| 1 |
#include "Python.h" |
| 2 |
#include "errcode.h" |
| 3 |
#include "pycore_token.h" |
| 4 |
|
| 5 |
#include "../lexer/state.h" |
| 6 |
|
| 7 |
|
| 8 |
/* ############## ERRORS ############## */ |
| 9 |
|
| 10 |
static int |
| 11 |
_syntaxerror_range(struct tok_state *tok, const char *format, |
| 12 |
int col_offset, int end_col_offset, |
| 13 |
va_list vargs) |
| 14 |
{ |
| 15 |
// In release builds, we don't want to overwrite a previous error, but in debug builds we |
| 16 |
// want to fail if we are not doing it so we can fix it. |
| 17 |
assert(tok->done != E_ERROR); |
| 18 |
if (tok->done == E_ERROR) { |
| 19 |
return ERRORTOKEN; |
| 20 |
} |
| 21 |
PyObject *errmsg, *errtext, *args; |
| 22 |
errmsg = PyUnicode_FromFormatV(format, vargs); |
| 23 |
if (!errmsg) { |
| 24 |
goto error; |
| 25 |
} |
| 26 |
|
| 27 |
errtext = PyUnicode_DecodeUTF8(tok->line_start, tok->cur - tok->line_start, |
| 28 |
"replace"); |
| 29 |
if (!errtext) { |
| 30 |
goto error; |
| 31 |
} |
| 32 |
|
| 33 |
if (col_offset == -1) { |
| 34 |
col_offset = (int)PyUnicode_GET_LENGTH(errtext); |
| 35 |
} |
| 36 |
if (end_col_offset == -1) { |
| 37 |
end_col_offset = col_offset; |
| 38 |
} |
| 39 |
|
| 40 |
Py_ssize_t line_len = strcspn(tok->line_start, "\n"); |
| 41 |
if (line_len != tok->cur - tok->line_start) { |
| 42 |
Py_DECREF(errtext); |
| 43 |
errtext = PyUnicode_DecodeUTF8(tok->line_start, line_len, |
| 44 |
"replace"); |
| 45 |
} |
| 46 |
if (!errtext) { |
| 47 |
goto error; |
| 48 |
} |
| 49 |
|
| 50 |
args = Py_BuildValue("(O(OiiNii))", errmsg, |
| 51 |
tok->filename ? tok->filename : Py_None, |
| 52 |
tok->lineno, col_offset, errtext, |
| 53 |
tok->lineno, end_col_offset); |
| 54 |
if (args) { |
| 55 |
PyErr_SetObject(PyExc_SyntaxError, args); |
| 56 |
Py_DECREF(args); |
| 57 |
} |
| 58 |
|
| 59 |
error: |
| 60 |
Py_XDECREF(errmsg); |
| 61 |
tok->done = E_ERROR; |
| 62 |
return ERRORTOKEN; |
| 63 |
} |
| 64 |
|
| 65 |
int |
| 66 |
_PyTokenizer_syntaxerror(struct tok_state *tok, const char *format, ...) |
| 67 |
{ |
| 68 |
// This errors are cleaned on startup. Todo: Fix it. |
| 69 |
va_list vargs; |
| 70 |
va_start(vargs, format); |
| 71 |
int ret = _syntaxerror_range(tok, format, -1, -1, vargs); |
| 72 |
va_end(vargs); |
| 73 |
return ret; |
| 74 |
} |
| 75 |
|
| 76 |
int |
| 77 |
_PyTokenizer_syntaxerror_known_range(struct tok_state *tok, |
| 78 |
int col_offset, int end_col_offset, |
| 79 |
const char *format, ...) |
| 80 |
{ |
| 81 |
va_list vargs; |
| 82 |
va_start(vargs, format); |
| 83 |
int ret = _syntaxerror_range(tok, format, col_offset, end_col_offset, vargs); |
| 84 |
va_end(vargs); |
| 85 |
