remove initialize functions from EUC_JP and Shift_JIS by using gperf

diff --git a/src/Makefile.am b/src/Makefile.am
index 68eb4b620e6c5e68aad88d735ffc8194e522a906..2154cfc115e39176cce9f19a41f14ba4aabd9c6c 100644 (file)
--- a/src/Makefile.am
+++ b/src/Makefile.am
@@ -14,7 +14,9 @@ libonig_la_SOURCES = regint.h regparse.h regenc.h st.h \
        unicode.c ascii.c utf8.c \
        utf16_be.c utf16_le.c \
        utf32_be.c utf32_le.c \
-       euc_jp.c sjis.c iso8859_1.c \
+       euc_jp.c euc_jp_prop.c \
+       sjis.c sjis_prop.c \
+       iso8859_1.c \
        iso8859_2.c  iso8859_3.c \
        iso8859_4.c  iso8859_5.c \
        iso8859_6.c  iso8859_7.c \

diff --git a/src/Makefile.in b/src/Makefile.in
index 86d2c6f36edbaa356ed97e227a089dbf56da980e..1a2cffc2efdae728a900f7cc7161e496f9066e76 100644 (file)
--- a/src/Makefile.in
+++ b/src/Makefile.in
@@ -127,12 +127,12 @@ am_libonig_la_OBJECTS = regerror.lo regparse.lo regext.lo regcomp.lo \
        regexec.lo reggnu.lo regenc.lo regsyntax.lo regtrav.lo \
        regversion.lo st.lo regposix.lo regposerr.lo unicode.lo \
        ascii.lo utf8.lo utf16_be.lo utf16_le.lo utf32_be.lo \
-       utf32_le.lo euc_jp.lo sjis.lo iso8859_1.lo iso8859_2.lo \
-       iso8859_3.lo iso8859_4.lo iso8859_5.lo iso8859_6.lo \
-       iso8859_7.lo iso8859_8.lo iso8859_9.lo iso8859_10.lo \
-       iso8859_11.lo iso8859_13.lo iso8859_14.lo iso8859_15.lo \
-       iso8859_16.lo euc_tw.lo euc_kr.lo big5.lo gb18030.lo koi8_r.lo \
-       cp1251.lo onig_init.lo
+       utf32_le.lo euc_jp.lo euc_jp_prop.lo sjis.lo sjis_prop.lo \
+       iso8859_1.lo iso8859_2.lo iso8859_3.lo iso8859_4.lo \
+       iso8859_5.lo iso8859_6.lo iso8859_7.lo iso8859_8.lo \
+       iso8859_9.lo iso8859_10.lo iso8859_11.lo iso8859_13.lo \
+       iso8859_14.lo iso8859_15.lo iso8859_16.lo euc_tw.lo euc_kr.lo \
+       big5.lo gb18030.lo koi8_r.lo cp1251.lo onig_init.lo
 libonig_la_OBJECTS = $(am_libonig_la_OBJECTS)
 AM_V_lt = $(am__v_lt_@AM_V@)
 am__v_lt_ = $(am__v_lt_@AM_DEFAULT_V@)
@@ -333,7 +333,9 @@ libonig_la_SOURCES = regint.h regparse.h regenc.h st.h \
        unicode.c ascii.c utf8.c \
        utf16_be.c utf16_le.c \
        utf32_be.c utf32_le.c \
-       euc_jp.c sjis.c iso8859_1.c \
+       euc_jp.c euc_jp_prop.c \
+       sjis.c sjis_prop.c \
+       iso8859_1.c \
        iso8859_2.c  iso8859_3.c \
        iso8859_4.c  iso8859_5.c \
        iso8859_6.c  iso8859_7.c \
@@ -446,6 +448,7 @@ distclean-compile:
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/big5.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/cp1251.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/euc_jp.Plo@am__quote@
+@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/euc_jp_prop.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/euc_kr.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/euc_tw.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/gb18030.Plo@am__quote@
@@ -479,6 +482,7 @@ distclean-compile:
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regtrav.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/regversion.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/sjis.Plo@am__quote@
+@AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/sjis_prop.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/st.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/unicode.Plo@am__quote@
 @AMDEP_TRUE@@am__include@ @am__quote@./$(DEPDIR)/utf16_be.Plo@am__quote@

diff --git a/src/euc_jp.c b/src/euc_jp.c
index 0fb21769522ff0d780f3ea41692441dac73cbf60..f0a09d9f3bb6d4c179c4a763720ef77616ba0156 100644 (file)
--- a/src/euc_jp.c
+++ b/src/euc_jp.c
@@ -173,11 +173,6 @@ is_allowed_reverse_match(const UChar* s, const UChar* end ARG_UNUSED)
 }
 
