XmlPrologTok to + return XML_TOK_NONE on the next call, which would then cause the + function to exit with *nextPtr set to s - that is what we want for other + tokens, but not for the BOM - we would rather like to skip it; + then, when this routine is entered the next time, XmlPrologTok will + return XML_TOK_INVALID, since the BOM is still in the buffer + */ + else if (tok == XML_TOK_BOM && next == end && !ps_finalBuffer) { + *nextPtr = next; + return XML_ERROR_NONE; + } + /* If we get this token, we have the start of what might be a + normal tag, but not a declaration (i.e. it doesn't begin with + "internalEventPtr); + eventEndPP = &(openInternalEntities->internalEventEndPtr); + } + + for (;;) { + int role; + XML_Bool handleDefault = XML_TRUE; + *eventPP = s; + *eventEndPP = next; + if (tok <= 0) { + if (haveMore && tok != XML_TOK_INVALID) { + *nextPtr = s; + return XML_ERROR_NONE; + } + switch (tok) { + case XML_TOK_INVALID: + *eventPP = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL: + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + return XML_ERROR_PARTIAL_CHAR; + case -XML_TOK_PROLOG_S: + tok = -tok; + break; + case XML_TOK_NONE: +#ifdef XML_DTD + /* for internal PE NOT referenced between declarations */ + if (enc != encoding && !openInternalEntities->betweenDecl) { + *nextPtr = s; + return XML_ERROR_NONE; + } + /* WFC: PE Between Declarations - must check that PE contains + complete markup, not only for external PEs, but also for + internal PEs if the reference occurs between declarations. + */ + if (isParamEntity || enc != encoding) { + if (XmlTokenRole(&prologState, XML_TOK_NONE, end, end, enc) + == XML_ROLE_ERROR) + return XML_ERROR_INCOMPLETE_PE; + *nextPtr = s; + return XML_ERROR_NONE; + } +#endif /* XML_DTD */ + return XML_ERROR_NO_ELEMENTS; + default: + tok = -tok; + next = end; + break; + } + } + role = XmlTokenRole(&prologState, tok, s, next, enc); + switch (role) { + case XML_ROLE_XML_DECL: + { + enum XML_Error result = processXmlDecl(parser, 0, s, next); + if (result != XML_ERROR_NONE) + return result; + enc = encoding; + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_DOCTYPE_NAME: + if (startDoctypeDeclHandler) { + doctypeName = poolStoreString(&tempPool, enc, s, next); + if (!doctypeName) + return XML_ERROR_NO_MEMORY; + poolFinish(&tempPool); + doctypePubid = NULL; + handleDefault = XML_FALSE; + } + doctypeSysid = NULL; /* always initialize to NULL */ + break; + case XML_ROLE_DOCTYPE_INTERNAL_SUBSET: + if (startDoctypeDeclHandler) { + startDoctypeDeclHandler(handlerArg, doctypeName, doctypeSysid, + doctypePubid, 1); + doctypeName = NULL; + poolClear(&tempPool); + handleDefault = XML_FALSE; + } + break; +#ifdef XML_DTD + case XML_ROLE_TEXT_DECL: + { + enum XML_Error result = processXmlDecl(parser, 1, s, next); + if (result != XML_ERROR_NONE) + return result; + enc = encoding; + handleDefault = XML_FALSE; + } + break; +#endif /* XML_DTD */ + case XML_ROLE_DOCTYPE_PUBLIC_ID: +#ifdef XML_DTD + useForeignDTD = XML_FALSE; + declEntity = (ENTITY *)lookup(parser, + &dtd->paramEntities, + externalSubsetName, + sizeof(ENTITY)); + if (!declEntity) + return XML_ERROR_NO_MEMORY; +#endif /* XML_DTD */ + dtd->hasParamEntityRefs = XML_TRUE; + if (startDoctypeDeclHandler) { + XML_Char *pubId; + if (!XmlIsPublicId(enc, s, next, eventPP)) + return XML_ERROR_PUBLICID; + pubId = poolStoreString(&tempPool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!pubId) + return XML_ERROR_NO_MEMORY; + normalizePublicId(pubId); + poolFinish(&tempPool); + doctypePubid = pubId; + handleDefault = XML_FALSE; + goto alreadyChecked; + } + /* fall through */ + case XML_ROLE_ENTITY_PUBLIC_ID: + if (!XmlIsPublicId(enc, s, next, eventPP)) + return XML_ERROR_PUBLICID; + alreadyChecked: + if (dtd->keepProcessing && declEntity) { + XML_Char *tem = poolStoreString(&dtd->pool, + enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!tem) + return XML_ERROR_NO_MEMORY; + normalizePublicId(tem); + declEntity->publicId = tem; + poolFinish(&dtd->pool); + if (entityDeclHandler) + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_DOCTYPE_CLOSE: + if (doctypeName) { + startDoctypeDeclHandler(handlerArg, doctypeName, + doctypeSysid, doctypePubid, 0); + poolClear(&tempPool); + handleDefault = XML_FALSE; + } + /* doctypeSysid will be non-NULL in the case of a previous + XML_ROLE_DOCTYPE_SYSTEM_ID, even if startDoctypeDeclHandler + was not set, indicating an external subset + */ +#ifdef XML_DTD + if (doctypeSysid || useForeignDTD) { + XML_Bool hadParamEntityRefs = dtd->hasParamEntityRefs; + dtd->hasParamEntityRefs = XML_TRUE; + if (paramEntityParsing && externalEntityRefHandler) { + ENTITY *entity = (ENTITY *)lookup(parser, + &dtd->paramEntities, + externalSubsetName, + sizeof(ENTITY)); + if (!entity) + return XML_ERROR_NO_MEMORY; + if (useForeignDTD) + entity->base = curBase; + dtd->paramEntityRead = XML_FALSE; + if (!externalEntityRefHandler(externalEntityRefHandlerArg, + 0, + entity->base, + entity->systemId, + entity->publicId)) + return XML_ERROR_EXTERNAL_ENTITY_HANDLING; + if (dtd->paramEntityRead) { + if (!dtd->standalone && + notStandaloneHandler && + !notStandaloneHandler(handlerArg)) + return XML_ERROR_NOT_STANDALONE; + } + /* if we didn't read the foreign DTD then this means that there + is no external subset and we must reset dtd->hasParamEntityRefs + */ + else if (!doctypeSysid) + dtd->hasParamEntityRefs = hadParamEntityRefs; + /* end of DTD - no need to update dtd->keepProcessing */ + } + useForeignDTD = XML_FALSE; + } +#endif /* XML_DTD */ + if (endDoctypeDeclHandler) { + endDoctypeDeclHandler(handlerArg); + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_INSTANCE_START: +#ifdef XML_DTD + /* if there is no DOCTYPE declaration then now is the + last chance to read the foreign DTD + */ + if (useForeignDTD) { + XML_Bool hadParamEntityRefs = dtd->hasParamEntityRefs; + dtd->hasParamEntityRefs = XML_TRUE; + if (paramEntityParsing && externalEntityRefHandler) { + ENTITY *entity = (ENTITY *)lookup(parser, &dtd->paramEntities, + externalSubsetName, + sizeof(ENTITY)); + if (!entity) + return XML_ERROR_NO_MEMORY; + entity->base = curBase; + dtd->paramEntityRead = XML_FALSE; + if (!externalEntityRefHandler(externalEntityRefHandlerArg, + 0, + entity->base, + entity->systemId, + entity->publicId)) + return XML_ERROR_EXTERNAL_ENTITY_HANDLING; + if (dtd->paramEntityRead) { + if (!dtd->standalone && + notStandaloneHandler && + !notStandaloneHandler(handlerArg)) + return XML_ERROR_NOT_STANDALONE; + } + /* if we didn't read the foreign DTD then this means that there + is no external subset and we must reset dtd->hasParamEntityRefs + */ + else + dtd->hasParamEntityRefs = hadParamEntityRefs; + /* end of DTD - no need to update dtd->keepProcessing */ + } + } +#endif /* XML_DTD */ + processor = contentProcessor; + return contentProcessor(parser, s, end, nextPtr); + case XML_ROLE_ATTLIST_ELEMENT_NAME: + declElementType = getElementType(parser, enc, s, next); + if (!declElementType) + return XML_ERROR_NO_MEMORY; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_NAME: + declAttributeId = getAttributeId(parser, enc, s, next); + if (!declAttributeId) + return XML_ERROR_NO_MEMORY; + declAttributeIsCdata = XML_FALSE; + declAttributeType = NULL; + declAttributeIsId = XML_FALSE; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_CDATA: + declAttributeIsCdata = XML_TRUE; + declAttributeType = atypeCDATA; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_ID: + declAttributeIsId = XML_TRUE; + declAttributeType = atypeID; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_IDREF: + declAttributeType = atypeIDREF; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_IDREFS: + declAttributeType = atypeIDREFS; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_ENTITY: + declAttributeType = atypeENTITY; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_ENTITIES: + declAttributeType = atypeENTITIES; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_NMTOKEN: + declAttributeType = atypeNMTOKEN; + goto checkAttListDeclHandler; + case XML_ROLE_ATTRIBUTE_TYPE_NMTOKENS: + declAttributeType = atypeNMTOKENS; + checkAttListDeclHandler: + if (dtd->keepProcessing && attlistDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_ATTRIBUTE_ENUM_VALUE: + case XML_ROLE_ATTRIBUTE_NOTATION_VALUE: + if (dtd->keepProcessing && attlistDeclHandler) { + const XML_Char *prefix; + if (declAttributeType) { + prefix = enumValueSep; + } + else { + prefix = (role == XML_ROLE_ATTRIBUTE_NOTATION_VALUE + ? notationPrefix + : enumValueStart); + } + if (!poolAppendString(&tempPool, prefix)) + return XML_ERROR_NO_MEMORY; + if (!poolAppend(&tempPool, enc, s, next)) + return XML_ERROR_NO_MEMORY; + declAttributeType = tempPool.start; + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_IMPLIED_ATTRIBUTE_VALUE: + case XML_ROLE_REQUIRED_ATTRIBUTE_VALUE: + if (dtd->keepProcessing) { + if (!defineAttribute(declElementType, declAttributeId, + declAttributeIsCdata, declAttributeIsId, + 0, parser)) + return XML_ERROR_NO_MEMORY; + if (attlistDeclHandler && declAttributeType) { + if (*declAttributeType == XML_T(ASCII_LPAREN) + || (*declAttributeType == XML_T(ASCII_N) + && declAttributeType[1] == XML_T(ASCII_O))) { + /* Enumerated or Notation type */ + if (!poolAppendChar(&tempPool, XML_T(ASCII_RPAREN)) + || !poolAppendChar(&tempPool, XML_T('\0'))) + return XML_ERROR_NO_MEMORY; + declAttributeType = tempPool.start; + poolFinish(&tempPool); + } + *eventEndPP = s; + attlistDeclHandler(handlerArg, declElementType->name, + declAttributeId->name, declAttributeType, + 0, role == XML_ROLE_REQUIRED_ATTRIBUTE_VALUE); + poolClear(&tempPool); + handleDefault = XML_FALSE; + } + } + break; + case XML_ROLE_DEFAULT_ATTRIBUTE_VALUE: + case XML_ROLE_FIXED_ATTRIBUTE_VALUE: + if (dtd->keepProcessing) { + const XML_Char *attVal; + enum XML_Error result = + storeAttributeValue(parser, enc, declAttributeIsCdata, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar, + &dtd->pool); + if (result) + return result; + attVal = poolStart(&dtd->pool); + poolFinish(&dtd->pool); + /* ID attributes aren't allowed to have a default */ + if (!defineAttribute(declElementType, declAttributeId, + declAttributeIsCdata, XML_FALSE, attVal, parser)) + return XML_ERROR_NO_MEMORY; + if (attlistDeclHandler && declAttributeType) { + if (*declAttributeType == XML_T(ASCII_LPAREN) + || (*declAttributeType == XML_T(ASCII_N) + && declAttributeType[1] == XML_T(ASCII_O))) { + /* Enumerated or Notation type */ + if (!poolAppendChar(&tempPool, XML_T(ASCII_RPAREN)) + || !poolAppendChar(&tempPool, XML_T('\0'))) + return XML_ERROR_NO_MEMORY; + declAttributeType = tempPool.start; + poolFinish(&tempPool); + } + *eventEndPP = s; + attlistDeclHandler(handlerArg, declElementType->name, + declAttributeId->name, declAttributeType, + attVal, + role == XML_ROLE_FIXED_ATTRIBUTE_VALUE); + poolClear(&tempPool); + handleDefault = XML_FALSE; + } + } + break; + case XML_ROLE_ENTITY_VALUE: + if (dtd->keepProcessing) { + enum XML_Error result = storeEntityValue(parser, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (declEntity) { + declEntity->textPtr = poolStart(&dtd->entityValuePool); + declEntity->textLen = (int)(poolLength(&dtd->entityValuePool)); + poolFinish(&dtd->entityValuePool); + if (entityDeclHandler) { + *eventEndPP = s; + entityDeclHandler(handlerArg, + declEntity->name, + declEntity->is_param, + declEntity->textPtr, + declEntity->textLen, + curBase, 0, 0, 0); + handleDefault = XML_FALSE; + } + } + else + poolDiscard(&dtd->entityValuePool); + if (result != XML_ERROR_NONE) + return result; + } + break; + case XML_ROLE_DOCTYPE_SYSTEM_ID: +#ifdef XML_DTD + useForeignDTD = XML_FALSE; +#endif /* XML_DTD */ + dtd->hasParamEntityRefs = XML_TRUE; + if (startDoctypeDeclHandler) { + doctypeSysid = poolStoreString(&tempPool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (doctypeSysid == NULL) + return XML_ERROR_NO_MEMORY; + poolFinish(&tempPool); + handleDefault = XML_FALSE; + } +#ifdef XML_DTD + else + /* use externalSubsetName to make doctypeSysid non-NULL + for the case where no startDoctypeDeclHandler is set */ + doctypeSysid = externalSubsetName; +#endif /* XML_DTD */ + if (!dtd->standalone +#ifdef XML_DTD + && !paramEntityParsing +#endif /* XML_DTD */ + && notStandaloneHandler + && !notStandaloneHandler(handlerArg)) + return XML_ERROR_NOT_STANDALONE; +#ifndef XML_DTD + break; +#else /* XML_DTD */ + if (!declEntity) { + declEntity = (ENTITY *)lookup(parser, + &dtd->paramEntities, + externalSubsetName, + sizeof(ENTITY)); + if (!declEntity) + return XML_ERROR_NO_MEMORY; + declEntity->publicId = NULL; + } + /* fall through */ +#endif /* XML_DTD */ + case XML_ROLE_ENTITY_SYSTEM_ID: + if (dtd->keepProcessing && declEntity) { + declEntity->systemId = poolStoreString(&dtd->pool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!declEntity->systemId) + return XML_ERROR_NO_MEMORY; + declEntity->base = curBase; + poolFinish(&dtd->pool); + if (entityDeclHandler) + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_ENTITY_COMPLETE: + if (dtd->keepProcessing && declEntity && entityDeclHandler) { + *eventEndPP = s; + entityDeclHandler(handlerArg, + declEntity->name, + declEntity->is_param, + 0,0, + declEntity->base, + declEntity->systemId, + declEntity->publicId, + 0); + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_ENTITY_NOTATION_NAME: + if (dtd->keepProcessing && declEntity) { + declEntity->notation = poolStoreString(&dtd->pool, enc, s, next); + if (!declEntity->notation) + return XML_ERROR_NO_MEMORY; + poolFinish(&dtd->pool); + if (unparsedEntityDeclHandler) { + *eventEndPP = s; + unparsedEntityDeclHandler(handlerArg, + declEntity->name, + declEntity->base, + declEntity->systemId, + declEntity->publicId, + declEntity->notation); + handleDefault = XML_FALSE; + } + else if (entityDeclHandler) { + *eventEndPP = s; + entityDeclHandler(handlerArg, + declEntity->name, + 0,0,0, + declEntity->base, + declEntity->systemId, + declEntity->publicId, + declEntity->notation); + handleDefault = XML_FALSE; + } + } + break; + case XML_ROLE_GENERAL_ENTITY_NAME: + { + if (XmlPredefinedEntityName(enc, s, next)) { + declEntity = NULL; + break; + } + if (dtd->keepProcessing) { + const XML_Char *name = poolStoreString(&dtd->pool, enc, s, next); + if (!name) + return XML_ERROR_NO_MEMORY; + declEntity = (ENTITY *)lookup(parser, &dtd->generalEntities, name, + sizeof(ENTITY)); + if (!declEntity) + return XML_ERROR_NO_MEMORY; + if (declEntity->name != name) { + poolDiscard(&dtd->pool); + declEntity = NULL; + } + else { + poolFinish(&dtd->pool); + declEntity->publicId = NULL; + declEntity->is_param = XML_FALSE; + /* if we have a parent parser or are reading an internal parameter + entity, then the entity declaration is not considered "internal" + */ + declEntity->is_internal = !