Repository navigation
Expand file tree
/
Copy pathstring_parser.c
More file actions
1448 lines (1286 loc) · 47.3 KB
/
Copy pathstring_parser.c
File metadata and controls
1448 lines (1286 loc) · 47.3 KB
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
#include <Python.h>
#include "tokenizer.h"
#include "pegen.h"
#include "string_parser.h"
//// STRING HANDLING FUNCTIONS ////
static int
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,
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 needed, in order for the SyntaxError to point to the token t,
since _PyPegen_raise_error uses p->tokens[p->fill - 1] for the
error location, if p->known_err_token is not set. */
p->known_err_token = t;
RAISE_SYNTAX_ERROR("invalid escape sequence \\%c", first_invalid_escape_char);
}
Py_DECREF(msg);
return -1;
}
Py_DECREF(msg);
return 0;
}
static PyObject *
decode_utf8(const char **sPtr, const char *end)
{
const char *s;
const char *t;
t = s = *sPtr;
while (s < end && (*s & 0x80)) {
s++;
}
*sPtr = s;
return PyUnicode_DecodeUTF8(t, s - t, NULL);
}
static PyObject *
decode_unicode_with_escapes(Parser *parser, const char *s, size_t len, Token *t)
{
PyObject *v;
PyObject *u;
char *buf;
char *p;
const char *end;
/* check for integer overflow */
if (len > SIZE_MAX / 6) {
return NULL;
}
/* "ä" (2 bytes) may become "\U000000E4" (10 bytes), or 1:5
"\ä" (3 bytes) may become "\u005c\U000000E4" (16 bytes), or ~1:6 */
u = PyBytes_FromStringAndSize((char *)NULL, len * 6);
if (u == NULL) {
return NULL;
}
p = buf = PyBytes_AsString(u);
end = s + len;
while (s < end) {
if (*s == '\\') {
*p++ = *s++;
if (s >= end || *s & 0x80) {
strcpy(p, "u005c");
p += 5;
if (s >= end) {
break;
}
}
}
if (*s & 0x80) {
PyObject *w;
int kind;
void *data;
Py_ssize_t w_len;
Py_ssize_t i;
w = decode_utf8(&s, end);
if (w == NULL) {
Py_DECREF(u);
return NULL;
}
kind = PyUnicode_KIND(w);
data = PyUnicode_DATA(w);
w_len = PyUnicode_GET_LENGTH(w);
for (i = 0; i < w_len; i++) {
Py_UCS4 chr = PyUnicode_READ(kind, data, i);
sprintf(p, "\\U%08x", chr);
p += 10;
}
/* Should be impossible to overflow */
assert(p - buf <= PyBytes_GET_SIZE(u));
Py_DECREF(w);
}
else {
*p++ = *s++;
}
}
len = p - buf;
s = buf;
const char *first_invalid_escape;
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, t) < 0) {
/* We have not decref u before because first_invalid_escape points
inside u. */
Py_XDECREF(u);
Py_DECREF(v);
return NULL;
}
}
Py_XDECREF(u);
return v;
}
static PyObject *
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);
if (result == NULL) {
return NULL;
}
if (first_invalid_escape != NULL) {
if (warn_invalid_escape_sequence(p, *first_invalid_escape, t) < 0) {
Py_DECREF(result);
return NULL;
}
}
return result;
}
/* s must include the bracketing quote characters, and r, b, u,
&/or f prefixes (if any), and embedded escape sequences (if any).
_PyPegen_parsestr parses it, and sets *result to decoded Python string object.
