Skip to content

Commit

Permalink
bpo-40334: Fix error location upon parsing an invalid string literal
Browse files Browse the repository at this point in the history 
When parsing a string with an invalid escape, the old parser used to
point to the beginning of the invalid string. This PR changes the new
parser to match that behavior, since it's currently pointing to the
end of the string (or to be more precise, to the beginning of the next
token).

Closes we-like-parsers#64.
  • Loading branch information
@lysnikolaou
lysnikolaou committed May 6, 2020
1 parent 091951a commit 8300034
Show file tree
Hide file tree
Showing 4 changed files with 24 additions and 19 deletions.
6 changes: 2 additions & 4 deletions Lib/test/test_string_literals.py
View file
Original file line number Diff line number Diff line change
Expand Up @@ -118,8 +118,7 @@ def test_eval_str_invalid_escape(self):
eval("'''\n\\z'''")
self.assertEqual(len(w), 1)
self.assertEqual(w[0].filename, '<string>')
if use_old_parser():
self.assertEqual(w[0].lineno, 1)
self.assertEqual(w[0].lineno, 1)

with warnings.catch_warnings(record=True) as w:
warnings.simplefilter('error', category=DeprecationWarning)
Expand All @@ -128,8 +127,7 @@ def test_eval_str_invalid_escape(self):
exc = cm.exception
self.assertEqual(w, [])
self.assertEqual(exc.filename, '<string>')
if use_old_parser():
self.assertEqual(exc.lineno, 1)
self.assertEqual(exc.lineno, 1)

def test_eval_str_raw(self):
self.assertEqual(eval(""" r'x' """), 'x')
Expand Down
33 changes: 20 additions & 13 deletions Parser/pegen/parse_string.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -12,20 +12,27 @@
// file (like "_PyPegen_raise_syntax_error").

static int
warn_invalid_escape_sequence(Parser *p, unsigned char first_invalid_escape_char)
warn_invalid_escape_sequence(Parser *p, unsigned char first_invalid_escape_char, Token *t)
{
PyObject *msg =
PyUnicode_FromFormat("invalid escape sequence \\%c", first_invalid_escape_char);
if (msg == NULL) {
return -1;
}
if (PyErr_WarnExplicitObject(PyExc_DeprecationWarning, msg, p->tok->filename,
p->tok->lineno, NULL, NULL) < 0) {
t->lineno, NULL, NULL) < 0) {
if (PyErr_ExceptionMatches(PyExc_DeprecationWarning)) {
/* Replace the DeprecationWarning exception with a SyntaxError
to get a more accurate error report */
PyErr_Clear();

/* This is a hack, in order for the SyntaxError to point to the token t,
since _PyPegen_raise_error always uses p->tokens[p->fill - 1] for the
error location. */
Token *old_end = p->tokens[p->fill - 1];
p->tokens[p->fill - 1] = t;
RAISE_SYNTAX_ERROR("invalid escape sequence \\%c", first_invalid_escape_char);
p->tokens[p->fill - 1] = old_end;
}
Py_DECREF(msg);
return -1;
Expand All @@ -47,7 +54,7 @@ decode_utf8(const char **sPtr, const char *end)
}

static PyObject *
decode_unicode_with_escapes(Parser *parser, const char *s, size_t len)
decode_unicode_with_escapes(Parser *parser, const char *s, size_t len, Token *t)
{
PyObject *v, *u;
char *buf;
Expand Down Expand Up @@ -110,7 +117,7 @@ decode_unicode_with_escapes(Parser *parser, const char *s, size_t len)
v = _PyUnicode_DecodeUnicodeEscape(s, len, NULL, &first_invalid_escape);

if (v != NULL && first_invalid_escape != NULL) {
if (warn_invalid_escape_sequence(parser, *first_invalid_escape) < 0) {
if (warn_invalid_escape_sequence(parser, *first_invalid_escape, t) < 0) {
/* We have not decref u before because first_invalid_escape points
inside u. */
Py_XDECREF(u);
Expand All @@ -123,7 +130,7 @@ decode_unicode_with_escapes(Parser *parser, const char *s, size_t len)
}

static PyObject *
decode_bytes_with_escapes(Parser *p, const char *s, Py_ssize_t len)
decode_bytes_with_escapes(Parser *p, const char *s, Py_ssize_t len, Token *t)
{
const char *first_invalid_escape;
PyObject *result = _PyBytes_DecodeEscape(s, len, NULL, &first_invalid_escape);
Expand All @@ -132,7 +139,7 @@ decode_bytes_with_escapes(Parser *p, const char *s, Py_ssize_t len)
}