 
-static const OnigCodePoint** PropertyList;
-static int PropertyListNum;
-static int PropertyListSize;
-static hash_table_type* PropertyNameTable;
-
 static const OnigCodePoint CR_Hiragana[] = {
   1,
   0xa4a1, 0xa4f3
@@ -190,37 +185,27 @@ static const OnigCodePoint CR_Katakana[] = {
   0xaab1, 0xaadd
 }; /* CR_Katakana */
 
-static int
-init_property_list(void)
-{
-  int r;
-
-  PROPERTY_LIST_ADD_PROP("Hiragana", CR_Hiragana);
-  PROPERTY_LIST_ADD_PROP("Katakana", CR_Katakana);
-
- end:
-  return r;
-}
-
-static int initialize(void)
-{
-  int r;
-
-  /* fprintf(stderr, "euc_jp: initialize called.\n"); */
-  r = init_property_list();
-  return r;
-}
+static const OnigCodePoint* PropertyList[] = {
+  CR_Hiragana,
+  CR_Katakana
+};
 
 static int
 property_name_to_ctype(OnigEncoding enc, UChar* p, UChar* end)
 {
-  hash_data_type ctype;
-
-  if (onig_st_lookup_strend(PropertyNameTable, p, end, &ctype) == 0) {
-    return onigenc_minimum_property_name_to_ctype(enc, p, end);
+  struct PropertyNameCtype* pc;
+  int len = end - p;
+  char q[32];
+
+  if (len < sizeof(q) - 1) {
+    xmemcpy(q, p, (size_t )len);
+    q[len] = '\0';
+    pc = euc_jp_lookup_property_name(q, len);
+    if (pc != 0)
+      return pc->ctype;
   }
 
-  return (int )ctype;
+  return ONIGERR_INVALID_CHAR_PROPERTY_NAME;
 }
 
 static int
@@ -237,7 +222,7 @@ is_code_ctype(OnigCodePoint code, unsigned int ctype)
   }
   else {
     ctype -= (ONIGENC_MAX_STD_CTYPE + 1);
-    if (ctype >= (unsigned int )PropertyListNum)
+    if (ctype >= (unsigned int )(sizeof(PropertyList)/sizeof(PropertyList[0])))
       return ONIGERR_TYPE_BUG;
 
     return onig_is_in_code_range((UChar* )PropertyList[ctype], code);
@@ -257,7 +242,7 @@ get_ctype_code_range(OnigCtype ctype, OnigCodePoint* sb_out,
     *sb_out = 0x80;
 
     ctype -= (ONIGENC_MAX_STD_CTYPE + 1);
-    if (ctype >= (OnigCtype )PropertyListNum)
+    if (ctype >= (OnigCtype )sizeof(PropertyList)/sizeof(PropertyList[0]))
       return ONIGERR_TYPE_BUG;
 
     *ranges = PropertyList[ctype];
@@ -283,5 +268,5 @@ OnigEncodingType OnigEncodingEUC_JP = {
   get_ctype_code_range,
   left_adjust_char_head,
   is_allowed_reverse_match,
-  initialize /* init */
+  NULL /* init */
 };

