Import UCDN into source tree

https://github.com/grigorig/ucdn

configure.ac

@@ -125,6 +125,14 @@ AM_CONDITIONAL(HAVE_HB_OLD, $have_hb_old)
 
 dnl ===========================================================================
 
+have_ucdn=true
+if $have_ucdn; then
+	AC_DEFINE(HAVE_UCDN, 1, [Have UCDN Unicode functions])
+fi
+AM_CONDITIONAL(HAVE_UCDN, $have_ucdn)
+
+dnl ===========================================================================
+
 PKG_CHECK_MODULES(GLIB, glib-2.0 >= 2.16, have_glib=true, have_glib=false)
 if $have_glib; then
 	AC_DEFINE(HAVE_GLIB, 1, [Have glib2 library])
@@ -245,8 +253,9 @@ Makefile
 harfbuzz.pc
 src/Makefile
 src/hb-version.h
-src/hb-old/Makefile
 src/hb-icu-le/Makefile
+src/hb-old/Makefile
+src/hb-ucdn/Makefile
 util/Makefile
 test/Makefile
 test/api/Makefile

src/Makefile.am

@@ -178,6 +178,13 @@ HBSOURCES += hb-icu-le.cc
 endif
 DIST_SUBDIRS += hb-icu-le
 
+if HAVE_UCDN
+SUBDIRS += hb-ucdn
+HBCFLAGS += -I$(srcdir)/hb-ucdn
+HBLIBS   += hb-ucdn/libhb-ucdn.la
+HBSOURCES += hb-ucdn.cc
+endif
+DIST_SUBDIRS += hb-ucdn
 
