[mirror_edk2.git] / MdeModulePkg / Universal / RegularExpressionDxe / Oniguruma / unicode.c

/**********************************************************************\r
  unicode.c -  Oniguruma (regular expression library)\r
**********************************************************************/\r
/*-\r
 * Copyright (c) 2002-2019  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>\r
 * All rights reserved.\r
 *\r
 * Redistribution and use in source and binary forms, with or without\r
 * modification, are permitted provided that the following conditions\r
 * are met:\r
 * 1. Redistributions of source code must retain the above copyright\r
 *    notice, this list of conditions and the following disclaimer.\r
 * 2. Redistributions in binary form must reproduce the above copyright\r
 *    notice, this list of conditions and the following disclaimer in the\r
 *    documentation and/or other materials provided with the distribution.\r
 *\r
 * THIS SOFTWARE IS PROVIDED BY THE AUTHOR AND CONTRIBUTORS ``AS IS'' AND\r
 * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE\r
 * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE\r
 * ARE DISCLAIMED.  IN NO EVENT SHALL THE AUTHOR OR CONTRIBUTORS BE LIABLE\r
 * FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL\r
 * DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS\r
 * OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)\r
 * HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT\r
 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY\r
 * OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF\r
 * SUCH DAMAGE.\r
 */\r
\r
#include "regint.h"\r
\r
struct PoolPropertyNameCtype {\r
  short int name;\r
  short int ctype;\r
};\r
\r
#define ONIGENC_IS_UNICODE_ISO_8859_1_CTYPE(code,ctype) \\r
  ((EncUNICODE_ISO_8859_1_CtypeTable[code] & CTYPE_TO_BIT(ctype)) != 0)\r
\r
static const unsigned short EncUNICODE_ISO_8859_1_CtypeTable[256] = {\r
  0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,\r
  0x4008, 0x428c, 0x4289, 0x4288, 0x4288, 0x4288, 0x4008, 0x4008,\r
  0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,\r
  0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008, 0x4008,\r
  0x4284, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,\r
  0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,\r
  0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0, 0x78b0,\r
  0x78b0, 0x78b0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x41a0,\r
  0x41a0, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x7ca2, 0x74a2,\r
  0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,\r
  0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2, 0x74a2,\r
  0x74a2, 0x74a2, 0x74a2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x51a0,\r
  0x41a0, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x78e2, 0x70e2,\r
  0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,\r
  0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2, 0x70e2,\r
  0x70e2, 0x70e2, 0x70e2, 0x41a0, 0x41a0, 0x41a0, 0x41a0, 0x4008,\r
  0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0288, 0x0008, 0x0008,\r
  0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,\r
  0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,\r
  0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008, 0x0008,\r
  0x0284, 0x01a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0, 0x00a0,\r
  0x00a0, 0x00a0, 0x30e2, 0x01a0, 0x00a0, 0x00a8, 0x00a0, 0x00a0,\r
  0x00a0, 0x00a0, 0x10a0, 0x10a0, 0x00a0, 0x30e2, 0x00a0, 0x01a0,\r
  0x00a0, 0x10a0, 0x30e2, 0x01a0, 0x10a0, 0x10a0, 0x10a0, 0x01a0,\r
  0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,\r
  0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2,\r
  0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x00a0,\r
  0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x34a2, 0x30e2,\r
  0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,\r
  0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2,\r
  0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x00a0,\r
  0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2, 0x30e2\r