diff --git a/src/euc_jp_prop.c b/src/euc_jp_prop.c
new file mode 100644 (file)
index 0000000..8436fa2
--- /dev/null
+++ b/src/euc_jp_prop.c
@@ -0,0 +1,158 @@
+/* ANSI-C code produced by gperf version 3.0.4 */
+/* Command-line: gperf -pt -T -L ANSI-C -N euc_jp_lookup_property_name --output-file euc_jp_prop.c euc_jp_prop.gperf  */
+/* Computed positions: -k'1,3' */
+
+#if !((' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \
+      && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \
+      && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \
+      && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \
+      && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \
+      && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \
+      && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \
+      && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \
+      && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \
+      && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \
+      && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \
+      && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \
+      && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \
+      && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \
+      && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \
+      && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \
+      && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \
+      && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \
+      && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \
+      && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \
+      && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \
+      && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \
+      && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126))
+/* The character set is not based on ISO-646.  */
+#error "gperf generated tables don't work with this execution character set. Please report a bug to <bug-gnu-gperf@gnu.org>."
+#endif
+
+#line 1 "euc_jp_prop.gperf"
+
+#include <string.h>
+#include "regenc.h"
+
+#define TOTAL_KEYWORDS 16
+#define MIN_WORD_LENGTH 4
+#define MAX_WORD_LENGTH 8
+#define MIN_HASH_VALUE 4
+#define MAX_HASH_VALUE 55
+/* maximum key range = 52, duplicates = 0 */
+
+#ifdef __GNUC__
+__inline
+#else
+#ifdef __cplusplus
+inline
+#endif
+#endif
+static unsigned int
+hash (register const char *str, register unsigned int len)
+{
+  static unsigned char asso_values[] =
+    {
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56,  0,  3, 10, 25, 56,
+      56, 30,  0, 56, 56,  0, 20, 56, 56, 56,
+      15, 56, 56, 20, 56, 30, 56,  0,  0, 56,
+      56, 56, 56, 56, 56, 56, 56, 15, 56, 56,
+      56, 56, 56, 25, 56, 10, 56, 56, 56, 56,
+       5, 56,  0, 56,  0, 56,  5, 56, 56, 20,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56
+    };
+  return len + asso_values[(unsigned char)str[2]] + asso_values[(unsigned char)str[0]];
+}
+
+#ifdef __GNUC__
+__inline
+#if defined __GNUC_STDC_INLINE__ || defined __GNUC_GNU_INLINE__
+__attribute__ ((__gnu_inline__))
+#endif
+#endif
+struct PropertyNameCtype *
+euc_jp_lookup_property_name (register const char *str, register unsigned int len)
+{
+  static struct PropertyNameCtype wordlist[] =
+    {
+      {""}, {""}, {""}, {""},
+#line 23 "euc_jp_prop.gperf"
+      {"Word",     12},
+#line 12 "euc_jp_prop.gperf"
+      {"Alpha",     1},
+      {""}, {""},
+#line 26 "euc_jp_prop.gperf"
+      {"Hiragana", 15},
+      {""},
+#line 24 "euc_jp_prop.gperf"
+      {"Alnum",    13},
+      {""}, {""},
+#line 27 "euc_jp_prop.gperf"
+      {"Katakana", 16},
+      {""},
+#line 25 "euc_jp_prop.gperf"
+      {"ASCII",    14},
+#line 22 "euc_jp_prop.gperf"
+      {"XDigit",   11},
+      {""}, {""}, {""},
+#line 14 "euc_jp_prop.gperf"
+      {"Cntrl",     3},
+      {""}, {""},
+#line 13 "euc_jp_prop.gperf"
+      {"Blank",     2},
+      {""},
+#line 19 "euc_jp_prop.gperf"
+      {"Punct",     8},
+      {""}, {""}, {""}, {""},
+#line 18 "euc_jp_prop.gperf"
+      {"Print",     7},
+      {""}, {""}, {""}, {""},
+#line 21 "euc_jp_prop.gperf"
+      {"Upper",    10},
+      {""}, {""}, {""}, {""},
+#line 20 "euc_jp_prop.gperf"
+      {"Space",     9},
+      {""}, {""}, {""}, {""},
+#line 17 "euc_jp_prop.gperf"
+      {"Lower",     6},
+      {""}, {""}, {""}, {""},
+#line 16 "euc_jp_prop.gperf"
+      {"Graph",     5},
+      {""}, {""}, {""}, {""},
+#line 15 "euc_jp_prop.gperf"
+      {"Digit",     4}
+    };
+
+  if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH)
+    {
+      register int key = hash (str, len);
+
+      if (key <= MAX_HASH_VALUE && key >= 0)
+        {
+          register const char *s = wordlist[key].name;
+
+          if (*str == *s && !strcmp (str + 1, s + 1))
+            return &wordlist[key];
+        }
+    }
+  return 0;
+}

