diff --git a/Nemesis b/Nemesis
--- a/Nemesis
+++ b/Nemesis
@@ -15,6 +15,9 @@
desc "start console"
task "i" (sh "ghci -isrc -Wall src/Text/HTML/Yuuko.hs")
+ desc "start main"
+ task "main" (sh "ghci -isrc -Wall src/Main.hs")
+
desc "put all .hs files in manifest"
task "manifest" $ do
sh "find . | grep 'hs$' > manifest"
diff --git a/changelog.md b/changelog.md
--- a/changelog.md
+++ b/changelog.md
@@ -1,3 +1,16 @@
+2009.10.23.2
+------------
+
+### Feature
+
+* add `yuuko` binary
+* add `yuuko_xml` lib for handling xml, e.g. rss
+* magic pure (with unsafePerformIO)
+
+### Fix
+
+* handle utf8 html
+
2009.10.23.1
------------
diff --git a/readme.md b/readme.md
--- a/readme.md
+++ b/readme.md
@@ -7,9 +7,10 @@
=======
- yuuko "//body" "hi"
+ > echo 'hi
there
' | yuuko //div
- > ["hi"]
+ hi
+ there
Tutorial
========
diff --git a/src/Main.hs b/src/Main.hs
new file mode 100644
--- /dev/null
+++ b/src/Main.hs
@@ -0,0 +1,11 @@
+module Main where
+
+import Text.HTML.Yuuko
+import System
+
+main :: IO ()
+main = do
+ args <- getArgs
+ case args of
+ [] -> putStrLn "\n usage: echo 'hi
' | yuuko //div\n"
+ expr:_ -> interact $ concatMap (++ "\n") . (yuuko expr)
diff --git a/src/Text/HTML/Yuuko.hs b/src/Text/HTML/Yuuko.hs
--- a/src/Text/HTML/Yuuko.hs
+++ b/src/Text/HTML/Yuuko.hs
@@ -2,23 +2,32 @@
import Text.XML.HXT.Arrow
import qualified Text.XML.HXT.Arrow.XPathSimple as XS
+import System.IO.Unsafe
-yuuko :: String -> String -> IO [String]
-yuuko expr s = do
+yuuko :: String -> String -> [String]
+yuuko = yuuko_base [(a_parse_html, v_1)]
+
+yuuko_xml :: String -> String -> [String]
+yuuko_xml = yuuko_base [(a_ignore_none_xml_contents, v_1)]
+
+yuuko_base :: [(String, String)] -> String -> String -> [String]
+yuuko_base args expr s = unsafePerformIO $
runX process
where
read_args =
[ (a_validate, v_0)
- , (a_parse_html, v_1)
- , (a_encoding, utf8)
, (a_issue_warnings, v_0)
]
+ ++ args
write_args =
- [ (a_encoding, utf8) ]
+ [ (a_encoding, utf8)
+ , (a_no_xml_pi, v_0 )
+ ]
process =
readString read_args s
>>> XS.getXPathTreesInDoc expr
>>> writeDocumentToString write_args
+
diff --git a/yuuko.cabal b/yuuko.cabal
--- a/yuuko.cabal
+++ b/yuuko.cabal
@@ -1,5 +1,5 @@
Name: yuuko
-Version: 2009.10.23.1
+Version: 2009.10.23.2
Build-type: Simple
Synopsis: A transcendental HTML parser gently wrapping the HXT library
Description:
@@ -21,6 +21,13 @@
category: Text
homepage: http://github.com/nfjinjing/yuuko
data-files: readme.md, changelog.md, known-issues.md, Nemesis
+
+Executable yuuko
+ ghc-options: -Wall
+ build-depends: base >= 4 && < 5, hxt, haskell98, utf8-prelude
+ hs-source-dirs: src/
+ main-is: Main.hs
+ other-modules: Text.HTML.Yuuko
library
ghc-options: -Wall