fluencelabs/musl
1
0
Fork 0
mirror of https://github.com/fluencelabs/musl synced 2025-04-25 15:22:15 +00:00
Code Issues Projects Releases Wiki Activity

handle ascii range individually in each iconv case

Browse Source 
short-circuiting low bytes before the switch precluded support for
character encodings that don't coincide with ascii in this range. this
limitation affected iso-2022 encodings, which use the esc byte to
introduce a shift sequence, and things like ebcdic.
This commit is contained in:
Rich Felker 2017-11-10 13:34:21 -05:00
parent bff59d13a8
commit 9eb6dd5165
1 changed files with 10 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

12
src/locale/iconv.c
View File

@ -193,8 +193,9 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
c = *(unsigned char *)*in; c = *(unsigned char *)*in;
l = 1; l = 1;
if (c >= 128 || type-UTF_32BE < 7U) switch (type) { switch (type) {
case UTF_8: case UTF_8:
if (c < 128) break; // optimization
l = mbrtowc_utf8(&wc, *in, *inb, &st); l = mbrtowc_utf8(&wc, *in, *inb, &st);
if (!l) l++; if (!l) l++;
else if (l == (size_t)-1) goto ilseq; else if (l == (size_t)-1) goto ilseq;
@ -202,7 +203,8 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
c = wc; c = wc;
break; break;
case US_ASCII: case US_ASCII:
goto ilseq; if (c >= 128) goto ilseq;
break;
case WCHAR_T: case WCHAR_T:
l = sizeof(wchar_t); l = sizeof(wchar_t);
if (*inb < l) goto starved; if (*inb < l) goto starved;
@ -234,6 +236,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
} }
break; break;
case SHIFT_JIS: case SHIFT_JIS:
if (c < 128) break;
if (c-0xa1 <= 0xdf-0xa1) { if (c-0xa1 <= 0xdf-0xa1) {
c += 0xff61-0xa1; c += 0xff61-0xa1;
break; break;
@ -257,6 +260,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
if (!c) goto ilseq; if (!c) goto ilseq;
break; break;
case EUC_JP: case EUC_JP:
if (c < 128) break;
l = 2; l = 2;
if (*inb < 2) goto starved; if (*inb < 2) goto starved;
d = *((unsigned char *)*in + 1); d = *((unsigned char *)*in + 1);
@ -273,9 +277,11 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
if (!c) goto ilseq; if (!c) goto ilseq;
break; break;
case GB2312: case GB2312:
if (c < 128) break;
if (c < 0xa1) goto ilseq; if (c < 0xa1) goto ilseq;
case GBK: case GBK:
case GB18030: case GB18030:
if (c < 128) break;
c -= 0x81; c -= 0x81;
if (c >= 126) goto ilseq; if (c >= 126) goto ilseq;
l = 2; l = 2;
@ -311,6 +317,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
c = gb18030[c][d]; c = gb18030[c][d];
break; break;
case BIG5: case BIG5:
if (c < 128) break;
l = 2; l = 2;
if (*inb < 2) goto starved; if (*inb < 2) goto starved;
d = *((unsigned char *)*in + 1); d = *((unsigned char *)*in + 1);
@ -348,6 +355,7 @@ size_t iconv(iconv_t cd, char **restrict in, size_t *restrict inb, char **restri
if (!c) goto ilseq; if (!c) goto ilseq;
break; break;
case EUC_KR: case EUC_KR:
if (c < 128) break;
l = 2; l = 2;
if (*inb < 2) goto starved; if (*inb < 2) goto starved;
d = *((unsigned char *)*in + 1); d = *((unsigned char *)*in + 1);