diff --git a/src/euc_jp_prop.gperf b/src/euc_jp_prop.gperf
new file mode 100644 (file)
index 0000000..2cec8cf
--- /dev/null
+++ b/src/euc_jp_prop.gperf
@@ -0,0 +1,27 @@
+%{
+#include <string.h>
+#include "regenc.h"
+%}
+
+struct PropertyNameCtype {
+  char *name;
+  int ctype;
+};
+
+%%
+Alpha,     1
+Blank,     2
+Cntrl,     3
+Digit,     4
+Graph,     5
+Lower,     6
+Print,     7
+Punct,     8
+Space,     9
+Upper,    10
+XDigit,   11
+Word,     12
+Alnum,    13
+ASCII,    14
+Hiragana, 15
+Katakana, 16

diff --git a/src/regenc.c b/src/regenc.c
index e4224a2f211e423b6d95fc45ddff485fc0bdfc83..e48010c46d667dd8e7816face8a974429c2d8d88 100644 (file)
--- a/src/regenc.c
+++ b/src/regenc.c
@@ -836,54 +836,3 @@ onigenc_with_ascii_strncmp(OnigEncoding enc, const UChar* p, const UChar* end,
   }
   return 0;
 }
-
-/* Property management */
-static int
-resize_property_list(int new_size, const OnigCodePoint*** plist, int* psize)
-{
-  int size;
-  const OnigCodePoint **list = *plist;
-
-  size = sizeof(OnigCodePoint*) * new_size;
-  if (IS_NULL(list)) {
-    list = (const OnigCodePoint** )xmalloc(size);
-  }
-  else {
-    list = (const OnigCodePoint** )xrealloc((void* )list, size);
-  }
-
-  if (IS_NULL(list)) return ONIGERR_MEMORY;
-
-  *plist = list;
-  *psize = new_size;
-
-  return 0;
-}
-
-extern int
-onigenc_property_list_add_property(UChar* name, const OnigCodePoint* prop,
-     hash_table_type **table, const OnigCodePoint*** plist, int *pnum,
-     int *psize)
-{
-#define PROP_INIT_SIZE     16
-
-  int r;
-
-  if (*psize <= *pnum) {
-    int new_size = (*psize == 0 ? PROP_INIT_SIZE : *psize * 2);
-    r = resize_property_list(new_size, plist, psize);
-    if (r != 0) return r;
-  }
-
-  (*plist)[*pnum] = prop;
-
-  if (ONIG_IS_NULL(*table)) {
-    *table = onig_st_init_strend_table_with_size(PROP_INIT_SIZE);
-    if (ONIG_IS_NULL(*table)) return ONIGERR_MEMORY;
-  }
-
-  *pnum = *pnum + 1;
-  onig_st_insert_strend(*table, name, name + strlen((char* )name),
-                       (hash_data_type )(*pnum + ONIGENC_MAX_STD_CTYPE));
-  return 0;
-}

diff --git a/src/regenc.h b/src/regenc.h
index ced76617d29f755110fae5e427ad3baf370d513f..659e96182b2b2062ba59e43bf1ea456437c70c03 100644 (file)
--- a/src/regenc.h
+++ b/src/regenc.h
@@ -4,7 +4,7 @@
   regenc.h -  Oniguruma (regular expression library)
 **********************************************************************/
 /*-
- * Copyright (c) 2002-2008  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
+ * Copyright (c) 2002-2016  K.Kosako  <sndgk393 AT ybb DOT ne DOT jp>
  * All rights reserved.
  *
  * Redistribution and use in source and binary forms, with or without
@@ -102,6 +102,10 @@ typedef struct {
   short int len;
 } PosixBracketEntryType;
 
+struct PropertyNameCtype {
+  char *name;
+  int ctype;
+};
 
 /* #define USE_CRNL_AS_LINE_TERMINATOR */
 #define USE_UNICODE_PROPERTIES
@@ -141,7 +145,8 @@ ONIG_EXTERN int onigenc_mb2_is_code_ctype P_((OnigEncoding enc, OnigCodePoint co
 ONIG_EXTERN int onigenc_mb4_code_to_mbclen P_((OnigCodePoint code));
 ONIG_EXTERN int onigenc_mb4_code_to_mbc P_((OnigEncoding enc, OnigCodePoint code, UChar *buf));
 ONIG_EXTERN int onigenc_mb4_is_code_ctype P_((OnigEncoding enc, OnigCodePoint code, unsigned int ctype));
-
+ONIG_EXTERN struct PropertyNameCtype* euc_jp_lookup_property_name P_((register const char *str, register unsigned int len));
+ONIG_EXTERN struct PropertyNameCtype* sjis_lookup_property_name P_((register const char *str, register unsigned int len));
 
