merge Oniguruma 4.4.5

git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@11234 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

ChangeLog

@@ -1,3 +1,17 @@
+Sat Oct 28 20:13:18 2006  K.Kosako  <sndgk393 AT ybb.ne.jp>
+
+	* oniguruma.h: Version 4.4.5
+
+	* regint.h: ditto.
+
+	* regerror.c: ditto.
+
+	* regexec.c: ditto.
+
+	* regcomp.c ditto.
+
+	* regparse.c ditto.
+
 Sat Oct 28 07:56:13 2006  Yukihiro Matsumoto  <matz@ruby-lang.org>
 
 	* marshal.c (r_object0): missing break.  [ruby-core:09345]
@@ -10926,7 +10940,7 @@ Fri Feb 18 04:06:41 2005  Yukihiro Matsumoto  <matz@ruby-lang.org>
 	* parse.y (fcall_gen): lvar(arg) will be evaluated as
 	  lvar.call(arg) when lvar is a defined local variable. [new]
 
-Thu Feb 17 22:15:34 2005  K.Kosako  <sndgk@ybb.ne.jp>
+Thu Feb 17 22:15:34 2005  K.Kosako  <sndgk393 AT ybb.ne.jp>
 
 	* ext/strscan/strscan.c: calls Oniguruma API directly.
 

oniguruma.h

@@ -36,7 +36,7 @@ extern "C" {
 #define ONIGURUMA
 #define ONIGURUMA_VERSION_MAJOR   4
 #define ONIGURUMA_VERSION_MINOR   4
-#define ONIGURUMA_VERSION_TEENY   4
+#define ONIGURUMA_VERSION_TEENY   5
 
 #ifdef __cplusplus
 # ifndef  HAVE_PROTOTYPES
@@ -715,6 +715,7 @@ struct re_registers {
 typedef struct re_registers   OnigRegion;
 
 typedef struct {
+  OnigEncoding enc;
   OnigUChar* par;
   OnigUChar* par_end;
 } OnigErrorInfo;

regcomp.c

@@ -5277,6 +5277,7 @@ onig_compile(regex_t* reg, const UChar* pattern, const UChar* pattern_end,
  err:
   if (IS_NOT_NULL(scan_env.error)) {
     if (IS_NOT_NULL(einfo)) {
+      einfo->enc     = scan_env.enc;
       einfo->par     = scan_env.error;
       einfo->par_end = scan_env.error_end;
     }
@@ -5470,8 +5471,6 @@ OnigOpInfoType OnigOpInfo[] = {
   { OP_ANYCHAR_ML_STAR_PEEK_NEXT, "anychar-ml*-peek-next", ARG_SPECIAL },
   { OP_WORD,                "word",            ARG_NON },
   { OP_NOT_WORD,            "not-word",        ARG_NON },
-  { OP_WORD_SB,             "word-sb",         ARG_NON },
-  { OP_WORD_MB,             "word-mb",         ARG_NON },
   { OP_WORD_BOUND,          "word-bound",      ARG_NON },
   { OP_NOT_WORD_BOUND,      "not-word-bound",  ARG_NON },
   { OP_WORD_BEGIN,          "word-begin",      ARG_NON },

regerror.c

@@ -183,6 +183,48 @@ onig_error_code_to_format(int code)
 }
 
 
+static int to_ascii(OnigEncoding enc, UChar *s, UChar *end,
+		    UChar buf[], int buf_size, int *is_over)
+{
+  int len;
+  UChar *p;
+  OnigCodePoint code;
+
+  if (ONIGENC_MBC_MINLEN(enc) > 1) {
+    p = s;
+    len = 0;
+    while (p < end) {
+      code = ONIGENC_MBC_TO_CODE(enc, p, end);
+      if (code >= 0x80) {
+	if (len + 5 <= buf_size) {
+	  sprintf((char* )(&(buf[len])), "\\%03o",
+		  (unsigned int)(code & 0377));
+	  len += 5;
+	}
+	else {
+	  break;
+	}
+      }
+      else {
+	buf[len++] = (UChar )code;
+      }
+
+      p += enc_len(enc, p);
+      if (len >= buf_size) break;
+    }
+
+    *is_over = ((p < end) ? 1 : 0);
+  }
+  else {
+    len = MIN((end - s), buf_size);
+    xmemcpy(buf, s, (size_t )len);
+    *is_over = ((buf_size < (end - s)) ? 1 : 0);
+  }
+
+  return len;
+}
+
+
 /* for ONIG_MAX_ERROR_MESSAGE_LEN */
 #define MAX_ERROR_PAR_LEN   30
 
@@ -198,7 +240,8 @@ onig_error_code_to_str(s, code, va_alist)
 {
   UChar *p, *q;
   OnigErrorInfo* einfo;
-  int len;
+  int len, is_over;
+  UChar parbuf[MAX_ERROR_PAR_LEN];
   va_list vargs;
 
   va_init_list(vargs, code);
@@ -212,23 +255,20 @@ onig_error_code_to_str(s, code, va_alist)
   case ONIGERR_INVALID_CHAR_IN_GROUP_NAME:
   case ONIGERR_INVALID_CHAR_PROPERTY_NAME:
     einfo = va_arg(vargs, OnigErrorInfo*);
-    len = einfo->par_end - einfo->par;
+    len = to_ascii(einfo->enc, einfo->par, einfo->par_end,
+		   parbuf, MAX_ERROR_PAR_LEN - 3, &is_over);
     q = onig_error_code_to_format(code);
     p = s;
     while (*q != '\0') {
       if (*q == '%') {
 	q++;
 	if (*q == 'n') { /* '%n': name */
-	  if (len > MAX_ERROR_PAR_LEN) {
+	  xmemcpy(p, parbuf, len);
-	    xmemcpy(p, einfo->par, MAX_ERROR_PAR_LEN - 3);
+	  p += len;
-	    p += (MAX_ERROR_PAR_LEN - 3);
+	  if (is_over != 0) {
 	    xmemcpy(p, "...", 3);
 	    p += 3;
 	  }
-	  else {
-	    xmemcpy(p, einfo->par, len);
-	    p += len;
-	  }
 	  q++;
 	}
 	else