};\r
\r
#include "st.h"\r
\r
#include "unicode_fold_data.c"\r
\r
extern int\r
onigenc_unicode_mbc_case_fold(OnigEncoding enc,\r
    OnigCaseFoldType flag ARG_UNUSED, const UChar** pp, const UChar* end,\r
    UChar* fold)\r
{\r
  const struct ByUnfoldKey* buk;\r
\r
  OnigCodePoint code;\r
  int i, len, rlen;\r
  const UChar *p = *pp;\r
\r
  code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
  len = enclen(enc, p);\r
  *pp += len;\r
\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  if ((flag & ONIGENC_CASE_FOLD_TURKISH_AZERI) != 0) {\r
    if (code == 0x0130) {\r
      return ONIGENC_CODE_TO_MBC(enc, 0x0069, fold);\r
    }\r
#if 0\r
    if (code == 0x0049) {\r
      return ONIGENC_CODE_TO_MBC(enc, 0x0131, fold);\r
    }\r
#endif\r
  }\r
#endif\r
\r
  buk = onigenc_unicode_unfold_key(code);\r
  if (buk != 0) {\r
    if (buk->fold_len == 1) {\r
      return ONIGENC_CODE_TO_MBC(enc, *FOLDS1_FOLD(buk->index), fold);\r
    }\r
    else {\r
      OnigCodePoint* addr;\r
\r
      FOLDS_FOLD_ADDR_BUK(buk, addr);\r
      rlen = 0;\r
      for (i = 0; i < buk->fold_len; i++) {\r
        OnigCodePoint c = addr[i];\r
        len = ONIGENC_CODE_TO_MBC(enc, c, fold);\r
        fold += len;\r
        rlen += len;\r
      }\r
      return rlen;\r
    }\r
  }\r
\r
  for (i = 0; i < len; i++) {\r
    *fold++ = *p++;\r
  }\r
  return len;\r
}\r
\r
static int\r
apply_case_fold1(int from, int to, OnigApplyAllCaseFoldFunc f, void* arg)\r
{\r
  int i, j, k, n, r;\r
\r
  for (i = from; i < to; ) {\r
    OnigCodePoint fold = *FOLDS1_FOLD(i);\r
    n = FOLDS1_UNFOLDS_NUM(i);\r
    for (j = 0; j < n; j++) {\r
      OnigCodePoint unfold = FOLDS1_UNFOLDS(i)[j];\r
\r
      r = (*f)(fold, &unfold, 1, arg);\r
      if (r != 0) return r;\r
      r = (*f)(unfold, &fold, 1, arg);\r
      if (r != 0) return r;\r
\r
      for (k = 0; k < j; k++) {\r
        OnigCodePoint unfold2 = FOLDS1_UNFOLDS(i)[k];\r
        r = (*f)(unfold, &unfold2, 1, arg);\r
        if (r != 0) return r;\r
        r = (*f)(unfold2, &unfold, 1, arg);\r
        if (r != 0) return r;\r
      }\r
    }\r
\r
    i = FOLDS1_NEXT_INDEX(i);\r
  }\r
\r
  return 0;\r
}\r
\r
static int\r
apply_case_fold2(int from, int to, OnigApplyAllCaseFoldFunc f, void* arg)\r
{\r
  int i, j, k, n, r;\r
\r
  for (i = from; i < to; ) {\r
    OnigCodePoint* fold = FOLDS2_FOLD(i);\r
    n = FOLDS2_UNFOLDS_NUM(i);\r
    for (j = 0; j < n; j++) {\r
      OnigCodePoint unfold = FOLDS2_UNFOLDS(i)[j];\r
\r
      r = (*f)(unfold, fold, 2, arg);\r
      if (r != 0) return r;\r
\r
      for (k = 0; k < j; k++) {\r
        OnigCodePoint unfold2 = FOLDS2_UNFOLDS(i)[k];\r
        r = (*f)(unfold, &unfold2, 1, arg);\r
        if (r != 0) return r;\r
        r = (*f)(unfold2, &unfold, 1, arg);\r
        if (r != 0) return r;\r
      }\r
    }\r
\r
    i = FOLDS2_NEXT_INDEX(i);\r
  }\r
\r
  return 0;\r
}\r
\r
static int\r
apply_case_fold3(int from, int to, OnigApplyAllCaseFoldFunc f, void* arg)\r
{\r
  int i, j, k, n, r;\r
\r
  for (i = from; i < to; ) {\r
    OnigCodePoint* fold = FOLDS3_FOLD(i);\r
    n = FOLDS3_UNFOLDS_NUM(i);\r
    for (j = 0; j < n; j++) {\r
      OnigCodePoint unfold = FOLDS3_UNFOLDS(i)[j];\r
\r
      r = (*f)(unfold, fold, 3, arg);\r
      if (r != 0) return r;\r
\r
      for (k = 0; k < j; k++) {\r
        OnigCodePoint unfold2 = FOLDS3_UNFOLDS(i)[k];\r
        r = (*f)(unfold, &unfold2, 1, arg);\r
        if (r != 0) return r;\r
        r = (*f)(unfold2, &unfold, 1, arg);\r
        if (r != 0) return r;\r
      }\r
    }\r
\r
    i = FOLDS3_NEXT_INDEX(i);\r
  }\r
\r
  return 0;\r
}\r
\r
extern int\r
onigenc_unicode_apply_all_case_fold(OnigCaseFoldType flag,\r
                                    OnigApplyAllCaseFoldFunc f, void* arg)\r
{\r
  int r;\r
\r
  r = apply_case_fold1(0, FOLDS1_NORMAL_END_INDEX, f, arg);\r
  if (r != 0) return r;\r
\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  if ((flag & ONIGENC_CASE_FOLD_TURKISH_AZERI) != 0) {\r
    code = 0x0131;\r
    r = (*f)(0x0049, &code, 1, arg);\r
    if (r != 0) return r;\r
    code = 0x0049;\r
