[ucdn] Update to Unicode 9.0.0

Fixes https://github.com/grigorig/ucdn/issues/12

src/hb-ucdn.cc

@@ -154,6 +154,12 @@ static const hb_script_t ucdn_script_translate[] =
     HB_SCRIPT_MULTANI,
     HB_SCRIPT_OLD_HUNGARIAN,
     HB_SCRIPT_SIGNWRITING,
+    HB_SCRIPT_ADLAM,
+    HB_SCRIPT_BHAIKSUKI,
+    HB_SCRIPT_MARCHEN,
+    HB_SCRIPT_NEWA,
+    HB_SCRIPT_OSAGE,
+    HB_SCRIPT_TANGUT,
 };
 
 static hb_unicode_combining_class_t

src/hb-ucdn/README

@@ -1,7 +1,6 @@
 Contents of this directory are derived from UCDN:
 
   https://github.com/grigorig/ucdn
-  https://github.com/behdad/ucdn
 
 The original README follows:
 

src/hb-ucdn/ucdn.c

@@ -16,6 +16,7 @@
 
 #include <stdio.h>
 #include <stdlib.h>
+#include <stdint.h>
 #include "ucdn.h"
 
 typedef struct {
@@ -24,14 +25,19 @@ typedef struct {
     unsigned char bidi_class;
     unsigned char mirrored;
     unsigned char east_asian_width;
-    unsigned char normalization_check;
     unsigned char script;
+    unsigned char linebreak_class;
 } UCDRecord;
 
 typedef struct {
     unsigned short from, to;
 } MirrorPair;
 
+typedef struct {
+  unsigned short from, to;
+  unsigned char type;
+} BracketPair;
+
 typedef struct {
     unsigned int start;
     short count, index;
@@ -108,6 +114,24 @@ static int compare_mp(const void *a, const void *b)
     return mpa->from - mpb->from;
 }
 
+static int compare_bp(const void *a, const void *b)
+{
+    BracketPair *bpa = (BracketPair *)a;
+    BracketPair *bpb = (BracketPair *)b;
+    return bpa->from - bpb->from;
+}
+
+static BracketPair *search_bp(uint32_t code)
+{
+    BracketPair bp = {0,0,2};
+    BracketPair *res;
+
+    bp.from = code;
+    res = bsearch(&bp, bracket_pairs, BIDI_BRACKET_LEN, sizeof(BracketPair),
+            compare_bp);
+    return res;
+}
+
 static int hangul_pair_decompose(uint32_t code, uint32_t *a, uint32_t *b)
 {
     int si = code - SBASE;
@@ -199,6 +223,42 @@ int ucdn_get_script(uint32_t code)
     return get_ucd_record(code)->script;
 }
 
+int ucdn_get_linebreak_class(uint32_t code)
+{
+    return get_ucd_record(code)->linebreak_class;
+}
+
+int ucdn_get_resolved_linebreak_class(uint32_t code)
+{
+    const UCDRecord *record = get_ucd_record(code);
+
+    switch (record->linebreak_class)
+    {
+    case UCDN_LINEBREAK_CLASS_AI:
+    case UCDN_LINEBREAK_CLASS_SG:
+    case UCDN_LINEBREAK_CLASS_XX:
+        return UCDN_LINEBREAK_CLASS_AL;
+
+    case UCDN_LINEBREAK_CLASS_SA:
+        if (record->category == UCDN_GENERAL_CATEGORY_MC ||
+                record->category == UCDN_GENERAL_CATEGORY_MN)
+            return UCDN_LINEBREAK_CLASS_CM;
+        return UCDN_LINEBREAK_CLASS_AL;
+
+    case UCDN_LINEBREAK_CLASS_CJ:
+        return UCDN_LINEBREAK_CLASS_NS;
+
+    case UCDN_LINEBREAK_CLASS_CB:
+        return UCDN_LINEBREAK_CLASS_B2;
+
+    case UCDN_LINEBREAK_CLASS_NL:
+        return UCDN_LINEBREAK_CLASS_BK;
+
+    default:
+        return record->linebreak_class;
+    }
+}
+
 uint32_t ucdn_mirror(uint32_t code)
 {
     MirrorPair mp = {0};
@@ -217,6 +277,24 @@ uint32_t ucdn_mirror(uint32_t code)
         return res->to;
 }
 
