summaryrefslogtreecommitdiff
path: root/src/GF/Text/UTF8.hs
diff options
context:
space:
mode:
authoraarne <aarne@cs.chalmers.se>2008-06-25 16:54:35 +0000
committeraarne <aarne@cs.chalmers.se>2008-06-25 16:54:35 +0000
commite9e80fc389365e24d4300d7d5390c7d833a96c50 (patch)
treef0b58473adaa670bd8fc52ada419d8cad470ee03 /src/GF/Text/UTF8.hs
parentb96b36f43de3e2f8b58d5f539daa6f6d47f25870 (diff)
changed names of resource-1.3; added a note on homepage on release
Diffstat (limited to 'src/GF/Text/UTF8.hs')
-rw-r--r--src/GF/Text/UTF8.hs48
1 files changed, 48 insertions, 0 deletions
diff --git a/src/GF/Text/UTF8.hs b/src/GF/Text/UTF8.hs
new file mode 100644
index 000000000..5e9687684
--- /dev/null
+++ b/src/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