return ret; |
| 86 |
} |
| 87 |
|
| 88 |
int |
| 89 |
_PyTokenizer_indenterror(struct tok_state *tok) |
| 90 |
{ |
| 91 |
tok->done = E_TABSPACE; |
| 92 |
tok->cur = tok->inp; |
| 93 |
return ERRORTOKEN; |
| 94 |
} |
| 95 |
|
| 96 |
char * |
| 97 |
_PyTokenizer_error_ret(struct tok_state *tok) /* XXX */ |
| 98 |
{ |
| 99 |
tok->decoding_erred = 1; |
| 100 |
if ((tok->fp != NULL || tok->readline != NULL) && tok->buf != NULL) {/* see _PyTokenizer_Free */ |
| 101 |
PyMem_Free(tok->buf); |
| 102 |
} |
| 103 |
tok->buf = tok->cur = tok->inp = NULL; |
| 104 |
tok->start = NULL; |
| 105 |
tok->end = NULL; |
| 106 |
tok->done = E_DECODE; |
| 107 |
return NULL; /* as if it were EOF */ |
| 108 |
} |
| 109 |
|
| 110 |
int |
| 111 |
_PyTokenizer_warn_invalid_escape_sequence(struct tok_state *tok, int first_invalid_escape_char) |
| 112 |
{ |
| 113 |
if (!tok->report_warnings) { |
| 114 |
return 0; |
| 115 |
} |
| 116 |
|
| 117 |
PyObject *msg = PyUnicode_FromFormat( |
| 118 |
"\"\\%c\" is an invalid escape sequence. " |
| 119 |
"Such sequences will not work in the future. " |
| 120 |
"Did you mean \"\\\\%c\"? A raw string is also an option.", |
| 121 |
(char) first_invalid_escape_char, |
| 122 |
(char) first_invalid_escape_char |
| 123 |
); |
| 124 |
|
| 125 |
if (msg == NULL) { |
| 126 |
return -1; |
| 127 |
} |
| 128 |
|
| 129 |
if (PyErr_WarnExplicitObject(PyExc_SyntaxWarning, msg, tok->filename, |
| 130 |
tok->lineno, NULL, NULL) < 0) { |
| 131 |
Py_DECREF(msg); |
| 132 |
|
| 133 |
if (PyErr_ExceptionMatches(PyExc_SyntaxWarning)) { |
| 134 |
/* Replace the SyntaxWarning exception with a SyntaxError |
| 135 |
to get a more accurate error report */ |
| 136 |
PyErr_Clear(); |
| 137 |
|
| 138 |
return _PyTokenizer_syntaxerror(tok, |
| 139 |
"\"\\%c\" is an invalid escape sequence. " |
| 140 |
"Did you mean \"\\\\%c\"? A raw string is also an option.", |
| 141 |
(char) first_invalid_escape_char, |
| 142 |
(char) first_invalid_escape_char); |
| 143 |
} |
| 144 |
|
| 145 |
return -1; |
| 146 |
} |
| 147 |
|
| 148 |
Py_DECREF(msg); |
| 149 |
return 0; |
| 150 |
} |
| 151 |
|
| 152 |
int |
| 153 |
_PyTokenizer_parser_warn(struct tok_state *tok, PyObject *category, const char *format, ...) |
| 154 |
{ |
| 155 |
if (!tok->report_warnings) { |
| 156 |
return 0; |
| 157 |
} |
| 158 |
|
| 159 |
PyObject *errmsg; |
| 160 |
va_list vargs; |
| 161 |
va_start(vargs, format); |
| 162 |
errmsg = PyUnicode_FromFormatV(format, vargs); |
| 163 |
va_end(vargs); |
| 164 |
if (!errmsg) { |
| 165 |
goto error; |
| 166 |
} |
| 167 |
|
| 168 |
if (PyErr_WarnExplicitObject(category, errmsg, tok->filename, |
| 169 |
tok->lineno, NULL, NULL) < 0) { |