regexec.c

@@ -388,18 +388,26 @@ typedef struct {
 
 #define STATE_CHECK_BUFF_MALLOC_THRESHOLD_SIZE  16
 
-#define STATE_CHECK_BUFF_INIT(msa, str_len, state_num) do { \
+#define STATE_CHECK_BUFF_INIT(msa, str_len, offset, state_num) do {	\
-  (msa).state_check_buff = (void* )0;\
   if ((state_num) > 0 && str_len >= STATE_CHECK_STRING_THRESHOLD_LEN) {\
     int size = ((int )((str_len) + 1) * (state_num) + 7) / 8;\
-    (msa).state_check_buff_size = size;			     \
+    if (size > 0 && offset < size && size < STATE_CHECK_BUFF_MAX_SIZE) {\
-    if (size > 0 && size < STATE_CHECK_BUFF_MAX_SIZE) {\
       if (size >= STATE_CHECK_BUFF_MALLOC_THRESHOLD_SIZE) \
         (msa).state_check_buff = (void* )xmalloc(size);\
       else \
         (msa).state_check_buff = (void* )xalloca(size);\
-      xmemset((msa).state_check_buff, 0, (size_t )size);\
+      xmemset(((char* )((msa).state_check_buff)+(offset)), 0, \
+              (size_t )(size - (offset))); \
+      (msa).state_check_buff_size = size;\
     }\
+    else {\
+      (msa).state_check_buff = (void* )0;\
+      (msa).state_check_buff_size = 0;\
+    }\
+  }\
+  else {\
+    (msa).state_check_buff = (void* )0;\
+    (msa).state_check_buff_size = 0;\
   }\
 } while (0)
 
@@ -410,7 +418,7 @@ typedef struct {
   }\
 } while (0);
 #else
-#define STATE_CHECK_BUFF_INIT(msa, str_len, state_num)
+#define STATE_CHECK_BUFF_INIT(msa, str_len, offset, state_num)
 #define MATCH_ARG_FREE(msa)  if ((msa).stack_p) xfree((msa).stack_p)
 #endif
 
@@ -3261,7 +3269,12 @@ onig_match(regex_t* reg, const UChar* str, const UChar* end, const UChar* at, On
 #endif /* USE_RECOMPILE_API && USE_MULTI_THREAD_SYSTEM */
 
   MATCH_ARG_INIT(msa, option, region, at);
-  STATE_CHECK_BUFF_INIT(msa, end - str, reg->num_comb_exp_check);
+#ifdef USE_COMBINATION_EXPLOSION_CHECK
+  {
+    int offset = at - str;
+    STATE_CHECK_BUFF_INIT(msa, end - str, offset, reg->num_comb_exp_check);
+  }
+#endif
 
   if (region
 #ifdef USE_POSIX_REGION_OPTION
@@ -3665,7 +3678,8 @@ onig_search(regex_t* reg, const UChar* str, const UChar* end,
 
       MATCH_ARG_INIT(msa, option, region, start);
 #ifdef USE_COMBINATION_EXPLOSION_CHECK
-      msa.state_check_buff = (void* )0;
+      msa.state_check_buff      = (void* )0;
+      msa.state_check_buff_size = 0;
 #endif
       MATCH_AND_RETURN_CHECK;
       goto mismatch;
@@ -3679,7 +3693,12 @@ onig_search(regex_t* reg, const UChar* str, const UChar* end,
 #endif
 
   MATCH_ARG_INIT(msa, option, region, orig_start);
-  STATE_CHECK_BUFF_INIT(msa, end - str, reg->num_comb_exp_check);
+#ifdef USE_COMBINATION_EXPLOSION_CHECK
+  {
+    int offset = (MIN(start, range) - str);
+    STATE_CHECK_BUFF_INIT(msa, end - str, offset, reg->num_comb_exp_check);
+  }
+#endif
 
   s = (UChar* )start;
   if (range > start) {   /* forward search */

regint.h

@@ -121,7 +121,7 @@
 #endif /* else NOT_RUBY */
 
 #define STATE_CHECK_STRING_THRESHOLD_LEN             7
-#define STATE_CHECK_BUFF_MAX_SIZE           0x08000000
+#define STATE_CHECK_BUFF_MAX_SIZE               0x8000
 
 #define THREAD_PASS_LIMIT_COUNT     8
 #define xmemset     memset
@@ -584,8 +584,6 @@ enum OpCode {
 
   OP_WORD,
   OP_NOT_WORD,
-  OP_WORD_SB,
-  OP_WORD_MB,
   OP_WORD_BOUND,
   OP_NOT_WORD_BOUND,
   OP_WORD_BEGIN,

regparse.c

@@ -4830,7 +4830,7 @@ onig_free_shared_cclass_table()
 {
   if (IS_NOT_NULL(OnigTypeCClassTable)) {
     onig_st_foreach(OnigTypeCClassTable, i_free_shared_class, 0);
-    xfree(OnigTypeCClassTable);
+    onig_st_free_table(OnigTypeCClassTable);
     OnigTypeCClassTable = NULL;
   }