If the string is an f-string, set *fstr and *fstrlen to the unparsed
string object. Return 0 if no errors occurred. */
int
_PyPegen_parsestr(Parser *p, int *bytesmode, int *rawmode, PyObject **result,
const char **fstr, Py_ssize_t *fstrlen, Token *t)
{
const char *s = PyBytes_AsString(t->bytes);
if (s == NULL) {
return -1;
}
size_t len;
int quote = Py_CHARMASK(*s);
int fmode = 0;
*bytesmode = 0;
*rawmode = 0;
*result = NULL;
*fstr = NULL;
if (Py_ISALPHA(quote)) {
while (!*bytesmode || !*rawmode) {
if (quote == 'b' || quote == 'B') {
quote =(unsigned char)*++s;
*bytesmode = 1;
}
else if (quote == 'u' || quote == 'U') {
quote = (unsigned char)*++s;
}
else if (quote == 'r' || quote == 'R') {
quote = (unsigned char)*++s;
*rawmode = 1;
}
else if (quote == 'f' || quote == 'F') {
quote = (unsigned char)*++s;
fmode = 1;
}
else {
break;
}
}
}
/* fstrings are only allowed in Python 3.6 and greater */
if (fmode && p->feature_version < 6) {
p->error_indicator = 1;
RAISE_SYNTAX_ERROR("Format strings are only supported in Python 3.6 and greater");
return -1;
}
if (fmode && *bytesmode) {
PyErr_BadInternalCall();
return -1;
}
if (quote != '\'' && quote != '\"') {
PyErr_BadInternalCall();
return -1;
}
/* Skip the leading quote char. */
s++;
len = strlen(s);
if (len > INT_MAX) {
PyErr_SetString(PyExc_OverflowError, "string to parse is too long");
return -1;
}
if (s[--len] != quote) {
/* Last quote char must match the first. */
PyErr_BadInternalCall();
return -1;
}
if (len >= 4 && s[0] == quote && s[1] == quote) {
/* A triple quoted string. We've already skipped one quote at
the start and one at the end of the string. Now skip the
two at the start. */
s += 2;
len -= 2;
/* And check that the last two match. */
if (s[--len] != quote || s[--len] != quote) {
PyErr_BadInternalCall();
return -1;
}
}
if (fmode) {
/* Just return the bytes. The caller will parse the resulting
string. */
*fstr = s;
*fstrlen = len;
return 0;
}
/* Not an f-string. */
/* Avoid invoking escape decoding routines if possible. */
*rawmode = *rawmode || strchr(s, '\\') == NULL;
if (*bytesmode) {
/* Disallow non-ASCII characters. */
const char *ch;
for (ch = s; *ch; ch++) {
if (Py_CHARMASK(*ch) >= 0x80) {
RAISE_SYNTAX_ERROR(
"bytes can only contain ASCII "
"literal characters.");
return -1;
}
}
if (*rawmode) {
*result = PyBytes_FromStringAndSize(s, len);
}
else {
*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, t);
}
}
return *result == NULL ? -1 : 0;
}
// FSTRING STUFF
static void fstring_shift_expr_locations(expr_ty n, int lineno, int col_offset);
static void fstring_shift_argument(expr_ty parent, arg_ty args, int lineno, int col_offset);
static inline void shift_expr(expr_ty parent, expr_ty n, int line, int col) {
if (n == NULL) {
return;
}
if (parent->lineno < n->lineno) {
col = 0;
}
fstring_shift_expr_locations(n, line, col);
}
static inline void shift_arg(expr_ty parent, arg_ty n, int line, int col) {
if (parent->lineno < n->lineno) {
col = 0;
}
fstring_shift_argument(parent, n, line, col);
}
static void fstring_shift_seq_locations(expr_ty parent, asdl_seq *seq, int lineno, int col_offset) {
for (Py_ssize_t i = 0, l = asdl_seq_LEN(seq); i < l; i++) {
expr_ty expr = asdl_seq_GET(seq, i);
if (expr == NULL){
continue;
}
shift_expr(parent, expr, lineno, col_offset);
}
}
static void fstring_shift_slice_locations(expr_ty parent, expr_ty slice, int lineno, int col_offset) {
switch (slice->kind) {
case Slice_kind:
if (slice->v.Slice.lower) {
shift_expr(parent, slice->v.Slice.lower, lineno, col_offset);
}
if (slice->v.Slice.upper) {
shift_expr(parent, slice->v.Slice.upper, lineno, col_offset);
}
if (slice->v.Slice.step) {
shift_expr(parent, slice->v.Slice.step, lineno, col_offset);
}
break;
case Tuple_kind:
fstring_shift_seq_locations(parent, slice->v.Tuple.elts, lineno, col_offset);
break;