| 170 |
if (PyErr_ExceptionMatches(category)) { |
| 171 |
/* Replace the DeprecationWarning exception with a SyntaxError |
| 172 |
to get a more accurate error report */ |
| 173 |
PyErr_Clear(); |
| 174 |
_PyTokenizer_syntaxerror(tok, "%U", errmsg); |
| 175 |
} |
| 176 |
goto error; |
| 177 |
} |
| 178 |
Py_DECREF(errmsg); |
| 179 |
return 0; |
| 180 |
|
| 181 |
error: |
| 182 |
Py_XDECREF(errmsg); |
| 183 |
tok->done = E_ERROR; |
| 184 |
return -1; |
| 185 |
} |
| 186 |
|
| 187 |
|
| 188 |
/* ############## STRING MANIPULATION ############## */ |
| 189 |
|
| 190 |
char * |
| 191 |
_PyTokenizer_new_string(const char *s, Py_ssize_t len, struct tok_state *tok) |
| 192 |
{ |
| 193 |
char* result = (char *)PyMem_Malloc(len + 1); |
| 194 |
if (!result) { |
| 195 |
tok->done = E_NOMEM; |
| 196 |
PyErr_NoMemory(); |
| 197 |
return NULL; |
| 198 |
} |
| 199 |
memcpy(result, s, len); |
| 200 |
result[len] = '\0'; |
| 201 |
return result; |
| 202 |
} |
| 203 |
|
| 204 |
PyObject * |
| 205 |
_PyTokenizer_translate_into_utf8(const char* str, const char* enc) { |
| 206 |
PyObject *utf8; |
| 207 |
PyObject* buf = PyUnicode_Decode(str, strlen(str), enc, NULL); |
| 208 |
if (buf == NULL) |
| 209 |
return NULL; |
| 210 |
utf8 = PyUnicode_AsUTF8String(buf); |
| 211 |
Py_DECREF(buf); |
| 212 |
return utf8; |
| 213 |
} |
| 214 |
|
| 215 |
char * |
| 216 |
_PyTokenizer_translate_newlines(const char *s, int exec_input, int preserve_crlf, |
| 217 |
struct tok_state *tok) { |
| 218 |
int skip_next_lf = 0; |
| 219 |
size_t needed_length = strlen(s) + 2, final_length; |
| 220 |
char *buf, *current; |
| 221 |
char c = '\0'; |
| 222 |
buf = PyMem_Malloc(needed_length); |
| 223 |
if (buf == NULL) { |
| 224 |
tok->done = E_NOMEM; |
| 225 |
PyErr_NoMemory(); |
| 226 |
return NULL; |
| 227 |
} |
| 228 |
for (current = buf; *s; s++, current++) { |
| 229 |
c = *s; |
| 230 |
if (skip_next_lf) { |
| 231 |
skip_next_lf = 0; |
| 232 |
if (c == '\n') { |
| 233 |
c = *++s; |
| 234 |
if (!c) |
| 235 |
break; |
| 236 |
} |
| 237 |
} |
| 238 |
if (!preserve_crlf && c == '\r') { |
| 239 |
skip_next_lf = 1; |
| 240 |
c = '\n'; |
| 241 |
} |
| 242 |
*current = c; |
| 243 |
} |
| 244 |
/* If this is exec input, add a newline to the end of the string if |
| 245 |
there isn't one already. */ |
| 246 |
if (exec_input && c != '\n' && c != '\0') { |
| 247 |
*current = '\n'; |
| 248 |
current++; |
| 249 |
} |
| 250 |
*current = '\0'; |
| 251 |
final_length = current - buf + 1; |
| 252 |
if (final_length < needed_length && final_length) { |
| 253 |
/* should never fail */ |
| 254 |
char* result = PyMem_Realloc(buf, final_length); |
| 255 |