+uint32_t ucdn_paired_bracket(uint32_t code)
+{
+    BracketPair *res = search_bp(code);
+    if (res == NULL)
+        return code;
+    else
+        return res->to;
+}
+
+int ucdn_paired_bracket_type(uint32_t code)
+{
+    BracketPair *res = search_bp(code);
+    if (res == NULL)
+        return UCDN_BIDI_PAIRED_BRACKET_TYPE_NONE;
+    else
+        return res->type;
+}
+
 int ucdn_decompose(uint32_t code, uint32_t *a, uint32_t *b)
 {
     const unsigned short *rec;

src/hb-ucdn/ucdn.h

@@ -17,6 +17,8 @@
 #ifndef UCDN_H
 #define UCDN_H
 
+
+
 #if defined(__GNUC__) && (__GNUC__ >= 4) && !defined(__MINGW32__)
 # define HB_BEGIN_VISIBILITY _Pragma ("GCC visibility push(hidden)")
 # define HB_END_VISIBILITY _Pragma ("GCC visibility pop")
@@ -58,6 +60,7 @@ typedef unsigned __int64 uint64_t;
 
 #endif
 
+
 #define UCDN_EAST_ASIAN_F 0
 #define UCDN_EAST_ASIAN_H 1
 #define UCDN_EAST_ASIAN_W 2
@@ -197,6 +200,53 @@ typedef unsigned __int64 uint64_t;
 #define UCDN_SCRIPT_MULTANI 129
 #define UCDN_SCRIPT_OLD_HUNGARIAN 130
 #define UCDN_SCRIPT_SIGNWRITING 131
+#define UCDN_SCRIPT_ADLAM 132
+#define UCDN_SCRIPT_BHAIKSUKI 133
+#define UCDN_SCRIPT_MARCHEN 134
+#define UCDN_SCRIPT_NEWA 135
+#define UCDN_SCRIPT_OSAGE 136
+#define UCDN_SCRIPT_TANGUT 137
+
+#define UCDN_LINEBREAK_CLASS_OP 0
+#define UCDN_LINEBREAK_CLASS_CL 1
+#define UCDN_LINEBREAK_CLASS_CP 2
+#define UCDN_LINEBREAK_CLASS_QU 3
+#define UCDN_LINEBREAK_CLASS_GL 4
+#define UCDN_LINEBREAK_CLASS_NS 5
+#define UCDN_LINEBREAK_CLASS_EX 6
+#define UCDN_LINEBREAK_CLASS_SY 7
+#define UCDN_LINEBREAK_CLASS_IS 8
+#define UCDN_LINEBREAK_CLASS_PR 9
+#define UCDN_LINEBREAK_CLASS_PO 10
+#define UCDN_LINEBREAK_CLASS_NU 11
+#define UCDN_LINEBREAK_CLASS_AL 12
+#define UCDN_LINEBREAK_CLASS_HL 13
+#define UCDN_LINEBREAK_CLASS_ID 14
+#define UCDN_LINEBREAK_CLASS_IN 15
+#define UCDN_LINEBREAK_CLASS_HY 16
+#define UCDN_LINEBREAK_CLASS_BA 17
+#define UCDN_LINEBREAK_CLASS_BB 18
+#define UCDN_LINEBREAK_CLASS_B2 19
+#define UCDN_LINEBREAK_CLASS_ZW 20
+#define UCDN_LINEBREAK_CLASS_CM 21
+#define UCDN_LINEBREAK_CLASS_WJ 22
+#define UCDN_LINEBREAK_CLASS_H2 23
+#define UCDN_LINEBREAK_CLASS_H3 24
+#define UCDN_LINEBREAK_CLASS_JL 25
+#define UCDN_LINEBREAK_CLASS_JV 26
+#define UCDN_LINEBREAK_CLASS_JT 27
+#define UCDN_LINEBREAK_CLASS_RI 28
+#define UCDN_LINEBREAK_CLASS_AI 29
+#define UCDN_LINEBREAK_CLASS_BK 30
+#define UCDN_LINEBREAK_CLASS_CB 31
+#define UCDN_LINEBREAK_CLASS_CJ 32
+#define UCDN_LINEBREAK_CLASS_CR 33
+#define UCDN_LINEBREAK_CLASS_LF 34
+#define UCDN_LINEBREAK_CLASS_NL 35
+#define UCDN_LINEBREAK_CLASS_SA 36
+#define UCDN_LINEBREAK_CLASS_SG 37
+#define UCDN_LINEBREAK_CLASS_SP 38
+#define UCDN_LINEBREAK_CLASS_XX 39
 