 
 # Put the library together

src/hb-ucdn.cc

@@ -0,0 +1,207 @@
+/*
+ * Copyright (C) 2012 Grigori Goronzy <greg@kinoho.net>
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#include "hb-private.hh"
+
+#include "hb-unicode-private.hh"
+
+HB_BEGIN_DECLS
+#include "ucdn.h"
+HB_END_DECLS
+
+static const hb_script_t ucdn_script_translate[] =
+{
+    HB_SCRIPT_COMMON,
+    HB_SCRIPT_LATIN,
+    HB_SCRIPT_GREEK,
+    HB_SCRIPT_CYRILLIC,
+    HB_SCRIPT_ARMENIAN,
+    HB_SCRIPT_HEBREW,
+    HB_SCRIPT_ARABIC,
+    HB_SCRIPT_SYRIAC,
+    HB_SCRIPT_THAANA,
+    HB_SCRIPT_DEVANAGARI,
+    HB_SCRIPT_BENGALI,
+    HB_SCRIPT_GURMUKHI,
+    HB_SCRIPT_GUJARATI,
+    HB_SCRIPT_ORIYA,
+    HB_SCRIPT_TAMIL,
+    HB_SCRIPT_TELUGU,
+    HB_SCRIPT_KANNADA,
+    HB_SCRIPT_MALAYALAM,
+    HB_SCRIPT_SINHALA,
+    HB_SCRIPT_THAI,
+    HB_SCRIPT_LAO,
+    HB_SCRIPT_TIBETAN,
+    HB_SCRIPT_MYANMAR,
+    HB_SCRIPT_GEORGIAN,
+    HB_SCRIPT_HANGUL,
+    HB_SCRIPT_ETHIOPIC,
+    HB_SCRIPT_CHEROKEE,
+    HB_SCRIPT_CANADIAN_ABORIGINAL,
+    HB_SCRIPT_OGHAM,
+    HB_SCRIPT_RUNIC,
+    HB_SCRIPT_KHMER,
+    HB_SCRIPT_MONGOLIAN,
+    HB_SCRIPT_HIRAGANA,
+    HB_SCRIPT_KATAKANA,
+    HB_SCRIPT_BOPOMOFO,
+    HB_SCRIPT_HAN,
+    HB_SCRIPT_YI,
+    HB_SCRIPT_OLD_ITALIC,
+    HB_SCRIPT_GOTHIC,
+    HB_SCRIPT_DESERET,
+    HB_SCRIPT_INHERITED,
+    HB_SCRIPT_TAGALOG,
+    HB_SCRIPT_HANUNOO,
+    HB_SCRIPT_BUHID,
+    HB_SCRIPT_TAGBANWA,
+    HB_SCRIPT_LIMBU,
+    HB_SCRIPT_TAI_LE,
+    HB_SCRIPT_LINEAR_B,
+    HB_SCRIPT_UGARITIC,
+    HB_SCRIPT_SHAVIAN,
+    HB_SCRIPT_OSMANYA,
+    HB_SCRIPT_CYPRIOT,
+    HB_SCRIPT_BRAILLE,
+    HB_SCRIPT_BUGINESE,
+    HB_SCRIPT_COPTIC,
+    HB_SCRIPT_NEW_TAI_LUE,
+    HB_SCRIPT_GLAGOLITIC,
+    HB_SCRIPT_TIFINAGH,
+    HB_SCRIPT_SYLOTI_NAGRI,
+    HB_SCRIPT_OLD_PERSIAN,
+    HB_SCRIPT_KHAROSHTHI,
+    HB_SCRIPT_BALINESE,
+    HB_SCRIPT_CUNEIFORM,
+    HB_SCRIPT_PHOENICIAN,
+    HB_SCRIPT_PHAGS_PA,
+    HB_SCRIPT_NKO,
+    HB_SCRIPT_SUNDANESE,
+    HB_SCRIPT_LEPCHA,
+    HB_SCRIPT_OL_CHIKI,
+    HB_SCRIPT_VAI,
+    HB_SCRIPT_SAURASHTRA,
+    HB_SCRIPT_KAYAH_LI,
+    HB_SCRIPT_REJANG,
+    HB_SCRIPT_LYCIAN,
+    HB_SCRIPT_CARIAN,
+    HB_SCRIPT_LYDIAN,
+    HB_SCRIPT_CHAM,
+    HB_SCRIPT_TAI_THAM,
+    HB_SCRIPT_TAI_VIET,
+    HB_SCRIPT_AVESTAN,
+    HB_SCRIPT_EGYPTIAN_HIEROGLYPHS,
+    HB_SCRIPT_SAMARITAN,
+    HB_SCRIPT_LISU,
+    HB_SCRIPT_BAMUM,
+    HB_SCRIPT_JAVANESE,
+    HB_SCRIPT_MEETEI_MAYEK,
+    HB_SCRIPT_IMPERIAL_ARAMAIC,
+    HB_SCRIPT_OLD_SOUTH_ARABIAN,
