diff options
| author | aarne <aarne@cs.chalmers.se> | 2008-05-21 09:26:44 +0000 |
|---|---|---|
| committer | aarne <aarne@cs.chalmers.se> | 2008-05-21 09:26:44 +0000 |
| commit | 055c0d0d5a5bb0dc75904fe53df7f2e4f5732a8f (patch) | |
| tree | 0e63fb68c69c8f6ad0f78893c63420f0a3600e1c /src-3.0/GF/Text/UTF8.hs | |
| parent | 915a1de71783ab8446b1af9e72c7ba7dfbc12d3f (diff) | |
GF/src is now for 2.9, and the new sources are in src-3.0 - keep it this way until the release of GF 3
Diffstat (limited to 'src-3.0/GF/Text/UTF8.hs')
| -rw-r--r-- | src-3.0/GF/Text/UTF8.hs | 48 |
1 files changed, 48 insertions, 0 deletions
diff --git a/src-3.0/GF/Text/UTF8.hs b/src-3.0/GF/Text/UTF8.hs new file mode 100644 index 000000000..5e9687684 --- /dev/null +++ b/src-3.0/GF/Text/UTF8.hs @@ -0,0 +1,48 @@ +---------------------------------------------------------------------- +-- | +-- Module : UTF8 +-- Maintainer : AR +-- Stability : (stable) +-- Portability : (portable) +-- +-- > CVS $Date: 2005/04/21 16:23:42 $ +-- > CVS $Author: bringert $ +-- > CVS $Revision: 1.5 $ +-- +-- From the Char module supplied with HBC. +-- code by Thomas Hallgren (Jul 10 1999) +----------------------------------------------------------------------------- + +module GF.Text.UTF8 (decodeUTF8, encodeUTF8) where + +-- | Take a Unicode string and encode it as a string +-- with the UTF8 method. +decodeUTF8 :: String -> String +decodeUTF8 "" = "" +decodeUTF8 (c:cs) | c < '\x80' = c : decodeUTF8 cs +decodeUTF8 (c:c':cs) | '\xc0' <= c && c <= '\xdf' && + '\x80' <= c' && c' <= '\xbf' = + toEnum ((fromEnum c `mod` 0x20) * 0x40 + fromEnum c' `mod` 0x40) : decodeUTF8 cs +decodeUTF8 (c:c':c'':cs) | '\xe0' <= c && c <= '\xef' && + '\x80' <= c' && c' <= '\xbf' && + '\x80' <= c'' && c'' <= '\xbf' = + toEnum ((fromEnum c `mod` 0x10 * 0x1000) + (fromEnum c' `mod` 0x40) * 0x40 + fromEnum c'' `mod` 0x40) : decodeUTF8 cs +decodeUTF8 s = s ---- AR workaround 22/6/2006 +----decodeUTF8 _ = error "UniChar.decodeUTF8: bad data" + +encodeUTF8 :: String -> String +encodeUTF8 "" = "" +encodeUTF8 (c:cs) = + if c > '\x0000' && c < '\x0080' then + c : encodeUTF8 cs + else if c < toEnum 0x0800 then + let i = fromEnum c + in toEnum (0xc0 + i `div` 0x40) : + toEnum (0x80 + i `mod` 0x40) : + encodeUTF8 cs + else + let i = fromEnum c + in toEnum (0xe0 + i `div` 0x1000) : + toEnum (0x80 + (i `mod` 0x1000) `div` 0x40) : + toEnum (0x80 + i `mod` 0x40) : + encodeUTF8 cs |