    r = (*f)(0x0131, &code, 1, arg);\r
    if (r != 0) return r;\r
\r
    code = 0x0130;\r
    r = (*f)(0x0069, &code, 1, arg);\r
    if (r != 0) return r;\r
    code = 0x0069;\r
    r = (*f)(0x0130, &code, 1, arg);\r
    if (r != 0) return r;\r
  }\r
  else {\r
#endif\r
    r = apply_case_fold1(FOLDS1_NORMAL_END_INDEX, FOLDS1_END_INDEX, f, arg);\r
    if (r != 0) return r;\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  }\r
#endif\r
\r
  if ((flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) == 0)\r
    return 0;\r
\r
  r = apply_case_fold2(0, FOLDS2_NORMAL_END_INDEX, f, arg);\r
  if (r != 0) return r;\r
\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  if ((flag & ONIGENC_CASE_FOLD_TURKISH_AZERI) == 0) {\r
#endif\r
    r = apply_case_fold2(FOLDS2_NORMAL_END_INDEX, FOLDS2_END_INDEX, f, arg);\r
    if (r != 0) return r;\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  }\r
#endif\r
\r
  r = apply_case_fold3(0, FOLDS3_NORMAL_END_INDEX, f, arg);\r
  if (r != 0) return r;\r
\r
  return 0;\r
}\r
\r
extern int\r
onigenc_unicode_get_case_fold_codes_by_str(OnigEncoding enc,\r
    OnigCaseFoldType flag, const OnigUChar* p, const OnigUChar* end,\r
    OnigCaseFoldCodeItem items[])\r
{\r
  int n, m, i, j, k, len;\r
  OnigCodePoint code, codes[3];\r
  const struct ByUnfoldKey* buk;\r
\r
  n = 0;\r
\r
  code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
  len = enclen(enc, p);\r
\r
#ifdef USE_UNICODE_CASE_FOLD_TURKISH_AZERI\r
  if ((flag & ONIGENC_CASE_FOLD_TURKISH_AZERI) != 0) {\r
    if (code == 0x0049) {\r
      items[0].byte_len = len;\r
      items[0].code_len = 1;\r
      items[0].code[0]  = 0x0131;\r
      return 1;\r
    }\r
    else if (code == 0x0130) {\r
      items[0].byte_len = len;\r
      items[0].code_len = 1;\r
      items[0].code[0]  = 0x0069;\r
      return 1;\r
    }\r
    else if (code == 0x0131) {\r
      items[0].byte_len = len;\r
      items[0].code_len = 1;\r
      items[0].code[0]  = 0x0049;\r
      return 1;\r
    }\r
    else if (code == 0x0069) {\r
      items[0].byte_len = len;\r
      items[0].code_len = 1;\r
      items[0].code[0]  = 0x0130;\r
      return 1;\r
    }\r
  }\r
#endif\r
\r
  buk = onigenc_unicode_unfold_key(code);\r
  if (buk != 0) {\r
    if (buk->fold_len == 1) {\r
      int un;\r
      items[0].byte_len = len;\r
      items[0].code_len = 1;\r
      items[0].code[0]  = *FOLDS1_FOLD(buk->index);\r
      n++;\r
\r
      un = FOLDS1_UNFOLDS_NUM(buk->index);\r
      for (i = 0; i < un; i++) {\r
        OnigCodePoint unfold = FOLDS1_UNFOLDS(buk->index)[i];\r
        if (unfold != code) {\r
          items[n].byte_len = len;\r
          items[n].code_len = 1;\r
          items[n].code[0]  = unfold;\r
          n++;\r
        }\r
      }\r
      code = items[0].code[0]; /* for multi-code to unfold search. */\r
    }\r
    else if ((flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) != 0) {\r
      OnigCodePoint cs[3][4];\r
      int fn, ncs[3];\r
\r
      if (buk->fold_len == 2) {\r
        m = FOLDS2_UNFOLDS_NUM(buk->index);\r
        for (i = 0; i < m; i++) {\r
          OnigCodePoint unfold = FOLDS2_UNFOLDS(buk->index)[i];\r
          if (unfold == code) continue;\r
\r
          items[n].byte_len = len;\r
          items[n].code_len = 1;\r
          items[n].code[0]  = unfold;\r
          n++;\r
        }\r
\r
        for (fn = 0; fn < 2; fn++) {\r
          int index;\r
          cs[fn][0] = FOLDS2_FOLD(buk->index)[fn];\r
          index = onigenc_unicode_fold1_key(&cs[fn][0]);\r
          if (index >= 0) {\r
            int m = FOLDS1_UNFOLDS_NUM(index);\r
            for (i = 0; i < m; i++) {\r
              cs[fn][i+1] = FOLDS1_UNFOLDS(index)[i];\r
            }\r
            ncs[fn] = m + 1;\r
          }\r
          else\r
            ncs[fn] = 1;\r
        }\r
\r
        for (i = 0; i < ncs[0]; i++) {\r
          for (j = 0; j < ncs[1]; j++) {\r
            items[n].byte_len = len;\r
            items[n].code_len = 2;\r
            items[n].code[0]  = cs[0][i];\r
            items[n].code[1]  = cs[1][j];\r
            n++;\r
          }\r
        }\r