default:
break;
}
}
static void fstring_shift_comprehension(expr_ty parent, comprehension_ty comp, int lineno, int col_offset) {
shift_expr(parent, comp->target, lineno, col_offset);
shift_expr(parent, comp->iter, lineno, col_offset);
fstring_shift_seq_locations(parent, comp->ifs, lineno, col_offset);
}
static void fstring_shift_argument(expr_ty parent, arg_ty arg, int lineno, int col_offset) {
if (arg->annotation != NULL){
shift_expr(parent, arg->annotation, lineno, col_offset);
}
arg->col_offset = arg->col_offset + col_offset;
arg->end_col_offset = arg->end_col_offset + col_offset;
arg->lineno = arg->lineno + lineno;
arg->end_lineno = arg->end_lineno + lineno;
}
static void fstring_shift_arguments(expr_ty parent, arguments_ty args, int lineno, int col_offset) {
for (Py_ssize_t i = 0, l = asdl_seq_LEN(args->posonlyargs); i < l; i++) {
arg_ty arg = asdl_seq_GET(args->posonlyargs, i);
shift_arg(parent, arg, lineno, col_offset);
}
for (Py_ssize_t i = 0, l = asdl_seq_LEN(args->args); i < l; i++) {
arg_ty arg = asdl_seq_GET(args->args, i);
shift_arg(parent, arg, lineno, col_offset);
}
if (args->vararg != NULL) {
shift_arg(parent, args->vararg, lineno, col_offset);
}
for (Py_ssize_t i = 0, l = asdl_seq_LEN(args->kwonlyargs); i < l; i++) {
arg_ty arg = asdl_seq_GET(args->kwonlyargs, i);
shift_arg(parent, arg, lineno, col_offset);
}
fstring_shift_seq_locations(parent, args->kw_defaults, lineno, col_offset);
if (args->kwarg != NULL) {
shift_arg(parent, args->kwarg, lineno, col_offset);
}
fstring_shift_seq_locations(parent, args->defaults, lineno, col_offset);
}
static void fstring_shift_children_locations(expr_ty node, int lineno, int col_offset) {
switch (node->kind) {
case BoolOp_kind:
fstring_shift_seq_locations(node, node->v.BoolOp.values, lineno, col_offset);
break;
case NamedExpr_kind:
shift_expr(node, node->v.NamedExpr.target, lineno, col_offset);
shift_expr(node, node->v.NamedExpr.value, lineno, col_offset);
break;
case BinOp_kind:
shift_expr(node, node->v.BinOp.left, lineno, col_offset);
shift_expr(node, node->v.BinOp.right, lineno, col_offset);
break;
case UnaryOp_kind:
shift_expr(node, node->v.UnaryOp.operand, lineno, col_offset);
break;
case Lambda_kind:
fstring_shift_arguments(node, node->v.Lambda.args, lineno, col_offset);
shift_expr(node, node->v.Lambda.body, lineno, col_offset);
break;
case IfExp_kind:
shift_expr(node, node->v.IfExp.test, lineno, col_offset);
shift_expr(node, node->v.IfExp.body, lineno, col_offset);
shift_expr(node, node->v.IfExp.orelse, lineno, col_offset);
break;
case Dict_kind:
fstring_shift_seq_locations(node, node->v.Dict.keys, lineno, col_offset);
fstring_shift_seq_locations(node, node->v.Dict.values, lineno, col_offset);
break;
case Set_kind:
fstring_shift_seq_locations(node, node->v.Set.elts, lineno, col_offset);
break;
case ListComp_kind:
shift_expr(node, node->v.ListComp.elt, lineno, col_offset);
for (Py_ssize_t i = 0, l = asdl_seq_LEN(node->v.ListComp.generators); i < l; i++) {
comprehension_ty comp = asdl_seq_GET(node->v.ListComp.generators, i);
fstring_shift_comprehension(node, comp, lineno, col_offset);
}
break;
case SetComp_kind:
shift_expr(node, node->v.SetComp.elt, lineno, col_offset);
for (Py_ssize_t i = 0, l = asdl_seq_LEN(node->v.SetComp.generators); i < l; i++) {
comprehension_ty comp = asdl_seq_GET(node->v.SetComp.generators, i);
fstring_shift_comprehension(node, comp, lineno, col_offset);
}
break;
case DictComp_kind:
shift_expr(node, node->v.DictComp.key, lineno, col_offset);
shift_expr(node, node->v.DictComp.value, lineno, col_offset);
for (Py_ssize_t i = 0, l = asdl_seq_LEN(node->v.DictComp.generators); i < l; i++) {
comprehension_ty comp = asdl_seq_GET(node->v.DictComp.generators, i);
fstring_shift_comprehension(node, comp, lineno, col_offset);
}
break;
case GeneratorExp_kind:
shift_expr(node, node->v.GeneratorExp.elt, lineno, col_offset);
for (Py_ssize_t i = 0, l = asdl_seq_LEN(node->v.GeneratorExp.generators); i < l; i++) {
comprehension_ty comp = asdl_seq_GET(node->v.GeneratorExp.generators, i);
fstring_shift_comprehension(node, comp, lineno, col_offset);
}
break;
case Await_kind:
shift_expr(node, node->v.Await.value, lineno, col_offset);
break;
case Yield_kind:
shift_expr(node, node->v.Yield.value, lineno, col_offset);
break;
case YieldFrom_kind:
shift_expr(node, node->v.YieldFrom.value, lineno, col_offset);
break;
case Compare_kind:
shift_expr(node, node->v.Compare.left, lineno, col_offset);
fstring_shift_seq_locations(node, node->v.Compare.comparators, lineno, col_offset);
break;
case Call_kind:
shift_expr(node, node->v.Call.func, lineno, col_offset);
fstring_shift_seq_locations(node, node->v.Call.args, lineno, col_offset);
for (Py_ssize_t i = 0, l = asdl_seq_LEN(node->v.Call.keywords); i < l; i++) {
keyword_ty keyword = asdl_seq_GET(node->v.Call.keywords, i);
shift_expr(node, keyword->value, lineno, col_offset);
}
break;
case Attribute_kind:
shift_expr(node, node->v.Attribute.value, lineno, col_offset);
break;
case Subscript_kind:
shift_expr(node, node->v.Subscript.value, lineno, col_offset);
fstring_shift_slice_locations(node, node->v.Subscript.slice, lineno, col_offset);
shift_expr(node, node->v.Subscript.slice, lineno, col_offset);
break;
case Starred_kind:
shift_expr(node, node->v.Starred.value, lineno, col_offset);
break;
case List_kind:
fstring_shift_seq_locations(node, node->v.List.elts, lineno, col_offset);
break;
case Tuple_kind:
fstring_shift_seq_locations(node, node->v.Tuple.elts, lineno, col_offset);
break;
case JoinedStr_kind:
fstring_shift_seq_locations(node, node->v.JoinedStr.values, lineno, col_offset);
break;
case FormattedValue_kind:
shift_expr(node, node->v.FormattedValue.value, lineno, col_offset);
if (node->v.FormattedValue.format_spec) {
shift_expr(node, node->v.FormattedValue.format_spec, lineno, col_offset);
}
break;
default:
return;
}
}
/* Shift locations for the given node and all its children by adding `lineno`
and `col_offset` to existing locations. Note that n is the already parsed
expression. */
static void fstring_shift_expr_locations(expr_ty n, int lineno, int col_offset)
{
n->col_offset = n->col_offset + col_offset;
// The following is needed, in order for nodes spanning across multiple lines
// to be shifted correctly. An example of such a node is a Call node, the closing
// parenthesis of which is not on the same line as its name.
if (n->lineno == n->end_lineno) {
n->end_col_offset = n->end_col_offset + col_offset;
}
fstring_shift_children_locations(n, lineno, col_offset);
n->lineno = n->lineno + lineno;
n->end_lineno = n->end_lineno + lineno;
}
/* Fix locations for the given node and its children.
`parent` is the enclosing node.
`n` is the node which locations are going to be fixed relative to parent.
`expr_str` is the child node's string representation, including braces.
*/
static void
fstring_fix_expr_location(Token *parent, expr_ty n, char *expr_str)
{
char *substr = NULL;
char *start;
int lines = 0;
int cols = 0;
if (parent && parent->bytes) {
char *parent_str = PyBytes_AsString(parent->bytes);
if (!parent_str) {
return;
}
substr = strstr(parent_str, expr_str);
if (substr) {
// The following is needed, in order to correctly shift the column
// offset, in the case that (disregarding any whitespace) a newline
// immediately follows the opening curly brace of the fstring expression.
int newline_after_brace = 1;
start = substr + 1;
while (start && *start != '}' && *start != '\n') {
if (*start != ' ' && *start != '\t' && *start != '\f') {
newline_after_brace = 0;
break;
}
start++;
}
// Account for the characters from the last newline character to our
// left until the beginning of substr.