+    HB_SCRIPT_INSCRIPTIONAL_PARTHIAN,
+    HB_SCRIPT_INSCRIPTIONAL_PAHLAVI,
+    HB_SCRIPT_OLD_TURKIC,
+    HB_SCRIPT_KAITHI,
+    HB_SCRIPT_BATAK,
+    HB_SCRIPT_BRAHMI,
+    HB_SCRIPT_MANDAIC,
+    HB_SCRIPT_CHAKMA,
+    HB_SCRIPT_MEROITIC_CURSIVE,
+    HB_SCRIPT_MEROITIC_HIEROGLYPHS,
+    HB_SCRIPT_MIAO,
+    HB_SCRIPT_SHARADA,
+    HB_SCRIPT_SORA_SOMPENG,
+    HB_SCRIPT_TAKRI,
+    HB_SCRIPT_UNKNOWN,
+};
+
+static hb_unicode_combining_class_t
+hb_ucdn_combining_class(hb_unicode_funcs_t *ufuncs, hb_codepoint_t unicode,
+        void *user_data)
+{
+    return (hb_unicode_combining_class_t) ucdn_get_combining_class(unicode);
+}
+
+static unsigned int
+hb_ucdn_eastasian_width(hb_unicode_funcs_t *ufuncs, hb_codepoint_t unicode,
+        void *user_data)
+{
+    int w = ucdn_get_east_asian_width(unicode);
+    return (w == UCDN_EAST_ASIAN_F || w == UCDN_EAST_ASIAN_W) ? 2 : 1;
+}
+
+static hb_unicode_general_category_t
+hb_ucdn_general_category(hb_unicode_funcs_t *ufuncs,
+        hb_codepoint_t unicode, void *user_data)
+{
+    return (hb_unicode_general_category_t)ucdn_get_general_category(unicode);
+}
+
+static hb_codepoint_t
+hb_ucdn_mirroring(hb_unicode_funcs_t *ufuncs, hb_codepoint_t unicode,
+        void *user_data)
+{
+    return ucdn_mirror(unicode);
+}
+
+static hb_script_t
+hb_ucdn_script(hb_unicode_funcs_t *ufuncs, hb_codepoint_t unicode,
+        void *user_data)
+{
+    return ucdn_script_translate[ucdn_get_script(unicode)];
+}
+
+static hb_bool_t
+hb_ucdn_compose(hb_unicode_funcs_t *ufuncs, hb_codepoint_t a,
+        hb_codepoint_t b, hb_codepoint_t *ab, void *user_data)
+{
+    return ucdn_compose(ab, a, b);
+}
+
+static hb_bool_t
+hb_ucdn_decompose(hb_unicode_funcs_t *ufuncs, hb_codepoint_t ab,
+        hb_codepoint_t *a, hb_codepoint_t *b, void *user_data)
+{
+    return ucdn_decompose(ab, a, b);
+}
+
+static unsigned int
+hb_ucdn_decompose_compatibility(hb_unicode_funcs_t *ufuncs, hb_codepoint_t u,
+        hb_codepoint_t *decomposed, void *user_data)
+{
+    return ucdn_compat_decompose(u, decomposed);
+}
+
+extern "C" HB_INTERNAL
+hb_unicode_funcs_t *
+hb_ucdn_get_unicode_funcs (void)
+{
+  static const hb_unicode_funcs_t _hb_ucdn_unicode_funcs = {
+    HB_OBJECT_HEADER_STATIC,
+
+    NULL, /* parent */
+    true, /* immutable */
+    {
+#define HB_UNICODE_FUNC_IMPLEMENT(name) hb_ucdn_##name,
+      HB_UNICODE_FUNCS_IMPLEMENT_CALLBACKS
+#undef HB_UNICODE_FUNC_IMPLEMENT
+    }
+  };
+
+  return const_cast<hb_unicode_funcs_t *> (&_hb_ucdn_unicode_funcs);
+}
+

