1 -----------------------------------------------------------------------------
3 -- Module : Text.Regex.Posix
4 -- Copyright : (c) The University of Glasgow 2002
5 -- License : BSD-style (see the file libraries/base/LICENSE)
7 -- Maintainer : libraries@haskell.org
8 -- Stability : experimental
9 -- Portability : non-portable (needs POSIX regexps)
11 -- Interface to the POSIX regular expression library.
13 -----------------------------------------------------------------------------
15 -- ToDo: should have an interface using PackedStrings.
17 module Text.Regex.Posix (
21 -- * Compiling a regular expression
22 regcomp, -- :: String -> Int -> IO Regex
24 -- ** Flags for regcomp
25 regExtended, -- (flag to regcomp) use extended regex syntax
26 regIgnoreCase, -- (flag to regcomp) ignore case when matching
27 regNewline, -- (flag to regcomp) '.' doesn't match newline
29 -- * Matching a regular expression
30 regexec, -- :: Regex -- pattern
31 -- -> String -- string to match
32 -- -> IO (Maybe (String, -- everything before match
33 -- String, -- matched portion
34 -- String, -- everything after match
35 -- [String])) -- subexpression matches
39 #include <sys/types.h>
47 -- | A compiled regular expression
48 newtype Regex = Regex (ForeignPtr CRegex)
50 -- -----------------------------------------------------------------------------
53 -- | Compiles a regular expression
55 :: String -- ^ The regular expression to compile
56 -> Int -- ^ Flags (summed together)
57 -> IO Regex -- ^ Returns: the compiled regular expression
58 regcomp pattern flags = do
60 regex_fptr <- mallocForeignPtrBytes (#const sizeof(regex_t))
62 regex_ptr <- mallocBytes (#const sizeof(regex_t))
63 regex_fptr <- newForeignPtr regex_ptr (regfree regex_ptr)
65 r <- withCString pattern $ \cstr ->
66 withForeignPtr regex_fptr $ \p ->
67 c_regcomp p cstr (fromIntegral flags)
69 addForeignPtrFinalizer regex_fptr ptr_regfree
72 then return (Regex regex_fptr)
73 else error "Text.Regex.Posix.regcomp: error in pattern" -- ToDo
76 regfree :: Ptr CRegex -> IO ()
82 -- -----------------------------------------------------------------------------
85 -- | Matches a regular expression against a string
86 regexec :: Regex -- ^ Compiled regular expression
87 -> String -- ^ String to match against
88 -> IO (Maybe (String, String, String, [String]))
89 -- ^ Returns: 'Nothing' if the regex did not match the
93 -- 'Just' (everything before match,
95 -- everything after match,
96 -- subexpression matches)
99 regexec (Regex regex_fptr) str = do
100 withCString str $ \cstr -> do
101 withForeignPtr regex_fptr $ \regex_ptr -> do
102 nsub <- (#peek regex_t, re_nsub) regex_ptr
103 let nsub_int = fromIntegral (nsub :: CSize)
104 allocaBytes ((1 + nsub_int) * (#const sizeof(regmatch_t))) $ \p_match -> do
105 -- add one because index zero covers the whole match
106 r <- c_regexec regex_ptr cstr (1 + nsub) p_match 0{-no flags for now-}
108 if (r /= 0) then return Nothing else do
110 (before,match,after) <- matched_parts str p_match
113 mapM (unpack str) $ take nsub_int $ tail $
114 iterate (`plusPtr` (#const sizeof(regmatch_t))) p_match
116 return (Just (before, match, after, sub_strs))
118 matched_parts :: String -> Ptr CRegMatch -> IO (String, String, String)
119 matched_parts string p_match = do
120 start <- (#peek regmatch_t, rm_so) p_match :: IO CInt
121 end <- (#peek regmatch_t, rm_eo) p_match :: IO CInt
122 let s = fromIntegral start; e = fromIntegral end
123 return ( take (s-1) string,
124 take (e-s) (drop s string),
127 unpack :: String -> Ptr CRegMatch -> IO (String)
128 unpack string p_match = do
129 start <- (#peek regmatch_t, rm_so) p_match :: IO CInt
130 end <- (#peek regmatch_t, rm_eo) p_match :: IO CInt
131 -- the subexpression may not have matched at all, perhaps because it
132 -- was optional. In this case, the offsets are set to -1.
133 if (start == -1) then return "" else do
134 return (take (fromIntegral (end-start)) (drop (fromIntegral start) string))
136 -- -----------------------------------------------------------------------------
137 -- The POSIX regex C interface
144 -- Return values from regexec
152 regIgnoreCase = REG_ICASE, \
156 -- Error codes from regcomp
174 foreign import ccall unsafe "regcomp"
175 c_regcomp :: Ptr CRegex -> CString -> CInt -> IO CInt
178 foreign import ccall unsafe "®free"
179 ptr_regfree :: FunPtr (Ptr CRegex -> IO ())
181 foreign import ccall unsafe "regfree"
182 c_regfree :: Ptr CRegex -> IO ()
183 #endif /* __HUGS__ */
185 foreign import ccall unsafe "regexec"
186 c_regexec :: Ptr CRegex -> CString -> CSize
187 -> Ptr CRegMatch -> CInt -> IO CInt