(parentParser || openInternalEntities); + if (entityDeclHandler) + handleDefault = XML_FALSE; + } + } + else { + poolDiscard(&dtd->pool); + declEntity = NULL; + } + } + break; + case XML_ROLE_PARAM_ENTITY_NAME: +#ifdef XML_DTD + if (dtd->keepProcessing) { + const XML_Char *name = poolStoreString(&dtd->pool, enc, s, next); + if (!name) + return XML_ERROR_NO_MEMORY; + declEntity = (ENTITY *)lookup(parser, &dtd->paramEntities, + name, sizeof(ENTITY)); + if (!declEntity) + return XML_ERROR_NO_MEMORY; + if (declEntity->name != name) { + poolDiscard(&dtd->pool); + declEntity = NULL; + } + else { + poolFinish(&dtd->pool); + declEntity->publicId = NULL; + declEntity->is_param = XML_TRUE; + /* if we have a parent parser or are reading an internal parameter + entity, then the entity declaration is not considered "internal" + */ + declEntity->is_internal = !(parentParser || openInternalEntities); + if (entityDeclHandler) + handleDefault = XML_FALSE; + } + } + else { + poolDiscard(&dtd->pool); + declEntity = NULL; + } +#else /* not XML_DTD */ + declEntity = NULL; +#endif /* XML_DTD */ + break; + case XML_ROLE_NOTATION_NAME: + declNotationPublicId = NULL; + declNotationName = NULL; + if (notationDeclHandler) { + declNotationName = poolStoreString(&tempPool, enc, s, next); + if (!declNotationName) + return XML_ERROR_NO_MEMORY; + poolFinish(&tempPool); + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_NOTATION_PUBLIC_ID: + if (!XmlIsPublicId(enc, s, next, eventPP)) + return XML_ERROR_PUBLICID; + if (declNotationName) { /* means notationDeclHandler != NULL */ + XML_Char *tem = poolStoreString(&tempPool, + enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!tem) + return XML_ERROR_NO_MEMORY; + normalizePublicId(tem); + declNotationPublicId = tem; + poolFinish(&tempPool); + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_NOTATION_SYSTEM_ID: + if (declNotationName && notationDeclHandler) { + const XML_Char *systemId + = poolStoreString(&tempPool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!systemId) + return XML_ERROR_NO_MEMORY; + *eventEndPP = s; + notationDeclHandler(handlerArg, + declNotationName, + curBase, + systemId, + declNotationPublicId); + handleDefault = XML_FALSE; + } + poolClear(&tempPool); + break; + case XML_ROLE_NOTATION_NO_SYSTEM_ID: + if (declNotationPublicId && notationDeclHandler) { + *eventEndPP = s; + notationDeclHandler(handlerArg, + declNotationName, + curBase, + 0, + declNotationPublicId); + handleDefault = XML_FALSE; + } + poolClear(&tempPool); + break; + case XML_ROLE_ERROR: + switch (tok) { + case XML_TOK_PARAM_ENTITY_REF: + /* PE references in internal subset are + not allowed within declarations. */ + return XML_ERROR_PARAM_ENTITY_REF; + case XML_TOK_XML_DECL: + return XML_ERROR_MISPLACED_XML_PI; + default: + return XML_ERROR_SYNTAX; + } +#ifdef XML_DTD + case XML_ROLE_IGNORE_SECT: + { + enum XML_Error result; + if (defaultHandler) + reportDefault(parser, enc, s, next); + handleDefault = XML_FALSE; + result = doIgnoreSection(parser, enc, &next, end, nextPtr, haveMore); + if (result != XML_ERROR_NONE) + return result; + else if (!next) { + processor = ignoreSectionProcessor; + return result; + } + } + break; +#endif /* XML_DTD */ + case XML_ROLE_GROUP_OPEN: + if (prologState.level >= groupSize) { + if (groupSize) { + char *temp = (char *)REALLOC(groupConnector, groupSize *= 2); + if (temp == NULL) + return XML_ERROR_NO_MEMORY; + groupConnector = temp; + if (dtd->scaffIndex) { + int *temp = (int *)REALLOC(dtd->scaffIndex, + groupSize * sizeof(int)); + if (temp == NULL) + return XML_ERROR_NO_MEMORY; + dtd->scaffIndex = temp; + } + } + else { + groupConnector = (char *)MALLOC(groupSize = 32); + if (!groupConnector) + return XML_ERROR_NO_MEMORY; + } + } + groupConnector[prologState.level] = 0; + if (dtd->in_eldecl) { + int myindex = nextScaffoldPart(parser); + if (myindex < 0) + return XML_ERROR_NO_MEMORY; + dtd->scaffIndex[dtd->scaffLevel] = myindex; + dtd->scaffLevel++; + dtd->scaffold[myindex].type = XML_CTYPE_SEQ; + if (elementDeclHandler) + handleDefault = XML_FALSE; + } + break; + case XML_ROLE_GROUP_SEQUENCE: + if (groupConnector[prologState.level] == ASCII_PIPE) + return XML_ERROR_SYNTAX; + groupConnector[prologState.level] = ASCII_COMMA; + if (dtd->in_eldecl && elementDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_GROUP_CHOICE: + if (groupConnector[prologState.level] == ASCII_COMMA) + return XML_ERROR_SYNTAX; + if (dtd->in_eldecl + && !groupConnector[prologState.level] + && (dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type + != XML_CTYPE_MIXED) + ) { + dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type + = XML_CTYPE_CHOICE; + if (elementDeclHandler) + handleDefault = XML_FALSE; + } + groupConnector[prologState.level] = ASCII_PIPE; + break; + case XML_ROLE_PARAM_ENTITY_REF: +#ifdef XML_DTD + case XML_ROLE_INNER_PARAM_ENTITY_REF: + dtd->hasParamEntityRefs = XML_TRUE; + if (!paramEntityParsing) + dtd->keepProcessing = dtd->standalone; + else { + const XML_Char *name; + ENTITY *entity; + name = poolStoreString(&dtd->pool, enc, + s + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!name) + return XML_ERROR_NO_MEMORY; + entity = (ENTITY *)lookup(parser, &dtd->paramEntities, name, 0); + poolDiscard(&dtd->pool); + /* first, determine if a check for an existing declaration is needed; + if yes, check that the entity exists, and that it is internal, + otherwise call the skipped entity handler + */ + if (prologState.documentEntity && + (dtd->standalone + ? !openInternalEntities + : !dtd->hasParamEntityRefs)) { + if (!entity) + return XML_ERROR_UNDEFINED_ENTITY; + else if (!entity->is_internal) + return XML_ERROR_ENTITY_DECLARED_IN_PE; + } + else if (!entity) { + dtd->keepProcessing = dtd->standalone; + /* cannot report skipped entities in declarations */ + if ((role == XML_ROLE_PARAM_ENTITY_REF) && skippedEntityHandler) { + skippedEntityHandler(handlerArg, name, 1); + handleDefault = XML_FALSE; + } + break; + } + if (entity->open) + return XML_ERROR_RECURSIVE_ENTITY_REF; + if (entity->textPtr) { + enum XML_Error result; + XML_Bool betweenDecl = + (role == XML_ROLE_PARAM_ENTITY_REF ? XML_TRUE : XML_FALSE); + result = processInternalEntity(parser, entity, betweenDecl); + if (result != XML_ERROR_NONE) + return result; + handleDefault = XML_FALSE; + break; + } + if (externalEntityRefHandler) { + dtd->paramEntityRead = XML_FALSE; + entity->open = XML_TRUE; + if (!externalEntityRefHandler(externalEntityRefHandlerArg, + 0, + entity->base, + entity->systemId, + entity->publicId)) { + entity->open = XML_FALSE; + return XML_ERROR_EXTERNAL_ENTITY_HANDLING; + } + entity->open = XML_FALSE; + handleDefault = XML_FALSE; + if (!dtd->paramEntityRead) { + dtd->keepProcessing = dtd->standalone; + break; + } + } + else { + dtd->keepProcessing = dtd->standalone; + break; + } + } +#endif /* XML_DTD */ + if (!dtd->standalone && + notStandaloneHandler && + !notStandaloneHandler(handlerArg)) + return XML_ERROR_NOT_STANDALONE; + break; + + /* Element declaration stuff */ + + case XML_ROLE_ELEMENT_NAME: + if (elementDeclHandler) { + declElementType = getElementType(parser, enc, s, next); + if (!declElementType) + return XML_ERROR_NO_MEMORY; + dtd->scaffLevel = 0; + dtd->scaffCount = 0; + dtd->in_eldecl = XML_TRUE; + handleDefault = XML_FALSE; + } + break; + + case XML_ROLE_CONTENT_ANY: + case XML_ROLE_CONTENT_EMPTY: + if (dtd->in_eldecl) { + if (elementDeclHandler) { + XML_Content * content = (XML_Content *) MALLOC(sizeof(XML_Content)); + if (!content) + return XML_ERROR_NO_MEMORY; + content->quant = XML_CQUANT_NONE; + content->name = NULL; + content->numchildren = 0; + content->children = NULL; + content->type = ((role == XML_ROLE_CONTENT_ANY) ? + XML_CTYPE_ANY : + XML_CTYPE_EMPTY); + *eventEndPP = s; + elementDeclHandler(handlerArg, declElementType->name, content); + handleDefault = XML_FALSE; + } + dtd->in_eldecl = XML_FALSE; + } + break; + + case XML_ROLE_CONTENT_PCDATA: + if (dtd->in_eldecl) { + dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel - 1]].type + = XML_CTYPE_MIXED; + if (elementDeclHandler) + handleDefault = XML_FALSE; + } + break; + + case XML_ROLE_CONTENT_ELEMENT: + quant = XML_CQUANT_NONE; + goto elementContent; + case XML_ROLE_CONTENT_ELEMENT_OPT: + quant = XML_CQUANT_OPT; + goto elementContent; + case XML_ROLE_CONTENT_ELEMENT_REP: + quant = XML_CQUANT_REP; + goto elementContent; + case XML_ROLE_CONTENT_ELEMENT_PLUS: + quant = XML_CQUANT_PLUS; + elementContent: + if (dtd->in_eldecl) { + ELEMENT_TYPE *el; + const XML_Char *name; + int nameLen; + const char *nxt = (quant == XML_CQUANT_NONE + ? next + : next - enc->minBytesPerChar); + int myindex = nextScaffoldPart(parser); + if (myindex < 0) + return XML_ERROR_NO_MEMORY; + dtd->scaffold[myindex].type = XML_CTYPE_NAME; + dtd->scaffold[myindex].quant = quant; + el = getElementType(parser, enc, s, nxt); + if (!el) + return XML_ERROR_NO_MEMORY; + name = el->name; + dtd->scaffold[myindex].name = name; + nameLen = 0; + for (; name[nameLen++]; ); + dtd->contentStringLen += nameLen; + if (elementDeclHandler) + handleDefault = XML_FALSE; + } + break; + + case XML_ROLE_GROUP_CLOSE: + quant = XML_CQUANT_NONE; + goto closeGroup; + case XML_ROLE_GROUP_CLOSE_OPT: + quant = XML_CQUANT_OPT; + goto closeGroup; + case XML_ROLE_GROUP_CLOSE_REP: + quant = XML_CQUANT_REP; + goto closeGroup; + case XML_ROLE_GROUP_CLOSE_PLUS: + quant = XML_CQUANT_PLUS; + closeGroup: + if (dtd->in_eldecl) { + if (elementDeclHandler) + handleDefault = XML_FALSE; + dtd->scaffLevel--; + dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel]].quant = quant; + if (dtd->scaffLevel == 0) { + if (!handleDefault) { + XML_Content *model = build_model(parser); + if (!model) + return XML_ERROR_NO_MEMORY; + *eventEndPP = s; + elementDeclHandler(handlerArg, declElementType->name, model); + } + dtd->in_eldecl = XML_FALSE; + dtd->contentStringLen = 0; + } + } + break; + /* End element declaration stuff */ + + case XML_ROLE_PI: + if (!reportProcessingInstruction(parser, enc, s, next)) + return XML_ERROR_NO_MEMORY; + handleDefault = XML_FALSE; + break; + case XML_ROLE_COMMENT: + if (!reportComment(parser, enc, s, next)) + return XML_ERROR_NO_MEMORY; + handleDefault = XML_FALSE; + break; + case XML_ROLE_NONE: + switch (tok) { + case XML_TOK_BOM: + handleDefault = XML_FALSE; + break; + } + break; + case XML_ROLE_DOCTYPE_NONE: + if (startDoctypeDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_ENTITY_NONE: + if (dtd->keepProcessing && entityDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_NOTATION_NONE: + if (notationDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_ATTLIST_NONE: + if (dtd->keepProcessing && attlistDeclHandler) + handleDefault = XML_FALSE; + break; + case XML_ROLE_ELEMENT_NONE: + if (elementDeclHandler) + handleDefault = XML_FALSE; + break; + } /* end of big switch */ + + if (handleDefault && defaultHandler) + reportDefault(parser, enc, s, next); + + switch (ps_parsing) { + case XML_SUSPENDED: + *nextPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: + s = next; + tok = XmlPrologTok(enc, s, end, &next); + } + } + /* not reached */ +} + +static enum XML_Error PTRCALL +epilogProcessor(XML_Parser parser, + const char *s, + const char *end, + const char **nextPtr) +{ + processor = epilogProcessor; + eventPtr = s; + for (;;) { + const char *next = NULL; + int tok = XmlPrologTok(encoding, s, end, &next); + eventEndPtr = next; + switch (tok) { + /* report partial linebreak - it might be the last token */ + case -XML_TOK_PROLOG_S: + if (defaultHandler) { + reportDefault(parser, encoding, s, next); + if (ps_parsing == XML_FINISHED) + return XML_ERROR_ABORTED; + } + *nextPtr = next; + return XML_ERROR_NONE; + case XML_TOK_NONE: + *nextPtr = s; + return XML_ERROR_NONE; + case XML_TOK_PROLOG_S: + if (defaultHandler) + reportDefault(parser, encoding, s, next); + break; + case XML_TOK_PI: + if (!reportProcessingInstruction(parser, encoding, s, next)) + return XML_ERROR_NO_MEMORY; + break; + case XML_TOK_COMMENT: + if (!reportComment(parser, encoding, s, next)) + return XML_ERROR_NO_MEMORY; + break; + case XML_TOK_INVALID: + eventPtr = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL: + if (!ps_finalBuffer) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_UNCLOSED_TOKEN; + case XML_TOK_PARTIAL_CHAR: + if (!ps_finalBuffer) { + *nextPtr = s; + return XML_ERROR_NONE; + } + return XML_ERROR_PARTIAL_CHAR; + default: + return XML_ERROR_JUNK_AFTER_DOC_ELEMENT; + } + eventPtr = s = next; + switch (ps_parsing) { + case XML_SUSPENDED: + *nextPtr = next; + return XML_ERROR_NONE; + case XML_FINISHED: + return XML_ERROR_ABORTED; + default: ; + } + } +} + +static enum XML_Error +processInternalEntity(XML_Parser parser, ENTITY *entity, + XML_Bool betweenDecl) +{ + const char *textStart, *textEnd; + const char *next; + enum XML_Error result; + OPEN_INTERNAL_ENTITY *openEntity; + + if (freeInternalEntities) { + openEntity = freeInternalEntities; + freeInternalEntities = openEntity->next; + } + else { + openEntity = (OPEN_INTERNAL_ENTITY *)MALLOC(sizeof(OPEN_INTERNAL_ENTITY)); + if (!openEntity) + return XML_ERROR_NO_MEMORY; + } + entity->open = XML_TRUE; + entity->processed = 0; + openEntity->next = openInternalEntities; + openInternalEntities = openEntity; + openEntity->entity = entity; + openEntity->startTagLevel = tagLevel; + openEntity->betweenDecl = betweenDecl; + openEntity->internalEventPtr = NULL; + openEntity->internalEventEndPtr = NULL; + textStart = (char *)entity->textPtr; + textEnd = (char *)(entity->textPtr + entity->textLen); + /* Set a safe default value in case 'next' does not get set */ + next = textStart; + +#ifdef XML_DTD + if (entity->is_param) { + int tok = XmlPrologTok(internalEncoding, textStart, textEnd, &next); + result = doProlog(parser, internalEncoding, textStart, textEnd, tok, + next, &next, XML_FALSE); + } + else +#endif /* XML_DTD */ + result = doContent(parser, tagLevel, internalEncoding, textStart, + textEnd, &next, XML_FALSE); + + if (result == XML_ERROR_NONE) { + if (textEnd != next && ps_parsing == XML_SUSPENDED) { + entity->processed = (int)(next - textStart); + processor = internalEntityProcessor; + } + else { + entity->open = XML_FALSE; + openInternalEntities = openEntity->next; + /* put openEntity back in list of free instances */ + openEntity->next = freeInternalEntities; + freeInternalEntities = openEntity; + } + } + return result; +} + +static enum XML_Error PTRCALL +internalEntityProcessor(XML_Parser parser, + const char *s, + const char *end, + const char **nextPtr) +{ + ENTITY *entity; + const char *textStart, *textEnd; + const char *next; + enum XML_Error result; + OPEN_INTERNAL_ENTITY *openEntity = openInternalEntities; + if (!openEntity) + return XML_ERROR_UNEXPECTED_STATE; + + entity = openEntity->entity; + textStart = ((char *)entity->textPtr) + entity->processed; + textEnd = (char *)(entity->textPtr + entity->textLen); + /* Set a safe default value in case 'next' does not get set */ + next = textStart; + +#ifdef XML_DTD + if (entity->is_param) { + int tok = XmlPrologTok(internalEncoding, textStart, textEnd, &next); + result = doProlog(parser, internalEncoding, textStart, textEnd, tok, + next, &next, XML_FALSE); + } + else +#endif /* XML_DTD */ + result = doContent(parser, openEntity->startTagLevel, internalEncoding, + textStart, textEnd, &next, XML_FALSE); + + if (result != XML_ERROR_NONE) + return result; + else if (textEnd != next && ps_parsing == XML_SUSPENDED) { + entity->processed = (int)(next - (char *)entity->textPtr); + return result; + } + else { + entity->open = XML_FALSE; + openInternalEntities = openEntity->next; + /* put openEntity back in list of free instances */ + openEntity->next = freeInternalEntities; + freeInternalEntities = openEntity; + } + +#ifdef XML_DTD + if (entity->is_param) { + int tok; + processor = prologProcessor; + tok = XmlPrologTok(encoding, s, end, &next); + return doProlog(parser, encoding, s, end, tok, next, nextPtr, + (XML_Bool)!ps_finalBuffer); + } + else +#endif /* XML_DTD */ + { + processor = contentProcessor; + /* see externalEntityContentProcessor vs contentProcessor */ + return doContent(parser, parentParser ? 1 : 0, encoding, s, end, + nextPtr, (XML_Bool)!ps_finalBuffer); + } +} + +static enum XML_Error PTRCALL +errorProcessor(XML_Parser parser, + const char *UNUSED_P(s), + const char *UNUSED_P(end), + const char **UNUSED_P(nextPtr)) +{ + return errorCode; +} + +static enum XML_Error +storeAttributeValue(XML_Parser parser, const ENCODING *enc, XML_Bool isCdata, + const char *ptr, const char *end, + STRING_POOL *pool) +{ + enum XML_Error result = appendAttributeValue(parser, enc, isCdata, ptr, + end, pool); + if (result) + return result; + if (!isCdata && poolLength(pool) && poolLastChar(pool) == 0x20) + poolChop(pool); + if (!poolAppendChar(pool, XML_T('\0'))) + return XML_ERROR_NO_MEMORY; + return XML_ERROR_NONE; +} + +static enum XML_Error +appendAttributeValue(XML_Parser parser, const ENCODING *enc, XML_Bool isCdata, + const char *ptr, const char *end, + STRING_POOL *pool) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + for (;;) { + const char *next; + int tok = XmlAttributeValueTok(enc, ptr, end, &next); + switch (tok) { + case XML_TOK_NONE: + return XML_ERROR_NONE; + case XML_TOK_INVALID: + if (enc == encoding) + eventPtr = next; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_PARTIAL: + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_INVALID_TOKEN; + case XML_TOK_CHAR_REF: + { + XML_Char buf[XML_ENCODE_MAX]; + int i; + int n = XmlCharRefNumber(enc, ptr); + if (n < 0) { + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_BAD_CHAR_REF; + } + if (!isCdata + && n == 0x20 /* space */ + && (poolLength(pool) == 0 || poolLastChar(pool) == 0x20)) + break; + n = XmlEncode(n, (ICHAR *)buf); + if (!n) { + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_BAD_CHAR_REF; + } + for (i = 0; i < n; i++) { + if (!poolAppendChar(pool, buf[i])) + return XML_ERROR_NO_MEMORY; + } + } + break; + case XML_TOK_DATA_CHARS: + if (!poolAppend(pool, enc, ptr, next)) + return XML_ERROR_NO_MEMORY; + break; + case XML_TOK_TRAILING_CR: + next = ptr + enc->minBytesPerChar; + /* fall through */ + case XML_TOK_ATTRIBUTE_VALUE_S: + case XML_TOK_DATA_NEWLINE: + if (!isCdata && (poolLength(pool) == 0 || poolLastChar(pool) == 0x20)) + break; + if (!poolAppendChar(pool, 0x20)) + return XML_ERROR_NO_MEMORY; + break; + case XML_TOK_ENTITY_REF: + { + const XML_Char *name; + ENTITY *entity; + char checkEntityDecl; + XML_Char ch = (XML_Char) XmlPredefinedEntityName(enc, + ptr + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (ch) { + if (!poolAppendChar(pool, ch)) + return XML_ERROR_NO_MEMORY; + break; + } + name = poolStoreString(&temp2Pool, enc, + ptr + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!name) + return XML_ERROR_NO_MEMORY; + entity = (ENTITY *)lookup(parser, &dtd->generalEntities, name, 0); + poolDiscard(&temp2Pool); + /* First, determine if a check for an existing declaration is needed; + if yes, check that the entity exists, and that it is internal. + */ + if (pool == &dtd->pool) /* are we called from prolog? */ + checkEntityDecl = +#ifdef XML_DTD + prologState.documentEntity && +#endif /* XML_DTD */ + (dtd->standalone + ? !openInternalEntities + : !dtd->hasParamEntityRefs); + else /* if (pool == &tempPool): we are called from content */ + checkEntityDecl = !dtd->hasParamEntityRefs || dtd->standalone; + if (checkEntityDecl) { + if (!entity) + return XML_ERROR_UNDEFINED_ENTITY; + else if (!entity->is_internal) + return XML_ERROR_ENTITY_DECLARED_IN_PE; + } + else if (!entity) { + /* Cannot report skipped entity here - see comments on + skippedEntityHandler. + if (skippedEntityHandler) + skippedEntityHandler(handlerArg, name, 0); + */ + /* Cannot call the default handler because this would be + out of sync with the call to the startElementHandler. + if ((pool == &tempPool) && defaultHandler) + reportDefault(parser, enc, ptr, next); + */ + break; + } + if (entity->open) { + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_RECURSIVE_ENTITY_REF; + } + if (entity->notation) { + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_BINARY_ENTITY_REF; + } + if (!entity->textPtr) { + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_ATTRIBUTE_EXTERNAL_ENTITY_REF; + } + else { + enum XML_Error result; + const XML_Char *textEnd = entity->textPtr + entity->textLen; + entity->open = XML_TRUE; + result = appendAttributeValue(parser, internalEncoding, isCdata, + (char *)entity->textPtr, + (char *)textEnd, pool); + entity->open = XML_FALSE; + if (result) + return result; + } + } + break; + default: + if (enc == encoding) + eventPtr = ptr; + return XML_ERROR_UNEXPECTED_STATE; + } + ptr = next; + } + /* not reached */ +} + +static enum XML_Error +storeEntityValue(XML_Parser parser, + const ENCODING *enc, + const char *entityTextPtr, + const char *entityTextEnd) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + STRING_POOL *pool = &(dtd->entityValuePool); + enum XML_Error result = XML_ERROR_NONE; +#ifdef XML_DTD + int oldInEntityValue = prologState.inEntityValue; + prologState.inEntityValue = 1; +#endif /* XML_DTD */ + /* never return Null for the value argument in EntityDeclHandler, + since this would indicate an external entity; therefore we + have to make sure that entityValuePool.start is not null */ + if (!pool->blocks) { + if (!poolGrow(pool)) + return XML_ERROR_NO_MEMORY; + } + + for (;;) { + const char *next; + int tok = XmlEntityValueTok(enc, entityTextPtr, entityTextEnd, &next); + switch (tok) { + case XML_TOK_PARAM_ENTITY_REF: +#ifdef XML_DTD + if (isParamEntity || enc != encoding) { + const XML_Char *name; + ENTITY *entity; + name = poolStoreString(&tempPool, enc, + entityTextPtr + enc->minBytesPerChar, + next - enc->minBytesPerChar); + if (!name) { + result = XML_ERROR_NO_MEMORY; + goto endEntityValue; + } + entity = (ENTITY *)lookup(parser, &dtd->paramEntities, name, 0); + poolDiscard(&tempPool); + if (!entity) { + /* not a well-formedness error - see XML 1.0: WFC Entity Declared */ + /* cannot report skipped entity here - see comments on + skippedEntityHandler + if (skippedEntityHandler) + skippedEntityHandler(handlerArg, name, 0); + */ + dtd->keepProcessing = dtd->standalone; + goto endEntityValue; + } + if (entity->open) { + if (enc == encoding) + eventPtr = entityTextPtr; + result = XML_ERROR_RECURSIVE_ENTITY_REF; + goto endEntityValue; + } + if (entity->systemId) { + if (externalEntityRefHandler) { + dtd->paramEntityRead = XML_FALSE; + entity->open = XML_TRUE; + if (!externalEntityRefHandler(externalEntityRefHandlerArg, + 0, + entity->base, + entity->systemId, + entity->publicId)) { + entity->open = XML_FALSE; + result = XML_ERROR_EXTERNAL_ENTITY_HANDLING; + goto endEntityValue; + } + entity->open = XML_FALSE; + if (!dtd->paramEntityRead) + dtd->keepProcessing = dtd->standalone; + } + else + dtd->keepProcessing = dtd->standalone; + } + else { + entity->open = XML_TRUE; + result = storeEntityValue(parser, + internalEncoding, + (char *)entity->textPtr, + (char *)(entity->textPtr + + entity->textLen)); + entity->open = XML_FALSE; + if (result) + goto endEntityValue; + } + break; + } +#endif /* XML_DTD */ + /* In the internal subset, PE references are not legal + within markup declarations, e.g entity values in this case. */ + eventPtr = entityTextPtr; + result = XML_ERROR_PARAM_ENTITY_REF; + goto endEntityValue; + case XML_TOK_NONE: + result = XML_ERROR_NONE; + goto endEntityValue; + case XML_TOK_ENTITY_REF: + case XML_TOK_DATA_CHARS: + if (!poolAppend(pool, enc, entityTextPtr, next)) { + result = XML_ERROR_NO_MEMORY; + goto endEntityValue; + } + break; + case XML_TOK_TRAILING_CR: + next = entityTextPtr + enc->minBytesPerChar; + /* fall through */ + case XML_TOK_DATA_NEWLINE: + if (pool->end == pool->ptr && !poolGrow(pool)) { + result = XML_ERROR_NO_MEMORY; + goto endEntityValue; + } + *(pool->ptr)++ = 0xA; + break; + case XML_TOK_CHAR_REF: + { + XML_Char buf[XML_ENCODE_MAX]; + int i; + int n = XmlCharRefNumber(enc, entityTextPtr); + if (n < 0) { + if (enc == encoding) + eventPtr = entityTextPtr; + result = XML_ERROR_BAD_CHAR_REF; + goto endEntityValue; + } + n = XmlEncode(n, (ICHAR *)buf); + if (!n) { + if (enc == encoding) + eventPtr = entityTextPtr; + result = XML_ERROR_BAD_CHAR_REF; + goto endEntityValue; + } + for (i = 0; i < n; i++) { + if (pool->end == pool->ptr && !poolGrow(pool)) { + result = XML_ERROR_NO_MEMORY; + goto endEntityValue; + } + *(pool->ptr)++ = buf[i]; + } + } + break; + case XML_TOK_PARTIAL: + if (enc == encoding) + eventPtr = entityTextPtr; + result = XML_ERROR_INVALID_TOKEN; + goto