src/hb-ucdn/Makefile.am

@@ -0,0 +1,18 @@
+## Process this file with automake to produce Makefile.in
+
+noinst_LTLIBRARIES = libhb-ucdn.la
+
+
+libhb_ucdn_la_SOURCES = \
+	ucdn.h \
+	ucdn.c \
+	unicodedata_db.h
+libhb_ucdn_CPPFLAGS = \
+	-I$(top_srcdir) \
+	-I$(top_srcdir)/src \
+	-I$(top_builddir)/src
+libhb_ucdn_la_LIBADD =
+
+EXTRA_DIST = README
+
+-include $(top_srcdir)/git.mk

src/hb-ucdn/README

@@ -0,0 +1,33 @@
+UCDN - Unicode Database and Normalization
+
+UCDN is a Unicode support library. Currently, it provides access
+to basic character properties contained in the Unicode Character
+Database and low-level normalization functions (pairwise canonical
+composition/decomposition and compatibility decomposition). More
+functionality might be provided in the future, such as additional
+properties, string normalization and encoding conversion.
+
+UCDN uses standard C89 with no particular dependencies or requirements
+except for stdint.h, and can be easily integrated into existing
+projects. However, it can also be used as a standalone library,
+and a CMake build script is provided for this. The first motivation
+behind UCDN development was to provide a standalone set of Unicode
+functions for the HarfBuzz OpenType shaping library. For this purpose,
+a HarfBuzz-specific wrapper is shipped along with it (hb-ucdn.h).
+
+UCDN is published under the ISC license, please see the license header
+in the C source code for more information. The makeunicodata.py script
+required for parsing Unicode database files is licensed under the
+PSF license, please see PYTHON-LICENSE for more information.
+
+UCDN was written by Grigori Goronzy <greg@kinoho.net>.
+
+How to Use
+
+Include ucdn.c, ucdn.h and unicodedata_db.h in your project. Now,
+just use the functions as documented in ucdn.h.
+
+In some cases, it might be necessary to regenerate the Unicode
+database file. The script makeunicodedata.py (Python 3.x required)
+fetches the appropriate files and dumps the compressed database into
+unicodedata_db.h.