if (result == NULL) { |
| 256 |
PyMem_Free(buf); |
| 257 |
} |
| 258 |
buf = result; |
| 259 |
} |
| 260 |
return buf; |
| 261 |
} |
| 262 |
|
| 263 |
/* ############## ENCODING STUFF ############## */ |
| 264 |
|
| 265 |
|
| 266 |
/* See whether the file starts with a BOM. If it does, |
| 267 |
invoke the set_readline function with the new encoding. |
| 268 |
Return 1 on success, 0 on failure. */ |
| 269 |
int |
| 270 |
_PyTokenizer_check_bom(int get_char(struct tok_state *), |
| 271 |
void unget_char(int, struct tok_state *), |
| 272 |
int set_readline(struct tok_state *, const char *), |
| 273 |
struct tok_state *tok) |
| 274 |
{ |
| 275 |
int ch1, ch2, ch3; |
| 276 |
ch1 = get_char(tok); |
| 277 |
tok->decoding_state = STATE_SEEK_CODING; |
| 278 |
if (ch1 == EOF) { |
| 279 |
return 1; |
| 280 |
} else if (ch1 == 0xEF) { |
| 281 |
ch2 = get_char(tok); |
| 282 |
if (ch2 != 0xBB) { |
| 283 |
unget_char(ch2, tok); |
| 284 |
unget_char(ch1, tok); |
| 285 |
return 1; |
| 286 |
} |
| 287 |
ch3 = get_char(tok); |
| 288 |
if (ch3 != 0xBF) { |
| 289 |
unget_char(ch3, tok); |
| 290 |
unget_char(ch2, tok); |
| 291 |
unget_char(ch1, tok); |
| 292 |
return 1; |
| 293 |
} |
| 294 |
} else { |
| 295 |
unget_char(ch1, tok); |
| 296 |
return 1; |
| 297 |
} |
| 298 |
if (tok->encoding != NULL) |
| 299 |
PyMem_Free(tok->encoding); |
| 300 |
tok->encoding = _PyTokenizer_new_string("utf-8", 5, tok); |
| 301 |
if (!tok->encoding) |
| 302 |
return 0; |
| 303 |
/* No need to set_readline: input is already utf-8 */ |
| 304 |
return 1; |
| 305 |
} |
| 306 |
|
| 307 |
static const char * |
| 308 |
get_normal_name(const char *s) /* for utf-8 and latin-1 */ |
| 309 |
{ |
| 310 |
char buf[13]; |
| 311 |
int i; |
| 312 |
for (i = 0; i < 12; i++) { |
| 313 |
int c = s[i]; |
| 314 |
if (c == '\0') |
| 315 |
break; |
| 316 |
else if (c == '_') |
| 317 |
buf[i] = '-'; |
| 318 |
else |
| 319 |
buf[i] = Py_TOLOWER(c); |
| 320 |
} |
| 321 |
buf[i] = '\0'; |
| 322 |
if (strcmp(buf, "utf-8") == 0 || |
| 323 |
strncmp(buf, "utf-8-", 6) == 0) |
| 324 |
return "utf-8"; |
| 325 |
else if (strcmp(buf, "latin-1") == 0 || |
| 326 |
strcmp(buf, "iso-8859-1") == 0 || |
| 327 |
strcmp(buf, "iso-latin-1") == 0 || |
| 328 |
strncmp(buf, "latin-1-", 8) == 0 || |
| 329 |
strncmp(buf, "iso-8859-1-", 11) == 0 || |
| 330 |
strncmp(buf, "iso-latin-1-", 12) == 0) |
| 331 |
return "iso-8859-1"; |
| 332 |
else |
| 333 |
return s; |
| 334 |
} |
| 335 |
|
| 336 |
/* Return the coding spec in S, or NULL if none is found. */ |
| 337 |
static int |
| 338 |
get_coding_spec(const char *s, char **spec, Py_ssize_t size, struct tok_state *tok) |
| 339 |
{ |
| 340 |