      }\r
      else { /* fold_len == 3 */\r
        m = FOLDS3_UNFOLDS_NUM(buk->index);\r
        for (i = 0; i < m; i++) {\r
          OnigCodePoint unfold = FOLDS3_UNFOLDS(buk->index)[i];\r
          if (unfold == code) continue;\r
\r
          items[n].byte_len = len;\r
          items[n].code_len = 1;\r
          items[n].code[0]  = unfold;\r
          n++;\r
        }\r
\r
        for (fn = 0; fn < 3; fn++) {\r
          int index;\r
          cs[fn][0] = FOLDS3_FOLD(buk->index)[fn];\r
          index = onigenc_unicode_fold1_key(&cs[fn][0]);\r
          if (index >= 0) {\r
            int m = FOLDS1_UNFOLDS_NUM(index);\r
            for (i = 0; i < m; i++) {\r
              cs[fn][i+1] = FOLDS1_UNFOLDS(index)[i];\r
            }\r
            ncs[fn] = m + 1;\r
          }\r
          else\r
            ncs[fn] = 1;\r
        }\r
\r
        for (i = 0; i < ncs[0]; i++) {\r
          for (j = 0; j < ncs[1]; j++) {\r
            for (k = 0; k < ncs[2]; k++) {\r
              items[n].byte_len = len;\r
              items[n].code_len = 3;\r
              items[n].code[0]  = cs[0][i];\r
              items[n].code[1]  = cs[1][j];\r
              items[n].code[2]  = cs[2][k];\r
              n++;\r
            }\r
          }\r
        }\r
      }\r
\r
      /* multi char folded code is not head of another folded multi char */\r
      return n;\r
    }\r
  }\r
  else {\r
    int index = onigenc_unicode_fold1_key(&code);\r
    if (index >= 0) {\r
      int m = FOLDS1_UNFOLDS_NUM(index);\r
      for (i = 0; i < m; i++) {\r
        items[n].byte_len = len;\r
        items[n].code_len = 1;\r
        items[n].code[0]  = FOLDS1_UNFOLDS(index)[i];\r
        n++;\r
      }\r
    }\r
  }\r
\r
  if ((flag & INTERNAL_ONIGENC_CASE_FOLD_MULTI_CHAR) == 0)\r
    return n;\r
\r
  p += len;\r
  if (p < end) {\r
    int clen;\r
    int index;\r
\r
    codes[0] = code;\r
    code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
\r
    buk = onigenc_unicode_unfold_key(code);\r
    if (buk != 0 && buk->fold_len == 1) {\r
      codes[1] = *FOLDS1_FOLD(buk->index);\r
    }\r
    else\r
      codes[1] = code;\r
\r
    clen = enclen(enc, p);\r
    len += clen;\r
\r
    index = onigenc_unicode_fold2_key(codes);\r
    if (index >= 0) {\r
      m = FOLDS2_UNFOLDS_NUM(index);\r
      for (i = 0; i < m; i++) {\r
        items[n].byte_len = len;\r
        items[n].code_len = 1;\r
        items[n].code[0]  = FOLDS2_UNFOLDS(index)[i];\r
        n++;\r
      }\r
    }\r
\r
    p += clen;\r
    if (p < end) {\r
      code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
      buk = onigenc_unicode_unfold_key(code);\r
      if (buk != 0 && buk->fold_len == 1) {\r
        codes[2] = *FOLDS1_FOLD(buk->index);\r
      }\r
      else\r
        codes[2] = code;\r
\r
      clen = enclen(enc, p);\r
      len += clen;\r
\r
      index = onigenc_unicode_fold3_key(codes);\r
      if (index >= 0) {\r
        m = FOLDS3_UNFOLDS_NUM(index);\r
        for (i = 0; i < m; i++) {\r
          items[n].byte_len = len;\r
          items[n].code_len = 1;\r
          items[n].code[0]  = FOLDS3_UNFOLDS(index)[i];\r
          n++;\r
        }\r
      }\r
    }\r
  }\r
\r
  return n;\r
}\r
\r
#ifdef USE_UNICODE_PROPERTIES\r
#include "unicode_property_data.c"\r
#else\r
#include "unicode_property_data_posix.c"\r
#endif\r
\r
\r
#ifdef USE_UNICODE_WORD_BREAK\r
\r
enum WB_TYPE {\r
  WB_Any = 0,\r
  WB_ALetter,\r
  WB_CR,\r
  WB_Double_Quote,\r
  WB_Extend,\r
  WB_ExtendNumLet,\r
  WB_Format,\r
  WB_Hebrew_Letter,\r
  WB_Katakana,\r
  WB_LF,\r
  WB_MidLetter,\r
  WB_MidNum,\r
  WB_MidNumLet,\r
  WB_Newline,\r
  WB_Numeric,\r
  WB_Regional_Indicator,\r
  WB_Single_Quote,\r
  WB_WSegSpace,\r
  WB_ZWJ,\r
};\r
\r
typedef struct {\r
  OnigCodePoint start;\r
  OnigCodePoint end;\r
  enum WB_TYPE  type;\r
} WB_RANGE_TYPE;\r
\r
#include "unicode_wb_data.c"\r
\r
static enum WB_TYPE\r
wb_get_type(OnigCodePoint code)\r
{\r
  OnigCodePoint low, high, x;\r
  enum WB_TYPE type;\r
\r
  for (low = 0, high = (OnigCodePoint )WB_RANGE_NUM; low < high; ) {\r
    x = (low + high) >> 1;\r