if (!newline_after_brace) {
start = substr;
while (start > parent_str && *start != '\n') {
start--;
}
cols += (int)(substr - start);
}
/* adjust the start based on the number of newlines encountered
before the f-string expression */
for (char* p = parent_str; p < substr; p++) {
if (*p == '\n') {
lines++;
}
}
}
}
fstring_shift_expr_locations(n, lines, cols);
}
/* Compile this expression in to an expr_ty. Add parens around the
expression, in order to allow leading spaces in the expression. */
static expr_ty
fstring_compile_expr(Parser *p, const char *expr_start, const char *expr_end,
Token *t)
{
expr_ty expr = NULL;
char *str;
Py_ssize_t len;
const char *s;
expr_ty result = NULL;
assert(expr_end >= expr_start);
assert(*(expr_start-1) == '{');
assert(*expr_end == '}' || *expr_end == '!' || *expr_end == ':' ||
*expr_end == '=');
/* If the substring is all whitespace, it's an error. We need to catch this
here, and not when we call PyParser_SimpleParseStringFlagsFilename,
because turning the expression '' in to '()' would go from being invalid
to valid. */
for (s = expr_start; s != expr_end; s++) {
char c = *s;
/* The Python parser ignores only the following whitespace
characters (\r already is converted to \n). */
if (!(c == ' ' || c == '\t' || c == '\n' || c == '\f')) {
break;
}
}
if (s == expr_end) {
RAISE_SYNTAX_ERROR("f-string: empty expression not allowed");
return NULL;
}
len = expr_end - expr_start;
/* Allocate 3 extra bytes: open paren, close paren, null byte. */
str = PyMem_Malloc(len + 3);
if (str == NULL) {
PyErr_NoMemory();
return NULL;
}
str[0] = '(';
memcpy(str+1, expr_start, len);
str[len+1] = ')';
str[len+2] = 0;
struct tok_state* tok = PyTokenizer_FromString(str, 1);
if (tok == NULL) {
PyMem_Free(str);
return NULL;
}
Py_INCREF(p->tok->filename);
tok->filename = p->tok->filename;
Parser *p2 = _PyPegen_Parser_New(tok, Py_fstring_input, p->flags, p->feature_version,
NULL, p->arena);
p2->starting_lineno = p->starting_lineno + p->tok->first_lineno - 1;
p2->starting_col_offset = p->tok->first_lineno == p->tok->lineno
? p->starting_col_offset + t->col_offset : 0;
expr = _PyPegen_run_parser(p2);
if (expr == NULL) {
goto exit;
}
/* Reuse str to find the correct column offset. */
str[0] = '{';
str[len+1] = '}';
fstring_fix_expr_location(t, expr, str);
result = expr;
exit:
PyMem_Free(str);
_PyPegen_Parser_Free(p2);
PyTokenizer_Free(tok);
return result;
}
/* Return -1 on error.
Return 0 if we reached the end of the literal.
Return 1 if we haven't reached the end of the literal, but we want
the caller to process the literal up to this point. Used for
doubled braces.
*/
static int
fstring_find_literal(Parser *p, const char **str, const char *end, int raw,
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
"\N{EULER CONSTANT}"), or the end of the string. */
const char *s = *str;
const char *literal_start = s;
int result = 0;
assert(*literal == NULL);
while (s < end) {
char ch = *s++;
if (!raw && ch == '\\' && s < end) {
ch = *s++;
if (ch == 'N') {
if (s < end && *s++ == '{') {
while (s < end && *s++ != '}') {
}
continue;
}
break;
}
if (ch == '{' && warn_invalid_escape_sequence(p, ch, t) < 0) {
return -1;
}
}
if (ch == '{' || ch == '}') {
/* Check for doubled braces, but only at the top level. If
we checked at every level, then f'{0:{3}}' would fail
with the two closing braces. */
if (recurse_lvl == 0) {
if (s < end && *s == ch) {
/* We're going to tell the caller that the literal ends
here, but that they should continue scanning. But also
skip over the second brace when we resume scanning. */
*str = s + 1;
result = 1;
goto done;
}
/* Where a single '{' is the start of a new expression, a
single '}' is not allowed. */
if (ch == '}') {
*str = s - 1;
RAISE_SYNTAX_ERROR("f-string: single '}' is not allowed");
return -1;
}
}
/* We're either at a '{', which means we're starting another
expression; or a '}', which means we're at the end of this
f-string (for a nested format_spec). */
s--;
break;
}
}
*str = s;
assert(s <= end);
assert(s == end || *s == '{' || *s == '}');
done:
if (literal_start != s) {
if (raw) {
*literal = PyUnicode_DecodeUTF8Stateful(literal_start,
s - literal_start,
NULL, NULL);
} else {
*literal = decode_unicode_with_escapes(p, literal_start,
s - literal_start, t);
}
if (!*literal) {
return -1;
}
}
return result;
}
/* Forward declaration because parsing is recursive. */
static expr_ty
fstring_parse(Parser *p, const char **str, const char *end, int raw, int recurse_lvl,
Token *first_token, Token* t, Token *last_token);
/* Parse the f-string at *str, ending at end. We know *str starts an
expression (so it must be a '{'). Returns the FormattedValue node, which
includes the expression, conversion character, format_spec expression, and
optionally the text of the expression (if = is used).