Py_ssize_t i; |
| 341 |
*spec = NULL; |
| 342 |
/* Coding spec must be in a comment, and that comment must be |
| 343 |
* the only statement on the source code line. */ |
| 344 |
for (i = 0; i < size - 6; i++) { |
| 345 |
if (s[i] == '#') |
| 346 |
break; |
| 347 |
if (s[i] != ' ' && s[i] != '\t' && s[i] != '\014') |
| 348 |
return 1; |
| 349 |
} |
| 350 |
for (; i < size - 6; i++) { /* XXX inefficient search */ |
| 351 |
const char* t = s + i; |
| 352 |
if (memcmp(t, "coding", 6) == 0) { |
| 353 |
const char* begin = NULL; |
| 354 |
t += 6; |
| 355 |
if (t[0] != ':' && t[0] != '=') |
| 356 |
continue; |
| 357 |
do { |
| 358 |
t++; |
| 359 |
} while (t[0] == ' ' || t[0] == '\t'); |
| 360 |
|
| 361 |
begin = t; |
| 362 |
while (Py_ISALNUM(t[0]) || |
| 363 |
t[0] == '-' || t[0] == '_' || t[0] == '.') |
| 364 |
t++; |
| 365 |
|
| 366 |
if (begin < t) { |
| 367 |
char* r = _PyTokenizer_new_string(begin, t - begin, tok); |
| 368 |
const char* q; |
| 369 |
if (!r) |
| 370 |
return 0; |
| 371 |
q = get_normal_name(r); |
| 372 |
if (r != q) { |
| 373 |
PyMem_Free(r); |
| 374 |
r = _PyTokenizer_new_string(q, strlen(q), tok); |
| 375 |
if (!r) |
| 376 |
return 0; |
| 377 |
} |
| 378 |
*spec = r; |
| 379 |
break; |
| 380 |
} |
| 381 |
} |
| 382 |
} |
| 383 |
return 1; |
| 384 |
} |
| 385 |
|
| 386 |
/* Check whether the line contains a coding spec. If it does, |
| 387 |
invoke the set_readline function for the new encoding. |
| 388 |
This function receives the tok_state and the new encoding. |
| 389 |
Return 1 on success, 0 on failure. */ |
| 390 |
int |
| 391 |
_PyTokenizer_check_coding_spec(const char* line, Py_ssize_t size, struct tok_state *tok, |
| 392 |
int set_readline(struct tok_state *, const char *)) |
| 393 |
{ |
| 394 |
char *cs; |
| 395 |
if (tok->cont_line) { |
| 396 |
/* It's a continuation line, so it can't be a coding spec. */ |
| 397 |
tok->decoding_state = STATE_NORMAL; |
| 398 |
return 1; |
| 399 |
} |
| 400 |
if (!get_coding_spec(line, &cs, size, tok)) { |
| 401 |
return 0; |
| 402 |
} |
| 403 |
if (!cs) { |
| 404 |
Py_ssize_t i; |
| 405 |
for (i = 0; i < size; i++) { |
| 406 |
if (line[i] == '#' || line[i] == '\n' || line[i] == '\r') |
| 407 |
break; |
| 408 |
if (line[i] != ' ' && line[i] != '\t' && line[i] != '\014') { |
| 409 |
/* Stop checking coding spec after a line containing |
| 410 |
* anything except a comment. */ |
| 411 |
tok->decoding_state = STATE_NORMAL; |
| 412 |
break; |
| 413 |
} |
| 414 |
} |
| 415 |
return 1; |
| 416 |
} |
| 417 |
tok->decoding_state = STATE_NORMAL; |
| 418 |
if (tok->encoding == NULL) { |
| 419 |
assert(tok->decoding_readline == NULL); |
| 420 |