 /* in enc/unicode.c */
 ONIG_EXTERN int onigenc_unicode_is_code_ctype P_((OnigCodePoint code, unsigned int ctype));

diff --git a/src/regint.h b/src/regint.h
index 69c6e887da1f0be3ddb79995445a8f3e23f2ed13..7a39839c968c26ef8d229085f436e6580b954143 100644 (file)
--- a/src/regint.h
+++ b/src/regint.h
@@ -767,15 +767,6 @@ extern hash_table_type* onig_st_init_strend_table_with_size P_((int size));
 extern int onig_st_lookup_strend P_((hash_table_type* table, const UChar* str_key, const UChar* end_key, hash_data_type *value));
 extern int onig_st_insert_strend P_((hash_table_type* table, const UChar* str_key, const UChar* end_key, hash_data_type value));
 
-/* encoding property management */
-#define PROPERTY_LIST_ADD_PROP(Name, CR) \
-  r = onigenc_property_list_add_property((UChar* )Name, CR,\
-             &PropertyNameTable, &PropertyList, &PropertyListNum,\
-             &PropertyListSize);\
-  if (r != 0) goto end
-
-extern int onigenc_property_list_add_property P_((UChar* name, const OnigCodePoint* prop, hash_table_type **table, const OnigCodePoint*** plist, int *pnum, int *psize));
-
 typedef int (*ONIGENC_INIT_PROPERTY_LIST_FUNC_TYPE)(void);
 
 #endif /* REGINT_H */

diff --git a/src/sjis.c b/src/sjis.c
index 58d216b38e41b85f9a135d32281fcd956bcc3088..bda8247cc2fda1fc4e4769fa9a74ee88027dfc48 100644 (file)
--- a/src/sjis.c
+++ b/src/sjis.c
@@ -206,11 +206,6 @@ is_allowed_reverse_match(const UChar* s, const UChar* end ARG_UNUSED)
 }
 
 
-static const OnigCodePoint** PropertyList;
-static int PropertyListNum;
-static int PropertyListSize;
-static hash_table_type* PropertyNameTable;
-
 static const OnigCodePoint CR_Hiragana[] = {
   1,
   0x829f, 0x82f1
@@ -224,38 +219,28 @@ static const OnigCodePoint CR_Katakana[] = {
   0x8380, 0x8396,
 }; /* CR_Katakana */
 
-static int
-init_property_list(void)
-{
-  int r;
-
-  PROPERTY_LIST_ADD_PROP("Hiragana", CR_Hiragana);
-  PROPERTY_LIST_ADD_PROP("Katakana", CR_Katakana);
-
- end:
-  return r;
-}
-
-static int initialize(void)
-{
-  int r;
-
-  /* fprintf(stderr, "sjis: initialize called.\n"); */
-  r = init_property_list();
-  return r;
-}
+static const OnigCodePoint* PropertyList[] = {
+  CR_Hiragana,
+  CR_Katakana
+};
 
 
 static int
 property_name_to_ctype(OnigEncoding enc, UChar* p, UChar* end)
 {
-  hash_data_type ctype;
-
-  if (onig_st_lookup_strend(PropertyNameTable, p, end, &ctype) == 0) {
-    return onigenc_minimum_property_name_to_ctype(enc, p, end);
+  struct PropertyNameCtype* pc;
+  int len = end - p;
+  char q[32];
+
+  if (len < sizeof(q) - 1) {
+    xmemcpy(q, p, (size_t )len);
+    q[len] = '\0';
+    pc = euc_jp_lookup_property_name(q, len);
+    if (pc != 0)
+      return pc->ctype;
   }
 