src/hb-ucdn/ucdn.c

@@ -0,0 +1,282 @@
+/*
+ * Copyright (C) 2012 Grigori Goronzy <greg@kinoho.net>
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#include <stdio.h>
+#include <stdlib.h>
+#include <stdint.h>
+#include "ucdn.h"
+
+typedef struct {
+    const unsigned char category;
+    const unsigned char combining;
+    const unsigned char bidi_class;
+    const unsigned char mirrored;
+    const unsigned char east_asian_width;
+    const unsigned char normalization_check;
+    const unsigned char script;
+} UCDRecord;
+
+typedef struct {
+    unsigned short from, to;
+} MirrorPair;
+
+typedef struct {
+    int start;
+    short count, index;
+} Reindex;
+
+#include "unicodedata_db.h"
+
+/* constants required for Hangul (de)composition */
+#define SBASE 0xAC00
+#define LBASE 0x1100
+#define VBASE 0x1161
+#define TBASE 0x11A7
+#define SCOUNT 11172
+#define LCOUNT 19
+#define VCOUNT 21
+#define TCOUNT 28
+#define NCOUNT (VCOUNT * TCOUNT)
+
+static UCDRecord *get_ucd_record(uint32_t code)
+{
+    int index, offset;
+
+    if (code >= 0x110000)
+        index = 0;
+    else {
+        index  = index0[code >> (SHIFT1+SHIFT2)] << SHIFT1;
+        offset = (code >> SHIFT2) & ((1<<SHIFT1) - 1);
+        index  = index1[index + offset] << SHIFT2;
+        offset = code & ((1<<SHIFT2) - 1);
+        index  = index2[index + offset];
+    }
+
+    return &ucd_records[index];
+}
+
+static unsigned short *get_decomp_record(uint32_t code)
+{
+    int index, offset;
+
+    if (code >= 0x110000)
+        index = 0;
+    else {
+        index  = decomp_index0[code >> (DECOMP_SHIFT1+DECOMP_SHIFT2)]
+            << DECOMP_SHIFT1;
+        offset = (code >> DECOMP_SHIFT2) & ((1<<DECOMP_SHIFT1) - 1);
+        index  = decomp_index1[index + offset] << DECOMP_SHIFT2;
+        offset = code & ((1<<DECOMP_SHIFT2) - 1);
+        index  = decomp_index2[index + offset];
+    }
+
+    return &decomp_data[index];
+}
+
+static int get_comp_index(uint32_t code, Reindex *idx)
+{
+    int i;
+
+    for (i = 0; idx[i].start; i++) {
+        Reindex *cur = &idx[i];
+        if (code < cur->start)
+            return -1;
+        if (code <= cur->start + cur->count) {
+            return cur->index + (code - cur->start);
+        }
+    }
+
+    return -1;
+}
+
+static int compare_mp(const void *a, const void *b)
+{
+    MirrorPair *mpa = (MirrorPair *)a;
+    MirrorPair *mpb = (MirrorPair *)b;
+    return mpa->from - mpb->from;
+}
+
+static int hangul_pair_decompose(uint32_t code, uint32_t *a, uint32_t *b)
+{
+    int si = code - SBASE;
+
+    if (si < 0 || si >= SCOUNT)
+        return 0;
+
+    if (si % TCOUNT) {
+        /* LV,T */
+        *a = SBASE + (si / TCOUNT) * TCOUNT;
+        *b = TBASE + (si % TCOUNT);
+        return 3;
+    } else {
+        /* L,V */
+        *a = LBASE + (si / NCOUNT);
+        *b = VBASE + (si % NCOUNT) / TCOUNT;
+        return 2;
+    }
+}
+
+static int hangul_pair_compose(uint32_t *code, uint32_t a, uint32_t b)
+{
+    if (b < VBASE || b >= (TBASE + TCOUNT))
+        return 0;
+
+    if ((a < LBASE || a >= (LBASE + LCOUNT))
+            && (a < SBASE || a >= (SBASE + SCOUNT)))
+        return 0;
+
+    if (a >= SBASE) {
+        /* LV,T */
+        *code = a + (b - TBASE);
+        return 3;
+    } else {
+        /* L,V */
+        int li = a - LBASE;
+        int vi = b - VBASE;
+        *code = SBASE + li * NCOUNT + vi * TCOUNT;
+        return 2;
+    }
+}
+
+static uint32_t decode_utf16(unsigned short **code_ptr)
+{
+    unsigned short *code = *code_ptr;
+
+    if ((code[0] & 0xd800) != 0xd800) {
+        *code_ptr += 1;
+        return (uint32_t)code[0];
+    } else {
+        *code_ptr += 2;
+        return 0x10000 + ((uint32_t)code[1] - 0xdc00) +
+            (((uint32_t)code[0] - 0xd800) << 10);
+    }
+}
+
+const char *ucdn_get_unicode_version(void)
+{
+    return UNIDATA_VERSION;
+}
+
+int ucdn_get_combining_class(uint32_t code)
+{
+    return get_ucd_record(code)->combining;
+}
+
+int ucdn_get_east_asian_width(uint32_t code)
+{
+    return get_ucd_record(code)->east_asian_width;
+}
+
+int ucdn_get_general_category(uint32_t code)
+{
+    return get_ucd_record(code)->category;
+}
+
+int ucdn_get_bidi_class(uint32_t code)
+{
+    return get_ucd_record(code)->bidi_class;
+}
+
+int ucdn_get_mirrored(uint32_t code)
+{
+    return get_ucd_record(code)->mirrored;
+}
+
+int ucdn_get_script(uint32_t code)
+{
+    return get_ucd_record(code)->script;
+}
+
+uint32_t ucdn_mirror(uint32_t code)
+{
+    MirrorPair mp = {0};
+    MirrorPair *res;
+
+    if (get_ucd_record(code)->mirrored == 0)
+        return code;
+
+    mp.from = code;
+    res = bsearch(&mp, mirror_pairs, BIDI_MIRROR_LEN, sizeof(MirrorPair),
+            compare_mp);
+
+    if (res == NULL)
+        return code;
+    else
+        return res->to;
+}
+
+int ucdn_decompose(uint32_t code, uint32_t *a, uint32_t *b)
+{
+    unsigned short *rec;
+    int len;
+
+    if (hangul_pair_decompose(code, a, b))
+        return 1;
+
+    rec = get_decomp_record(code);
+    len = rec[0] >> 8;
+
+    if ((rec[0] & 0xff) != 0 || len == 0)
+        return 0;
+
+    rec++;
+    *a = decode_utf16(&rec);
+    if (len > 1)
+        *b = decode_utf16(&rec);
+    else
+        *b = 0;
+
+    return 1;
+}
+
+int ucdn_compose(uint32_t *code, uint32_t a, uint32_t b)
+{
+    int l, r, index, indexi, offset;
+
+    if (hangul_pair_compose(code, a, b))
+        return 1;
+
+    l = get_comp_index(a, nfc_first);
+    r = get_comp_index(b, nfc_last);
+
+    if (l < 0 || r < 0)
+        return 0;
+
+    indexi = l * TOTAL_LAST + r;
+    index  = comp_index0[indexi >> (COMP_SHIFT1+COMP_SHIFT2)] << COMP_SHIFT1;
+    offset = (indexi >> COMP_SHIFT2) & ((1<<COMP_SHIFT1) - 1);
+    index  = comp_index1[index + offset] << COMP_SHIFT2;
+    offset = indexi & ((1<<COMP_SHIFT2) - 1);
+    *code  = comp_data[index + offset];
+
+    return *code != 0;
+}
+
+int ucdn_compat_decompose(uint32_t code, uint32_t *decomposed)
+{
+    int i, len;
+    unsigned short *rec = get_decomp_record(code);
+    len = rec[0] >> 8;
+
+    if (len == 0)
+        return 0;
+
+    rec++;
+    for (i = 0; i < len; i++)
+        decomposed[i] = decode_utf16(&rec);
+
+    return len;
+}

