msyds/gf-core
1
0
Fork 0
forked from GitHub/gf-core
Code Pull Requests Activity

Telugu transliteration

Browse Source
This commit is contained in:
aarne
2008-09-23 08:49:58 +00:00
parent 9470a689c9
commit cbde244ac1
2 changed files with 16 additions and 0 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
src/GF/Command/Commands.hs
View File

@@ -450,6 +450,7 @@ allCommands enc pgf = Map.fromList [
options = [ options = [
("arabic", "Arabic"), ("arabic", "Arabic"),
("devanagari","Devanagari"), ("devanagari","Devanagari"),
("telugu", "Telugu"),
("thai", "Thai") ("thai", "Thai")
] ]
}), }),
@@ -594,6 +595,7 @@ stringOpOptions = [
("from_cp1251","decode from cp1251 (Cyrillic used in Bulgarian resource)"), ("from_cp1251","decode from cp1251 (Cyrillic used in Bulgarian resource)"),
("from_arabic","from unicode to GF Arabic transliteration"), ("from_arabic","from unicode to GF Arabic transliteration"),
("from_devanagari","from unicode to GF Devanagari transliteration"), ("from_devanagari","from unicode to GF Devanagari transliteration"),
("from_thai","from unicode to GF Telugu transliteration"),
("from_thai","from unicode to GF Thai transliteration"), ("from_thai","from unicode to GF Thai transliteration"),
("from_utf8","decode from utf8"), ("from_utf8","decode from utf8"),
("lextext","text-like lexer"), ("lextext","text-like lexer"),
@@ -603,6 +605,7 @@ stringOpOptions = [
("to_arabic","from GF Arabic transliteration to unicode"), ("to_arabic","from GF Arabic transliteration to unicode"),
("to_devanagari","from GF Devanagari transliteration to unicode"), ("to_devanagari","from GF Devanagari transliteration to unicode"),
("to_html","wrap in a html file with linebreaks"), ("to_html","wrap in a html file with linebreaks"),
("to_telugu","from GF Telugu transliteration to unicode"),
("to_thai","from GF Thai transliteration to unicode"), ("to_thai","from GF Thai transliteration to unicode"),
("to_utf8","encode to utf8"), ("to_utf8","encode to utf8"),
("unlextext","text-like unlexer"), ("unlextext","text-like unlexer"),

13
src/GF/Text/Transliterations.hs
View File

@@ -27,6 +27,7 @@ transliteration :: String -> Maybe Transliteration
transliteration s = case s of transliteration s = case s of
"arabic" -> Just transArabic "arabic" -> Just transArabic
"devanagari" -> Just transDevanagari "devanagari" -> Just transDevanagari
"telugu" -> Just transTelugu
"thai" -> Just transThai "thai" -> Just transThai
---- "urdu" -> Just transUrdu ---- "urdu" -> Just transUrdu
_ -> Nothing _ -> Nothing
@@ -114,3 +115,15 @@ transArabic = mkTransliteration allTrans allCodes where
allCodes = [0x0621..0x062f] ++ [0x0630..0x063a] ++ allCodes = [0x0621..0x062f] ++ [0x0630..0x063a] ++
[0x0641..0x064f] ++ [0x0650..0x0657] [0x0641..0x064f] ++ [0x0650..0x0657]
transTelugu :: Transliteration
transTelugu = mkTransliteration allTrans allCodes where
allTrans = words $
"- c1 c2 c3 - A A: I I: U U: R_ L_ - E E: " ++
"A' - O O: A_ k k. g g. n. c c. j j. n' T " ++
"T. d d. N t t. d d. n - p p. b b. m y " ++
"r R l L - v s' S s h - - - c5 a: i " ++
"i: u u: r_ r. - e e: a' - o o: a_ c6 - - " ++
"- - - - - c7 c8 z Z - - - - - - - " ++
"R+ L+ l+ l* - - n0 n1 n2 n3 n4 n5 n6 n7 n8 n9 "
allCodes = [0x0c00 .. 0x0c7f]