drew
/
acadia-newlib
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Cygwin: glob: perform correct utf-32 -> multibyte conversion 

g_Ctoc, converting the UTF-32 filenames to multibyte, still
used UTF-16 to  multibyte conversion.  Introduce a wirtomb
helper and fix that.

Signed-off-by: Corinna Vinschen <corinna@vinschen.de>
This commit is contained in:
Corinna Vinschen 2023-02-22 10:03:03 +01:00
parent 4349a1e4db
commit f3f20038c4
3 changed files with 24 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
winsup/cygwin/glob.cc
View File

@ -1024,7 +1024,7 @@ g_Ctoc(const Char *str, char *buf, size_t len)
memset(&mbs, 0, sizeof(mbs)); memset(&mbs, 0, sizeof(mbs));
while (len >= (size_t) MB_CUR_MAX) { while (len >= (size_t) MB_CUR_MAX) {
clen = wcrtomb(buf, *str, &mbs); clen = wirtomb(buf, *str, &mbs);
if (clen == (size_t)-1) if (clen == (size_t)-1)
return (1); return (1);
if (*str == L'\0') if (*str == L'\0')

4
winsup/cygwin/local_includes/wchar.h
View File

@ -44,6 +44,10 @@ extern wctomb_f __utf8_wctomb;
for surrogate pairs, plus a wchar_t NUL. */ for surrogate pairs, plus a wchar_t NUL. */
void wcintowcs (wchar_t *, wint_t *, size_t); void wcintowcs (wchar_t *, wint_t *, size_t);
/* replacement function for wcrtomb, converting a UTF-32 char to a
multibyte string. */
size_t wirtomb (char *, wint_t, mbstate_t *);
/* replacement function for mbrtowc, returning a wint_t representing /* replacement function for mbrtowc, returning a wint_t representing
a UTF-32 value. Defined in strfuncs.cc */ a UTF-32 value. Defined in strfuncs.cc */
extern size_t mbrtowi (wint_t *, const char *, size_t, mbstate_t *); extern size_t mbrtowi (wint_t *, const char *, size_t, mbstate_t *);

19
winsup/cygwin/strfuncs.cc
View File

@ -129,6 +129,25 @@ wcintowcs (wchar_t *dest, wint_t *src, size_t len)
*dest = '\0'; *dest = '\0';
} }
/* replacement function for wcrtomb, converting a UTF-32 char to a
multibyte string. */
extern "C" size_t
wirtomb (char *s, wint_t wi, mbstate_t *ps)
{
wchar_t wc[3] = { (wchar_t) wi, '\0', '\0' };
const wchar_t *wcp = wc;
size_t nwc = 1;
if (wi >= 0x10000)
{
wi -= 0x10000;
wc[0] = (wi >> 10) + 0xd800;
wc[1] = (wi & 0x3ff) + 0xdc00;
nwc = 2;
}
return wcsnrtombs (s, &wcp, nwc, SIZE_MAX, ps);
}
/* replacement function for mbrtowc, returning a wint_t representing /* replacement function for mbrtowc, returning a wint_t representing
a UTF-32 value. */ a UTF-32 value. */
extern "C" size_t extern "C" size_t