    if (code > WB_RANGES[x].end)\r
      low = x + 1;\r
    else\r
      high = x;\r
  }\r
\r
  type = (low < (OnigCodePoint )WB_RANGE_NUM &&\r
          code >= WB_RANGES[low].start) ?\r
    WB_RANGES[low].type : WB_Any;\r
\r
  return type;\r
}\r
\r
#define IS_WB_IGNORE_TAIL(t)  ((t) == WB_Extend || (t) == WB_Format || (t) == WB_ZWJ)\r
#define IS_WB_AHLetter(t)     ((t) == WB_ALetter || (t) == WB_Hebrew_Letter)\r
#define IS_WB_MidNumLetQ(t)   ((t) == WB_MidNumLet || (t) == WB_Single_Quote)\r
\r
static int\r
wb_get_next_main_code(OnigEncoding enc, UChar* p, const UChar* end,\r
                      OnigCodePoint* rcode, enum WB_TYPE* rtype)\r
{\r
  OnigCodePoint code;\r
  enum WB_TYPE type;\r
\r
  while (TRUE) {\r
    p += enclen(enc, p);\r
    if (p >= end) break;\r
\r
    code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
    type = wb_get_type(code);\r
    if (! IS_WB_IGNORE_TAIL(type)) {\r
      *rcode = code;\r
      *rtype = type;\r
      return 1;\r
    }\r
  }\r
\r
  return 0;\r
}\r
\r
extern int\r
onigenc_wb_is_break_position(OnigEncoding enc, UChar* p, UChar* prev,\r
                             const UChar* start, const UChar* end)\r
{\r
  int r;\r
  UChar* pp;\r
  OnigCodePoint cfrom;\r
  OnigCodePoint cfrom2;\r
  OnigCodePoint cto;\r
  OnigCodePoint cto2;\r
  enum WB_TYPE from;\r
  enum WB_TYPE from2;\r
  enum WB_TYPE to;\r
  enum WB_TYPE to2;\r
\r
  /* WB1: sot / Any */\r
  if (p == start) return TRUE;\r
  /* WB2: Any / eot */\r
  if (p == end)   return TRUE;\r
\r
  if (IS_NULL(prev)) {\r
    prev = onigenc_get_prev_char_head(enc, start, p);\r
    if (IS_NULL(prev)) return TRUE;\r
  }\r
\r
  cfrom = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
  cto   = ONIGENC_MBC_TO_CODE(enc, p, end);\r
\r
  from = wb_get_type(cfrom);\r
  to   = wb_get_type(cto);\r
\r
  /* short cut */\r
  if (from == 0 && to == 0) goto WB999;\r
\r
  /* WB3: CR + LF */\r
  if (from == WB_CR && to == WB_LF) return FALSE;\r
\r
  /* WB3a: (Newline|CR|LF) /  */\r
  if (from == WB_Newline || from == WB_CR || from == WB_LF) return TRUE;\r
  /* WB3b: / (Newline|CR|LF) */\r
  if (to == WB_Newline || to == WB_CR || to == WB_LF) return TRUE;\r
\r
  /* WB3c: ZWJ + {Extended_Pictographic} */\r
  if (from == WB_ZWJ) {\r
    if (onigenc_unicode_is_code_ctype(cto, PROP_INDEX_EXTENDEDPICTOGRAPHIC))\r
      return FALSE;\r
  }\r
\r
  /* WB3d: WSegSpace + WSegSpace */\r
  if (from == WB_WSegSpace && to == WB_WSegSpace) return FALSE;\r
\r
  /* WB4:  X (Extend|Format|ZWJ)* -> X */\r
  if (IS_WB_IGNORE_TAIL(to)) return FALSE;\r
  if (IS_WB_IGNORE_TAIL(from)) {\r
    while ((pp = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
      prev = pp;\r
      cfrom = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
      from = wb_get_type(cfrom);\r
      if (! IS_WB_IGNORE_TAIL(from))\r
        break;\r
    }\r
  }\r
\r
  if (IS_WB_AHLetter(from)) {\r
    /* WB5: AHLetter + AHLetter */\r
    if (IS_WB_AHLetter(to)) return FALSE;\r
\r
    /* WB6: AHLetter + (MidLetter | MidNumLetQ) AHLetter */\r
    if (to == WB_MidLetter || IS_WB_MidNumLetQ(to)) {\r
      r = wb_get_next_main_code(enc, p, end, &cto2, &to2);\r
      if (r == 1) {\r
        if (IS_WB_AHLetter(to2)) return FALSE;\r
      }\r
    }\r
  }\r
\r
  /* WB7: AHLetter (MidLetter | MidNumLetQ) + AHLetter */\r
  if (from == WB_MidLetter || IS_WB_MidNumLetQ(from)) {\r
    if (IS_WB_AHLetter(to)) {\r
      from2 = WB_Any;\r
      while ((pp = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
        prev = pp;\r
        cfrom2 = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
        from2 = wb_get_type(cfrom2);\r
        if (! IS_WB_IGNORE_TAIL(from2))\r
          break;\r
      }\r
\r
      if (IS_WB_AHLetter(from2)) return FALSE;\r
    }\r
  }\r
\r
  if (from == WB_Hebrew_Letter) {\r
    /* WB7a: Hebrew_Letter + Single_Quote */\r
    if (to == WB_Single_Quote) return FALSE;\r
\r
    /* WB7b: Hebrew_Letter + Double_Quote Hebrew_Letter */\r
    if (to == WB_Double_Quote) {\r
      r = wb_get_next_main_code(enc, p, end, &cto2, &to2);\r
      if (r == 1) {\r
