diff --git a/CMark.hsc b/CMark.hsc --- a/CMark.hsc +++ b/CMark.hsc @@ -26,13 +26,15 @@ import Foreign import Foreign.C.Types -import Foreign.C.String (CString, withCString) +import Foreign.C.String (CString) import qualified System.IO.Unsafe as Unsafe import GHC.Generics (Generic) import Data.Data (Data) import Data.Typeable (Typeable) -import Data.Text (Text, empty, unpack) +import Data.Text (Text, empty) import qualified Data.Text.Foreign as TF +import qualified Data.ByteString as B +import Data.Text.Encoding (encodeUtf8) import Control.Applicative ((<$>), (<*>)) #include @@ -55,14 +57,14 @@ -- | Convert CommonMark formatted text to a structured 'Node' tree, -- which can be transformed or rendered using Haskell code. commonmarkToNode :: [CMarkOption] -> Text -> Node -commonmarkToNode opts s = io $ do +commonmarkToNode opts s = Unsafe.unsafePerformIO $ do nptr <- TF.withCStringLen s $! \(ptr, len) -> c_cmark_parse_document ptr len (combineOptions opts) fptr <- newForeignPtr c_cmark_node_free nptr withForeignPtr fptr toNode nodeToCommonmark :: [CMarkOption] -> Int -> Node -> Text -nodeToCommonmark opts width node = io $ do +nodeToCommonmark opts width node = Unsafe.unsafePerformIO $ do nptr <- fromNode node fptr <- newForeignPtr c_cmark_node_free nptr withForeignPtr fptr $ \ptr -> do @@ -73,14 +75,15 @@ -> [CMarkOption] -> Text -> Text -commonmarkToX renderer opts s = io $ TF.withCStringLen s $ \(ptr, len) -> do - let opts' = combineOptions opts - nptr <- c_cmark_parse_document ptr len opts' - fptr <- newForeignPtr c_cmark_node_free nptr - withForeignPtr fptr $ \p -> do - str <- renderer p opts' - t <- TF.peekCStringLen $! (str, c_strlen str) - return t +commonmarkToX renderer opts s = Unsafe.unsafePerformIO $ + TF.withCStringLen s $ \(ptr, len) -> do + let opts' = combineOptions opts + nptr <- c_cmark_parse_document ptr len opts' + fptr <- newForeignPtr c_cmark_node_free nptr + withForeignPtr fptr $ \p -> do + str <- renderer p opts' + t <- TF.peekCStringLen $! (str, c_strlen str) + return t type NodePtr = Ptr () @@ -263,12 +266,12 @@ BLOCK_QUOTE -> c_cmark_node_new (#const CMARK_NODE_BLOCK_QUOTE) HTML literal -> do n <- c_cmark_node_new (#const CMARK_NODE_HTML) - c_cmark_node_set_literal n (fromtext literal) + c_cmark_node_set_literal n =<< fromtext literal return n CODE_BLOCK info literal -> do n <- c_cmark_node_new (#const CMARK_NODE_CODE_BLOCK) - c_cmark_node_set_literal n (fromtext literal) - c_cmark_node_set_fence_info n (fromtext info) + c_cmark_node_set_literal n =<< fromtext literal + c_cmark_node_set_fence_info n =<< fromtext info return n LIST attr -> do n <- c_cmark_node_new (#const CMARK_NODE_LIST) @@ -290,41 +293,38 @@ STRONG -> c_cmark_node_new (#const CMARK_NODE_STRONG) LINK url title -> do n <- c_cmark_node_new (#const CMARK_NODE_LINK) - c_cmark_node_set_url n (fromtext url) - c_cmark_node_set_title n (fromtext title) + c_cmark_node_set_url n =<< fromtext url + c_cmark_node_set_title n =<< fromtext title return n IMAGE url title -> do n <- c_cmark_node_new (#const CMARK_NODE_IMAGE) - c_cmark_node_set_url n (fromtext url) - c_cmark_node_set_title n (fromtext title) + c_cmark_node_set_url n =<< fromtext url + c_cmark_node_set_title n =<< fromtext title return n TEXT literal -> do n <- c_cmark_node_new (#const CMARK_NODE_TEXT) - c_cmark_node_set_literal n (fromtext literal) + c_cmark_node_set_literal n =<< fromtext literal return n CODE literal -> do n <- c_cmark_node_new (#const CMARK_NODE_CODE) - c_cmark_node_set_literal n (fromtext literal) + c_cmark_node_set_literal n =<< fromtext literal return n INLINE_HTML literal -> do n <- c_cmark_node_new (#const