src/hb-ucdn/ucdn.h

@@ -0,0 +1,290 @@
+/*
+ * Copyright (C) 2012 Grigori Goronzy <greg@kinoho.net>
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#ifndef UCDN_H
+#define UCDN_H
+
+#include <stdint.h>
+
+#define UCDN_EAST_ASIAN_F 0
+#define UCDN_EAST_ASIAN_H 1
+#define UCDN_EAST_ASIAN_W 2
+#define UCDN_EAST_ASIAN_NA 3
+#define UCDN_EAST_ASIAN_A 4
+#define UCDN_EAST_ASIAN_N 5
+
+#define UCDN_SCRIPT_COMMON 0
+#define UCDN_SCRIPT_LATIN 1
+#define UCDN_SCRIPT_GREEK 2
+#define UCDN_SCRIPT_CYRILLIC 3
+#define UCDN_SCRIPT_ARMENIAN 4
+#define UCDN_SCRIPT_HEBREW 5
+#define UCDN_SCRIPT_ARABIC 6
+#define UCDN_SCRIPT_SYRIAC 7
+#define UCDN_SCRIPT_THAANA 8
+#define UCDN_SCRIPT_DEVANAGARI 9
+#define UCDN_SCRIPT_BENGALI 10
+#define UCDN_SCRIPT_GURMUKHI 11
+#define UCDN_SCRIPT_GUJARATI 12
+#define UCDN_SCRIPT_ORIYA 13
+#define UCDN_SCRIPT_TAMIL 14
+#define UCDN_SCRIPT_TELUGU 15
+#define UCDN_SCRIPT_KANNADA 16
+#define UCDN_SCRIPT_MALAYALAM 17
+#define UCDN_SCRIPT_SINHALA 18
+#define UCDN_SCRIPT_THAI 19
+#define UCDN_SCRIPT_LAO 20
+#define UCDN_SCRIPT_TIBETAN 21
+#define UCDN_SCRIPT_MYANMAR 22
+#define UCDN_SCRIPT_GEORGIAN 23
+#define UCDN_SCRIPT_HANGUL 24
+#define UCDN_SCRIPT_ETHIOPIC 25
+#define UCDN_SCRIPT_CHEROKEE 26
+#define UCDN_SCRIPT_CANADIAN_ABORIGINAL 27
+#define UCDN_SCRIPT_OGHAM 28
+#define UCDN_SCRIPT_RUNIC 29
+#define UCDN_SCRIPT_KHMER 30
+#define UCDN_SCRIPT_MONGOLIAN 31
+#define UCDN_SCRIPT_HIRAGANA 32
+#define UCDN_SCRIPT_KATAKANA 33
+#define UCDN_SCRIPT_BOPOMOFO 34
+#define UCDN_SCRIPT_HAN 35
+#define UCDN_SCRIPT_YI 36
+#define UCDN_SCRIPT_OLD_ITALIC 37
+#define UCDN_SCRIPT_GOTHIC 38
+#define UCDN_SCRIPT_DESERET 39
+#define UCDN_SCRIPT_INHERITED 40
+#define UCDN_SCRIPT_TAGALOG 41
+#define UCDN_SCRIPT_HANUNOO 42
+#define UCDN_SCRIPT_BUHID 43
+#define UCDN_SCRIPT_TAGBANWA 44
+#define UCDN_SCRIPT_LIMBU 45
+#define UCDN_SCRIPT_TAI_LE 46
+#define UCDN_SCRIPT_LINEAR_B 47
+#define UCDN_SCRIPT_UGARITIC 48
+#define UCDN_SCRIPT_SHAVIAN 49
+#define UCDN_SCRIPT_OSMANYA 50
+#define UCDN_SCRIPT_CYPRIOT 51
+#define UCDN_SCRIPT_BRAILLE 52
+#define UCDN_SCRIPT_BUGINESE 53
+#define UCDN_SCRIPT_COPTIC 54
+#define UCDN_SCRIPT_NEW_TAI_LUE 55
+#define UCDN_SCRIPT_GLAGOLITIC 56
+#define UCDN_SCRIPT_TIFINAGH 57
+#define UCDN_SCRIPT_SYLOTI_NAGRI 58
+#define UCDN_SCRIPT_OLD_PERSIAN 59
+#define UCDN_SCRIPT_KHAROSHTHI 60
+#define UCDN_SCRIPT_BALINESE 61
+#define UCDN_SCRIPT_CUNEIFORM 62
+#define UCDN_SCRIPT_PHOENICIAN 63
+#define UCDN_SCRIPT_PHAGS_PA 64
+#define UCDN_SCRIPT_NKO 65
+#define UCDN_SCRIPT_SUNDANESE 66
+#define UCDN_SCRIPT_LEPCHA 67
+#define UCDN_SCRIPT_OL_CHIKI 68
+#define UCDN_SCRIPT_VAI 69
+#define UCDN_SCRIPT_SAURASHTRA 70
+#define UCDN_SCRIPT_KAYAH_LI 71
+#define UCDN_SCRIPT_REJANG 72
+#define UCDN_SCRIPT_LYCIAN 73
+#define UCDN_SCRIPT_CARIAN 74
+#define UCDN_SCRIPT_LYDIAN 75
+#define UCDN_SCRIPT_CHAM 76
+#define UCDN_SCRIPT_TAI_THAM 77
+#define UCDN_SCRIPT_TAI_VIET 78
+#define UCDN_SCRIPT_AVESTAN 79