        if (to2 == WB_Hebrew_Letter) return FALSE;\r
      }\r
    }\r
  }\r
\r
  /* WB7c: Hebrew_Letter Double_Quote + Hebrew_Letter */\r
  if (from == WB_Double_Quote) {\r
    if (to == WB_Hebrew_Letter) {\r
      from2 = WB_Any;\r
      while ((pp = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
        prev = pp;\r
        cfrom2 = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
        from2 = wb_get_type(cfrom2);\r
        if (! IS_WB_IGNORE_TAIL(from2))\r
          break;\r
      }\r
\r
      if (from2 == WB_Hebrew_Letter) return FALSE;\r
    }\r
  }\r
\r
  if (to == WB_Numeric) {\r
    /* WB8: Numeric + Numeric */\r
    if (from == WB_Numeric) return FALSE;\r
\r
    /* WB9: AHLetter + Numeric */\r
    if (IS_WB_AHLetter(from)) return FALSE;\r
\r
    /* WB11: Numeric (MidNum | MidNumLetQ) + Numeric */\r
    if (from == WB_MidNum || IS_WB_MidNumLetQ(from)) {\r
      from2 = WB_Any;\r
      while ((pp = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
        prev = pp;\r
        cfrom2 = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
        from2 = wb_get_type(cfrom2);\r
        if (! IS_WB_IGNORE_TAIL(from2))\r
          break;\r
      }\r
\r
      if (from2 == WB_Numeric) return FALSE;\r
    }\r
  }\r
\r
  if (from == WB_Numeric) {\r
    /* WB10: Numeric + AHLetter */\r
    if (IS_WB_AHLetter(to)) return FALSE;\r
\r
    /* WB12: Numeric + (MidNum | MidNumLetQ) Numeric */\r
    if (to == WB_MidNum || IS_WB_MidNumLetQ(to)) {\r
      r = wb_get_next_main_code(enc, p, end, &cto2, &to2);\r
      if (r == 1) {\r
        if (to2 == WB_Numeric) return FALSE;\r
      }\r
    }\r
  }\r
\r
  /* WB13: Katakana + Katakana */\r
  if (from == WB_Katakana && to == WB_Katakana) return FALSE;\r
\r
  /* WB13a: (AHLetter | Numeric | Katakana | ExtendNumLet) + ExtendNumLet */\r
  if (IS_WB_AHLetter(from) || from == WB_Numeric || from == WB_Katakana\r
      || from == WB_ExtendNumLet) {\r
    if (to == WB_ExtendNumLet) return FALSE;\r
  }\r
\r
  /* WB13b: ExtendNumLet + (AHLetter | Numeric | Katakana) */\r
  if (from == WB_ExtendNumLet) {\r
    if (IS_WB_AHLetter(to) || to == WB_Numeric || to == WB_Katakana)\r
      return FALSE;\r
  }\r
\r
\r
  /* WB15:   sot (RI RI)* RI + RI */\r
  /* WB16: [^RI] (RI RI)* RI + RI */\r
  if (from == WB_Regional_Indicator && to == WB_Regional_Indicator) {\r
    int n = 0;\r
    while ((prev = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
      cfrom2 = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
      from2  = wb_get_type(cfrom2);\r
      if (from2 != WB_Regional_Indicator)\r
        break;\r
\r
      n++;\r
    }\r
    if ((n % 2) == 0) return FALSE;\r
  }\r
\r
 WB999:\r
  /* WB999: Any / Any */\r
  return TRUE;\r
}\r
\r
#endif /* USE_UNICODE_WORD_BREAK */\r
\r
\r
#ifdef USE_UNICODE_EXTENDED_GRAPHEME_CLUSTER\r
\r
enum EGCB_BREAK_TYPE {\r
  EGCB_NOT_BREAK = 0,\r
  EGCB_BREAK     = 1,\r
  EGCB_BREAK_UNDEF_GB11  = 2,\r
  EGCB_BREAK_UNDEF_RI_RI = 3\r
};\r
\r
enum EGCB_TYPE {\r
  EGCB_Other   = 0,\r
  EGCB_CR      = 1,\r
  EGCB_LF      = 2,\r
  EGCB_Control = 3,\r
  EGCB_Extend  = 4,\r
  EGCB_Prepend = 5,\r
  EGCB_Regional_Indicator = 6,\r
  EGCB_SpacingMark = 7,\r
  EGCB_ZWJ         = 8,\r
#if 0\r
  /* obsoleted */\r
  EGCB_E_Base         = 9,\r
  EGCB_E_Base_GAZ     = 10,\r
  EGCB_E_Modifier     = 11,\r
  EGCB_Glue_After_Zwj = 12,\r
#endif\r
  EGCB_L   = 13,\r
  EGCB_LV  = 14,\r
  EGCB_LVT = 15,\r
  EGCB_T   = 16,\r
  EGCB_V   = 17\r
};\r
\r
typedef struct {\r
  OnigCodePoint  start;\r
  OnigCodePoint  end;\r
  enum EGCB_TYPE type;\r
} EGCB_RANGE_TYPE;\r
\r
#include "unicode_egcb_data.c"\r
\r
static enum EGCB_TYPE\r
egcb_get_type(OnigCodePoint code)\r
{\r
  OnigCodePoint low, high, x;\r
  enum EGCB_TYPE type;\r
\r
  for (low = 0, high = (OnigCodePoint )EGCB_RANGE_NUM; low < high; ) {\r
    x = (low + high) >> 1;\r
    if (code > EGCB_RANGES[x].end)\r
      low = x + 1;\r
    else\r
      high = x;\r
  }\r
\r
  type = (low < (OnigCodePoint )EGCB_RANGE_NUM &&\r
          code >= EGCB_RANGES[low].start) ?\r
    EGCB_RANGES[low].type : EGCB_Other;\r
\r
  return type;\r
}\r
\r