-  return (int )ctype;
+  return ONIGERR_INVALID_CHAR_PROPERTY_NAME;
 }
 
 static int
@@ -272,7 +257,7 @@ is_code_ctype(OnigCodePoint code, unsigned int ctype)
   }
   else {
     ctype -= (ONIGENC_MAX_STD_CTYPE + 1);
-    if (ctype >= (unsigned int )PropertyListNum)
+    if (ctype >= (unsigned int )(sizeof(PropertyList)/sizeof(PropertyList[0])))
       return ONIGERR_TYPE_BUG;
 
     return onig_is_in_code_range((UChar* )PropertyList[ctype], code);
@@ -292,7 +277,7 @@ get_ctype_code_range(OnigCtype ctype, OnigCodePoint* sb_out,
     *sb_out = 0x80;
 
     ctype -= (ONIGENC_MAX_STD_CTYPE + 1);
-    if (ctype >= (OnigCtype )PropertyListNum)
+    if (ctype >= (OnigCtype )(sizeof(PropertyList)/sizeof(PropertyList[0])))
       return ONIGERR_TYPE_BUG;
 
     *ranges = PropertyList[ctype];
@@ -317,5 +302,5 @@ OnigEncodingType OnigEncodingSJIS = {
   get_ctype_code_range,
   left_adjust_char_head,
   is_allowed_reverse_match,
-  initialize /* init */
+  NULL /* init */
 };