+#define UCDN_SCRIPT_EGYPTIAN_HIEROGLYPHS 80
+#define UCDN_SCRIPT_SAMARITAN 81
+#define UCDN_SCRIPT_LISU 82
+#define UCDN_SCRIPT_BAMUM 83
+#define UCDN_SCRIPT_JAVANESE 84
+#define UCDN_SCRIPT_MEETEI_MAYEK 85
+#define UCDN_SCRIPT_IMPERIAL_ARAMAIC 86
+#define UCDN_SCRIPT_OLD_SOUTH_ARABIAN 87
+#define UCDN_SCRIPT_INSCRIPTIONAL_PARTHIAN 88
+#define UCDN_SCRIPT_INSCRIPTIONAL_PAHLAVI 89
+#define UCDN_SCRIPT_OLD_TURKIC 90
+#define UCDN_SCRIPT_KAITHI 91
+#define UCDN_SCRIPT_BATAK 92
+#define UCDN_SCRIPT_BRAHMI 93
+#define UCDN_SCRIPT_MANDAIC 94
+#define UCDN_SCRIPT_CHAKMA 95
+#define UCDN_SCRIPT_MEROITIC_CURSIVE 96
+#define UCDN_SCRIPT_MEROITIC_HIEROGLYPHS 97
+#define UCDN_SCRIPT_MIAO 98
+#define UCDN_SCRIPT_SHARADA 99
+#define UCDN_SCRIPT_SORA_SOMPENG 100
+#define UCDN_SCRIPT_TAKRI 101
+#define UCDN_SCRIPT_UNKNOWN 102
+
+#define UCDN_GENERAL_CATEGORY_CC 0
+#define UCDN_GENERAL_CATEGORY_CF 1
+#define UCDN_GENERAL_CATEGORY_CN 2
+#define UCDN_GENERAL_CATEGORY_CO 3
+#define UCDN_GENERAL_CATEGORY_CS 4
+#define UCDN_GENERAL_CATEGORY_LL 5
+#define UCDN_GENERAL_CATEGORY_LM 6
+#define UCDN_GENERAL_CATEGORY_LO 7
+#define UCDN_GENERAL_CATEGORY_LT 8
+#define UCDN_GENERAL_CATEGORY_LU 9
+#define UCDN_GENERAL_CATEGORY_MC 10
+#define UCDN_GENERAL_CATEGORY_ME 11
+#define UCDN_GENERAL_CATEGORY_MN 12
+#define UCDN_GENERAL_CATEGORY_ND 13
+#define UCDN_GENERAL_CATEGORY_NL 14
+#define UCDN_GENERAL_CATEGORY_NO 15
+#define UCDN_GENERAL_CATEGORY_PC 16
+#define UCDN_GENERAL_CATEGORY_PD 17
+#define UCDN_GENERAL_CATEGORY_PE 18
+#define UCDN_GENERAL_CATEGORY_PF 19
+#define UCDN_GENERAL_CATEGORY_PI 20
+#define UCDN_GENERAL_CATEGORY_PO 21
+#define UCDN_GENERAL_CATEGORY_PS 22
+#define UCDN_GENERAL_CATEGORY_SC 23
+#define UCDN_GENERAL_CATEGORY_SK 24
+#define UCDN_GENERAL_CATEGORY_SM 25
+#define UCDN_GENERAL_CATEGORY_SO 26
+#define UCDN_GENERAL_CATEGORY_ZL 27
+#define UCDN_GENERAL_CATEGORY_ZP 28
+#define UCDN_GENERAL_CATEGORY_ZS 29
+
+#define UCDN_BIDI_CLASS_L 0
+#define UCDN_BIDI_CLASS_LRE 1
+#define UCDN_BIDI_CLASS_LRO 2
+#define UCDN_BIDI_CLASS_R 3
+#define UCDN_BIDI_CLASS_AL 4
+#define UCDN_BIDI_CLASS_RLE 5
+#define UCDN_BIDI_CLASS_RLO 6
+#define UCDN_BIDI_CLASS_PDF 7
+#define UCDN_BIDI_CLASS_EN 8
+#define UCDN_BIDI_CLASS_ES 9
+#define UCDN_BIDI_CLASS_ET 10
+#define UCDN_BIDI_CLASS_AN 11
+#define UCDN_BIDI_CLASS_CS 12
+#define UCDN_BIDI_CLASS_NSM 13
+#define UCDN_BIDI_CLASS_BN 14
+#define UCDN_BIDI_CLASS_B 15
+#define UCDN_BIDI_CLASS_S 16
+#define UCDN_BIDI_CLASS_WS 17
+#define UCDN_BIDI_CLASS_ON 18
+
+/**
+ * Return version of the Unicode database.
+ *
+ * @return Unicode database version
+ */
+const char *ucdn_get_unicode_version(void);
+
+/**
+ * Get combining class of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return combining class value, as defined in UAX#44
+ */
+int ucdn_get_combining_class(uint32_t code);
+
+/**
+ * Get east-asian width of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_EAST_ASIAN_* and as defined in UAX#11.
+ */