CMARK_NODE_INLINE_HTML) - c_cmark_node_set_literal n (fromtext literal) + c_cmark_node_set_literal n =<< fromtext literal return n SOFTBREAK -> c_cmark_node_new (#const CMARK_NODE_SOFTBREAK) LINEBREAK -> c_cmark_node_new (#const CMARK_NODE_LINEBREAK) mapM_ (\child -> fromNode child >>= c_cmark_node_append_child node) children return node -io :: IO a -> a -io = Unsafe.unsafePerformIO - totext :: CString -> IO Text totext str | str == nullPtr = return empty | otherwise = TF.peekCStringLen (str, c_strlen str) -fromtext :: Text -> CString -fromtext t = io $! withCString (unpack t) return +fromtext :: Text -> IO CString +fromtext t = B.useAsCString (encodeUtf8 t) return foreign import ccall "string.h strlen" c_strlen :: CString -> Int diff --git a/README.md b/README.md --- a/README.md +++ b/README.md @@ -3,8 +3,8 @@ This package provides Haskell bindings for [libcmark], the reference parser for [CommonMark], a fully specified variant of Markdown. -It includes sources for [libcmark] (as of commit f40e098, implementing -version 0.18 of the spec) and does not require prior installation of +It includes sources for [libcmark] (version 0.19.0, implementing +version 0.19 of the spec) and does not require prior installation of the C library. cmark provides the following advantages over existing Markdown diff --git a/cbits/cmark_version.h b/cbits/cmark_version.h --- a/cbits/cmark_version.h +++ b/cbits/cmark_version.h @@ -1,7 +1,7 @@ #ifndef CMARK_VERSION_H #define CMARK_VERSION_H -#define CMARK_VERSION ((0 << 16) | (18 << 8) | 1) -#define CMARK_VERSION_STRING "0.18.1" +#define CMARK_VERSION ((0 << 16) | (19 << 8) | 0) +#define CMARK_VERSION_STRING "0.19.0" #endif diff --git a/cbits/commonmark.c b/cbits/commonmark.c --- a/cbits/commonmark.c +++ b/cbits/commonmark.c @@ -49,35 +49,35 @@ static inline bool needs_escaping(escaping escape, - int32_t c, - unsigned char next_c, - struct render_state *state) + int32_t c, + unsigned char next_c, + struct render_state *state) { if (escape == NORMAL) { return (c == '*' || c == '_' || c == '[' || c == ']' || - c == '<' || c == '>' || c == '\\' || c == '`' || - (c == '&' && isalpha(next_c)) || - (c == '!' && next_c == '[') || - (state->begin_line && - (c == '-' || c == '+' || c == '#' || c == '=')) || - (c == '#' && (isspace(next_c) || next_c == '\0')) || - ((c == '.' || c == ')') && - isdigit(state->buffer->ptr[state->buffer->size - 1]))); + c == '<' || c == '>' || c == '\\' || c == '`' || + (c == '&' && isalpha(next_c)) || + (c == '!' && next_c == '[') || + (state->begin_line && + (c == '-' || c == '+' || c == '#' || c == '=')) || + (c == '#' && (isspace(next_c) || next_c == '\0')) || + ((c == '.' || c == ')') && + isdigit(state->buffer->ptr[state->buffer->size - 1]))); } else if (escape == TITLE) { return (c == '`' || c == '<' || c == '>' || c == '"' || - c == '\\'); + c == '\\'); } else if (escape == URL) { return (c == '`' || c == '<' || c == '>' || isspace(c) || - c == '\\' || c == ')' || c == '('); + c == '\\' || c == ')' || c == '('); } else { return false; } } static inline void out(struct render_state *state, - cmark_chunk str, - bool wrap, - escaping escape) + cmark_chunk str, + bool wrap, + escaping escape) { unsigned char* source = str.data; int length = str.len; @@ -100,7 +100,7 @@ cmark_strbuf_putc(state->buffer, '\n'); if (state->need_cr > 1) { cmark_strbuf_put(state->buffer, state->prefix->ptr, - state->prefix->size); + state->prefix->size); } } state->column = 0; @@ -111,12 +111,15 @@ while (i < length) { if (state->begin_line) { cmark_strbuf_put(state->buffer, state->prefix->ptr, - state->prefix->size); + state->prefix->size); // note: this assumes