* include/ruby/oniguruma.h (OnigEncodingTypeST): add precise_ret

  argument for mbc_to_code.
  (ONIGENC_MBC_TO_CODE): provide NULL for precise_ret.
  (ONIGENC_MBC_PRECISE_CODEPOINT): defined.

* include/ruby/encoding.h (rb_enc_mbc_precise_codepoint): defined.

* regenc.h (onigenc_single_byte_mbc_to_code): precise_ret argument
  added.
  (onigenc_mbn_mbc_to_code): ditto.

* regenc.c (onigenc_single_byte_mbc_to_code): precise_ret argument
  added.
  (onigenc_mbn_mbc_to_code): ditto.

* string.c (count_utf8_lead_bytes_with_word): removed.
  (str_utf8_nth): removed.
  (str_utf8_offset): removed.
  (str_strlen): UTF-8 codepoint oriented optimization removed.
  (rb_str_substr): ditto.
  (enc_succ_char): use rb_enc_mbc_precise_codepoint.
  (enc_pred_char): ditto.
  (rb_str_succ): ditto.

* encoding.c (rb_enc_ascget): check length with
  rb_enc_mbc_precise_codepoint.
  (rb_enc_codepoint): use rb_enc_mbc_precise_codepoint.

* regexec.c (string_cmp_ic): add text_end argument.
  (match_at): check end of character after exact string matches.

* enc/utf_8.c (graphme_table): defined for extended graphme cluster
  boundary.
  (grapheme_cmp): defined.
  (get_grapheme_properties): defined.
  (grapheme_boundary_p): defined.
  (MAX_BYTES_LENGTH): defined.
  (comb_char_enc_len): defined.
  (mbc_to_code0): extracted from mbc_to_code.
  (mbc_to_code): use mbc_to_code0.
  (left_adjust_combchar_head): defined.
  (utf_8): use a extended graphme cluster as a unit.

* enc/unicode.c (onigenc_unicode_mbc_case_fold): use
  ONIGENC_MBC_PRECISE_CODEPOINT to extract codepoints.
  (onigenc_unicode_get_case_fold_codes_by_str): ditto.

* enc/euc_jp.c (mbc_to_code): follow mbc_to_code field change.
  use onigenc_mbn_mbc_to_code.

* enc/shift_jis.c (mbc_to_code): ditto.

* enc/emacs_mule.c (mbc_to_code): ditto.

* enc/gbk.c (gbk_mbc_to_code): follow mbc_to_code field and
  onigenc_mbn_mbc_to_code change.

* enc/cp949.c (cp949_mbc_to_code): ditto.

* enc/big5.c (big5_mbc_to_code): ditto.

* enc/euc_tw.c (euctw_mbc_to_code): ditto.

* enc/euc_kr.c (euckr_mbc_to_code): ditto.

* enc/gb18030.c (gb18030_mbc_to_code): ditto.

* enc/utf_32be.c (utf32be_mbc_to_code): follow mbc_to_code field
  change.

* enc/utf_16be.c (utf16be_mbc_to_code): ditto.

* enc/utf_32le.c (utf32le_mbc_to_code): ditto.

* enc/utf_16le.c (utf16le_mbc_to_code): ditto.



git-svn-id: svn+ssh://ci.ruby-lang.org/ruby/trunk@19389 b2dd03c8-39d4-4d8f-98ff-823fe69b080e

1 files changed, 14 insertions, 3 deletions

diff --git a/regenc.c b/regenc.c
index 634afd8883..f25bc80ad8 100644
--- a/regenc.c
+++ b/regenc.c
@@ -615,8 +615,11 @@ onigenc_single_byte_mbc_enc_len(const UChar* p ARG_UNUSED, const UChar* e ARG_UN
 
 extern OnigCodePoint
 onigenc_single_byte_mbc_to_code(const UChar* p, const UChar* end ARG_UNUSED,
+                                int *precise_ret,
 				OnigEncoding enc ARG_UNUSED)
 {
+  if (precise_ret)
+    *precise_ret = ONIGENC_CONSTRUCT_MBCLEN_CHARFOUND(1);
   return (OnigCodePoint )(*p);
 }
 
@@ -668,12 +671,20 @@ onigenc_ascii_is_code_ctype(OnigCodePoint code, unsigned int ctype,
 }
 
 extern OnigCodePoint
-onigenc_mbn_mbc_to_code(OnigEncoding enc, const UChar* p, const UChar* end)
+onigenc_mbn_mbc_to_code(OnigEncoding enc, const UChar* p, const UChar* end, int *precise_ret)
 {
-  int c, i, len;
+  int c, i, len, ret;
   OnigCodePoint n;
 
-  len = enclen(enc, p, end);
+  ret = ONIGENC_PRECISE_MBC_ENC_LEN(enc, p, end);
+  if (precise_ret)
+    *precise_ret = ret;
+  if (ONIGENC_MBCLEN_CHARFOUND_P(ret))
+    len = ONIGENC_MBCLEN_CHARFOUND_LEN(ret);
+  else if (ONIGENC_MBCLEN_NEEDMORE_P(ret))
+    len = end-p+ONIGENC_MBCLEN_NEEDMORE_LEN(ret);
+  else
+    len = 1;
   n = (OnigCodePoint )(*p++);
   if (len == 1) return n;