endEntityValue; + case XML_TOK_INVALID: + if (enc == encoding) + eventPtr = next; + result = XML_ERROR_INVALID_TOKEN; + goto endEntityValue; + default: + if (enc == encoding) + eventPtr = entityTextPtr; + result = XML_ERROR_UNEXPECTED_STATE; + goto endEntityValue; + } + entityTextPtr = next; + } +endEntityValue: +#ifdef XML_DTD + prologState.inEntityValue = oldInEntityValue; +#endif /* XML_DTD */ + return result; +} + +static void FASTCALL +normalizeLines(XML_Char *s) +{ + XML_Char *p; + for (;; s++) { + if (*s == XML_T('\0')) + return; + if (*s == 0xD) + break; + } + p = s; + do { + if (*s == 0xD) { + *p++ = 0xA; + if (*++s == 0xA) + s++; + } + else + *p++ = *s++; + } while (*s); + *p = XML_T('\0'); +} + +static int +reportProcessingInstruction(XML_Parser parser, const ENCODING *enc, + const char *start, const char *end) +{ + const XML_Char *target; + XML_Char *data; + const char *tem; + if (!processingInstructionHandler) { + if (defaultHandler) + reportDefault(parser, enc, start, end); + return 1; + } + start += enc->minBytesPerChar * 2; + tem = start + XmlNameLength(enc, start); + target = poolStoreString(&tempPool, enc, start, tem); + if (!target) + return 0; + poolFinish(&tempPool); + data = poolStoreString(&tempPool, enc, + XmlSkipS(enc, tem), + end - enc->minBytesPerChar*2); + if (!data) + return 0; + normalizeLines(data); + processingInstructionHandler(handlerArg, target, data); + poolClear(&tempPool); + return 1; +} + +static int +reportComment(XML_Parser parser, const ENCODING *enc, + const char *start, const char *end) +{ + XML_Char *data; + if (!commentHandler) { + if (defaultHandler) + reportDefault(parser, enc, start, end); + return 1; + } + data = poolStoreString(&tempPool, + enc, + start + enc->minBytesPerChar * 4, + end - enc->minBytesPerChar * 3); + if (!data) + return 0; + normalizeLines(data); + commentHandler(handlerArg, data); + poolClear(&tempPool); + return 1; +} + +static void +reportDefault(XML_Parser parser, const ENCODING *enc, + const char *s, const char *end) +{ + if (MUST_CONVERT(enc, s)) { + enum XML_Convert_Result convert_res; + const char **eventPP; + const char **eventEndPP; + if (enc == encoding) { + eventPP = &eventPtr; + eventEndPP = &eventEndPtr; + } + else { + eventPP = &(openInternalEntities->internalEventPtr); + eventEndPP = &(openInternalEntities->internalEventEndPtr); + } + do { + ICHAR *dataPtr = (ICHAR *)dataBuf; + convert_res = XmlConvert(enc, &s, end, &dataPtr, (ICHAR *)dataBufEnd); + *eventEndPP = s; + defaultHandler(handlerArg, dataBuf, (int)(dataPtr - (ICHAR *)dataBuf)); + *eventPP = s; + } while ((convert_res != XML_CONVERT_COMPLETED) && (convert_res != XML_CONVERT_INPUT_INCOMPLETE)); + } + else + defaultHandler(handlerArg, (XML_Char *)s, (int)((XML_Char *)end - (XML_Char *)s)); +} + + +static int +defineAttribute(ELEMENT_TYPE *type, ATTRIBUTE_ID *attId, XML_Bool isCdata, + XML_Bool isId, const XML_Char *value, XML_Parser parser) +{ + DEFAULT_ATTRIBUTE *att; + if (value || isId) { + /* The handling of default attributes gets messed up if we have + a default which duplicates a non-default. */ + int i; + for (i = 0; i < type->nDefaultAtts; i++) + if (attId == type->defaultAtts[i].id) + return 1; + if (isId && !type->idAtt && !attId->xmlns) + type->idAtt = attId; + } + if (type->nDefaultAtts == type->allocDefaultAtts) { + if (type->allocDefaultAtts == 0) { + type->allocDefaultAtts = 8; + type->defaultAtts = (DEFAULT_ATTRIBUTE *)MALLOC(type->allocDefaultAtts + * sizeof(DEFAULT_ATTRIBUTE)); + if (!type->defaultAtts) + return 0; + } + else { + DEFAULT_ATTRIBUTE *temp; + int count = type->allocDefaultAtts * 2; + temp = (DEFAULT_ATTRIBUTE *) + REALLOC(type->defaultAtts, (count * sizeof(DEFAULT_ATTRIBUTE))); + if (temp == NULL) + return 0; + type->allocDefaultAtts = count; + type->defaultAtts = temp; + } + } + att = type->defaultAtts + type->nDefaultAtts; + att->id = attId; + att->value = value; + att->isCdata = isCdata; + if (!isCdata) + attId->maybeTokenized = XML_TRUE; + type->nDefaultAtts += 1; + return 1; +} + +static int +setElementTypePrefix(XML_Parser parser, ELEMENT_TYPE *elementType) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + const XML_Char *name; + for (name = elementType->name; *name; name++) { + if (*name == XML_T(ASCII_COLON)) { + PREFIX *prefix; + const XML_Char *s; + for (s = elementType->name; s != name; s++) { + if (!poolAppendChar(&dtd->pool, *s)) + return 0; + } + if (!poolAppendChar(&dtd->pool, XML_T('\0'))) + return 0; + prefix = (PREFIX *)lookup(parser, &dtd->prefixes, poolStart(&dtd->pool), + sizeof(PREFIX)); + if (!prefix) + return 0; + if (prefix->name == poolStart(&dtd->pool)) + poolFinish(&dtd->pool); + else + poolDiscard(&dtd->pool); + elementType->prefix = prefix; + + } + } + return 1; +} + +static ATTRIBUTE_ID * +getAttributeId(XML_Parser parser, const ENCODING *enc, + const char *start, const char *end) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + ATTRIBUTE_ID *id; + const XML_Char *name; + if (!poolAppendChar(&dtd->pool, XML_T('\0'))) + return NULL; + name = poolStoreString(&dtd->pool, enc, start, end); + if (!name) + return NULL; + /* skip quotation mark - its storage will be re-used (like in name[-1]) */ + ++name; + id = (ATTRIBUTE_ID *)lookup(parser, &dtd->attributeIds, name, sizeof(ATTRIBUTE_ID)); + if (!id) + return NULL; + if (id->name != name) + poolDiscard(&dtd->pool); + else { + poolFinish(&dtd->pool); + if (!ns) + ; + else if (name[0] == XML_T(ASCII_x) + && name[1] == XML_T(ASCII_m) + && name[2] == XML_T(ASCII_l) + && name[3] == XML_T(ASCII_n) + && name[4] == XML_T(ASCII_s) + && (name[5] == XML_T('\0') || name[5] == XML_T(ASCII_COLON))) { + if (name[5] == XML_T('\0')) + id->prefix = &dtd->defaultPrefix; + else + id->prefix = (PREFIX *)lookup(parser, &dtd->prefixes, name + 6, sizeof(PREFIX)); + id->xmlns = XML_TRUE; + } + else { + int i; + for (i = 0; name[i]; i++) { + /* attributes without prefix are *not* in the default namespace */ + if (name[i] == XML_T(ASCII_COLON)) { + int j; + for (j = 0; j < i; j++) { + if (!poolAppendChar(&dtd->pool, name[j])) + return NULL; + } + if (!poolAppendChar(&dtd->pool, XML_T('\0'))) + return NULL; + id->prefix = (PREFIX *)lookup(parser, &dtd->prefixes, poolStart(&dtd->pool), + sizeof(PREFIX)); + if (!id->prefix) + return NULL; + if (id->prefix->name == poolStart(&dtd->pool)) + poolFinish(&dtd->pool); + else + poolDiscard(&dtd->pool); + break; + } + } + } + } + return id; +} + +#define CONTEXT_SEP XML_T(ASCII_FF) + +static const XML_Char * +getContext(XML_Parser parser) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + HASH_TABLE_ITER iter; + XML_Bool needSep = XML_FALSE; + + if (dtd->defaultPrefix.binding) { + int i; + int len; + if (!poolAppendChar(&tempPool, XML_T(ASCII_EQUALS))) + return NULL; + len = dtd->defaultPrefix.binding->uriLen; + if (namespaceSeparator) + len--; + for (i = 0; i < len; i++) + if (!poolAppendChar(&tempPool, dtd->defaultPrefix.binding->uri[i])) + return NULL; + needSep = XML_TRUE; + } + + hashTableIterInit(&iter, &(dtd->prefixes)); + for (;;) { + int i; + int len; + const XML_Char *s; + PREFIX *prefix = (PREFIX *)hashTableIterNext(&iter); + if (!prefix) + break; + if (!prefix->binding) + continue; + if (needSep && !poolAppendChar(&tempPool, CONTEXT_SEP)) + return NULL; + for (s = prefix->name; *s; s++) + if (!poolAppendChar(&tempPool, *s)) + return NULL; + if (!poolAppendChar(&tempPool, XML_T(ASCII_EQUALS))) + return NULL; + len = prefix->binding->uriLen; + if (namespaceSeparator) + len--; + for (i = 0; i < len; i++) + if (!poolAppendChar(&tempPool, prefix->binding->uri[i])) + return NULL; + needSep = XML_TRUE; + } + + + hashTableIterInit(&iter, &(dtd->generalEntities)); + for (;;) { + const XML_Char *s; + ENTITY *e = (ENTITY *)hashTableIterNext(&iter); + if (!e) + break; + if (!e->open) + continue; + if (needSep && !poolAppendChar(&tempPool, CONTEXT_SEP)) + return NULL; + for (s = e->name; *s; s++) + if (!poolAppendChar(&tempPool, *s)) + return 0; + needSep = XML_TRUE; + } + + if (!poolAppendChar(&tempPool, XML_T('\0'))) + return NULL; + return tempPool.start; +} + +static XML_Bool +setContext(XML_Parser parser, const XML_Char *context) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + const XML_Char *s = context; + + while (*context != XML_T('\0')) { + if (*s == CONTEXT_SEP || *s == XML_T('\0')) { + ENTITY *e; + if (!poolAppendChar(&tempPool, XML_T('\0'))) + return XML_FALSE; + e = (ENTITY *)lookup(parser, &dtd->generalEntities, poolStart(&tempPool), 0); + if (e) + e->open = XML_TRUE; + if (*s != XML_T('\0')) + s++; + context = s; + poolDiscard(&tempPool); + } + else if (*s == XML_T(ASCII_EQUALS)) { + PREFIX *prefix; + if (poolLength(&tempPool) == 0) + prefix = &dtd->defaultPrefix; + else { + if (!poolAppendChar(&tempPool, XML_T('\0'))) + return XML_FALSE; + prefix = (PREFIX *)lookup(parser, &dtd->prefixes, poolStart(&tempPool), + sizeof(PREFIX)); + if (!prefix) + return XML_FALSE; + if (prefix->name == poolStart(&tempPool)) { + prefix->name = poolCopyString(&dtd->pool, prefix->name); + if (!prefix->name) + return XML_FALSE; + } + poolDiscard(&tempPool); + } + for (context = s + 1; + *context != CONTEXT_SEP && *context != XML_T('\0'); + context++) + if (!poolAppendChar(&tempPool, *context)) + return XML_FALSE; + if (!poolAppendChar(&tempPool, XML_T('\0'))) + return XML_FALSE; + if (addBinding(parser, prefix, NULL, poolStart(&tempPool), + &inheritedBindings) != XML_ERROR_NONE) + return XML_FALSE; + poolDiscard(&tempPool); + if (*context != XML_T('\0')) + ++context; + s = context; + } + else { + if (!poolAppendChar(&tempPool, *s)) + return XML_FALSE; + s++; + } + } + return XML_TRUE; +} + +static void FASTCALL +normalizePublicId(XML_Char *publicId) +{ + XML_Char *p = publicId; + XML_Char *s; + for (s = publicId; *s; s++) { + switch (*s) { + case 0x20: + case 0xD: + case 0xA: + if (p != publicId && p[-1] != 0x20) + *p++ = 0x20; + break; + default: + *p++ = *s; + } + } + if (p != publicId && p[-1] == 0x20) + --p; + *p = XML_T('\0'); +} + +static DTD * +dtdCreate(const XML_Memory_Handling_Suite *ms) +{ + DTD *p = (DTD *)ms->malloc_fcn(sizeof(DTD)); + if (p == NULL) + return p; + poolInit(&(p->pool), ms); + poolInit(&(p->entityValuePool), ms); + hashTableInit(&(p->generalEntities), ms); + hashTableInit(&(p->elementTypes), ms); + hashTableInit(&(p->attributeIds), ms); + hashTableInit(&(p->prefixes), ms); +#ifdef XML_DTD + p->paramEntityRead = XML_FALSE; + hashTableInit(&(p->paramEntities), ms); +#endif /* XML_DTD */ + p->defaultPrefix.name = NULL; + p->defaultPrefix.binding = NULL; + + p->in_eldecl = XML_FALSE; + p->scaffIndex = NULL; + p->scaffold = NULL; + p->scaffLevel = 0; + p->scaffSize = 0; + p->scaffCount = 0; + p->contentStringLen = 0; + + p->keepProcessing = XML_TRUE; + p->hasParamEntityRefs = XML_FALSE; + p->standalone = XML_FALSE; + return p; +} + +static void +dtdReset(DTD *p, const XML_Memory_Handling_Suite *ms) +{ + HASH_TABLE_ITER iter; + hashTableIterInit(&iter, &(p->elementTypes)); + for (;;) { + ELEMENT_TYPE *e = (ELEMENT_TYPE *)hashTableIterNext(&iter); + if (!e) + break; + if (e->allocDefaultAtts != 0) + ms->free_fcn(e->defaultAtts); + } + hashTableClear(&(p->generalEntities)); +#ifdef XML_DTD + p->paramEntityRead = XML_FALSE; + hashTableClear(&(p->paramEntities)); +#endif /* XML_DTD */ + hashTableClear(&(p->elementTypes)); + hashTableClear(&(p->attributeIds)); + hashTableClear(&(p->prefixes)); + poolClear(&(p->pool)); + poolClear(&(p->entityValuePool)); + p->defaultPrefix.name = NULL; + p->defaultPrefix.binding = NULL; + + p->in_eldecl = XML_FALSE; + + ms->free_fcn(p->scaffIndex); + p->scaffIndex = NULL; + ms->free_fcn(p->scaffold); + p->scaffold = NULL; + + p->scaffLevel = 0; + p->scaffSize = 0; + p->scaffCount = 0; + p->contentStringLen = 0; + + p->keepProcessing = XML_TRUE; + p->hasParamEntityRefs = XML_FALSE; + p->standalone = XML_FALSE; +} + +static void +dtdDestroy(DTD *p, XML_Bool isDocEntity, const XML_Memory_Handling_Suite *ms) +{ + HASH_TABLE_ITER iter; + hashTableIterInit(&iter, &(p->elementTypes)); + for (;;) { + ELEMENT_TYPE *e = (ELEMENT_TYPE *)hashTableIterNext(&iter); + if (!e) + break; + if (e->allocDefaultAtts != 0) + ms->free_fcn(e->defaultAtts); + } + hashTableDestroy(&(p->generalEntities)); +#ifdef XML_DTD + hashTableDestroy(&(p->paramEntities)); +#endif /* XML_DTD */ + hashTableDestroy(&(p->elementTypes)); + hashTableDestroy(&(p->attributeIds)); + hashTableDestroy(&(p->prefixes)); + poolDestroy(&(p->pool)); + poolDestroy(&(p->entityValuePool)); + if (isDocEntity) { + ms->free_fcn(p->scaffIndex); + ms->free_fcn(p->scaffold); + } + ms->free_fcn(p); +} + +/* Do a deep copy of the DTD. Return 0 for out of memory, non-zero otherwise. + The new DTD has already been initialized. +*/ +static int +dtdCopy(XML_Parser oldParser, DTD *newDtd, const DTD *oldDtd, const XML_Memory_Handling_Suite *ms) +{ + HASH_TABLE_ITER iter; + + /* Copy the prefix table. */ + + hashTableIterInit(&iter, &(oldDtd->prefixes)); + for (;;) { + const XML_Char *name; + const PREFIX *oldP = (PREFIX *)hashTableIterNext(&iter); + if (!oldP) + break; + name = poolCopyString(&(newDtd->pool), oldP->name); + if (!name) + return 0; + if (!lookup(oldParser, &(newDtd->prefixes), name, sizeof(PREFIX))) + return 0; + } + + hashTableIterInit(&iter, &(oldDtd->attributeIds)); + + /* Copy the attribute id table. */ + + for (;;) { + ATTRIBUTE_ID *newA; + const XML_Char *name; + const ATTRIBUTE_ID *oldA = (ATTRIBUTE_ID *)hashTableIterNext(&iter); + + if (!oldA) + break; + /* Remember to allocate the scratch byte before the name. */ + if (!poolAppendChar(&(newDtd->pool), XML_T('\0'))) + return 0; + name = poolCopyString(&(newDtd->pool), oldA->name); + if (!name) + return 0; + ++name; + newA = (ATTRIBUTE_ID *)lookup(oldParser, &(newDtd->attributeIds), name, + sizeof(ATTRIBUTE_ID)); + if (!newA) + return 0; + newA->maybeTokenized = oldA->maybeTokenized; + if (oldA->prefix) { + newA->xmlns = oldA->xmlns; + if (oldA->prefix == &oldDtd->defaultPrefix) + newA->prefix = &newDtd->defaultPrefix; + else + newA->prefix = (PREFIX *)lookup(oldParser, &(newDtd->prefixes), + oldA->prefix->name, 0); + } + } + + /* Copy the element type table. */ + + hashTableIterInit(&iter, &(oldDtd->elementTypes)); + + for (;;) { + int i; + ELEMENT_TYPE *newE; + const XML_Char *name; + const ELEMENT_TYPE *oldE = (ELEMENT_TYPE *)hashTableIterNext(&iter); + if (!oldE) + break; + name = poolCopyString(&(newDtd->pool), oldE->name); + if (!name) + return 0; + newE = (ELEMENT_TYPE *)lookup(oldParser, &(newDtd->elementTypes), name, + sizeof(ELEMENT_TYPE)); + if (!newE) + return 0; + if (oldE->nDefaultAtts) { + newE->defaultAtts = (DEFAULT_ATTRIBUTE *) + ms->malloc_fcn(oldE->nDefaultAtts * sizeof(DEFAULT_ATTRIBUTE)); + if (!newE->defaultAtts) { + return 0; + } + } + if (oldE->idAtt) + newE->idAtt = (ATTRIBUTE_ID *) + lookup(oldParser, &(newDtd->attributeIds), oldE->idAtt->name, 0); + newE->allocDefaultAtts = newE->nDefaultAtts = oldE->nDefaultAtts; + if (oldE->prefix) + newE->prefix = (PREFIX *)lookup(oldParser, &(newDtd->prefixes), + oldE->prefix->name, 0); + for (i = 0; i < newE->nDefaultAtts; i++) { + newE->defaultAtts[i].id = (ATTRIBUTE_ID *) + lookup(oldParser, &(newDtd->attributeIds), oldE->defaultAtts[i].id->name, 0); + newE->defaultAtts[i].isCdata = oldE->defaultAtts[i].isCdata; + if (oldE->defaultAtts[i].value) { + newE->defaultAtts[i].value + = poolCopyString(&(newDtd->pool), oldE->defaultAtts[i].value); + if (!newE->defaultAtts[i].value) + return 0; + } + else + newE->defaultAtts[i].value = NULL; + } + } + + /* Copy the entity tables. */ + if (!copyEntityTable(oldParser, + &(newDtd->generalEntities), + &(newDtd->pool), + &(oldDtd->generalEntities))) + return 0; + +#ifdef XML_DTD + if (!copyEntityTable(oldParser, + &(newDtd->paramEntities), + &(newDtd->pool), + &(oldDtd->paramEntities))) + return 0; + newDtd->paramEntityRead = oldDtd->paramEntityRead; +#endif /* XML_DTD */ + + newDtd->keepProcessing = oldDtd->keepProcessing; + newDtd->hasParamEntityRefs = oldDtd->hasParamEntityRefs; + newDtd->standalone = oldDtd->standalone; + + /* Don't want deep copying for scaffolding */ + newDtd->in_eldecl = oldDtd->in_eldecl; + newDtd->scaffold = oldDtd->scaffold; + newDtd->contentStringLen = oldDtd->contentStringLen; + newDtd->scaffSize = oldDtd->scaffSize; + newDtd->scaffLevel = oldDtd->scaffLevel; + newDtd->scaffIndex = oldDtd->scaffIndex; + + return 1; +} /* End dtdCopy */ + +static int +copyEntityTable(XML_Parser oldParser, + HASH_TABLE *newTable, + STRING_POOL *newPool, + const HASH_TABLE *oldTable) +{ + HASH_TABLE_ITER iter; + const XML_Char *cachedOldBase = NULL; + const XML_Char *cachedNewBase = NULL; + + hashTableIterInit(&iter, oldTable); + + for (;;) { + ENTITY *newE; + const XML_Char *name; + const ENTITY *oldE = (ENTITY *)hashTableIterNext(&iter); + if (!oldE) + break; + name = poolCopyString(newPool, oldE->name); + if (!name) + return 0; + newE = (ENTITY *)lookup(oldParser, newTable, name, sizeof(ENTITY)); + if (!newE) + return 0; + if (oldE->systemId) { + const XML_Char *tem = poolCopyString(newPool, oldE->systemId); + if (!tem) + return 0; + newE->systemId = tem; + if (oldE->base) { + if (oldE->base == cachedOldBase) + newE->base = cachedNewBase; + else { + cachedOldBase = oldE->base; + tem = poolCopyString(newPool, cachedOldBase); + if (!tem) + return 0; + cachedNewBase = newE->base = tem; + } + } + if (oldE->publicId) { + tem = poolCopyString(newPool, oldE->publicId); + if (!tem) + return 0; + newE->publicId = tem; + } + } + else { + const XML_Char *tem = poolCopyStringN(newPool, oldE->textPtr, + oldE->textLen); + if (!tem) + return 0; + newE->textPtr = tem; + newE->textLen = oldE->textLen; + } + if (oldE->notation) { + const XML_Char *tem = poolCopyString(newPool, oldE->notation); + if (!tem) + return 0; + newE->notation = tem; + } + newE->is_param = oldE->is_param; + newE->is_internal = oldE->is_internal; + } + return 1; +} + +#define INIT_POWER 6 + +static XML_Bool FASTCALL +keyeq(KEY s1, KEY s2) +{ + for (; *s1 == *s2; s1++, s2++) + if (*s1 == 0) + return XML_TRUE; + return XML_FALSE; +} + +static size_t +keylen(KEY s) +{ + size_t len = 0; + for (; *s; s++, len++); + return len; +} + +static void +copy_salt_to_sipkey(XML_Parser parser, struct sipkey * key) +{ + key->k[0] = 0; + key->k[1] = get_hash_secret_salt(parser); +} + +static unsigned long FASTCALL +hash(XML_Parser parser, KEY s) +{ + struct siphash state; + struct sipkey key; + (void)sip_tobin; + (void)sip24_valid; + copy_salt_to_sipkey(parser, &key); + sip24_init(&state, &key); + sip24_update(&state, s, keylen(s) * sizeof(XML_Char)); + return (unsigned long)sip24_final(&state); +} + +static NAMED * +lookup(XML_Parser parser, HASH_TABLE *table, KEY name, size_t createSize) +{ + size_t i; + if (table->size == 0) { + size_t tsize; + if (!createSize) + return NULL; + table->power = INIT_POWER; + /* table->size is a power of 2 */ + table->size = (size_t)1 << INIT_POWER; + tsize = table->size * sizeof(NAMED *); + table->v = (NAMED **)table->mem->malloc_fcn(tsize); + if (!table->v) { + table->size = 0; + return NULL; + } + memset(table->v, 0, tsize); + i = hash(parser, name) & ((unsigned long)table->size - 1); + } + else { + unsigned long h = hash(parser, name); + unsigned long mask = (unsigned long)table->size - 1; + unsigned char step = 0; + i = h & mask; + while (table->v[i]) { + if (keyeq(name, table->v[i]->name)) + return table->v[i]; + if (!step) + step = PROBE_STEP(h, mask, table->power); + i < step ? (i += table->size - step) : (i -= step); + } + if (!createSize) + return NULL; + + /* check for overflow (table is half full) */ + if (table->used >> (table->power - 1)) { + unsigned char newPower = table->power + 1; + size_t newSize = (size_t)1 << newPower; + unsigned long newMask = (unsigned long)newSize - 1; + size_t tsize = newSize * sizeof(NAMED *); + NAMED **newV = (NAMED **)table->mem->malloc_fcn(tsize); + if (!newV) + return NULL; + memset(newV, 0, tsize); + for (i = 0; i < table->size; i++) + if (table->v[i]) { + unsigned long newHash = hash(parser, table->v[i]->name); + size_t j = newHash & newMask; + step = 0; + while (newV[j]) { + if (!step) + step = PROBE_STEP(newHash, newMask, newPower); + j < step ? (j += newSize - step) : (j -= step); + } + newV[j] = table->v[i]; + } + table->mem->free_fcn(table->v); + table->v = newV; + table->power = newPower; + table->size = newSize; + i = h & newMask; + step = 0; + while (table->v[i]) { + if (!step) + step = PROBE_STEP(h, newMask, newPower); + i < step ? (i += newSize - step) : (i -= step); + } + } + } + table->v[i] = (NAMED *)table->mem->malloc_fcn(createSize); + if (!table->v[i]) + return NULL; + memset(table->v[i], 0, createSize); + table->v[i]->name = name; + (table->used)++; + return table->v[i]; +} + +static void FASTCALL +hashTableClear(HASH_TABLE *table) +{ + size_t i; + for (i = 0; i < table->size; i++) { + table->mem->free_fcn(table->v[i]); + table->v[i] = NULL; + } + table->used = 0; +} + +static void FASTCALL +hashTableDestroy(HASH_TABLE *table) +{ + size_t i; + for (i = 0; i < table->size; i++) + table->mem->free_fcn(table->v[i]); + table->mem->free_fcn(table->v); +} + +static void FASTCALL +hashTableInit(HASH_TABLE *p, const XML_Memory_Handling_Suite *ms) +{ + p->power = 0; + p->size = 0; + p->used = 0; + p->v = NULL; + p->mem = ms; +} + +static void FASTCALL +hashTableIterInit(HASH_TABLE_ITER *iter, const HASH_TABLE *table) +{ + iter->p = table->v; + iter->end = iter->p + table->size; +} + +static NAMED * FASTCALL +hashTableIterNext(HASH_TABLE_ITER *iter) +{ + while (iter->p != iter->end) { + NAMED *tem = *(iter->p)++; + if (tem) + return tem; + } + return NULL; +} + +static void FASTCALL +poolInit(STRING_POOL *pool, const XML_Memory_Handling_Suite *ms) +{ + pool->blocks = NULL; + pool->freeBlocks = NULL; + pool->start = NULL; + pool->ptr = NULL; + pool->end = NULL; + pool->mem = ms; +} + +static void FASTCALL +poolClear(STRING_POOL *pool) +{ + if (!pool->freeBlocks) + pool->freeBlocks = pool->blocks; + else { + BLOCK *p = pool->blocks; + while (p) { + BLOCK *tem = p->next; + p->next = pool->freeBlocks; + pool->freeBlocks = p; + p = tem; + } + } + pool->blocks = NULL; + pool->start = NULL; + pool->ptr = NULL; + pool->end = NULL; +} + +static void FASTCALL +poolDestroy(STRING_POOL *pool) +{ + BLOCK *p = pool->blocks; + while (p) { + BLOCK *tem = p->next; + pool->mem->free_fcn(p); + p = tem; + } + p = pool->freeBlocks; + while (p) { + BLOCK *tem = p->next; + pool->mem->free_fcn(p); + p = tem; + } +} + +static XML_Char * +poolAppend(STRING_POOL *pool, const ENCODING *enc, + const char *ptr, const char *end) +{ + if (!pool->ptr && !poolGrow(pool)) + return NULL; + for (;;) { + const enum XML_Convert_Result convert_res = XmlConvert(enc, &ptr, end, (ICHAR **)&(pool->ptr), (ICHAR *)pool->end); + if ((convert_res == XML_CONVERT_COMPLETED) || (convert_res == XML_CONVERT_INPUT_INCOMPLETE)) + break; + if (!poolGrow(pool)) + return NULL; + } + return pool->start; +} + +static const XML_Char * FASTCALL +poolCopyString(STRING_POOL *pool, const XML_Char *s) +{ + do { + if (!poolAppendChar(pool, *s)) + return NULL; + } while (*s++); + s = pool->start; + poolFinish(pool); + return s; +} + +static const XML_Char * +poolCopyStringN(STRING_POOL *pool, const XML_Char *s, int n) +{ + if (!pool->ptr && !poolGrow(pool)) + return NULL; + for (; n > 0; --n, s++) { + if (!poolAppendChar(pool, *s)) + return NULL; + } + s = pool->start; + poolFinish(pool); + return s; +} + +static const XML_Char * FASTCALL +poolAppendString(STRING_POOL *pool, const XML_Char *s) +{ + while (*s) { + if (!poolAppendChar(pool, *s)) + return NULL; + s++; + } + return pool->start; +} + +static XML_Char * +poolStoreString(STRING_POOL *pool, const ENCODING *enc, + const char *ptr, const char *end) +{ + if (!poolAppend(pool, enc, ptr, end)) + return NULL; + if (pool->ptr == pool->end && !poolGrow(pool)) + return NULL; + *(pool->ptr)++ = 0; + return pool->start; +} + +static size_t +poolBytesToAllocateFor(int blockSize) +{ + /* Unprotected math would be: + ** return offsetof(BLOCK, s) + blockSize * sizeof(XML_Char); + ** + ** Detect overflow, avoiding _signed_ overflow undefined behavior + ** For a + b * c we check b * c in isolation first, so that addition of a + ** on top has no chance of making us accept a small non-negative number + */ + const size_t stretch = sizeof(XML_Char); /* can be 4 bytes */ + + if (blockSize <= 0) + return 0; + + if (blockSize > (int)(INT_MAX / stretch)) + return 0; + + { + const int stretchedBlockSize = blockSize * (int)stretch; + const int bytesToAllocate = (int)( + offsetof(BLOCK, s) + (unsigned)stretchedBlockSize); + if (bytesToAllocate < 0) + return 0; + + return (size_t)bytesToAllocate; + } +} + +static XML_Bool FASTCALL +poolGrow(STRING_POOL *pool) +{ + if (pool->freeBlocks) { + if (pool->start == 0) { + pool->blocks = pool->freeBlocks; + pool->freeBlocks = pool->freeBlocks->next; + pool->blocks->next = NULL; + pool->start = pool->blocks->s; + pool->end = pool->start + pool->blocks->size; + pool->ptr = pool->start; + return XML_TRUE; + } + if (pool->end - pool->start < pool->freeBlocks->size) { + BLOCK *tem = pool->freeBlocks->next; + pool->freeBlocks->next = pool->blocks; + pool->blocks = pool->freeBlocks; + pool->freeBlocks = tem; + memcpy(pool->blocks->s, pool->start, + (pool->end - pool->start) * sizeof(XML_Char)); + pool->ptr = pool->blocks->s + (pool->ptr - pool->start); + pool->start = pool->blocks->s; + pool->end = pool->start + pool->blocks->size; + return XML_TRUE; + } + } + if (pool->blocks && pool->start == pool->blocks->s) { + BLOCK *temp; + int blockSize = (int)((unsigned)(pool->end - pool->start)*2U); + size_t bytesToAllocate; + + if (blockSize < 0) + return XML_FALSE; + + bytesToAllocate = poolBytesToAllocateFor(blockSize); + if (bytesToAllocate == 0) + return XML_FALSE; + + temp = (BLOCK *) + pool->mem->realloc_fcn(pool->blocks, (unsigned)bytesToAllocate); + if (temp == NULL) + return XML_FALSE; + pool->blocks = temp; + pool->blocks->size = blockSize; + pool->ptr = pool->blocks->s + (pool->ptr - pool->start); + pool->start = pool->blocks->s; + pool->end = pool->start + blockSize; + } + else { + BLOCK *tem; + int blockSize = (int)(pool->end - pool->start); + size_t bytesToAllocate; + + if (blockSize < 0) + return XML_FALSE; + + if (blockSize < INIT_BLOCK_SIZE) + blockSize = INIT_BLOCK_SIZE; + else { + /* Detect overflow, avoiding _signed_ overflow undefined behavior */ + if ((int)((unsigned)blockSize * 2U) < 0) { + return XML_FALSE; + } + blockSize *= 2; + } + + bytesToAllocate = poolBytesToAllocateFor(blockSize); + if (bytesToAllocate == 0) + return XML_FALSE; + + tem = (BLOCK *)pool->mem->malloc_fcn(bytesToAllocate); + if (!tem) + return XML_FALSE; + tem->size = blockSize; + tem->next = pool->blocks; + pool->blocks = tem; + if (pool->ptr != pool->start) + memcpy(tem->s, pool->start, + (pool->ptr - pool->start) * sizeof(XML_Char)); + pool->ptr = tem->s + (pool->ptr - pool->start); + pool->start = tem->s; + pool->end = tem->s + blockSize; + } + return XML_TRUE; +} + +static int FASTCALL +nextScaffoldPart(XML_Parser parser) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + CONTENT_SCAFFOLD * me; + int next; + + if (!dtd->scaffIndex) { + dtd->scaffIndex = (int *)MALLOC(groupSize * sizeof(int)); + if (!dtd->scaffIndex) + return -1; + dtd->scaffIndex[0] = 0; + } + + if (dtd->scaffCount >= dtd->scaffSize) { + CONTENT_SCAFFOLD *temp; + if (dtd->scaffold) { + temp = (CONTENT_SCAFFOLD *) + REALLOC(dtd->scaffold, dtd->scaffSize * 2 * sizeof(CONTENT_SCAFFOLD)); + if (temp == NULL) + return -1; + dtd->scaffSize *= 2; + } + else { + temp = (CONTENT_SCAFFOLD *)MALLOC(INIT_SCAFFOLD_ELEMENTS + * sizeof(CONTENT_SCAFFOLD)); + if (temp == NULL) + return -1; + dtd->scaffSize = INIT_SCAFFOLD_ELEMENTS; + } + dtd->scaffold = temp; + } + next = dtd->scaffCount++; + me = &dtd->scaffold[next]; + if (dtd->scaffLevel) { + CONTENT_SCAFFOLD *parent = &dtd->scaffold[dtd->scaffIndex[dtd->scaffLevel-1]]; + if (parent->lastchild) { + dtd->scaffold[parent->lastchild].nextsib = next; + } + if (!parent->childcnt) + parent->firstchild = next; + parent->lastchild = next; + parent->childcnt++; + } + me->firstchild = me->lastchild = me->childcnt = me->nextsib = 0; + return next; +} + +static void +build_node(XML_Parser parser, + int src_node, + XML_Content *dest, + XML_Content **contpos, + XML_Char **strpos) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + dest->type = dtd->scaffold[src_node].type; + dest->quant = dtd->scaffold[src_node].quant; + if (dest->type == XML_CTYPE_NAME) { + const XML_Char *src; + dest->name = *strpos; + src = dtd->scaffold[src_node].name; + for (;;) { + *(*strpos)++ = *src; + if (!*src) + break; + src++; + } + dest->numchildren = 0; + dest->children = NULL; + } + else { + unsigned int i; + int cn; + dest->numchildren = dtd->scaffold[src_node].childcnt; + dest->children = *contpos; + *contpos += dest->numchildren; + for (i = 0, cn = dtd->scaffold[src_node].firstchild; + i < dest->numchildren; + i++, cn = dtd->scaffold[cn].nextsib) { + build_node(parser, cn, &(dest->children[i]), contpos, strpos); + } + dest->name = NULL; + } +} + +static XML_Content * +build_model (XML_Parser parser) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + XML_Content *ret; + XML_Content *cpos; + XML_Char * str; + int allocsize = (dtd->scaffCount * sizeof(XML_Content) + + (dtd->contentStringLen * sizeof(XML_Char))); + + ret = (XML_Content *)MALLOC(allocsize); + if (!ret) + return NULL; + + str = (XML_Char *) (&ret[dtd->scaffCount]); + cpos = &ret[1]; + + build_node(parser, 0, ret, &cpos, &str); + return ret; +} + +static ELEMENT_TYPE * +getElementType(XML_Parser parser, + const ENCODING *enc, + const char *ptr, + const char *end) +{ + DTD * const dtd = _dtd; /* save one level of indirection */ + const XML_Char *name = poolStoreString(&dtd->pool, enc, ptr, end); + ELEMENT_TYPE *ret; + + if (!name) + return NULL; + ret = (ELEMENT_TYPE *) lookup(parser, &dtd->elementTypes, name, sizeof(ELEMENT_TYPE)); if (!ret) return NULL; if (ret->name != name) diff --git a/cbits/xmlrole.c b/cbits/xmlrole.c --- a/cbits/xmlrole.c +++ b/cbits/xmlrole.c @@ -4,19 +4,13 @@ #include -#ifdef COMPILED_FROM_DSP +#ifdef _WIN32 #include "winconfig.h" -#elif defined(MACOS_CLASSIC) -#include "macconfig.h" -#elif defined(__amigaos4__) -#include "amigaconfig.h" -#elif defined(__WATCOMC__) -#include "watcomconfig.h" #else #ifdef HAVE_EXPAT_CONFIG_H #include #endif -#endif /* ndef COMPILED_FROM_DSP */ +#endif /* ndef _WIN32 */ #include "expat_external.h" #include "internal.h" @@ -195,9 +189,9 @@ static int PTRCALL prolog2(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -216,9 +210,9 @@ static int PTRCALL doctype0(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -264,9 +258,9 @@ static int PTRCALL doctype2(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -281,9 +275,9 @@ static int PTRCALL doctype3(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -298,9 +292,9 @@ static int PTRCALL doctype4(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -318,9 +312,9 @@ static int PTRCALL doctype5(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -437,9 +431,9 @@ static int PTRCALL entity0(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -457,9 +451,9 @@ static int PTRCALL entity1(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -502,9 +496,9 @@ static int PTRCALL entity3(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -519,9 +513,9 @@ static int PTRCALL entity4(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -559,9 +553,9 @@ static int PTRCALL entity6(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -605,9 +599,9 @@ static int PTRCALL entity8(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -622,9 +616,9 @@ static int PTRCALL entity9(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -639,9 +633,9 @@ static int PTRCALL entity10(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -656,9 +650,9 @@ static int PTRCALL notation0(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -697,9 +691,9 @@ static int PTRCALL notation2(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -714,9 +708,9 @@ static int PTRCALL notation3(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -732,9 +726,9 @@ static int PTRCALL notation4(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -753,9 +747,9 @@ static int PTRCALL attlist0(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -771,9 +765,9 @@ static int PTRCALL attlist1(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -833,9 +827,9 @@ static int PTRCALL attlist3(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -852,9 +846,9 @@ static int PTRCALL attlist4(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -872,9 +866,9 @@ static int PTRCALL attlist5(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -889,9 +883,9 @@ static int PTRCALL attlist6(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -906,9 +900,9 @@ static int PTRCALL attlist7(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -967,9 +961,9 @@ static int PTRCALL attlist9(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -984,9 +978,9 @@ static int PTRCALL element0(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1072,9 +1066,9 @@ static int PTRCALL element3(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1097,9 +1091,9 @@ static int PTRCALL element4(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1115,9 +1109,9 @@ static int PTRCALL element5(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1136,9 +1130,9 @@ static int PTRCALL element6(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1166,9 +1160,9 @@ static int PTRCALL element7(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1240,9 +1234,9 @@ static int PTRCALL condSect1(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1258,9 +1252,9 @@ static int PTRCALL condSect2(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1277,9 +1271,9 @@ static int PTRCALL declClose(PROLOG_STATE *state, int tok, - const char *ptr, - const char *end, - const ENCODING *enc) + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { switch (tok) { case XML_TOK_PROLOG_S: @@ -1292,11 +1286,11 @@ } static int PTRCALL -error(PROLOG_STATE *state, - int tok, - const char *ptr, - const char *end, - const ENCODING *enc) +error(PROLOG_STATE *UNUSED_P(state), + int UNUSED_P(tok), + const char *UNUSED_P(ptr), + const char *UNUSED_P(end), + const ENCODING *UNUSED_P(enc)) { return XML_ROLE_NONE; } diff --git a/cbits/xmltok.c b/cbits/xmltok.c --- a/cbits/xmltok.c +++ b/cbits/xmltok.c @@ -4,19 +4,13 @@ #include -#ifdef COMPILED_FROM_DSP +#ifdef _WIN32 #include "winconfig.h" -#elif defined(MACOS_CLASSIC) -#include "macconfig.h" -#elif defined(__amigaos4__) -#include "amigaconfig.h" -#elif defined(__WATCOMC__) -#include "watcomconfig.h" #else #ifdef HAVE_EXPAT_CONFIG_H #include #endif -#endif /* ndef COMPILED_FROM_DSP */ +#endif /* ndef _WIN32 */ #include "expat_external.h" #include "internal.h" @@ -46,7 +40,7 @@ #define VTABLE VTABLE1, PREFIX(toUtf8), PREFIX(toUtf16) #define UCS2_GET_NAMING(pages, hi, lo) \ - (namingBitmap[(pages[hi] << 3) + ((lo) >> 5)] & (1 << ((lo) & 0x1F))) + (namingBitmap[(pages[hi] << 3) + ((lo) >> 5)] & (1u << ((lo) & 0x1F))) /* A 2 byte UTF-8 representation splits the characters 11 bits between the bottom 5 and 6 bits of the bytes. We need 8 bits to index into @@ -56,7 +50,7 @@ (namingBitmap[((pages)[(((byte)[0]) >> 2) & 7] << 3) \ + ((((byte)[0]) & 3) << 1) \ + ((((byte)[1]) >> 5) & 1)] \ - & (1 << (((byte)[1]) & 0x1F))) + & (1u << (((byte)[1]) & 0x1F))) /* A 3 byte UTF-8 representation splits the characters 16 bits between the bottom 4, 6 and 6 bits of the bytes. We need 8 bits to index @@ -69,7 +63,7 @@ << 3) \ + ((((byte)[1]) & 3) << 1) \ + ((((byte)[2]) >> 5) & 1)] \ - & (1 << (((byte)[2]) & 0x1F))) + & (1u << (((byte)[2]) & 0x1F))) #define UTF8_GET_NAMING(pages, p, n) \ ((n) == 2 \ @@ -122,19 +116,19 @@ ((*p) == 0xF4 ? (p)[1] > 0x8F : ((p)[1] & 0xC0) == 0xC0))) static int PTRFASTCALL -isNever(const ENCODING *enc, const char *p) +isNever(const ENCODING *UNUSED_P(enc), const char *UNUSED_P(p)) { return 0; } static int PTRFASTCALL -utf8_isName2(const ENCODING *enc, const char *p) +utf8_isName2(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_GET_NAMING2(namePages, (const unsigned char *)p); } static int PTRFASTCALL -utf8_isName3(const ENCODING *enc, const char *p) +utf8_isName3(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_GET_NAMING3(namePages, (const unsigned char *)p); } @@ -142,13 +136,13 @@ #define utf8_isName4 isNever static int PTRFASTCALL -utf8_isNmstrt2(const ENCODING *enc, const char *p) +utf8_isNmstrt2(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_GET_NAMING2(nmstrtPages, (const unsigned char *)p); } static int PTRFASTCALL -utf8_isNmstrt3(const ENCODING *enc, const char *p) +utf8_isNmstrt3(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_GET_NAMING3(nmstrtPages, (const unsigned char *)p); } @@ -156,19 +150,19 @@ #define utf8_isNmstrt4 isNever static int PTRFASTCALL -utf8_isInvalid2(const ENCODING *enc, const char *p) +utf8_isInvalid2(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_INVALID2((const unsigned char *)p); } static int PTRFASTCALL -utf8_isInvalid3(const ENCODING *enc, const char *p) +utf8_isInvalid3(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_INVALID3((const unsigned char *)p); } static int PTRFASTCALL -utf8_isInvalid4(const ENCODING *enc, const char *p) +utf8_isInvalid4(const ENCODING *UNUSED_P(enc), const char *p) { return UTF8_INVALID4((const unsigned char *)p); } @@ -222,6 +216,17 @@ E ## isInvalid3, \ E ## isInvalid4 +#define NULL_VTABLE \ + /* isName2 */ NULL, \ + /* isName3 */ NULL, \ + /* isName4 */ NULL, \ + /* isNmstrt2 */ NULL, \ + /* isNmstrt3 */ NULL, \ + /* isNmstrt4 */ NULL, \ + /* isInvalid2 */ NULL, \ + /* isInvalid3 */ NULL, \ + /* isInvalid4 */ NULL + static int FASTCALL checkCharRefNumber(int); #include "xmltok_impl.h" @@ -318,39 +323,89 @@ UTF8_cval4 = 0xf0 }; -static void PTRCALL -utf8_toUtf8(const ENCODING *enc, +void +align_limit_to_full_utf8_characters(const char * from, const char ** fromLimRef) +{ + const char * fromLim = *fromLimRef; + size_t walked = 0; + for (; fromLim > from; fromLim--, walked++) { + const unsigned char prev = (unsigned char)fromLim[-1]; + if ((prev & 0xf8u) == 0xf0u) { /* 4-byte character, lead by 0b11110xxx byte */ + if (walked + 1 >= 4) { + fromLim += 4 - 1; + break; + } else { + walked = 0; + } + } else if ((prev & 0xf0u) == 0xe0u) { /* 3-byte character, lead by 0b1110xxxx byte */ + if (walked + 1 >= 3) { + fromLim += 3 - 1; + break; + } else { + walked = 0; + } + } else if ((prev & 0xe0u) == 0xc0u) { /* 2-byte character, lead by 0b110xxxxx byte */ + if (walked + 1 >= 2) { + fromLim += 2 - 1; + break; + } else { + walked = 0; + } + } else if ((prev & 0x80u) == 0x00u) { /* 1-byte character, matching 0b0xxxxxxx */ + break; + } + } + *fromLimRef = fromLim; +} + +static enum XML_Convert_Result PTRCALL +utf8_toUtf8(const ENCODING *UNUSED_P(enc), const char **fromP, const char *fromLim, char **toP, const char *toLim) { char *to; const char *from; - if (fromLim - *fromP > toLim - *toP) { - /* Avoid copying partial characters. */ - for (fromLim = *fromP + (toLim - *toP); fromLim > *fromP; fromLim--) - if (((unsigned char)fromLim[-1] & 0xc0) != 0x80) - break; - } - for (to = *toP, from = *fromP; from != fromLim; from++, to++) + const char *fromLimInitial = fromLim; + + /* Avoid copying partial characters. */ + align_limit_to_full_utf8_characters(*fromP, &fromLim); + + for (to = *toP, from = *fromP; (from < fromLim) && (to < toLim); from++, to++) *to = *from; *fromP = from; *toP = to; + + if (fromLim < fromLimInitial) + return XML_CONVERT_INPUT_INCOMPLETE; + else if ((to == toLim) && (from < fromLim)) + return XML_CONVERT_OUTPUT_EXHAUSTED; + else + return XML_CONVERT_COMPLETED; } -static void PTRCALL +static enum XML_Convert_Result PTRCALL utf8_toUtf16(const ENCODING *enc, const char **fromP, const char *fromLim, unsigned short **toP, const unsigned short *toLim) { + enum XML_Convert_Result res = XML_CONVERT_COMPLETED; unsigned short *to = *toP; const char *from = *fromP; - while (from != fromLim && to != toLim) { + while (from < fromLim && to < toLim) { switch (((struct normal_encoding *)enc)->type[(unsigned char)*from]) { case BT_LEAD2: + if (fromLim - from < 2) { + res = XML_CONVERT_INPUT_INCOMPLETE; + goto after; + } *to++ = (unsigned short)(((from[0] & 0x1f) << 6) | (from[1] & 0x3f)); from += 2; break; case BT_LEAD3: + if (fromLim - from < 3) { + res = XML_CONVERT_INPUT_INCOMPLETE; + goto after; + } *to++ = (unsigned short)(((from[0] & 0xf) << 12) | ((from[1] & 0x3f) << 6) | (from[2] & 0x3f)); from += 3; @@ -358,8 +413,14 @@ case BT_LEAD4: { unsigned long n; - if (to + 1 == toLim) + if (toLim - to < 2) { + res = XML_CONVERT_OUTPUT_EXHAUSTED; goto after; + } + if (fromLim - from < 4) { + res = XML_CONVERT_INPUT_INCOMPLETE; + goto after; + } n = ((from[0] & 0x7) << 18) | ((from[1] & 0x3f) << 12) | ((from[2] & 0x3f) << 6) | (from[3] & 0x3f); n -= 0x10000; @@ -374,9 +435,12 @@ break; } } + if (from < fromLim) + res = XML_CONVERT_OUTPUT_EXHAUSTED; after: *fromP = from; *toP = to; + return res; } #ifdef XML_NS @@ -425,38 +489,43 @@ STANDARD_VTABLE(sb_) NORMAL_VTABLE(utf8_) }; -static void PTRCALL -latin1_toUtf8(const ENCODING *enc, +static enum XML_Convert_Result PTRCALL +latin1_toUtf8(const ENCODING *UNUSED_P(enc), const char **fromP, const char *fromLim, char **toP, const char *toLim) { for (;;) { unsigned char c; if (*fromP == fromLim) - break; + return XML_CONVERT_COMPLETED; c = (unsigned char)**fromP; if (c & 0x80) { if (toLim - *toP < 2) - break; + return XML_CONVERT_OUTPUT_EXHAUSTED; *(*toP)++ = (char)((c >> 6) | UTF8_cval2); *(*toP)++ = (char)((c & 0x3f) | 0x80); (*fromP)++; } else { if (*toP == toLim) - break; + return XML_CONVERT_OUTPUT_EXHAUSTED; *(*toP)++ = *(*fromP)++; } } } -static void PTRCALL -latin1_toUtf16(const ENCODING *enc, +static enum XML_Convert_Result PTRCALL +latin1_toUtf16(const ENCODING *UNUSED_P(enc), const char **fromP, const char *fromLim, unsigned short **toP, const unsigned short *toLim) { - while (*fromP != fromLim && *toP != toLim) + while (*fromP < fromLim && *toP < toLim) *(*toP)++ = (unsigned char)*(*fromP)++; + + if ((*toP == toLim) && (*fromP < fromLim)) + return XML_CONVERT_OUTPUT_EXHAUSTED; + else + return XML_CONVERT_COMPLETED; } #ifdef XML_NS @@ -467,7 +536,7 @@ #include "asciitab.h" #include "latin1tab.h" }, - STANDARD_VTABLE(sb_) + STANDARD_VTABLE(sb_) NULL_VTABLE }; #endif @@ -480,16 +549,21 @@ #undef BT_COLON #include "latin1tab.h" }, - STANDARD_VTABLE(sb_) + STANDARD_VTABLE(sb_) NULL_VTABLE }; -static void PTRCALL -ascii_toUtf8(const ENCODING *enc, +static enum XML_Convert_Result PTRCALL +ascii_toUtf8(const ENCODING *UNUSED_P(enc), const char **fromP, const char *fromLim, char **toP, const char *toLim) { - while (*fromP != fromLim && *toP != toLim) + while (*fromP < fromLim && *toP < toLim) *(*toP)++ = *(*fromP)++; + + if ((*toP == toLim) && (*fromP < fromLim)) + return XML_CONVERT_OUTPUT_EXHAUSTED; + else + return XML_CONVERT_COMPLETED; } #ifdef XML_NS @@ -500,7 +574,7 @@ #include "asciitab.h" /* BT_NONXML == 0 */ }, - STANDARD_VTABLE(sb_) + STANDARD_VTABLE(sb_) NULL_VTABLE }; #endif @@ -513,7 +587,7 @@ #undef BT_COLON /* BT_NONXML == 0 */ }, - STANDARD_VTABLE(sb_) + STANDARD_VTABLE(sb_) NULL_VTABLE }; static int PTRFASTCALL @@ -536,13 +610,14 @@ } #define DEFINE_UTF16_TO_UTF8(E) \ -static void PTRCALL \ -E ## toUtf8(const ENCODING *enc, \ +static enum XML_Convert_Result PTRCALL \ +E ## toUtf8(const ENCODING *UNUSED_P(enc), \ const char **fromP, const char *fromLim, \ char **toP, const char *toLim) \ { \ - const char *from; \ - for (from = *fromP; from != fromLim; from += 2) { \ + const char *from = *fromP; \ + fromLim = from + (((fromLim - from) >> 1) << 1); /* shrink to even */ \ + for (; from < fromLim; from += 2) { \ int plane; \ unsigned char lo2; \ unsigned char lo = GET_LO(from); \ @@ -552,7 +627,7 @@ if (lo < 0x80) { \ if (*toP == toLim) { \ *fromP = from; \ - return; \ + return XML_CONVERT_OUTPUT_EXHAUSTED; \ } \ *(*toP)++ = lo; \ break; \ @@ -562,7 +637,7 @@ case 0x4: case 0x5: case 0x6: case 0x7: \ if (toLim - *toP < 2) { \ *fromP = from; \ - return; \ + return XML_CONVERT_OUTPUT_EXHAUSTED; \ } \ *(*toP)++ = ((lo >> 6) | (hi << 2) | UTF8_cval2); \ *(*toP)++ = ((lo & 0x3f) | 0x80); \ @@ -570,7 +645,7 @@ default: \ if (toLim - *toP < 3) { \ *fromP = from; \ - return; \ + return XML_CONVERT_OUTPUT_EXHAUSTED; \ } \ /* 16 bits divided 4, 6, 6 amongst 3 bytes */ \ *(*toP)++ = ((hi >> 4) | UTF8_cval3); \ @@ -580,8 +655,12 @@ case 0xD8: case 0xD9: case 0xDA: case 0xDB: \ if (toLim - *toP < 4) { \ *fromP = from; \ - return; \ + return XML_CONVERT_OUTPUT_EXHAUSTED; \ } \ + if (fromLim - from < 4) { \ + *fromP = from; \ + return XML_CONVERT_INPUT_INCOMPLETE; \ + } \ plane = (((hi & 0x3) << 2) | ((lo >> 6) & 0x3)) + 1; \ *(*toP)++ = ((plane >> 2) | UTF8_cval4); \ *(*toP)++ = (((lo >> 2) & 0xF) | ((plane & 0x3) << 4) | 0x80); \ @@ -596,20 +675,32 @@ } \ } \ *fromP = from; \ + if (from < fromLim) \ + return XML_CONVERT_INPUT_INCOMPLETE; \ + else \ + return XML_CONVERT_COMPLETED; \ } #define DEFINE_UTF16_TO_UTF16(E) \ -static void PTRCALL \ -E ## toUtf16(const ENCODING *enc, \ +static enum XML_Convert_Result PTRCALL \ +E ## toUtf16(const ENCODING *UNUSED_P(enc), \ const char **fromP, const char *fromLim, \ unsigned short **toP, const unsigned short *toLim) \ { \ + enum XML_Convert_Result res = XML_CONVERT_COMPLETED; \ + fromLim = *fromP + (((fromLim - *fromP) >> 1) << 1); /* shrink to even */ \ /* Avoid copying first half only of surrogate */ \ if (fromLim - *fromP > ((toLim - *toP) << 1) \ - && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) \ + && (GET_HI(fromLim - 2) & 0xF8) == 0xD8) { \ fromLim -= 2; \ - for (; *fromP != fromLim && *toP != toLim; *fromP += 2) \ + res = XML_CONVERT_INPUT_INCOMPLETE; \ + } \ + for (; *fromP < fromLim && *toP < toLim; *fromP += 2) \ *(*toP)++ = (GET_HI(*fromP) << 8) | GET_LO(*fromP); \ + if ((*toP == toLim) && (*fromP < fromLim)) \ + return XML_CONVERT_OUTPUT_EXHAUSTED; \ + else \ + return res; \ } #define SET2(ptr, ch) \ @@ -726,7 +817,7 @@ #include "asciitab.h" #include "latin1tab.h" }, - STANDARD_VTABLE(little2_) + STANDARD_VTABLE(little2_) NULL_VTABLE }; #endif @@ -745,7 +836,7 @@ #undef BT_COLON #include "latin1tab.h" }, - STANDARD_VTABLE(little2_) + STANDARD_VTABLE(little2_) NULL_VTABLE }; #if BYTEORDER != 4321 @@ -758,7 +849,7 @@ #include "iasciitab.h" #include "latin1tab.h" }, - STANDARD_VTABLE(little2_) + STANDARD_VTABLE(little2_) NULL_VTABLE }; #endif @@ -771,7 +862,7 @@ #undef BT_COLON #include "latin1tab.h" }, - STANDARD_VTABLE(little2_) + STANDARD_VTABLE(little2_) NULL_VTABLE }; #endif @@ -867,7 +958,7 @@ #include "asciitab.h" #include "latin1tab.h" }, - STANDARD_VTABLE(big2_) + STANDARD_VTABLE(big2_) NULL_VTABLE }; #endif @@ -886,7 +977,7 @@ #undef BT_COLON #include "latin1tab.h" }, - STANDARD_VTABLE(big2_) + STANDARD_VTABLE(big2_) NULL_VTABLE }; #if BYTEORDER != 1234 @@ -899,7 +990,7 @@ #include "iasciitab.h" #include "latin1tab.h" }, - STANDARD_VTABLE(big2_) + STANDARD_VTABLE(big2_) NULL_VTABLE }; #endif @@ -912,7 +1003,7 @@ #undef BT_COLON #include "latin1tab.h" }, - STANDARD_VTABLE(big2_) + STANDARD_VTABLE(big2_) NULL_VTABLE }; #endif @@ -938,7 +1029,7 @@ } static void PTRCALL -initUpdatePosition(const ENCODING *enc, const char *ptr, +initUpdatePosition(const ENCODING *UNUSED_P(enc), const char *ptr, const char *end, POSITION *pos) { normal_updatePosition(&utf8_encoding.enc, ptr, end, pos); @@ -1288,7 +1379,7 @@ return (c & ~0xFFFF) || checkCharRefNumber(c) < 0; } -static void PTRCALL +static enum XML_Convert_Result PTRCALL unknown_toUtf8(const ENCODING *enc, const char **fromP, const char *fromLim, char **toP, const char *toLim) @@ -1299,21 +1390,21 @@ const char *utf8; int n; if (*fromP == fromLim) - break; + return XML_CONVERT_COMPLETED; utf8 = uenc->utf8[(unsigned char)**fromP]; n = *utf8++; if (n == 0) { int c = uenc->convert(uenc->userData, *fromP); n = XmlUtf8Encode(c, buf); if (n > toLim - *toP) - break; + return XML_CONVERT_OUTPUT_EXHAUSTED; utf8 = buf; *fromP += (AS_NORMAL_ENCODING(enc)->type[(unsigned char)**fromP] - (BT_LEAD2 - 2)); } else { if (n > toLim - *toP) - break; + return XML_CONVERT_OUTPUT_EXHAUSTED; (*fromP)++; } do { @@ -1322,13 +1413,13 @@ } } -static void PTRCALL +static enum XML_Convert_Result PTRCALL unknown_toUtf16(const ENCODING *enc, const char **fromP, const char *fromLim, unsigned short **toP, const unsigned short *toLim) { const struct unknown_encoding *uenc = AS_UNKNOWN_ENCODING(enc); - while (*fromP != fromLim && *toP != toLim) { + while (*fromP < fromLim && *toP < toLim) { unsigned short c = uenc->utf16[(unsigned char)**fromP]; if (c == 0) { c = (unsigned short) @@ -1340,12 +1431,17 @@ (*fromP)++; *(*toP)++ = c; } + + if ((*toP == toLim) && (*fromP < fromLim)) + return XML_CONVERT_OUTPUT_EXHAUSTED; + else + return XML_CONVERT_COMPLETED; } ENCODING * XmlInitUnknownEncoding(void *mem, int *table, - CONVERTER convert, + CONVERTER convert, void *userData) { int i; @@ -1503,7 +1599,7 @@ { const ENCODING **encPtr; - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; encPtr = enc->encPtr; if (ptr + 1 == end) { @@ -1639,7 +1735,7 @@ ENCODING * XmlInitUnknownEncodingNS(void *mem, int *table, - CONVERTER convert, + CONVERTER convert, void *userData) { ENCODING *enc = XmlInitUnknownEncoding(mem, table, convert, userData); diff --git a/cbits/xmltok.h b/cbits/xmltok.h --- a/cbits/xmltok.h +++ b/cbits/xmltok.h @@ -130,6 +130,12 @@ const char *, const char **); +enum XML_Convert_Result { + XML_CONVERT_COMPLETED = 0, + XML_CONVERT_INPUT_INCOMPLETE = 1, + XML_CONVERT_OUTPUT_EXHAUSTED = 2 /* and therefore potentially input remaining as well */ +}; + struct encoding { SCANNER scanners[XML_N_STATES]; SCANNER literalScanners[XML_N_LITERAL_TYPES]; @@ -158,12 +164,12 @@ const char *ptr, const char *end, const char **badPtr); - void (PTRCALL *utf8Convert)(const ENCODING *enc, + enum XML_Convert_Result (PTRCALL *utf8Convert)(const ENCODING *enc, const char **fromP, const char *fromLim, char **toP, const char *toLim); - void (PTRCALL *utf16Convert)(const ENCODING *enc, + enum XML_Convert_Result (PTRCALL *utf16Convert)(const ENCODING *enc, const char **fromP, const char *fromLim, unsigned short **toP, diff --git a/cbits/xmltok_impl.c b/cbits/xmltok_impl.c --- a/cbits/xmltok_impl.c +++ b/cbits/xmltok_impl.c @@ -87,27 +87,45 @@ #define PREFIX(ident) ident #endif + +#define HAS_CHARS(enc, ptr, end, count) \ + (end - ptr >= count * MINBPC(enc)) + +#define HAS_CHAR(enc, ptr, end) \ + HAS_CHARS(enc, ptr, end, 1) + +#define REQUIRE_CHARS(enc, ptr, end, count) \ + { \ + if (! HAS_CHARS(enc, ptr, end, count)) { \ + return XML_TOK_PARTIAL; \ + } \ + } + +#define REQUIRE_CHAR(enc, ptr, end) \ + REQUIRE_CHARS(enc, ptr, end, 1) + + /* ptr points to character following " */ switch (BYTE_TYPE(enc, ptr + MINBPC(enc))) { case BT_S: case BT_CR: case BT_LF: case BT_PERCNT: @@ -175,7 +191,7 @@ } static int PTRCALL -PREFIX(checkPiTarget)(const ENCODING *enc, const char *ptr, +PREFIX(checkPiTarget)(const ENCODING *UNUSED_P(enc), const char *ptr, const char *end, int *tokPtr) { int upper = 0; @@ -225,15 +241,14 @@ { int tok; const char *target = ptr; - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) default: *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_S: case BT_CR: case BT_LF: @@ -242,13 +257,12 @@ return XML_TOK_INVALID; } ptr += MINBPC(enc); - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { INVALID_CASES(ptr, nextTokPtr) case BT_QUEST: ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_GT)) { *nextTokPtr = ptr + MINBPC(enc); return tok; @@ -266,8 +280,7 @@ return XML_TOK_INVALID; } ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_GT)) { *nextTokPtr = ptr + MINBPC(enc); return tok; @@ -282,15 +295,14 @@ } static int PTRCALL -PREFIX(scanCdataSection)(const ENCODING *enc, const char *ptr, +PREFIX(scanCdataSection)(const ENCODING *UNUSED_P(enc), const char *ptr, const char *end, const char **nextTokPtr) { static const char CDATA_LSQB[] = { ASCII_C, ASCII_D, ASCII_A, ASCII_T, ASCII_A, ASCII_LSQB }; int i; /* CDATA[ */ - if (end - ptr < 6 * MINBPC(enc)) - return XML_TOK_PARTIAL; + REQUIRE_CHARS(enc, ptr, end, 6); for (i = 0; i < 6; i++, ptr += MINBPC(enc)) { if (!CHAR_MATCHES(enc, ptr, CDATA_LSQB[i])) { *nextTokPtr = ptr; @@ -305,7 +317,7 @@ PREFIX(cdataSectionTok)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; if (MINBPC(enc) > 1) { size_t n = end - ptr; @@ -319,13 +331,11 @@ switch (BYTE_TYPE(enc, ptr)) { case BT_RSQB: ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (!CHAR_MATCHES(enc, ptr, ASCII_RSQB)) break; ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (!CHAR_MATCHES(enc, ptr, ASCII_GT)) { ptr -= MINBPC(enc); break; @@ -334,8 +344,7 @@ return XML_TOK_CDATA_SECT_CLOSE; case BT_CR: ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (BYTE_TYPE(enc, ptr) == BT_LF) ptr += MINBPC(enc); *nextTokPtr = ptr; @@ -348,7 +357,7 @@ ptr += MINBPC(enc); break; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { #define LEAD_CASE(n) \ case BT_LEAD ## n: \ @@ -383,19 +392,18 @@ PREFIX(scanEndTag)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) default: *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_S: case BT_CR: case BT_LF: - for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) { + for (ptr += MINBPC(enc); HAS_CHAR(enc, ptr, end); ptr += MINBPC(enc)) { switch (BYTE_TYPE(enc, ptr)) { case BT_S: case BT_CR: case BT_LF: break; @@ -432,7 +440,7 @@ PREFIX(scanHexCharRef)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr != end) { + if (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { case BT_DIGIT: case BT_HEX: @@ -441,7 +449,7 @@ *nextTokPtr = ptr; return XML_TOK_INVALID; } - for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) { + for (ptr += MINBPC(enc); HAS_CHAR(enc, ptr, end); ptr += MINBPC(enc)) { switch (BYTE_TYPE(enc, ptr)) { case BT_DIGIT: case BT_HEX: @@ -464,7 +472,7 @@ PREFIX(scanCharRef)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr != end) { + if (HAS_CHAR(enc, ptr, end)) { if (CHAR_MATCHES(enc, ptr, ASCII_x)) return PREFIX(scanHexCharRef)(enc, ptr + MINBPC(enc), end, nextTokPtr); switch (BYTE_TYPE(enc, ptr)) { @@ -474,7 +482,7 @@ *nextTokPtr = ptr; return XML_TOK_INVALID; } - for (ptr += MINBPC(enc); ptr != end; ptr += MINBPC(enc)) { + for (ptr += MINBPC(enc); HAS_CHAR(enc, ptr, end); ptr += MINBPC(enc)) { switch (BYTE_TYPE(enc, ptr)) { case BT_DIGIT: break; @@ -496,8 +504,7 @@ PREFIX(scanRef)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) case BT_NUM: @@ -506,7 +513,7 @@ *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_SEMI: @@ -529,7 +536,7 @@ #ifdef XML_NS int hadColon = 0; #endif - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) #ifdef XML_NS @@ -540,8 +547,7 @@ } hadColon = 1; ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) default: @@ -555,8 +561,7 @@ int t; ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); t = BYTE_TYPE(enc, ptr); if (t == BT_EQUALS) break; @@ -579,8 +584,7 @@ #endif for (;;) { ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); open = BYTE_TYPE(enc, ptr); if (open == BT_QUOT || open == BT_APOS) break; @@ -598,8 +602,7 @@ /* in attribute value */ for (;;) { int t; - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); t = BYTE_TYPE(enc, ptr); if (t == open) break; @@ -624,8 +627,7 @@ } } ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { case BT_S: case BT_CR: @@ -642,8 +644,7 @@ /* ptr points to closing quote */ for (;;) { ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) case BT_S: case BT_CR: case BT_LF: @@ -655,8 +656,7 @@ case BT_SOL: sol: ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (!CHAR_MATCHES(enc, ptr, ASCII_GT)) { *nextTokPtr = ptr; return XML_TOK_INVALID; @@ -688,13 +688,12 @@ #ifdef XML_NS int hadColon; #endif - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) case BT_EXCL: - if ((ptr += MINBPC(enc)) == end) - return XML_TOK_PARTIAL; + ptr += MINBPC(enc); + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { case BT_MINUS: return PREFIX(scanComment)(enc, ptr + MINBPC(enc), end, nextTokPtr); @@ -716,7 +715,7 @@ hadColon = 0; #endif /* we have a start-tag */ - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) #ifdef XML_NS @@ -727,8 +726,7 @@ } hadColon = 1; ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) default: @@ -740,7 +738,7 @@ case BT_S: case BT_CR: case BT_LF: { ptr += MINBPC(enc); - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) case BT_GT: @@ -765,8 +763,7 @@ case BT_SOL: sol: ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); if (!CHAR_MATCHES(enc, ptr, ASCII_GT)) { *nextTokPtr = ptr; return XML_TOK_INVALID; @@ -785,7 +782,7 @@ PREFIX(contentTok)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; if (MINBPC(enc) > 1) { size_t n = end - ptr; @@ -803,7 +800,7 @@ return PREFIX(scanRef)(enc, ptr + MINBPC(enc), end, nextTokPtr); case BT_CR: ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return XML_TOK_TRAILING_CR; if (BYTE_TYPE(enc, ptr) == BT_LF) ptr += MINBPC(enc); @@ -814,12 +811,12 @@ return XML_TOK_DATA_NEWLINE; case BT_RSQB: ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return XML_TOK_TRAILING_RSQB; if (!CHAR_MATCHES(enc, ptr, ASCII_RSQB)) break; ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return XML_TOK_TRAILING_RSQB; if (!CHAR_MATCHES(enc, ptr, ASCII_GT)) { ptr -= MINBPC(enc); @@ -832,7 +829,7 @@ ptr += MINBPC(enc); break; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { #define LEAD_CASE(n) \ case BT_LEAD ## n: \ @@ -845,12 +842,12 @@ LEAD_CASE(2) LEAD_CASE(3) LEAD_CASE(4) #undef LEAD_CASE case BT_RSQB: - if (ptr + MINBPC(enc) != end) { + if (HAS_CHARS(enc, ptr, end, 2)) { if (!CHAR_MATCHES(enc, ptr + MINBPC(enc), ASCII_RSQB)) { ptr += MINBPC(enc); break; } - if (ptr + 2*MINBPC(enc) != end) { + if (HAS_CHARS(enc, ptr, end, 3)) { if (!CHAR_MATCHES(enc, ptr + 2*MINBPC(enc), ASCII_GT)) { ptr += MINBPC(enc); break; @@ -884,8 +881,7 @@ PREFIX(scanPercent)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) - return -XML_TOK_PERCENT; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) case BT_S: case BT_LF: case BT_CR: case BT_PERCNT: @@ -895,7 +891,7 @@ *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_SEMI: @@ -913,15 +909,14 @@ PREFIX(scanPoundName)(const ENCODING *enc, const char *ptr, const char *end, const char **nextTokPtr) { - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { CHECK_NMSTRT_CASES(enc, ptr, end, nextTokPtr) default: *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_CR: case BT_LF: case BT_S: @@ -941,7 +936,7 @@ const char *ptr, const char *end, const char **nextTokPtr) { - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { int t = BYTE_TYPE(enc, ptr); switch (t) { INVALID_CASES(ptr, nextTokPtr) @@ -950,7 +945,7 @@ ptr += MINBPC(enc); if (t != open) break; - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return -XML_TOK_LITERAL; *nextTokPtr = ptr; switch (BYTE_TYPE(enc, ptr)) { @@ -973,7 +968,7 @@ const char **nextTokPtr) { int tok; - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; if (MINBPC(enc) > 1) { size_t n = end - ptr; @@ -992,8 +987,7 @@ case BT_LT: { ptr += MINBPC(enc); - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); switch (BYTE_TYPE(enc, ptr)) { case BT_EXCL: return PREFIX(scanDecl)(enc, ptr + MINBPC(enc), end, nextTokPtr); @@ -1021,7 +1015,7 @@ case BT_S: case BT_LF: for (;;) { ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) break; switch (BYTE_TYPE(enc, ptr)) { case BT_S: case BT_LF: @@ -1048,11 +1042,10 @@ return XML_TOK_OPEN_BRACKET; case BT_RSQB: ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return -XML_TOK_CLOSE_BRACKET; if (CHAR_MATCHES(enc, ptr, ASCII_RSQB)) { - if (ptr + MINBPC(enc) == end) - return XML_TOK_PARTIAL; + REQUIRE_CHARS(enc, ptr, end, 2); if (CHAR_MATCHES(enc, ptr + MINBPC(enc), ASCII_GT)) { *nextTokPtr = ptr + 2*MINBPC(enc); return XML_TOK_COND_SECT_CLOSE; @@ -1065,7 +1058,7 @@ return XML_TOK_OPEN_PAREN; case BT_RPAR: ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return -XML_TOK_CLOSE_PAREN; switch (BYTE_TYPE(enc, ptr)) { case BT_AST: @@ -1141,7 +1134,7 @@ *nextTokPtr = ptr; return XML_TOK_INVALID; } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) case BT_GT: case BT_RPAR: case BT_COMMA: @@ -1154,8 +1147,7 @@ ptr += MINBPC(enc); switch (tok) { case XML_TOK_NAME: - if (ptr == end) - return XML_TOK_PARTIAL; + REQUIRE_CHAR(enc, ptr, end); tok = XML_TOK_PREFIXED_NAME; switch (BYTE_TYPE(enc, ptr)) { CHECK_NAME_CASES(enc, ptr, end, nextTokPtr) @@ -1204,10 +1196,12 @@ const char *end, const char **nextTokPtr) { const char *start; - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; + else if (! HAS_CHAR(enc, ptr, end)) + return XML_TOK_PARTIAL; start = ptr; - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { #define LEAD_CASE(n) \ case BT_LEAD ## n: ptr += n; break; @@ -1232,7 +1226,7 @@ case BT_CR: if (ptr == start) { ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return XML_TOK_TRAILING_CR; if (BYTE_TYPE(enc, ptr) == BT_LF) ptr += MINBPC(enc); @@ -1262,10 +1256,12 @@ const char *end, const char **nextTokPtr) { const char *start; - if (ptr == end) + if (ptr >= end) return XML_TOK_NONE; + else if (! HAS_CHAR(enc, ptr, end)) + return XML_TOK_PARTIAL; start = ptr; - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { #define LEAD_CASE(n) \ case BT_LEAD ## n: ptr += n; break; @@ -1294,7 +1290,7 @@ case BT_CR: if (ptr == start) { ptr += MINBPC(enc); - if (ptr == end) + if (! HAS_CHAR(enc, ptr, end)) return XML_TOK_TRAILING_CR; if (BYTE_TYPE(enc, ptr) == BT_LF) ptr += MINBPC(enc); @@ -1326,15 +1322,15 @@ end = ptr + n; } } - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { INVALID_CASES(ptr, nextTokPtr) case BT_LT: - if ((ptr += MINBPC(enc)) == end) - return XML_TOK_PARTIAL; + ptr += MINBPC(enc); + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_EXCL)) { - if ((ptr += MINBPC(enc)) == end) - return XML_TOK_PARTIAL; + ptr += MINBPC(enc); + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_LSQB)) { ++level; ptr += MINBPC(enc); @@ -1342,11 +1338,11 @@ } break; case BT_RSQB: - if ((ptr += MINBPC(enc)) == end) - return XML_TOK_PARTIAL; + ptr += MINBPC(enc); + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_RSQB)) { - if ((ptr += MINBPC(enc)) == end) - return XML_TOK_PARTIAL; + ptr += MINBPC(enc); + REQUIRE_CHAR(enc, ptr, end); if (CHAR_MATCHES(enc, ptr, ASCII_GT)) { ptr += MINBPC(enc); if (level == 0) { @@ -1373,7 +1369,7 @@ { ptr += MINBPC(enc); end -= MINBPC(enc); - for (; ptr != end; ptr += MINBPC(enc)) { + for (; HAS_CHAR(enc, ptr, end); ptr += MINBPC(enc)) { switch (BYTE_TYPE(enc, ptr)) { case BT_DIGIT: case BT_HEX: @@ -1521,7 +1517,7 @@ } static int PTRFASTCALL -PREFIX(charRefNumber)(const ENCODING *enc, const char *ptr) +PREFIX(charRefNumber)(const ENCODING *UNUSED_P(enc), const char *ptr) { int result = 0; /* skip &# */ @@ -1565,7 +1561,7 @@ } static int PTRCALL -PREFIX(predefinedEntityName)(const ENCODING *enc, const char *ptr, +PREFIX(predefinedEntityName)(const ENCODING *UNUSED_P(enc), const char *ptr, const char *end) { switch ((end - ptr)/MINBPC(enc)) { @@ -1683,11 +1679,11 @@ } static int PTRCALL -PREFIX(nameMatchesAscii)(const ENCODING *enc, const char *ptr1, +PREFIX(nameMatchesAscii)(const ENCODING *UNUSED_P(enc), const char *ptr1, const char *end1, const char *ptr2) { for (; *ptr2; ptr1 += MINBPC(enc), ptr2++) { - if (ptr1 == end1) + if (end1 - ptr1 < MINBPC(enc)) return 0; if (!CHAR_MATCHES(enc, ptr1, *ptr2)) return 0; @@ -1744,7 +1740,7 @@ const char *end, POSITION *pos) { - while (ptr != end) { + while (HAS_CHAR(enc, ptr, end)) { switch (BYTE_TYPE(enc, ptr)) { #define LEAD_CASE(n) \ case BT_LEAD ## n: \ @@ -1760,7 +1756,7 @@ case BT_CR: pos->lineNumber++; ptr += MINBPC(enc); - if (ptr != end && BYTE_TYPE(enc, ptr) == BT_LF) + if (HAS_CHAR(enc, ptr, end) && BYTE_TYPE(enc, ptr) == BT_LF) ptr += MINBPC(enc); pos->columnNumber = (XML_Size)-1; break; diff --git a/hexpat.cabal b/hexpat.cabal --- a/hexpat.cabal +++ b/hexpat.cabal @@ -1,6 +1,6 @@ Cabal-Version: >= 1.6 Name: hexpat -Version: 0.20.10 +Version: 0.20.11 Synopsis: XML parser/formatter based on expat Description: This package provides a general purpose Haskell XML library using Expat to @@ -50,7 +50,8 @@ performance; 0.20.2 make parseSaxG lazier; 0.20.3 minor build issues; 0.20.4 remove dependency on extensible-exceptions; 0.20.5 bump text upper bound; 0.20.6 bump text again to include 1.1.x.x; 0.20.7 bump text again for 1.2.x.x; 0.20.8 bump utf8-string dep; - 0.20.9 bump deepseq dep/ghc-7.10 compatibility.; 0.20.10 increase dependency upper bounds + 0.20.9 bump deepseq dep/ghc-7.10 compatibility.; 0.20.10 increase dependency upper bounds; + 0.20.11 update to libexpat-2.2.1 which includes several security fixes. Category: XML License: BSD3 License-File: LICENSE @@ -95,21 +96,29 @@ test/hexpat-leak/build.sh test/readRoads.hs, test/ROADS.xml, - cbits/amigaconfig.h, - cbits/asciitab.h, - cbits/expat.h, - cbits/internal.h, - cbits/macconfig.h, - cbits/utf8tab.h, - cbits/xmlrole.h, - cbits/xmltok_impl.h, - cbits/ascii.h, - cbits/expat_external.h, - cbits/iasciitab.h, - cbits/latin1tab.h, - cbits/nametab.h, cbits/winconfig.h, cbits/xmltok.h + cbits/winconfig.h + cbits/xmltok_ns.c + cbits/internal.h + cbits/utf8tab.h + cbits/siphash.h + cbits/latin1tab.h + cbits/xmltok.h + cbits/expat.h + cbits/xmltok.c + cbits/iasciitab.h + cbits/asciitab.h + cbits/README + cbits/xmlparse.c + cbits/xmltok_impl.h + cbits/xmltok_impl.c + cbits/xmlrole.c + cbits/xmlrole.h + cbits/expat_external.h + cbits/ascii.h + cbits/nametab.h + Build-Type: Simple Stability: beta source-repository head