prefix is ascii: state->column = state->prefix->size; } len = utf8proc_iterate(source + i, length - i, &c); + if (len == -1) { // error condition + return; // return without rendering rest of string + } nextc = source[i + len]; if (c == 32 && wrap) { if (!state->begin_line) { @@ -124,7 +127,7 @@ state->column += 1; state->begin_line = false; state->last_breakable = state->buffer->size - - 1; + 1; // skip following spaces while (source[i + 1] == ' ') { i++; @@ -167,7 +170,7 @@ // add newline, prefix, and remainder cmark_strbuf_putc(state->buffer, '\n'); cmark_strbuf_put(state->buffer, state->prefix->ptr, - state->prefix->size); + state->prefix->size); cmark_strbuf_put(state->buffer, remainder.data, remainder.len); state->column = state->prefix->size + remainder.len; cmark_chunk_free(&remainder); @@ -254,8 +257,8 @@ } cmark_consolidate_text_nodes(node); return (strncmp(url, - (char*)node->as.literal.data, - node->as.literal.len) == 0); + (char*)node->as.literal.data, + node->as.literal.len) == 0); } // if node is a block node, returns node. @@ -265,7 +268,7 @@ { while (node && (node->type < CMARK_NODE_FIRST_BLOCK || - node->type > CMARK_NODE_LAST_BLOCK)) { + node->type > CMARK_NODE_LAST_BLOCK)) { node = node->parent; } return node; @@ -284,7 +287,7 @@ bool entering = (ev_type == CMARK_EVENT_ENTER); const char *info; const char *title; - char listmarker[64]; + cmark_strbuf listmarker = GH_BUF_INIT; char *emph_delim; int marker_width; @@ -293,14 +296,14 @@ // a following list. if (!(node->type == CMARK_NODE_ITEM && node->prev == NULL && entering)) { - tmp = get_containing_block(node); - state->in_tight_list_item = - (tmp->type == CMARK_NODE_ITEM && - cmark_node_get_list_tight(tmp->parent)) || - (tmp && - tmp->parent && - tmp->parent->type == CMARK_NODE_ITEM && - cmark_node_get_list_tight(tmp->parent->parent)); + tmp = get_containing_block(node); + state->in_tight_list_item = + (tmp->type == CMARK_NODE_ITEM && + cmark_node_get_list_tight(tmp->parent)) || + (tmp && + tmp->parent && + tmp->parent->type == CMARK_NODE_ITEM && + cmark_node_get_list_tight(tmp->parent->parent)); } switch (node->type) { @@ -316,7 +319,7 @@ cmark_strbuf_puts(state->prefix, "> "); } else { cmark_strbuf_truncate(state->prefix, - state->prefix->size - 2); + state->prefix->size - 2); blankline(state); } break; @@ -347,11 +350,12 @@ // we ensure a width of at least 4 so // we get nice transition from single digits // to double - snprintf(listmarker, 63, "%d%s%s", list_number, - list_delim == CMARK_PAREN_DELIM ? - ")" : ".", - list_number < 10 ? " " : " "); - marker_width = strlen(listmarker); + cmark_strbuf_printf(&listmarker, + "%d%s%s", list_number, + list_delim == CMARK_PAREN_DELIM ? + ")" : ".", + list_number < 10 ? " " : " "); + marker_width = listmarker.size; } if (entering) { if (cmark_node_get_list_type(node->parent) == @@ -359,17 +363,18 @@ lit(state, "* ", false); cmark_strbuf_puts(state->prefix, " "); } else { - lit(state, listmarker, false); - for (i=marker_width; i--;) { + lit(state, (char *)listmarker.ptr, false); + for (i = marker_width; i--;) { cmark_strbuf_putc(state->prefix, ' '); } } } else { cmark_strbuf_truncate(state->prefix, - state->prefix->size - - marker_width); + state->prefix->size - + marker_width); cr(state); } + cmark_strbuf_free(&listmarker); break; case CMARK_NODE_HEADER: @@ -403,7 +408,7 @@ cmark_strbuf_puts(state->prefix, " "); out(state, node->as.code.literal, false, LITERAL); cmark_strbuf_truncate(state->prefix, - state->prefix->size - 4); + state->prefix->size - 4); } else { numticks = longest_backtick_sequence(code) + 1; if (numticks < 3) { @@ -512,7 +517,7 @@ if (entering) { lit(state, "<", false); if (strncmp(cmark_node_get_url(node), - "mailto:", 7) == 0) { + "mailto:", 7) == 0) { lit(state, (char *)cmark_node_get_url(node) + 7, false); @@ -577,9 +582,10 @@ if (CMARK_OPT_HARDBREAKS & options) { width = 0; } - struct render_state state = - { options, &commonmark, &prefix, 0, width, - 0, 0, true, false, false}; + struct render_state state = { + options, &commonmark, &prefix, 0, width, + 0, 0, true, false, false + }; cmark_node *cur; cmark_event_type ev_type; cmark_iter *iter = cmark_iter_new(root); diff --git a/cbits/inlines.c b/cbits/inlines.c --- a/cbits/inlines.c +++ b/cbits/inlines.c @@ -301,8 +301,10 @@ !utf8proc_is_space(after_char) && !utf8proc_is_punctuation(after_char)); if (c == '_') { - *can_open = left_flanking && !right_flanking; - *can_close = right_flanking && !left_flanking; + *can_open = left_flanking && + (!right_flanking || utf8proc_is_punctuation(before_char)); + *can_close = right_flanking && + (!left_flanking || utf8proc_is_punctuation(after_char)); } else if (c == '\'' || c == '"') { *can_open = left_flanking && !right_flanking; *can_close = right_flanking; @@ -636,7 +638,7 @@ } cmark_strbuf_unescape(&buf); - return cmark_strbuf_detach(&buf); + return buf.size == 0 ? NULL : cmark_strbuf_detach(&buf); } unsigned char *cmark_clean_title(cmark_chunk *title) @@ -660,7 +662,7 @@ } cmark_strbuf_unescape(&buf); - return cmark_strbuf_detach(&buf); + return buf.size == 0 ? NULL : cmark_strbuf_detach(&buf); } // Parse an autolink or HTML tag. diff --git a/cbits/man.c b/cbits/man.c --- a/cbits/man.c +++ b/cbits/man.c @@ -20,6 +20,9 @@ while (i < length) { len = utf8proc_iterate(source + i, length - i, &c); + if (len == -1) { // error condition + return; // return without rendering anything + } switch(c) { case 46: if (beginLine) { diff --git a/cbits/utf8.c b/cbits/utf8.c --- a/cbits/utf8.c +++ b/cbits/utf8.c @@ -172,8 +172,7 @@ case 3: uc = ((str[0] & 0x0F) << 12) + ((str[1] & 0x3F) << 6) + (str[2] & 0x3F); - if (uc < 0x800 || (uc >= 0xD800 && uc < 0xE000) || - (uc >= 0xFDD0 && uc < 0xFDF0)) uc = -1; + if (uc < 0x800 || (uc >= 0xD800 && uc < 0xE000)) uc = -1; break; case 4: uc = ((str[0] & 0x07) << 18) + ((str[1] & 0x3F) << 12) @@ -182,7 +181,7 @@ break; } - if (uc < 0 || ((uc & 0xFFFF) >= 0xFFFE)) + if (uc < 0) return -1; *dst = uc; diff --git a/changelog b/changelog --- a/changelog +++ b/changelog @@ -1,3 +1,11 @@ +cmark 0.3.2 (28 Apr 2015) + + * Updated libcmark C sources to 0.19.0. + * Fixes to avoid warning for earlier ghc versions. + * Refactored, removing the `io` convenience function. + * Removed unnecessary `unsafePerformIO` on `fromtext`. + * `fromtext`: use `ByteString.useAsCString` + explicit `encodeUtf8`. + cmark 0.3.1 (29 Mar 2015) * Added nodeToCommonmark. diff --git a/cmark.cabal b/cmark.cabal --- a/cmark.cabal +++ b/cmark.cabal @@ -1,14 +1,13 @@ name: cmark -version: 0.3.1 +version: 0.3.2 synopsis: Fast, accurate CommonMark (Markdown) parser and renderer description: This package provides Haskell bindings for , the reference parser for , a fully specified variant of Markdown. It includes sources for - libcmark (as of commit f40e098, implementing version 0.18 - of the CommonMark spec) and does not require prior - installation of the C library. + libcmark (0.19.0) and does not require prior installation of the + C library. . cmark provides the following advantages over existing Markdown libraries for Haskell: @@ -118,7 +117,7 @@ library exposed-modules: CMark - build-depends: base >=4.5 && < 4.9, text + build-depends: base >=4.5 && < 4.9, text, bytestring if impl(ghc < 7.6) build-depends: ghc-prim >= 0.2 default-language: Haskell2010