diff options
| author | Ammar Askar <ammar_askar@hotmail.com> | 2018-09-24 21:12:49 (GMT) |
|---|---|---|
| committer | Guido van Rossum <guido@python.org> | 2018-09-24 21:12:49 (GMT) |
| commit | 025eb98dc0c1dc27404df6c544fc2944e0fa9f3a (patch) | |
| tree | ad93cb6963abd43430766fa85b629b2d5896889b /Parser | |
| parent | 223e501fb9c2b6ae21b96054e20c4c31d94a5d96 (diff) | |
| download | cpython-025eb98dc0c1dc27404df6c544fc2944e0fa9f3a.zip cpython-025eb98dc0c1dc27404df6c544fc2944e0fa9f3a.tar.gz cpython-025eb98dc0c1dc27404df6c544fc2944e0fa9f3a.tar.bz2 | |
bpo-34683: Make SyntaxError column offsets consistently 1-indexed (gh-9338)
Also point to start of tokens in parsing errors.
Fixes bpo-34683
Diffstat (limited to 'Parser')
| -rw-r--r-- | Parser/parsetok.c | 8 |
1 files changed, 6 insertions, 2 deletions
diff --git a/Parser/parsetok.c b/Parser/parsetok.c index a1580e6..fc878d8 100644 --- a/Parser/parsetok.c +++ b/Parser/parsetok.c @@ -187,6 +187,7 @@ parsetok(struct tok_state *tok, grammar *g, int start, perrdetail *err_ret, parser_state *ps; node *n; int started = 0; + int col_offset; if ((ps = PyParser_New(g, start)) == NULL) { err_ret->error = E_NOMEM; @@ -203,7 +204,7 @@ parsetok(struct tok_state *tok, grammar *g, int start, perrdetail *err_ret, int type; size_t len; char *str; - int col_offset; + col_offset = -1; type = PyTokenizer_Get(tok, &a, &b); if (type == ERRORTOKEN) { @@ -321,7 +322,10 @@ parsetok(struct tok_state *tok, grammar *g, int start, perrdetail *err_ret, if (tok->buf != NULL) { size_t len; assert(tok->cur - tok->buf < INT_MAX); - err_ret->offset = (int)(tok->cur - tok->buf); + /* if we've managed to parse a token, point the offset to its start, + * else use the current reading position of the tokenizer + */ + err_ret->offset = col_offset != -1 ? col_offset + 1 : ((int)(tok->cur - tok->buf)); len = tok->inp - tok->buf; err_ret->text = (char *) PyObject_MALLOC(len + 1); if (err_ret->text != NULL) { |