#define IS_CONTROL_CR_LF(code)   ((code) <= EGCB_Control && (code) >= EGCB_CR)\r
#define IS_HANGUL(code)          ((code) >= EGCB_L)\r
\r
/* GB1 and GB2 are outside of this function. */\r
static enum EGCB_BREAK_TYPE\r
unicode_egcb_is_break_2code(OnigCodePoint from_code, OnigCodePoint to_code)\r
{\r
  enum EGCB_TYPE from;\r
  enum EGCB_TYPE to;\r
\r
  from = egcb_get_type(from_code);\r
  to   = egcb_get_type(to_code);\r
\r
  /* short cut */\r
  if (from == 0 && to == 0) goto GB999;\r
\r
  /* GB3 */\r
  if (from == EGCB_CR && to == EGCB_LF) return EGCB_NOT_BREAK;\r
  /* GB4 */\r
  if (IS_CONTROL_CR_LF(from)) return EGCB_BREAK;\r
  /* GB5 */\r
  if (IS_CONTROL_CR_LF(to)) return EGCB_BREAK;\r
\r
  if (IS_HANGUL(from) && IS_HANGUL(to)) {\r
    /* GB6 */\r
    if (from == EGCB_L && to != EGCB_T) return EGCB_NOT_BREAK;\r
    /* GB7 */\r
    if ((from == EGCB_LV || from == EGCB_V)\r
        && (to == EGCB_V || to == EGCB_T)) return EGCB_NOT_BREAK;\r
\r
    /* GB8 */\r
    if ((to == EGCB_T) && (from == EGCB_LVT || from == EGCB_T))\r
      return EGCB_NOT_BREAK;\r
\r
    goto GB999;\r
  }\r
\r
  /* GB9 */\r
  if (to == EGCB_Extend || to == EGCB_ZWJ) return EGCB_NOT_BREAK;\r
\r
  /* GB9a */\r
  if (to == EGCB_SpacingMark) return EGCB_NOT_BREAK;\r
  /* GB9b */\r
  if (from == EGCB_Prepend) return EGCB_NOT_BREAK;\r
\r
  /* GB10 removed */\r
\r
  /* GB11 */\r
  if (from == EGCB_ZWJ) {\r
    if (onigenc_unicode_is_code_ctype(to_code, PROP_INDEX_EXTENDEDPICTOGRAPHIC))\r
      return EGCB_BREAK_UNDEF_GB11;\r
\r
    goto GB999;\r
  }\r
\r
  /* GB12, GB13 */\r
  if (from == EGCB_Regional_Indicator && to == EGCB_Regional_Indicator) {\r
    return EGCB_BREAK_UNDEF_RI_RI;\r
  }\r
\r
 GB999:\r
  return EGCB_BREAK;\r
}\r
\r
#endif /* USE_UNICODE_EXTENDED_GRAPHEME_CLUSTER */\r
\r
extern int\r
onigenc_egcb_is_break_position(OnigEncoding enc, UChar* p, UChar* prev,\r
                               const UChar* start, const UChar* end)\r
{\r
  OnigCodePoint from;\r
  OnigCodePoint to;\r
#ifdef USE_UNICODE_EXTENDED_GRAPHEME_CLUSTER\r
  enum EGCB_BREAK_TYPE btype;\r
  enum EGCB_TYPE type;\r
#endif\r
\r
  /* GB1 and GB2 */\r
  if (p == start) return 1;\r
  if (p == end)   return 1;\r
\r
  if (IS_NULL(prev)) {\r
    prev = onigenc_get_prev_char_head(enc, start, p);\r
    if (IS_NULL(prev)) return 1;\r
  }\r
\r
  from = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
  to   = ONIGENC_MBC_TO_CODE(enc, p, end);\r
\r
#ifdef USE_UNICODE_EXTENDED_GRAPHEME_CLUSTER\r
  if (! ONIGENC_IS_UNICODE_ENCODING(enc)) {\r
    return from != 0x000d || to != 0x000a;\r
  }\r
\r
  btype = unicode_egcb_is_break_2code(from, to);\r
  switch (btype) {\r
  case EGCB_NOT_BREAK:\r
    return 0;\r
    break;\r
  case EGCB_BREAK:\r
    return 1;\r
    break;\r
\r
  case EGCB_BREAK_UNDEF_GB11:\r
    while ((prev = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
      from = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
      if (onigenc_unicode_is_code_ctype(from, PROP_INDEX_EXTENDEDPICTOGRAPHIC))\r
        return 0;\r
\r
      type = egcb_get_type(from);\r
      if (type != EGCB_Extend)\r
        break;\r
    }\r
    break;\r
\r
  case EGCB_BREAK_UNDEF_RI_RI:\r
    {\r
      int n = 0;\r
      while ((prev = onigenc_get_prev_char_head(enc, start, prev)) != NULL) {\r
        from = ONIGENC_MBC_TO_CODE(enc, prev, end);\r
        type = egcb_get_type(from);\r
        if (type != EGCB_Regional_Indicator)\r
          break;\r
\r
        n++;\r
      }\r
      if ((n % 2) == 0) return 0;\r
    }\r
    break;\r
  }\r
\r
  return 1;\r
\r
#else\r
  return from != 0x000d || to != 0x000a;\r
#endif /* USE_UNICODE_EXTENDED_GRAPHEME_CLUSTER */\r
}\r
\r
\r
#define USER_DEFINED_PROPERTY_MAX_NUM  20\r
\r
typedef struct {\r
  int ctype;\r
  OnigCodePoint* ranges;\r
} UserDefinedPropertyValue;\r
\r
static int UserDefinedPropertyNum;\r
static UserDefinedPropertyValue\r
UserDefinedPropertyRanges[USER_DEFINED_PROPERTY_MAX_NUM];\r
static st_table* UserDefinedPropertyTable;\r
\r
extern int\r
onig_unicode_define_user_property(const char* name, OnigCodePoint* ranges)\r
{\r