 #define UCDN_GENERAL_CATEGORY_CC 0
 #define UCDN_GENERAL_CATEGORY_CF 1
@@ -253,6 +303,10 @@ typedef unsigned __int64 uint64_t;
 #define UCDN_BIDI_CLASS_FSI 21
 #define UCDN_BIDI_CLASS_PDI 22
 
+#define UCDN_BIDI_PAIRED_BRACKET_TYPE_OPEN 0
+#define UCDN_BIDI_PAIRED_BRACKET_TYPE_CLOSE 1
+#define UCDN_BIDI_PAIRED_BRACKET_TYPE_NONE 2
+
 /**
  * Return version of the Unicode database.
  *
@@ -301,6 +355,27 @@ int ucdn_get_bidi_class(uint32_t code);
  */
 int ucdn_get_script(uint32_t code);
 
+/**
+ * Get unresolved linebreak class of a codepoint. This does not take
+ * rule LB1 of UAX#14 into account. See ucdn_get_resolved_linebreak_class()
+ * for resolved linebreak classes.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_LINEBREAK_* and as defined in UAX#14.
+ */
+int ucdn_get_linebreak_class(uint32_t code);
+
+/**
+ * Get resolved linebreak class of a codepoint. This resolves characters
+ * in the AI, SG, XX, SA and CJ classes according to rule LB1 of UAX#14.
+ * In addition the CB class is resolved as the equivalent B2 class and
+ * the NL class is resolved as the equivalent BK class.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_LINEBREAK_* and as defined in UAX#14.
+ */
+int ucdn_get_resolved_linebreak_class(uint32_t code);
+
 /**
  * Check if codepoint can be mirrored.
  *
@@ -318,6 +393,25 @@ int ucdn_get_mirrored(uint32_t code);
  */
 uint32_t ucdn_mirror(uint32_t code);
 
+/**
+ * Get paired bracket for a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return paired bracket codepoint or the original codepoint if no
+ * paired bracket character exists
+ */
+uint32_t ucdn_paired_bracket(uint32_t code);
+
+/**
+ * Get paired bracket type for a codepoint.
+ *
+ * @param code Unicode codepoint
+ * @return value according to UCDN_BIDI_PAIRED_BRACKET_TYPE_* and as defined
+ * in UAX#9.
+ *
+ */
+int ucdn_paired_bracket_type(uint32_t code);
+
 /**
  * Pairwise canonical decomposition of a codepoint. This includes
  * Hangul Jamo decomposition (see chapter 3.12 of the Unicode core
@@ -359,6 +453,8 @@ int ucdn_compat_decompose(uint32_t code, uint32_t *decomposed);
  */
 int ucdn_compose(uint32_t *code, uint32_t a, uint32_t b);
 
-HB_END_HEADER
+#ifdef __cplusplus
+}
+#endif
 
 #endif