Note that I don't do a perfect job here: I don't make sure that a
closing brace doesn't match an opening paren, for example. It
doesn't need to error on all invalid expressions, just correctly
find the end of all valid ones. Any errors inside the expression
will be caught when we parse it later.
*expression is set to the expression. For an '=' "debug" expression,
*expr_text is set to the debug text (the original text of the expression,
including the '=' and any whitespace around it, as a string object). If
not a debug expression, *expr_text set to NULL. */
static int
fstring_find_expr(Parser *p, const char **str, const char *end, int raw, int recurse_lvl,
PyObject **expr_text, expr_ty *expression, Token *first_token,
Token *t, Token *last_token)
{
/* Return -1 on error, else 0. */
const char *expr_start;
const char *expr_end;
expr_ty simple_expression;
expr_ty format_spec = NULL; /* Optional format specifier. */
int conversion = -1; /* The conversion char. Use default if not
specified, or !r if using = and no format
spec. */
/* 0 if we're not in a string, else the quote char we're trying to
match (single or double quote). */
char quote_char = 0;
/* If we're inside a string, 1=normal, 3=triple-quoted. */
int string_type = 0;
/* Keep track of nesting level for braces/parens/brackets in
expressions. */
Py_ssize_t nested_depth = 0;
char parenstack[MAXLEVEL];
*expr_text = NULL;
/* Can only nest one level deep. */
if (recurse_lvl >= 2) {
RAISE_SYNTAX_ERROR("f-string: expressions nested too deeply");
goto error;
}
/* The first char must be a left brace, or we wouldn't have gotten
here. Skip over it. */
assert(**str == '{');
*str += 1;
expr_start = *str;
for (; *str < end; (*str)++) {
char ch;
/* Loop invariants. */
assert(nested_depth >= 0);
assert(*str >= expr_start && *str < end);
if (quote_char) {
assert(string_type == 1 || string_type == 3);
} else {
assert(string_type == 0);
}
ch = **str;
/* Nowhere inside an expression is a backslash allowed. */
if (ch == '\\') {
/* Error: can't include a backslash character, inside
parens or strings or not. */
RAISE_SYNTAX_ERROR(
"f-string expression part "
"cannot include a backslash");
goto error;
}
if (quote_char) {
/* We're inside a string. See if we're at the end. */
/* This code needs to implement the same non-error logic
as tok_get from tokenizer.c, at the letter_quote
label. To actually share that code would be a
nightmare. But, it's unlikely to change and is small,
so duplicate it here. Note we don't need to catch all
of the errors, since they'll be caught when parsing the
expression. We just need to match the non-error
cases. Thus we can ignore \n in single-quoted strings,
for example. Or non-terminated strings. */
if (ch == quote_char) {
/* Does this match the string_type (single or triple
quoted)? */
if (string_type == 3) {
if (*str+2 < end && *(*str+1) == ch && *(*str+2) == ch) {
/* We're at the end of a triple quoted string. */
*str += 2;
string_type = 0;
quote_char = 0;
continue;
}
} else {
/* We're at the end of a normal string. */
quote_char = 0;
string_type = 0;
continue;
}
}
} else if (ch == '\'' || ch == '"') {
/* Is this a triple quoted string? */
if (*str+2 < end && *(*str+1) == ch && *(*str+2) == ch) {
string_type = 3;
*str += 2;
} else {
/* Start of a normal string. */
string_type = 1;
}
/* Start looking for the end of the string. */
quote_char = ch;
} else if (ch == '[' || ch == '{' || ch == '(') {
if (nested_depth >= MAXLEVEL) {
RAISE_SYNTAX_ERROR("f-string: too many nested parenthesis");
goto error;
}
parenstack[nested_depth] = ch;
