1 -----------------------------------------------------------------------------
3 -- Module : Text.Regex.Posix
4 -- Copyright : (c) The University of Glasgow 2001
5 -- License : BSD-style (see the file libraries/core/LICENSE)
7 -- Maintainer : libraries@haskell.org
8 -- Stability : experimental
9 -- Portability : non-portable (only on platforms that provide POSIX regexps)
11 -- $Id: Posix.hsc,v 1.5 2002/02/14 14:08:50 simonmar Exp $
13 -- Interface to the POSIX regular expression library.
14 -- ToDo: should have an interface using PackedStrings.
16 -----------------------------------------------------------------------------
18 module Text.Regex.Posix (
21 regcomp, -- :: String -> Int -> IO Regex
23 regexec, -- :: Regex -- pattern
24 -- -> String -- string to match
25 -- -> IO (Maybe (String, -- everything before match
26 -- String, -- matched portion
27 -- String, -- everything after match
28 -- [String])) -- subexpression matches
30 regExtended, -- (flag to regcomp) use extended regex syntax
31 regIgnoreCase, -- (flag to regcomp) ignore case when matching
32 regNewline -- (flag to regcomp) '.' doesn't match newline
42 newtype Regex = Regex (ForeignPtr CRegex)
44 -- -----------------------------------------------------------------------------
47 regcomp :: String -> Int -> IO Regex
48 regcomp pattern flags = do
49 regex_ptr <- mallocBytes (#const sizeof(regex_t))
50 regex_fptr <- newForeignPtr regex_ptr (regfree regex_ptr)
51 r <- withCString pattern $ \cstr ->
52 withForeignPtr regex_fptr $ \p ->
53 c_regcomp p cstr (fromIntegral flags)
55 then return (Regex regex_fptr)
56 else error "Text.Regex.Posix.regcomp: error in pattern" -- ToDo
58 regfree :: Ptr CRegex -> IO ()
63 -- -----------------------------------------------------------------------------
66 regexec :: Regex -- pattern
67 -> String -- string to match
68 -> IO (Maybe (String, -- everything before match
69 String, -- matched portion
70 String, -- everything after match
71 [String])) -- subexpression matches
73 regexec (Regex regex_fptr) str = do
74 withCString str $ \cstr -> do
75 withForeignPtr regex_fptr $ \regex_ptr -> do
76 nsub <- (#peek regex_t, re_nsub) regex_ptr
77 let nsub_int = fromIntegral (nsub :: CSize)
78 allocaBytes ((1 + nsub_int) * (#const sizeof(regmatch_t))) $ \p_match -> do
79 -- add one because index zero covers the whole match
80 r <- c_regexec regex_ptr cstr (1 + nsub) p_match 0{-no flags for now-}
82 if (r /= 0) then return Nothing else do
84 (before,match,after) <- matched_parts str p_match
87 mapM (unpack str) $ take nsub_int $ tail $
88 iterate (`plusPtr` (#const sizeof(regmatch_t))) p_match
90 return (Just (before, match, after, sub_strs))
92 matched_parts :: String -> Ptr CRegMatch -> IO (String, String, String)
93 matched_parts string p_match = do
94 start <- (#peek regmatch_t, rm_so) p_match :: IO CInt
95 end <- (#peek regmatch_t, rm_eo) p_match :: IO CInt
96 let s = fromIntegral start; e = fromIntegral end
97 return ( take (s-1) string,
98 take (e-s) (drop s string),
101 unpack :: String -> Ptr CRegMatch -> IO (String)
102 unpack string p_match = do
103 start <- (#peek regmatch_t, rm_so) p_match :: IO CInt
104 end <- (#peek regmatch_t, rm_eo) p_match :: IO CInt
105 -- the subexpression may not have matched at all, perhaps because it
106 -- was optional. In this case, the offsets are set to -1.
107 if (start == -1) then return "" else do
108 return (take (fromIntegral (end-start)) (drop (fromIntegral start) string))
110 -- -----------------------------------------------------------------------------
111 -- The POSIX regex C interface
118 -- Return values from regexec
126 regIgnoreCase = REG_ICASE, \
130 -- Error codes from regcomp
148 foreign import ccall unsafe "regcomp"
149 c_regcomp :: Ptr CRegex -> CString -> CInt -> IO CInt
151 foreign import ccall unsafe "regfree"
152 c_regfree :: Ptr CRegex -> IO ()
154 foreign import ccall unsafe "regexec"
155 c_regexec :: Ptr CRegex -> CString -> CSize
156 -> Ptr CRegMatch -> CInt -> IO CInt