mirrors/cpython
1
0
Fork 0
mirror of https://github.com/python/cpython.git synced 2025-08-30 05:35:08 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions 8

Change %s argument for PyUnicode_FromFormat to

Browse source 
be UTF-8. Fixes #1070.
This commit is contained in:
Martin v. Löwis 2007-08-31 11:01:23 +00:00
parent bef9d21d14
commit 90d1fcd101
1 changed files with 46 additions and 2 deletions
Download patch file Download diff file Expand all files Collapse all files

48
Objects/unicodeobject.c
View file

@ -621,8 +621,39 @@ PyUnicode_FromFormatV(const char *format, va_list vargs)
abuffersize = width; abuffersize = width;
break; break;
case 's': case 's':
n += strlen(va_arg(count, char*)); {
/* UTF-8 */
unsigned char*s;
s = va_arg(count, unsigned char*);
while (*s) {
if (*s < 128) {
n++; s++;
} else if (*s < 0xc0) {
/* invalid UTF-8 */
n++; s++;
} else if (*s < 0xc0) {
n++;
s++; if(!*s)break;
s++;
} else if (*s < 0xe0) {
n++;
s++; if(!*s)break;
s++; if(!*s)break;
s++;
} else {
#ifdef Py_UNICODE_WIDE
n++;
#else
n+=2;
#endif
s++; if(!*s)break;
s++; if(!*s)break;
s++; if(!*s)break;
s++;
}
}
break; break;
}
case 'U': case 'U':
{ {
PyObject *obj = va_arg(count, PyObject *); PyObject *obj = va_arg(count, PyObject *);
@ -775,9 +806,22 @@ PyUnicode_FromFormatV(const char *format, va_list vargs)
appendstring(realbuffer); appendstring(realbuffer);
break; break;
case 's': case 's':
{
/* Parameter must be UTF-8 encoded.
In case of encoding errors, use
the replacement character. */
PyObject *u;
p = va_arg(vargs, char*); p = va_arg(vargs, char*);
appendstring(p); u = PyUnicode_DecodeUTF8(p, strlen(p),
"replace");
if (!u)
goto fail;
Py_UNICODE_COPY(s, PyUnicode_AS_UNICODE(u),
PyUnicode_GET_SIZE(u));
s += PyUnicode_GET_SIZE(u);
Py_DECREF(u);
break; break;
}
case 'U': case 'U':
{ {
PyObject *obj = va_arg(vargs, PyObject *); PyObject *obj = va_arg(vargs, PyObject *);