nested_depth++;
} else if (ch == '#') {
/* Error: can't include a comment character, inside parens
or not. */
RAISE_SYNTAX_ERROR("f-string expression part cannot include '#'");
goto error;
} else if (nested_depth == 0 &&
(ch == '!' || ch == ':' || ch == '}' ||
ch == '=' || ch == '>' || ch == '<')) {
/* See if there's a next character. */
if (*str+1 < end) {
char next = *(*str+1);
/* For "!=". since '=' is not an allowed conversion character,
nothing is lost in this test. */
if ((ch == '!' && next == '=') || /* != */
(ch == '=' && next == '=') || /* == */
(ch == '<' && next == '=') || /* <= */
(ch == '>' && next == '=') /* >= */
) {
*str += 1;
continue;
}
/* Don't get out of the loop for these, if they're single
chars (not part of 2-char tokens). If by themselves, they
don't end an expression (unlike say '!'). */
if (ch == '>' || ch == '<') {
continue;
}
}
/* Normal way out of this loop. */
break;
} else if (ch == ']' || ch == '}' || ch == ')') {
if (!nested_depth) {
RAISE_SYNTAX_ERROR("f-string: unmatched '%c'", ch);
goto error;
}
nested_depth--;
int opening = (unsigned char)parenstack[nested_depth];
if (!((opening == '(' && ch == ')') ||
(opening == '[' && ch == ']') ||
(opening == '{' && ch == '}')))
{
RAISE_SYNTAX_ERROR(
"f-string: closing parenthesis '%c' "
"does not match opening parenthesis '%c'",
ch, opening);
goto error;
}
} else {
/* Just consume this char and loop around. */
}
}
expr_end = *str;
/* If we leave this loop in a string or with mismatched parens, we
don't care. We'll get a syntax error when compiling the
expression. But, we can produce a better error message, so
let's just do that.*/
if (quote_char) {
RAISE_SYNTAX_ERROR("f-string: unterminated string");
goto error;
}
if (nested_depth) {
int opening = (unsigned char)parenstack[nested_depth - 1];
RAISE_SYNTAX_ERROR("f-string: unmatched '%c'", opening);
goto error;
}
if (*str >= end) {
goto unexpected_end_of_string;
}
/* Compile the expression as soon as possible, so we show errors
related to the expression before errors related to the
conversion or format_spec. */
simple_expression = fstring_compile_expr(p, expr_start, expr_end, t);
if (!simple_expression) {
goto error;
}
/* Check for =, which puts the text value of the expression in
expr_text. */
if (**str == '=') {
if (p->feature_version < 8) {
RAISE_SYNTAX_ERROR("f-string: self documenting expressions are "
"only supported in Python 3.8 and greater");
goto error;
}
*str += 1;
/* Skip over ASCII whitespace. No need to test for end of string
here, since we know there's at least a trailing quote somewhere
ahead. */
while (Py_ISSPACE(**str)) {
*str += 1;
}
/* Set *expr_text to the text of the expression. */
*expr_text = PyUnicode_FromStringAndSize(expr_start, *str-expr_start);
if (!*expr_text) {
goto error;
}
}
/* Check for a conversion char, if present. */
if (**str == '!') {
*str += 1;
if (*str >= end) {
goto unexpected_end_of_string;
}
conversion = (unsigned char)**str;
*str += 1;
/* Validate the conversion. */
if (!(conversion == 's' || conversion == 'r' || conversion == 'a')) {
RAISE_SYNTAX_ERROR(
"f-string: invalid conversion character: "
"expected 's', 'r', or 'a'");
goto error;
}
}
/* Check for the format spec, if present. */
if (*str >= end) {
goto unexpected_end_of_string;
}
if (**str == ':') {
*str += 1;
if (*str >= end) {
goto unexpected_end_of_string;
}
/* Parse the format spec. */
format_spec = fstring_parse(p, str, end, raw, recurse_lvl+1,
first_token, t, last_token);
if (!format_spec) {
goto error;
}
}
if (*str >= end || **str != '}') {
goto unexpected_end_of_string;
}