if (first_invalid_escape != NULL) {
if (warn_invalid_escape_sequence(p, *first_invalid_escape) < 0) {
if (warn_invalid_escape_sequence(p, *first_invalid_escape, t) < 0) {
Py_DECREF(result);
return NULL;
}
Expand All @@ -147,7 +154,7 @@ decode_bytes_with_escapes(Parser *p, const char *s, Py_ssize_t len)
string object. Return 0 if no errors occurred. */
int
_PyPegen_parsestr(Parser *p, const char *s, int *bytesmode, int *rawmode, PyObject **result,
const char **fstr, Py_ssize_t *fstrlen)
const char **fstr, Py_ssize_t *fstrlen, Token *t)
{
size_t len;
int quote = Py_CHARMASK(*s);
Expand Down Expand Up @@ -245,15 +252,15 @@ _PyPegen_parsestr(Parser *p, const char *s, int *bytesmode, int *rawmode, PyObje
*result = PyBytes_FromStringAndSize(s, len);
}
else {
*result = decode_bytes_with_escapes(p, s, len);
*result = decode_bytes_with_escapes(p, s, len, t);
}
}
else {
if (*rawmode) {
*result = PyUnicode_DecodeUTF8Stateful(s, len, NULL, NULL);
}
else {
*result = decode_unicode_with_escapes(p, s, len);
*result = decode_unicode_with_escapes(p, s, len, t);
}
}
return *result == NULL ? -1 : 0;
Expand Down Expand Up @@ -637,7 +644,7 @@ fstring_compile_expr(Parser *p, const char *expr_start, const char *expr_end,
*/
static int
fstring_find_literal(Parser *p, const char **str, const char *end, int raw,
PyObject **literal, int recurse_lvl)
PyObject **literal, int recurse_lvl, Token *t)
{
/* Get any literal string. It ends when we hit an un-doubled left
brace (which isn't part of a unicode name escape such as
Expand All @@ -660,7 +667,7 @@ fstring_find_literal(Parser *p, const char **str, const char *end, int raw,
}
break;
}
if (ch == '{' && warn_invalid_escape_sequence(p, ch) < 0) {
if (ch == '{' && warn_invalid_escape_sequence(p, ch, t) < 0) {
return -1;
}
}
Expand Down Expand Up @@ -704,7 +711,7 @@ fstring_find_literal(Parser *p, const char **str, const char *end, int raw,
NULL, NULL);
else
*literal = decode_unicode_with_escapes(p, literal_start,
s - literal_start);
s - literal_start, t);
if (!*literal)
return -1;
}
Expand Down Expand Up @@ -1041,7 +1048,7 @@ fstring_find_literal_and_expr(Parser *p, const char **str, const char *end, int
assert(*literal == NULL && *expression == NULL);

/* Get any literal string. */
result = fstring_find_literal(p, str, end, raw, literal, recurse_lvl);
result = fstring_find_literal(p, str, end, raw, literal, recurse_lvl, t);
if (result < 0)
goto error;

Expand Down
2 changes: 1 addition & 1 deletion Parser/pegen/parse_string.h
View file
Original file line number Diff line number Diff line change
Expand Up @@ -35,7 +35,7 @@ typedef struct {

void _PyPegen_FstringParser_Init(FstringParser *);
int _PyPegen_parsestr(Parser *, const char *, int *, int *, PyObject **,
const char **, Py_ssize_t *);
const char **, Py_ssize_t *, Token *);
int _PyPegen_FstringParser_ConcatFstring(Parser *, FstringParser *, const char **,
const char *, int, int, Token *, Token *,
Token *);
Expand Down
2 changes: 1 addition & 1 deletion Parser/pegen/pegen.c
View file
Original file line number Diff line number Diff line change
Expand Up @@ -1977,7 +1977,7 @@ _PyPegen_concatenate_strings(Parser *p, asdl_seq *strings)
goto error;
}

if (_PyPegen_parsestr(p, this_str, &this_bytesmode, &this_rawmode, &s, &fstr, &fstrlen) != 0) {
if (_PyPegen_parsestr(p, this_str, &this_bytesmode, &this_rawmode, &s, &fstr, &fstrlen, t) != 0) {
goto error;
}

Expand Down

0 comments on commit 8300034

Please sign in to comment.