  UserDefinedPropertyValue* e;\r
  int r;\r
  int i;\r
  int n;\r
  int len;\r
  int c;\r
  char* s;\r
  UChar* uname;\r
\r
  if (UserDefinedPropertyNum >= USER_DEFINED_PROPERTY_MAX_NUM)\r
    return ONIGERR_TOO_MANY_USER_DEFINED_OBJECTS;\r
\r
  len = (int )strlen_s(name,MAX_STRING_SIZE);\r
  if (len >= PROPERTY_NAME_MAX_SIZE)\r
    return ONIGERR_TOO_LONG_PROPERTY_NAME;\r
\r
  s = (char* )xmalloc(len + 1);\r
  if (s == 0)\r
    return ONIGERR_MEMORY;\r
\r
  uname = (UChar* )name;\r
  n = 0;\r
  for (i = 0; i < len; i++) {\r
    c = uname[i];\r
    if (c < 0x20 || c >= 0x80) {\r
      xfree(s);\r
      return ONIGERR_INVALID_CHAR_PROPERTY_NAME;\r
    }\r
\r
    if (c != ' ' && c != '-' && c != '_') {\r
      s[n] = c;\r
      n++;\r
    }\r
  }\r
  s[n] = '\0';\r
\r
  if (UserDefinedPropertyTable == 0) {\r
    UserDefinedPropertyTable = onig_st_init_strend_table_with_size(10);\r
    if (IS_NULL(UserDefinedPropertyTable)) {\r
      xfree(s);\r
      return ONIGERR_MEMORY;\r
    }\r
  }\r
\r
  e = UserDefinedPropertyRanges + UserDefinedPropertyNum;\r
  e->ctype = CODE_RANGES_NUM + UserDefinedPropertyNum;\r
  e->ranges = ranges;\r
  r = onig_st_insert_strend(UserDefinedPropertyTable,\r
                            (const UChar* )s, (const UChar* )s + n,\r
                            (hash_data_type )((void* )e));\r
  if (r < 0) return r;\r
\r
  UserDefinedPropertyNum++;\r
  return 0;\r
}\r
\r
extern int\r
onigenc_unicode_is_code_ctype(OnigCodePoint code, unsigned int ctype)\r
{\r
  if (\r
#ifdef USE_UNICODE_PROPERTIES\r
      ctype <= ONIGENC_MAX_STD_CTYPE &&\r
#endif\r
      code < 256) {\r
    return ONIGENC_IS_UNICODE_ISO_8859_1_CTYPE(code, ctype);\r
  }\r
\r
  if (ctype >= CODE_RANGES_NUM) {\r
    int index = ctype - CODE_RANGES_NUM;\r
    if (index < UserDefinedPropertyNum)\r
      return onig_is_in_code_range((UChar* )UserDefinedPropertyRanges[index].ranges, code);\r
    else\r
      return ONIGERR_TYPE_BUG;\r
  }\r
\r
  return onig_is_in_code_range((UChar* )CodeRanges[ctype], code);\r
}\r
\r
\r
extern int\r
onigenc_unicode_ctype_code_range(OnigCtype ctype, const OnigCodePoint* ranges[])\r
{\r
  if (ctype >= CODE_RANGES_NUM) {\r
    int index = ctype - CODE_RANGES_NUM;\r
    if (index < UserDefinedPropertyNum) {\r
      *ranges = UserDefinedPropertyRanges[index].ranges;\r
      return 0;\r
    }\r
    else\r
      return ONIGERR_TYPE_BUG;\r
  }\r
\r
  *ranges = CodeRanges[ctype];\r
  return 0;\r
}\r
\r
extern int\r
onigenc_utf16_32_get_ctype_code_range(OnigCtype ctype, OnigCodePoint* sb_out,\r
                                      const OnigCodePoint* ranges[])\r
{\r
  *sb_out = 0x00;\r
  return onigenc_unicode_ctype_code_range(ctype, ranges);\r
}\r
\r
extern int\r
onigenc_unicode_property_name_to_ctype(OnigEncoding enc, UChar* name, UChar* end)\r
{\r
  int len;\r
  UChar *p;\r
  OnigCodePoint code;\r
  const struct PoolPropertyNameCtype* pc;\r
  char buf[PROPERTY_NAME_MAX_SIZE];\r
\r
  p = name;\r
  len = 0;\r
  while (p < end) {\r
    code = ONIGENC_MBC_TO_CODE(enc, p, end);\r
    if (code >= 0x80)\r
      return ONIGERR_INVALID_CHAR_PROPERTY_NAME;\r
\r
    if (code != ' ' && code != '-' && code != '_') {\r
      buf[len++] = (char )code;\r
      if (len >= PROPERTY_NAME_MAX_SIZE)\r
        return ONIGERR_INVALID_CHAR_PROPERTY_NAME;\r
    }\r
\r
    p += enclen(enc, p);\r
  }\r
\r
  buf[len] = 0;\r
\r
  if (UserDefinedPropertyTable != 0) {\r
    UserDefinedPropertyValue* e;\r
    e = (UserDefinedPropertyValue* )NULL;\r
    onig_st_lookup_strend(UserDefinedPropertyTable,\r
                          (const UChar* )buf, (const UChar* )buf + len,\r
                          (hash_data_type* )((void* )(&e)));\r
    if (e != 0) {\r
      return e->ctype;\r
    }\r
  }\r
\r
  pc = unicode_lookup_property_name(buf, len);\r
  if (pc != 0) {\r
    /* fprintf(stderr, "LOOKUP: %s: %d\n", buf, pc->ctype); */\r
#ifndef USE_UNICODE_PROPERTIES\r
    if (pc->ctype > ONIGENC_MAX_STD_CTYPE)\r
      return ONIGERR_INVALID_CHAR_PROPERTY_NAME;\r
#endif\r
\r
    return (int )pc->ctype;\r
  }\r
\r
  return ONIGERR_INVALID_CHAR_PROPERTY_NAME;\r
}\r