xeno 0.2 → 0.3

raw patch · 7 files changed

+72/−17 lines, 7 files

Files

• CHANGELOG.markdown +3/−0
• CONTRIBUTORS.md +12/−0
• README.md +6/−1
• src/Xeno/DOM.hs +11/−3
• src/Xeno/SAX.hs +20/−4
• test/Main.hs +14/−6
• xeno.cabal +6/−3

@@ -1,2 +1,5 @@+0.2+Added CDATA support (Rembane)	+ 0.1 First Hackage release

@@ -0,0 +1,12 @@+# Contributors++Original author : Chris Done ( chrisdone )++Current maintainer : Marco Zocca ( ocramz )++## In chronological order :++Andreas Ekeroot ( Rembane )+Kirill Zaborsky ( qrilka )++

@@ -16,7 +16,7 @@   [cheats like Hexml does](http://neilmitchell.blogspot.co.uk/2016/12/new-xml-parser-hexml.html)   (doesn't expand entities, or most of the XML standard). * It's written in pure Haskell.-* CDATA is supported now.+* CDATA is supported as of version 0.2.   ## Performance goals@@ -187,3 +187,8 @@ The `process` is marked as INLINE, which means use-sites of it will inline, and your particular monad's type will be potentially erased for great performance.+++## Contributors++See CONTRIBUTORS.md

@@ -126,10 +126,18 @@   case spork node of     Left e -> Left e     Right r ->-      case r ! 0 of-        0x00 -> Right (Node str 0 r)-        _ -> Left XenoExpectRootNode+      case findRootNode r of+        Just n -> Right n+        Nothing -> Left XenoExpectRootNode   where+    findRootNode r = go 0+      where+        go n = case r UV.!? n of+          Just 0x0 -> Just (Node str n r)+          -- skipping text assuming that it contains only white space+          -- characters+          Just 0x1 -> go (n+3)+          _ -> Nothing     node =       runST         (do nil <- UMV.new 1000

@@ -262,10 +262,17 @@ -- | Basically @findIndex (not . isNameChar)@, but doesn't allocate. parseName :: ByteString -> Int -> Int parseName str index =+  if not (isNameChar1 (s_index str index))+     then index+     else parseName' str (index + 1)++-- | Basically @findIndex (not . isNameChar)@, but doesn't allocate.+parseName' :: ByteString -> Int -> Int+parseName' str index =   if not (isNameChar (s_index str index))      then index-     else parseName str (index + 1)-{-# INLINE parseName #-}+     else parseName' str (index + 1)+{-# INLINE parseName' #-}  -- | Get index of an element starting from offset. elemIndexFrom :: Word8 -> ByteString -> Int -> Maybe Int@@ -282,10 +289,19 @@ isSpaceChar c = c == 32 || (c <= 10 && c >= 9) || c == 13 {-# INLINE isSpaceChar #-} --- | Is the character a valid tag name constituent?+-- | Is the character a valid first tag/attribute name constituent?+-- 'a'-'z', 'A'-'Z', '_', ':'+isNameChar1 :: Word8 -> Bool+isNameChar1 c =+  (c >= 97 && c <= 122) || (c >= 65 && c <= 90) || c == 95 || c == 58+{-# INLINE isNameChar1 #-}++-- | Is the character a valid tag/attribute name constituent?+-- isNameChar1 + '-', '.', '0'-'9' isNameChar :: Word8 -> Bool isNameChar c =-  (c >= 97 && c <= 122) || (c >= 65 && c <= 90) || c == 95 || c == 45 || c == 58+  (c >= 97 && c <= 122) || (c >= 65 && c <= 90) || c == 95 || c == 58 ||+  c == 45 || c == 46 || (c >= 48 && c <= 57) {-# INLINE isNameChar #-}  -- | Char for '\''.

@@ -9,12 +9,7 @@ import           Xeno.SAX import           Xeno.DOM import           Xeno.Types--- import Control.Monad.Catch -import Data.Either---- main = print "hello!"- main :: IO () main = hspec spec @@ -24,7 +19,7 @@     "hexml tests"     (do mapM_           (\(v, i) -> it (show i) (shouldBe (Xeno.SAX.validate i) v))-          hexml_examples_sax+          (hexml_examples_sax  ++ extra_examples_sax)         mapM_           (\(v, i) -> it (show i) (shouldBe (either (Left . show) (Right . id) (contents <$> Xeno.DOM.parse i)) v))           cdata_tests@@ -41,6 +36,12 @@           (shouldBe              (attributes (head (children $ fromRightE doc)))              [("id", "1"), ("extra", "2")])++        it "xml prologue test" $ do+          let docWithPrologue = "<?xml version=\"1.1\"?>\n<greeting>Hello, world!</greeting>"+              parsedRoot = fromRightE $ Xeno.DOM.parse docWithPrologue+          name parsedRoot `shouldBe` "greeting"+         -- If this works without crashing we're happy.         let nsdoc = "<ns:tag os:attr=\"Namespaced attribute value\">Content.</ns:tag>"         it@@ -59,6 +60,13 @@     ,(True, "<test></more>") -- SAX doesn't care about tag balancing     ,(False, "<test")     ,(True, "<?xml version=\"1.1\"?>\n<greeting>Hello, world!</greeting>")+    ]++extra_examples_sax :: [(Bool, ByteString)]+extra_examples_sax =+    [(True, "<some-example/>")+    ,(True, "<a numeric1=\"attribute\"/>")+    ,(True, "<also.a.dot></also.a.dot>")     ]  -- | We want to make sure that the parser doesn't jump out of the CDATA

@@ -1,5 +1,5 @@ name: xeno-version: 0.2+version: 0.3 synopsis: A fast event-based XML parser in pure Haskell description: Please see README.md           build-type: Simple@@ -12,7 +12,8 @@ maintainer: Marco Zocca (zocca.marco gmail) tested-with:         GHC == 8.0.1 extra-source-files:  README.md-                     CHANGELOG.markdown                                 +                     CHANGELOG.markdown+                     CONTRIBUTORS.md                               source-repository head   type:     git@@ -34,6 +35,7 @@   default-language: Haskell2010  test-suite xeno-test+              type: exitcode-stdio-1.0   hs-source-dirs: test   main-is: Main.hs@@ -50,7 +52,8 @@     build-depends: libxml   ghc-options: -Wall -threaded -rtsopts -with-rtsopts=-N -O2   if flag(libxml2)-     ghc-options: -DLIBXML2+     cpp-options: -DLIBXML2   +     -- ghc-options: -DLIBXML2  -- Hackage started complaining about this   default-language: Haskell2010  benchmark xeno-memory-bench