diff --git a/src/sjis_prop.c b/src/sjis_prop.c
new file mode 100644 (file)
index 0000000..83a7b45
--- /dev/null
+++ b/src/sjis_prop.c
@@ -0,0 +1,158 @@
+/* ANSI-C code produced by gperf version 3.0.4 */
+/* Command-line: gperf -pt -T -L ANSI-C -N sjis_lookup_property_name --output-file sjis_prop.c sjis_prop.gperf  */
+/* Computed positions: -k'1,3' */
+
+#if !((' ' == 32) && ('!' == 33) && ('"' == 34) && ('#' == 35) \
+      && ('%' == 37) && ('&' == 38) && ('\'' == 39) && ('(' == 40) \
+      && (')' == 41) && ('*' == 42) && ('+' == 43) && (',' == 44) \
+      && ('-' == 45) && ('.' == 46) && ('/' == 47) && ('0' == 48) \
+      && ('1' == 49) && ('2' == 50) && ('3' == 51) && ('4' == 52) \
+      && ('5' == 53) && ('6' == 54) && ('7' == 55) && ('8' == 56) \
+      && ('9' == 57) && (':' == 58) && (';' == 59) && ('<' == 60) \
+      && ('=' == 61) && ('>' == 62) && ('?' == 63) && ('A' == 65) \
+      && ('B' == 66) && ('C' == 67) && ('D' == 68) && ('E' == 69) \
+      && ('F' == 70) && ('G' == 71) && ('H' == 72) && ('I' == 73) \
+      && ('J' == 74) && ('K' == 75) && ('L' == 76) && ('M' == 77) \
+      && ('N' == 78) && ('O' == 79) && ('P' == 80) && ('Q' == 81) \
+      && ('R' == 82) && ('S' == 83) && ('T' == 84) && ('U' == 85) \
+      && ('V' == 86) && ('W' == 87) && ('X' == 88) && ('Y' == 89) \
+      && ('Z' == 90) && ('[' == 91) && ('\\' == 92) && (']' == 93) \
+      && ('^' == 94) && ('_' == 95) && ('a' == 97) && ('b' == 98) \
+      && ('c' == 99) && ('d' == 100) && ('e' == 101) && ('f' == 102) \
+      && ('g' == 103) && ('h' == 104) && ('i' == 105) && ('j' == 106) \
+      && ('k' == 107) && ('l' == 108) && ('m' == 109) && ('n' == 110) \
+      && ('o' == 111) && ('p' == 112) && ('q' == 113) && ('r' == 114) \
+      && ('s' == 115) && ('t' == 116) && ('u' == 117) && ('v' == 118) \
+      && ('w' == 119) && ('x' == 120) && ('y' == 121) && ('z' == 122) \
+      && ('{' == 123) && ('|' == 124) && ('}' == 125) && ('~' == 126))
+/* The character set is not based on ISO-646.  */
+#error "gperf generated tables don't work with this execution character set. Please report a bug to <bug-gnu-gperf@gnu.org>."
+#endif
+
+#line 1 "sjis_prop.gperf"
+
+#include <string.h>
+#include "regenc.h"
+
+#define TOTAL_KEYWORDS 16
+#define MIN_WORD_LENGTH 4
+#define MAX_WORD_LENGTH 8
+#define MIN_HASH_VALUE 4
+#define MAX_HASH_VALUE 55
+/* maximum key range = 52, duplicates = 0 */
+
+#ifdef __GNUC__
+__inline
+#else
+#ifdef __cplusplus
+inline
+#endif
+#endif
+static unsigned int
+hash (register const char *str, register unsigned int len)
+{
+  static unsigned char asso_values[] =
+    {
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56,  0,  3, 10, 25, 56,
+      56, 30,  0, 56, 56,  0, 20, 56, 56, 56,
+      15, 56, 56, 20, 56, 30, 56,  0,  0, 56,
+      56, 56, 56, 56, 56, 56, 56, 15, 56, 56,
+      56, 56, 56, 25, 56, 10, 56, 56, 56, 56,
+       5, 56,  0, 56,  0, 56,  5, 56, 56, 20,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56, 56, 56, 56, 56,
+      56, 56, 56, 56, 56, 56
+    };
+  return len + asso_values[(unsigned char)str[2]] + asso_values[(unsigned char)str[0]];
+}
+
+#ifdef __GNUC__
+__inline
+#if defined __GNUC_STDC_INLINE__ || defined __GNUC_GNU_INLINE__
+__attribute__ ((__gnu_inline__))
+#endif
+#endif
+struct PropertyNameCtype *
+sjis_lookup_property_name (register const char *str, register unsigned int len)
+{
+  static struct PropertyNameCtype wordlist[] =
+    {
+      {""}, {""}, {""}, {""},
+#line 23 "sjis_prop.gperf"
+      {"Word",     12},
+#line 12 "sjis_prop.gperf"
+      {"Alpha",     1},
+      {""}, {""},
+#line 26 "sjis_prop.gperf"
+      {"Hiragana", 15},
+      {""},
+#line 24 "sjis_prop.gperf"
+      {"Alnum",    13},
+      {""}, {""},
+#line 27 "sjis_prop.gperf"
+      {"Katakana", 16},
+      {""},
+#line 25 "sjis_prop.gperf"
+      {"ASCII",    14},
+#line 22 "sjis_prop.gperf"
+      {"XDigit",   11},
+      {""}, {""}, {""},
+#line 14 "sjis_prop.gperf"
+      {"Cntrl",     3},
+      {""}, {""},
+#line 13 "sjis_prop.gperf"
+      {"Blank",     2},
+      {""},
+#line 19 "sjis_prop.gperf"
+      {"Punct",     8},
+      {""}, {""}, {""}, {""},
+#line 18 "sjis_prop.gperf"
+      {"Print",     7},
+      {""}, {""}, {""}, {""},
+#line 21 "sjis_prop.gperf"
+      {"Upper",    10},
+      {""}, {""}, {""}, {""},
+#line 20 "sjis_prop.gperf"
+      {"Space",     9},
+      {""}, {""}, {""}, {""},
+#line 17 "sjis_prop.gperf"
+      {"Lower",     6},
+      {""}, {""}, {""}, {""},
+#line 16 "sjis_prop.gperf"
+      {"Graph",     5},
+      {""}, {""}, {""}, {""},
+#line 15 "sjis_prop.gperf"
+      {"Digit",     4}
+    };
+
+  if (len <= MAX_WORD_LENGTH && len >= MIN_WORD_LENGTH)
+    {
+      register int key = hash (str, len);
+
+      if (key <= MAX_HASH_VALUE && key >= 0)
+        {
+          register const char *s = wordlist[key].name;
+
+          if (*str == *s && !strcmp (str + 1, s + 1))
+            return &wordlist[key];
+        }
+    }
+  return 0;
+}

diff --git a/src/sjis_prop.gperf b/src/sjis_prop.gperf
new file mode 100644 (file)
index 0000000..2cec8cf
--- /dev/null
+++ b/src/sjis_prop.gperf
@@ -0,0 +1,27 @@
+%{
+#include <string.h>
+#include "regenc.h"
+%}
+
+struct PropertyNameCtype {
+  char *name;
+  int ctype;
+};
+
+%%
+Alpha,     1
+Blank,     2
+Cntrl,     3
+Digit,     4
+Graph,     5
+Lower,     6
+Print,     7
+Punct,     8
+Space,     9
+Upper,    10
+XDigit,   11
+Word,     12
+Alnum,    13
+ASCII,    14
+Hiragana, 15
+Katakana, 16