merge from GNU sed
Patches applied: * bonzini@gnu.org--2004b/sed--stable--4.1--patch-67 do not mention Estonian locale in BUGS * bonzini@gnu.org--2004b/sed--stable--4.1--patch-70 hack to speed up UTF-8 processing git-archimport-id: bonzini@gnu.org--2004b/sed--ssed--3.61--patch-30
This commit is contained in:
1 parent
202623b7d6
commit
ae382af78e
4 files changed
+51
-17
No files matched your search
@@ -1,3 +1,10 @@
|
||||
2006-08-03 Paolo Bonzini <bonzini@gnu.org>
|
||||
|
||||
* sed/execute.c (str_append): Use is_utf8 to skip useless work.
|
||||
* sed/mbcs.c (initialize_mbcs): Look for a UTF-8 locale.
|
||||
(is_utf8): New.
|
||||
* sed/sed.h (is_utf8): New.
|
||||
|
||||
2006-07-13 Paolo Bonzini <bonzini@gnu.org>
|
||||
|
||||
* bootstrap.sh.in: Add a few autoconfy tests.
|
||||
|
||||
+18
-17
@@ -229,25 +229,26 @@ str_append(to, string, length)
|
||||
to->length = new_length;
|
||||
|
||||
#ifdef HAVE_MBRTOWC
|
||||
if (mb_cur_max == 1)
|
||||
return;
|
||||
if (mb_cur_max > 1 && !is_utf8)
|
||||
while (length)
|
||||
{
|
||||
size_t n = MBRLEN (string, length, &to->mbstate);
|
||||
|
||||
while (length)
|
||||
{
|
||||
int n = MBRLEN (string, length, &to->mbstate);
|
||||
/* An invalid sequence is treated like a singlebyte character. */
|
||||
if (n == (size_t) -1)
|
||||
{
|
||||
memset (&to->mbstate, 0, sizeof (to->mbstate));
|
||||
n = 1;
|
||||
}
|
||||
|
||||
/* An invalid sequence is treated like a singlebyte character. */
|
||||
if (n == -1)
|
||||
{
|
||||
memset (&to->mbstate, 0, sizeof (to->mbstate));
|
||||
n = 1;
|
||||
}
|
||||
|
||||
if (n > 0)
|
||||
length -= n;
|
||||
else
|
||||
break;
|
||||
}
|
||||
if (n > 0)
|
||||
{
|
||||
string += n;
|
||||
length -= n;
|
||||
}
|
||||
else
|
||||
break;
|
||||
}
|
||||
#endif
|
||||
}
|
||||
|
||||
|
||||
+25
@@ -18,7 +18,12 @@
|
||||
#include "sed.h"
|
||||
#include <stdlib.h>
|
||||
|
||||
#ifdef HAVE_LANGINFO_CODESET
|
||||
#include <langinfo.h>
|
||||
#endif
|
||||
|
||||
int mb_cur_max;
|
||||
bool is_utf8;
|
||||
|
||||
#ifdef HAVE_MBRTOWC
|
||||
/* Add a byte to the multibyte character represented by the state
|
||||
@@ -47,6 +52,26 @@ int brlen (ch, cur_stat)
|
||||
void
|
||||
initialize_mbcs ()
|
||||
{
|
||||
/* For UTF-8, we know that the encoding is stateless. */
|
||||
const char *codeset_name;
|
||||
|
||||
#ifdef HAVE_LANGINFO_CODESET
|
||||
codeset_name = nl_langinfo (CODESET);
|
||||
#else
|
||||
codeset_name = getenv ("LC_ALL");
|
||||
if (codeset_name == NULL || codeset_name[0] == '\0')
|
||||
codeset_name = getenv ("LC_CTYPE");
|
||||
if (codeset_name == NULL || codeset_name[0] == '\0')
|
||||
codeset_name = getenv ("LANG");
|
||||
if (codeset_name == NULL)
|
||||
codeset_name = "";
|
||||
else if (strchr (codeset_name, '.') != NULL)
|
||||
codeset_name = strchr (codeset_name, '.') + 1;
|
||||
#endif
|
||||
|
||||
is_utf8 = (strcasecmp (codeset_name, "UTF-8") == 0
|
||||
|| strcasecmp (codeset_name, "UTF8") == 0);
|
||||
|
||||
#ifdef HAVE_MBRTOWC
|
||||
mb_cur_max = MB_CUR_MAX;
|
||||
#else
|
||||
|
||||
Reference in new issue
Block a user