if (strcmp(cs, "utf-8") != 0 && !set_readline(tok, cs)) { |
| 421 |
_PyTokenizer_error_ret(tok); |
| 422 |
PyErr_Format(PyExc_SyntaxError, "encoding problem: %s", cs); |
| 423 |
PyMem_Free(cs); |
| 424 |
return 0; |
| 425 |
} |
| 426 |
tok->encoding = cs; |
| 427 |
} else { /* then, compare cs with BOM */ |
| 428 |
if (strcmp(tok->encoding, cs) != 0) { |
| 429 |
tok->line_start = line; |
| 430 |
tok->cur = (char *)line; |
| 431 |
assert(size <= INT_MAX); |
| 432 |
_PyTokenizer_syntaxerror_known_range(tok, 0, (int)size, |
| 433 |
"encoding problem: %s with BOM", cs); |
| 434 |
PyMem_Free(cs); |
| 435 |
_PyTokenizer_error_ret(tok); |
| 436 |
return 0; |
| 437 |
} |
| 438 |
PyMem_Free(cs); |
| 439 |
} |
| 440 |
return 1; |
| 441 |
} |
| 442 |
|
| 443 |
/* Check whether the characters at s start a valid |
| 444 |
UTF-8 sequence. Return the number of characters forming |
| 445 |
the sequence if yes, 0 if not. The special cases match |
| 446 |
those in stringlib/codecs.h:utf8_decode. |
| 447 |
*/ |
| 448 |
static int |
| 449 |
valid_utf8(const unsigned char* s) |
| 450 |
{ |
| 451 |
int expected = 0; |
| 452 |
int length; |
| 453 |
if (*s < 0x80) { |
| 454 |
/* single-byte code */ |
| 455 |
return 1; |
| 456 |
} |
| 457 |
else if (*s < 0xE0) { |
| 458 |
/* \xC2\x80-\xDF\xBF -- 0080-07FF */ |
| 459 |
if (*s < 0xC2) { |
| 460 |
/* invalid sequence |
| 461 |
\x80-\xBF -- continuation byte |
| 462 |
\xC0-\xC1 -- fake 0000-007F */ |
| 463 |
return 0; |
| 464 |
} |
| 465 |
expected = 1; |
| 466 |
} |
| 467 |
else if (*s < 0xF0) { |
| 468 |
/* \xE0\xA0\x80-\xEF\xBF\xBF -- 0800-FFFF */ |
| 469 |
if (*s == 0xE0 && *(s + 1) < 0xA0) { |
| 470 |
/* invalid sequence |
| 471 |
\xE0\x80\x80-\xE0\x9F\xBF -- fake 0000-0800 */ |
| 472 |
return 0; |
| 473 |
} |
| 474 |
else if (*s == 0xED && *(s + 1) >= 0xA0) { |
| 475 |
/* Decoding UTF-8 sequences in range \xED\xA0\x80-\xED\xBF\xBF |
| 476 |
will result in surrogates in range D800-DFFF. Surrogates are |
| 477 |
not valid UTF-8 so they are rejected. |
| 478 |
See https://www.unicode.org/versions/Unicode5.2.0/ch03.pdf |
| 479 |
(table 3-7) and http://www.rfc-editor.org/rfc/rfc3629.txt */ |
| 480 |
return 0; |
| 481 |
} |
| 482 |
expected = 2; |
| 483 |
} |
| 484 |
else if (*s < 0xF5) { |
| 485 |
/* \xF0\x90\x80\x80-\xF4\x8F\xBF\xBF -- 10000-10FFFF */ |
| 486 |
if (*(s + 1) < 0x90 ? *s == 0xF0 : *s == 0xF4) { |
| 487 |
/* invalid sequence -- one of: |
| 488 |
\xF0\x80\x80\x80-\xF0\x8F\xBF\xBF -- fake 0000-FFFF |
| 489 |
\xF4\x90\x80\x80- -- 110000- overflow */ |
| 490 |
return 0; |
| 491 |
} |
| 492 |
expected = 3; |
| 493 |
} |
| 494 |
else { |
| 495 |
/* invalid start byte */ |
