haskell/encoding
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Fix chr calling on big numbers in UTF8

Browse Source 
Ignore-this: 3a7650c4b75691b79b082db6bddd5b8b

darcs-hash:20090813022455-a4fee-1989c74eb7f7a9daa87f902da138c45185ea520e
This commit is contained in:
Henning Guenther 2009-08-12 19:24:55 -07:00
parent c4a1e63869
commit 1ad8755a80
1 changed files with 10 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

16
Data/Encoding/UTF8.hs
View File

@ -69,11 +69,13 @@ instance Encoding UTF8 where
v2 = w2 .&. 0x3F v2 = w2 .&. 0x3F
v3 = w3 .&. 0x3F v3 = w3 .&. 0x3F
v4 = w4 .&. 0x3F v4 = w4 .&. 0x3F
return $ chr $ v = ((fromIntegral v1) `shiftL` 18)
((fromIntegral v1) `shiftL` 18)
.|. ((fromIntegral v2) `shiftL` 12) .|. ((fromIntegral v2) `shiftL` 12)
.|. ((fromIntegral v3) `shiftL` 6) .|. ((fromIntegral v3) `shiftL` 6)
.|. (fromIntegral v4) .|. (fromIntegral v4)
if v <= 0x10FFFF
then return $ chr v
else throwException (IllegalRepresentation [w1,w2,w3,w4])
| otherwise -> throwException (IllegalCharacter w1) | otherwise -> throwException (IllegalCharacter w1)
decodeChar UTF8Strict = do decodeChar UTF8Strict = do
w1 <- fetchWord8 w1 <- fetchWord8
@ -109,13 +111,15 @@ instance Encoding UTF8 where
v2 = w2 .&. 0x3F v2 = w2 .&. 0x3F
v3 = w3 .&. 0x3F v3 = w3 .&. 0x3F
v4 = w4 .&. 0x3F v4 = w4 .&. 0x3F
if v1 == 0 && v2 < 0x10 v = ((fromIntegral v1) `shiftL` 18)
then throwException (IllegalRepresentation [w1,w2,w3,w4])
else return $ chr $
((fromIntegral v1) `shiftL` 18)
.|. ((fromIntegral v2) `shiftL` 12) .|. ((fromIntegral v2) `shiftL` 12)
.|. ((fromIntegral v3) `shiftL` 6) .|. ((fromIntegral v3) `shiftL` 6)
.|. (fromIntegral v4) .|. (fromIntegral v4)
if v1 == 0 && v2 < 0x10
then throwException (IllegalRepresentation [w1,w2,w3,w4])
else (if v <= 0x10FFFF
then return $ chr v
else throwException (IllegalRepresentation [w1,w2,w3,w4]))
| otherwise -> throwException (IllegalCharacter w1) | otherwise -> throwException (IllegalCharacter w1)
where where
invalidExtend wrd = wrd .&. 0xC0 /= 0x80 invalidExtend wrd = wrd .&. 0xC0 /= 0x80