Check newly created consistency using _PyUnicode_CheckConsistency(str, 1)

* In debug mode, fill the string data with invalid characters
 * Simplify also reference counting in PyCodec_BackslashReplaceErrors()
   and PyCodec_XMLCharRefReplaceError()
This commit is contained in:
Victor Stinner 2012-04-27 13:55:39 +02:00
parent 990eff0776
commit 8f825060f1
10 changed files with 31 additions and 14 deletions

View File

@ -246,6 +246,7 @@ ascii_escape_unicode(PyObject *pystr)
} }
} }
output[chars++] = '"'; output[chars++] = '"';
assert(_PyUnicode_CheckConsistency(rval, 1));
return rval; return rval;
} }

View File

@ -397,6 +397,7 @@ MD5_hexdigest(MD5object *self, PyObject *unused)
c = (digest[i] & 0xf); c = (digest[i] & 0xf);
hex_digest[j++] = Py_hexdigits[c]; hex_digest[j++] = Py_hexdigits[c];
} }
assert(_PyUnicode_CheckConsistency(retval, 1));
return retval; return retval;
} }

View File

@ -373,6 +373,7 @@ SHA1_hexdigest(SHA1object *self, PyObject *unused)
c = (digest[i] & 0xf); c = (digest[i] & 0xf);
hex_digest[j++] = Py_hexdigits[c]; hex_digest[j++] = Py_hexdigits[c];
} }
assert(_PyUnicode_CheckConsistency(retval, 1));
return retval; return retval;
} }

View File

@ -466,6 +466,7 @@ SHA256_hexdigest(SHAobject *self, PyObject *unused)
c = (digest[i] & 0xf); c = (digest[i] & 0xf);
hex_digest[j++] = Py_hexdigits[c]; hex_digest[j++] = Py_hexdigits[c];
} }
assert(_PyUnicode_CheckConsistency(retval, 1));
return retval; return retval;
} }

View File

@ -532,6 +532,7 @@ SHA512_hexdigest(SHAobject *self, PyObject *unused)
c = (digest[i] & 0xf); c = (digest[i] & 0xf);
hex_digest[j++] = Py_hexdigits[c]; hex_digest[j++] = Py_hexdigits[c];
} }
assert(_PyUnicode_CheckConsistency(retval, 1));
return retval; return retval;
} }

View File

@ -626,6 +626,7 @@ PyBytes_Repr(PyObject *obj, int smartquotes)
*p++ = c; *p++ = c;
} }
*p++ = quote; *p++ = quote;
assert(_PyUnicode_CheckConsistency(v, 1));
return v; return v;
} }

View File

@ -967,7 +967,7 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
PyObject *obj; PyObject *obj;
PyCompactUnicodeObject *unicode; PyCompactUnicodeObject *unicode;
void *data; void *data;
int kind_state; enum PyUnicode_Kind kind;
int is_sharing, is_ascii; int is_sharing, is_ascii;
Py_ssize_t char_size; Py_ssize_t char_size;
Py_ssize_t struct_size; Py_ssize_t struct_size;
@ -986,17 +986,17 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
is_sharing = 0; is_sharing = 0;
struct_size = sizeof(PyCompactUnicodeObject); struct_size = sizeof(PyCompactUnicodeObject);
if (maxchar < 128) { if (maxchar < 128) {
kind_state = PyUnicode_1BYTE_KIND; kind = PyUnicode_1BYTE_KIND;
char_size = 1; char_size = 1;
is_ascii = 1; is_ascii = 1;
struct_size = sizeof(PyASCIIObject); struct_size = sizeof(PyASCIIObject);
} }
else if (maxchar < 256) { else if (maxchar < 256) {
kind_state = PyUnicode_1BYTE_KIND; kind = PyUnicode_1BYTE_KIND;
char_size = 1; char_size = 1;
} }
else if (maxchar < 65536) { else if (maxchar < 65536) {
kind_state = PyUnicode_2BYTE_KIND; kind = PyUnicode_2BYTE_KIND;
char_size = 2; char_size = 2;
if (sizeof(wchar_t) == 2) if (sizeof(wchar_t) == 2)
is_sharing = 1; is_sharing = 1;
@ -1007,7 +1007,7 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
"invalid maximum character passed to PyUnicode_New"); "invalid maximum character passed to PyUnicode_New");
return NULL; return NULL;
} }
kind_state = PyUnicode_4BYTE_KIND; kind = PyUnicode_4BYTE_KIND;
char_size = 4; char_size = 4;
if (sizeof(wchar_t) == 4) if (sizeof(wchar_t) == 4)
is_sharing = 1; is_sharing = 1;
@ -1041,7 +1041,7 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
_PyUnicode_LENGTH(unicode) = size; _PyUnicode_LENGTH(unicode) = size;
_PyUnicode_HASH(unicode) = -1; _PyUnicode_HASH(unicode) = -1;
_PyUnicode_STATE(unicode).interned = 0; _PyUnicode_STATE(unicode).interned = 0;
_PyUnicode_STATE(unicode).kind = kind_state; _PyUnicode_STATE(unicode).kind = kind;
_PyUnicode_STATE(unicode).compact = 1; _PyUnicode_STATE(unicode).compact = 1;
_PyUnicode_STATE(unicode).ready = 1; _PyUnicode_STATE(unicode).ready = 1;
_PyUnicode_STATE(unicode).ascii = is_ascii; _PyUnicode_STATE(unicode).ascii = is_ascii;
@ -1049,19 +1049,19 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
((char*)data)[size] = 0; ((char*)data)[size] = 0;
_PyUnicode_WSTR(unicode) = NULL; _PyUnicode_WSTR(unicode) = NULL;
} }
else if (kind_state == PyUnicode_1BYTE_KIND) { else if (kind == PyUnicode_1BYTE_KIND) {
((char*)data)[size] = 0; ((char*)data)[size] = 0;
_PyUnicode_WSTR(unicode) = NULL; _PyUnicode_WSTR(unicode) = NULL;
_PyUnicode_WSTR_LENGTH(unicode) = 0; _PyUnicode_WSTR_LENGTH(unicode) = 0;
unicode->utf8 = NULL; unicode->utf8 = NULL;
unicode->utf8_length = 0; unicode->utf8_length = 0;
} }
else { else {
unicode->utf8 = NULL; unicode->utf8 = NULL;
unicode->utf8_length = 0; unicode->utf8_length = 0;
if (kind_state == PyUnicode_2BYTE_KIND) if (kind == PyUnicode_2BYTE_KIND)
((Py_UCS2*)data)[size] = 0; ((Py_UCS2*)data)[size] = 0;
else /* kind_state == PyUnicode_4BYTE_KIND */ else /* kind == PyUnicode_4BYTE_KIND */
((Py_UCS4*)data)[size] = 0; ((Py_UCS4*)data)[size] = 0;
if (is_sharing) { if (is_sharing) {
_PyUnicode_WSTR_LENGTH(unicode) = size; _PyUnicode_WSTR_LENGTH(unicode) = size;
@ -1072,6 +1072,13 @@ PyUnicode_New(Py_ssize_t size, Py_UCS4 maxchar)
_PyUnicode_WSTR(unicode) = NULL; _PyUnicode_WSTR(unicode) = NULL;
} }
} }
#ifdef Py_DEBUG
/* Fill the data with invalid characters to detect bugs earlier.
_PyUnicode_CheckConsistency(str, 1) detects invalid characters,
at least for ASCII and UCS-4 strings. U+00FF is invalid in ASCII
and U+FFFFFFFF is an invalid character in Unicode 6.0. */
memset(data, 0xff, size * kind);
#endif
assert(_PyUnicode_CheckConsistency((PyObject*)unicode, 0)); assert(_PyUnicode_CheckConsistency((PyObject*)unicode, 0));
return obj; return obj;
} }

