From 2b6f1b59eaf243f5e173ba684a982342d97c5948 Mon Sep 17 00:00:00 2001
From: Bryan O'Sullivan <bos@serpentine.com>
Date: Wed, 15 Sep 2010 04:33:40 +0000
Subject: [PATCH] Combining class, and digit conversion

--HG--
extra : convert_revision : f801550777325b50474f1ff21cc32572db3bc631
---
 Data/Text/ICU/Char.hs | 30 ++++++++++++++++++++++++++++++
 cbits/text_icu.c      | 11 +++++++++++
 include/hs_text_icu.h |  2 ++
 3 files changed, 43 insertions(+)

diff --git a/Data/Text/ICU/Char.hs b/Data/Text/ICU/Char.hs
index a44e354..0a1e69b 100644
--- a/Data/Text/ICU/Char.hs
+++ b/Data/Text/ICU/Char.hs
@@ -17,14 +17,18 @@ module Data.Text.ICU.Char
     , Direction(..)
     -- * Functions
     , blockCode
+    , combiningClass
+    , digitToInt
     , direction
     , isMirrored
     , mirror
     ) where
 
 import Data.Char (chr, ord)
+import Data.Int (Int32)
 import Data.Text.ICU.Internal (UBool, UChar32, asBool)
 import Data.Typeable (Typeable)
+import Data.Word (Word8)
 import Foreign.C.Types (CInt)
 
 -- | The language directional property of a character set.
@@ -264,6 +268,26 @@ mirror :: Char -> Char
 mirror = chr . fromIntegral . u_charMirror . fromIntegral . ord
 {-# INLINE mirror #-}
 
+combiningClass :: Char -> Int
+combiningClass = fromIntegral . u_getCombiningClass . fromIntegral . ord
+{-# INLINE combiningClass #-}
+
+-- | Return the decimal digit value of a decimal digit character.
+-- Such characters have the general category @Nd@ (decimal digit
+-- numbers) and a @Numeric_Type@ of @Decimal@.
+--
+-- No digit values are returned for any Han characters, because Han
+-- number characters are often used with a special Chinese-style
+-- number format (with characters for powers of 10 in between) instead
+-- of in decimal-positional notation.  Unicode 4 explicitly assigns
+-- Han number characters a @Numeric_Type@ of @Numeric@ instead of
+-- @Decimal@.
+digitToInt :: Char -> Maybe Int
+digitToInt c
+    | i == -1   = Nothing
+    | otherwise = Just $! fromIntegral i
+  where i = u_charDigitValue . fromIntegral . ord $ c
+
 type UBlockCode = CInt
 type UCharDirection = CInt
 
@@ -278,3 +302,9 @@ foreign import ccall unsafe "hs_text_icu.h __hs_u_isMirrored" u_isMirrored
 
 foreign import ccall unsafe "hs_text_icu.h __hs_u_charMirror" u_charMirror
     :: UChar32 -> UChar32
+
+foreign import ccall unsafe "hs_text_icu.h __hs_u_getCombingingClass" u_getCombiningClass
+    :: UChar32 -> Word8
+
+foreign import ccall unsafe "hs_text_icu.h __hs_u_charDigitValue" u_charDigitValue
+    :: UChar32 -> Int32
diff --git a/cbits/text_icu.c b/cbits/text_icu.c
index 458db1e..b546af6 100644
--- a/cbits/text_icu.c
+++ b/cbits/text_icu.c
@@ -295,3 +295,14 @@ UChar32 __hs_u_charMirror(UChar32 c)
 {
     return u_charMirror(c);
 }
+
+uint8_t __hs_u_getCombiningClass(UChar32 c)
+{
+    return u_getCombiningClass(c);
+}
+
+int32_t __hs_u_charDigitValue(UChar32 c)
+{
+    return u_charDigitValue(c);
+}
+
diff --git a/include/hs_text_icu.h b/include/hs_text_icu.h
index 12990d9..f9e1ad5 100644
--- a/include/hs_text_icu.h
+++ b/include/hs_text_icu.h
@@ -36,6 +36,8 @@ UBlockCode __hs_ublock_getCode(UChar32 c);
 UCharDirection __hs_u_charDirection(UChar32 c);
 UBool __hs_u_isMirrored(UChar32 c);
 UChar32 __hs_u_charMirror(UChar32 c);
+uint8_t __hs_u_getCombiningClass(UChar32 c);
+int32_t __hs_u_charDigitValue(UChar32 c);
 
 /* ucol.h */