| 496 |
return 0; |
| 497 |
} |
| 498 |
length = expected + 1; |
| 499 |
for (int i = 1; i <= expected; i++) { |
| 500 |
if (s[i] < 0x80 || s[i] >= 0xC0) { |
| 501 |
return 0; |
| 502 |
} |
| 503 |
} |
| 504 |
return length; |
| 505 |
} |
| 506 |
|
| 507 |
int |
| 508 |
_PyTokenizer_ensure_utf8(const char *line, struct tok_state *tok, int lineno) |
| 509 |
{ |
| 510 |
const char *badchar = NULL; |
| 511 |
const char *c; |
| 512 |
int length; |
| 513 |
int col_offset = 0; |
| 514 |
const char *line_start = line; |
| 515 |
for (c = line; *c; c += length) { |
| 516 |
if (!(length = valid_utf8((const unsigned char *)c))) { |
| 517 |
badchar = c; |
| 518 |
break; |
| 519 |
} |
| 520 |
col_offset++; |
| 521 |
if (*c == '\n') { |
| 522 |
lineno++; |
| 523 |
col_offset = 0; |
| 524 |
line_start = c + 1; |
| 525 |
} |
| 526 |
} |
| 527 |
if (badchar) { |
| 528 |
tok->lineno = lineno; |
| 529 |
tok->line_start = line_start; |
| 530 |
tok->cur = (char *)badchar; |
| 531 |
_PyTokenizer_syntaxerror_known_range(tok, |
| 532 |
col_offset + 1, col_offset + 1, |
| 533 |
"Non-UTF-8 code starting with '\\x%.2x'" |
| 534 |
"%s%V on line %i, " |
| 535 |
"but no encoding declared; " |
| 536 |
"see https://peps.python.org/pep-0263/ for details", |
| 537 |
(unsigned char)*badchar, |
| 538 |
tok->filename ? " in file " : "", tok->filename, "", |
| 539 |
lineno); |
| 540 |
return 0; |
| 541 |
} |
| 542 |
return 1; |
| 543 |
} |
| 544 |
|
| 545 |
|
| 546 |
/* ############## DEBUGGING STUFF ############## */ |
| 547 |
|
| 548 |
#ifdef Py_DEBUG |
| 549 |
void |
| 550 |
_PyTokenizer_print_escape(FILE *f, const char *s, Py_ssize_t size) |
| 551 |
{ |
| 552 |
if (s == NULL) { |
| 553 |
fputs("NULL", f); |
| 554 |
return; |
| 555 |
} |
| 556 |
putc('"', f); |
| 557 |
while (size-- > 0) { |
| 558 |
unsigned char c = *s++; |
| 559 |
switch (c) { |
| 560 |
case '\n': fputs("\\n", f); break; |
| 561 |
case '\r': fputs("\\r", f); break; |
| 562 |
case '\t': fputs("\\t", f); break; |
| 563 |
case '\f': fputs("\\f", f); break; |
| 564 |
case '\'': fputs("\\'", f); break; |
| 565 |
case '"': fputs("\\\"", f); break; |
| 566 |
default: |
| 567 |
if (0x20 <= c && c <= 0x7f) |
| 568 |
putc(c, f); |
| 569 |
else |
| 570 |
fprintf(f, "\\x%02x", c); |
| 571 |
} |
| 572 |
} |
| 573 |
putc('"', f); |
| 574 |
} |
| 575 |
|
| 576 |
void |
| 577 |
_PyTokenizer_tok_dump(int type, char *start, char *end) |
| 578 |
{ |
| 579 |
fprintf(stderr, "%s", _PyParser_TokenNames[type]); |
| 580 |
if (type == NAME || type == NUMBER || type == STRING || type == OP) |
| 581 |
fprintf(stderr, "(%.*s)", (int)(end - start), start); |
| 582 |
} |
| 583 |
#endif |