View File

@ -534,6 +534,7 @@ PyObject *PyCodec_ReplaceErrors(PyObject *exc)
data = PyUnicode_DATA(res); data = PyUnicode_DATA(res);
for (i = 0; i < len; ++i) for (i = 0; i < len; ++i)
PyUnicode_WRITE(kind, data, i, '?'); PyUnicode_WRITE(kind, data, i, '?');
assert(_PyUnicode_CheckConsistency(res, 1));
return Py_BuildValue("(Nn)", res, end); return Py_BuildValue("(Nn)", res, end);
} }
else if (PyObject_IsInstance(exc, PyExc_UnicodeDecodeError)) { else if (PyObject_IsInstance(exc, PyExc_UnicodeDecodeError)) {
@ -559,6 +560,7 @@ PyObject *PyCodec_ReplaceErrors(PyObject *exc)
data = PyUnicode_DATA(res); data = PyUnicode_DATA(res);
for (i=0; i < len; i++) for (i=0; i < len; i++)
PyUnicode_WRITE(kind, data, i, Py_UNICODE_REPLACEMENT_CHARACTER); PyUnicode_WRITE(kind, data, i, Py_UNICODE_REPLACEMENT_CHARACTER);
assert(_PyUnicode_CheckConsistency(res, 1));
return Py_BuildValue("(Nn)", res, end); return Py_BuildValue("(Nn)", res, end);
} }
else { else {
@ -652,8 +654,8 @@ PyObject *PyCodec_XMLCharRefReplaceErrors(PyObject *exc)
} }
*outp++ = ';'; *outp++ = ';';
} }
restuple = Py_BuildValue("(On)", res, end); assert(_PyUnicode_CheckConsistency(res, 1));
Py_DECREF(res); restuple = Py_BuildValue("(Nn)", res, end);
Py_DECREF(object); Py_DECREF(object);
return restuple; return restuple;
} }
@ -720,8 +722,8 @@ PyObject *PyCodec_BackslashReplaceErrors(PyObject *exc)
*outp++ = Py_hexdigits[c&0xf]; *outp++ = Py_hexdigits[c&0xf];
} }
restuple = Py_BuildValue("(On)", res, end); assert(_PyUnicode_CheckConsistency(res, 1));
Py_DECREF(res); restuple = Py_BuildValue("(Nn)", res, end);
Py_DECREF(object); Py_DECREF(object);
return restuple; return restuple;
} }

View File

@ -263,6 +263,7 @@ _Py_Mangle(PyObject *privateobj, PyObject *ident)
Py_DECREF(result); Py_DECREF(result);
return NULL; return NULL;
} }
assert(_PyUnicode_CheckConsistency(result, 1));
return result; return result;
} }

View File

@ -992,6 +992,7 @@ make_source_pathname(PyObject *path)
(j = dot0-right)); (j = dot0-right));
PyUnicode_WRITE(kind, data, i+j, 'p'); PyUnicode_WRITE(kind, data, i+j, 'p');
PyUnicode_WRITE(kind, data, i+j+1, 'y'); PyUnicode_WRITE(kind, data, i+j+1, 'y');
assert(_PyUnicode_CheckConsistency(result, 1));
return result; return result;
} }