/* String search routines for GNU Emacs.
Copyright (C) 1985, 1986, 1987, 1993, 1994, 1997, 1998, 1999, 2001, 2002,
- 2003, 2004, 2005, 2006, 2007, 2008, 2009
+ 2003, 2004, 2005, 2006, 2007, 2008, 2009, 2010
Free Software Foundation, Inc.
This file is part of GNU Emacs.
#include <config.h>
+#include <setjmp.h>
#include "lisp.h"
#include "syntax.h"
#include "category.h"
static void set_search_regs P_ ((EMACS_INT, EMACS_INT));
static void save_search_regs P_ ((void));
-static EMACS_INT simple_search P_ ((int, unsigned char *, int, int,
+static EMACS_INT simple_search P_ ((int, unsigned char *, int, int,
Lisp_Object, EMACS_INT, EMACS_INT,
EMACS_INT, EMACS_INT));
-static EMACS_INT boyer_moore P_ ((int, unsigned char *, int, int,
+static EMACS_INT boyer_moore P_ ((int, unsigned char *, int, int,
Lisp_Object, Lisp_Object,
EMACS_INT, EMACS_INT,
EMACS_INT, EMACS_INT, int));
int i;
for (i = 0; i < REGEXP_CACHE_SIZE; ++i)
- /* It's tempting to compare with the syntax-table we've actually changd,
- but it's not sufficient because char-table inheritance mewans that
+ /* It's tempting to compare with the syntax-table we've actually changed,
+ but it's not sufficient because char-table inheritance means that
modifying one syntax-table can change others at the same time. */
if (!EQ (searchbufs[i].syntax_table, Qt))
searchbufs[i].regexp = Qnil;
if (regp)
re_set_registers (&cp->buf, regp, regp->num_regs, regp->start, regp->end);
- /* The compiled pattern can be used both for mulitbyte and unibyte
+ /* The compiled pattern can be used both for multibyte and unibyte
target. But, we have to tell which the pattern is used for. */
cp->buf.target_multibyte = multibyte;
return val;
}
\f
-/* Match REGEXP atainst the characters after POS to LIMIT, and return
+/* Match REGEXP against the characters after POS to LIMIT, and return
the number of matched characters. If STRING is non-nil, match
against the characters in it. In that case, POS and LIMIT are
indices into the string. This function doesn't modify the match
unsigned char *p1, *p2;
EMACS_INT s1, s2;
EMACS_INT len;
-
+
if (STRINGP (string))
{
if (pos_byte < 0)
base_pat++;
}
- c = STRING_CHAR_AND_LENGTH (base_pat, len_byte, in_charlen);
+ c = STRING_CHAR_AND_LENGTH (base_pat, in_charlen);
if (NILP (trt))
{
EMACS_INT this_pos = pos;
EMACS_INT this_pos_byte = pos_byte;
int this_len = len;
- int this_len_byte = len_byte;
unsigned char *p = pat;
if (pos + len > lim || pos_byte + len_byte > lim_byte)
goto stop;
int charlen, buf_charlen;
int pat_ch, buf_ch;
- pat_ch = STRING_CHAR_AND_LENGTH (p, this_len_byte, charlen);
+ pat_ch = STRING_CHAR_AND_LENGTH (p, charlen);
buf_ch = STRING_CHAR_AND_LENGTH (BYTE_POS_ADDR (this_pos_byte),
- ZV_BYTE - this_pos_byte,
buf_charlen);
TRANSLATE (buf_ch, trt, buf_ch);
if (buf_ch != pat_ch)
break;
- this_len_byte -= charlen;
this_len--;
p += charlen;
while (1)
{
/* Try matching at position POS. */
- EMACS_INT this_pos = pos - len;
- EMACS_INT this_pos_byte;
+ EMACS_INT this_pos = pos;
+ EMACS_INT this_pos_byte = pos_byte;
int this_len = len;
- int this_len_byte = len_byte;
- unsigned char *p = pat;
+ const unsigned char *p = pat + len_byte;
- if (this_pos < lim || (pos_byte - len_byte) < lim_byte)
+ if (this_pos - len < lim || (pos_byte - len_byte) < lim_byte)
goto stop;
- this_pos_byte = CHAR_TO_BYTE (this_pos);
- match_byte = pos_byte - this_pos_byte;
while (this_len > 0)
{
- int charlen, buf_charlen;
+ int charlen;
int pat_ch, buf_ch;
- pat_ch = STRING_CHAR_AND_LENGTH (p, this_len_byte, charlen);
- buf_ch = STRING_CHAR_AND_LENGTH (BYTE_POS_ADDR (this_pos_byte),
- ZV_BYTE - this_pos_byte,
- buf_charlen);
+ DEC_BOTH (this_pos, this_pos_byte);
+ PREV_CHAR_BOUNDARY (p, pat);
+ pat_ch = STRING_CHAR (p);
+ buf_ch = STRING_CHAR (BYTE_POS_ADDR (this_pos_byte));
TRANSLATE (buf_ch, trt, buf_ch);
if (buf_ch != pat_ch)
break;
- this_len_byte -= charlen;
this_len--;
- p += charlen;
- this_pos_byte += buf_charlen;
- this_pos++;
}
if (this_len == 0)
{
- pos -= len;
- pos_byte -= match_byte;
+ match_byte = pos_byte - this_pos_byte;
+ pos = this_pos;
+ pos_byte = this_pos_byte;
break;
}
while (! (CHAR_HEAD_P (*charstart)))
charstart--;
- ch = STRING_CHAR (charstart, ptr - charstart + 1);
+ ch = STRING_CHAR (charstart);
if (char_base != (ch & ~0x3F))
ch = -1;
}
/* Note that we don't have to increment POS. */
c = SREF (newtext, pos_byte++);
if (buf_multibyte)
- c = unibyte_char_to_multibyte (c);
+ MAKE_CHAR_MULTIBYTE (c);
}
/* Either set ADD_STUFF and ADD_LEN to the text to put in SUBSTED,
{
c = SREF (newtext, pos_byte++);
if (buf_multibyte)
- c = unibyte_char_to_multibyte (c);
+ MAKE_CHAR_MULTIBYTE (c);
}
if (c == '&')
}
searchbuf_head = &searchbufs[0];
- Qsearch_failed = intern ("search-failed");
+ Qsearch_failed = intern_c_string ("search-failed");
staticpro (&Qsearch_failed);
- Qinvalid_regexp = intern ("invalid-regexp");
+ Qinvalid_regexp = intern_c_string ("invalid-regexp");
staticpro (&Qinvalid_regexp);
Fput (Qsearch_failed, Qerror_conditions,
- Fcons (Qsearch_failed, Fcons (Qerror, Qnil)));
+ pure_cons (Qsearch_failed, pure_cons (Qerror, Qnil)));
Fput (Qsearch_failed, Qerror_message,
- build_string ("Search failed"));
+ make_pure_c_string ("Search failed"));
Fput (Qinvalid_regexp, Qerror_conditions,
- Fcons (Qinvalid_regexp, Fcons (Qerror, Qnil)));
+ pure_cons (Qinvalid_regexp, pure_cons (Qerror, Qnil)));
Fput (Qinvalid_regexp, Qerror_message,
- build_string ("Invalid regexp"));
+ make_pure_c_string ("Invalid regexp"));
last_thing_searched = Qnil;
staticpro (&last_thing_searched);