+int ucdn_get_east_asian_width(uint32_t code);
+
+/**
+ * Get general category of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_GENERAL_CATEGORY_* and as defined in
+ * UAX#44.
+ */
+int ucdn_get_general_category(uint32_t code);
+
+/**
+ * Get bidirectional class of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_BIDI_CLASS_* and as defined in UAX#44.
+ */
+int ucdn_get_bidi_class(uint32_t code);
+
+/**
+ * Get script of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_SCRIPT_* and as defined in UAX#24.
+ */
+int ucdn_get_script(uint32_t code);
+
+/**
+ * Check if codepoint can be mirrored.
+ *
+ * @param code Unicode codepoint
+ * @return 1 if mirrored character exists, otherwise 0
+ */
+int ucdn_get_mirrored(uint32_t code);
+
+/**
+ * Mirror a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return mirrored codepoint or the original codepoint if no
+ * mirrored character exists
+ */
+uint32_t ucdn_mirror(uint32_t code);
+
+/**
+ * Pairwise canonical decomposition of a codepoint. This includes
+ * Hangul Jamo decomposition (see chapter 3.12 of the Unicode core
+ * specification).
+ *
+ * Hangul is decomposed into L and V jamos for LV forms, and an
+ * LV precomposed syllable and a T jamo for LVT forms.
+ *
+ * @param code Unicode codepoint
+ * @param a filled with first codepoint of decomposition
+ * @param b filled with second codepoint of decomposition, or 0
+ * @return success
+ */
+int ucdn_decompose(uint32_t code, uint32_t *a, uint32_t *b);
+
+/**
+ * Compatibility decomposition of a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @param decomposed filled with decomposition, must be able to hold 18
+ * characters
+ * @return length of decomposition or 0 in case none exists
+ */
+int ucdn_compat_decompose(uint32_t code, uint32_t *decomposed);
+
+/**
+ * Pairwise canonical composition of two codepoints. This includes
+ * Hangul Jamo composition (see chapter 3.12 of the Unicode core
+ * specification).
+ *
+ * Hangul composition expects either L and V jamos, or an LV
+ * precomposed syllable and a T jamo. This is exactly the inverse
+ * of pairwise Hangul decomposition.
+ *
+ * @param code filled with composition
+ * @param a first codepoint
+ * @param b second codepoint
+ * @return success
+ */
+int ucdn_compose(uint32_t *code, uint32_t a, uint32_t b);
+
+#endif

src/hb-unicode.cc

@@ -112,6 +112,7 @@ hb_unicode_decompose_compatibility_nil (hb_unicode_funcs_t *ufuncs     HB_UNUSED
 #define HB_UNICODE_FUNCS_IMPLEMENT_SET \
   HB_UNICODE_FUNCS_IMPLEMENT (glib) \
   HB_UNICODE_FUNCS_IMPLEMENT (icu) \
+  HB_UNICODE_FUNCS_IMPLEMENT (ucdn) \
   HB_UNICODE_FUNCS_IMPLEMENT (nil) \
   /* ^--- Add new callbacks before nil */
 
@@ -134,6 +135,8 @@ hb_unicode_funcs_get_default (void)
   HB_UNICODE_FUNCS_IMPLEMENT(glib)
 #elif defined(HAVE_ICU)
   HB_UNICODE_FUNCS_IMPLEMENT(icu)
+#elif defined(HAVE_UCDN)
+  HB_UNICODE_FUNCS_IMPLEMENT(ucdn)
 #else
 #define HB_UNICODE_FUNCS_NIL 1
   HB_UNICODE_FUNCS_IMPLEMENT(nil)