aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/tools/python3/src/Parser/pegen_errors.c
blob: fb9fa2909717f96d57853f44b47c09086481e262 (plain) (blame)
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
#include <Python.h>
#include <errcode.h>

#include "tokenizer.h"
#include "pegen.h"

// TOKENIZER ERRORS

void
_PyPegen_raise_tokenizer_init_error(PyObject *filename)
{
    if (!(PyErr_ExceptionMatches(PyExc_LookupError)
          || PyErr_ExceptionMatches(PyExc_SyntaxError)
          || PyErr_ExceptionMatches(PyExc_ValueError)
          || PyErr_ExceptionMatches(PyExc_UnicodeDecodeError))) {
        return;
    }
    PyObject *errstr = NULL;
    PyObject *tuple = NULL;
    PyObject *type;
    PyObject *value;
    PyObject *tback;
    PyErr_Fetch(&type, &value, &tback);
    errstr = PyObject_Str(value);
    if (!errstr) {
        goto error;
    }

    PyObject *tmp = Py_BuildValue("(OiiO)", filename, 0, -1, Py_None);
    if (!tmp) {
        goto error;
    }

    tuple = PyTuple_Pack(2, errstr, tmp);
    Py_DECREF(tmp);
    if (!value) {
        goto error;
    }
    PyErr_SetObject(PyExc_SyntaxError, tuple);

error:
    Py_XDECREF(type);
    Py_XDECREF(value);
    Py_XDECREF(tback);
    Py_XDECREF(errstr);
    Py_XDECREF(tuple);
}

static inline void
raise_unclosed_parentheses_error(Parser *p) {
       int error_lineno = p->tok->parenlinenostack[p->tok->level-1];
       int error_col = p->tok->parencolstack[p->tok->level-1];
       RAISE_ERROR_KNOWN_LOCATION(p, PyExc_SyntaxError,
                                  error_lineno, error_col, error_lineno, -1,
                                  "'%c' was never closed",
                                  p->tok->parenstack[p->tok->level-1]);
}

int
_Pypegen_tokenizer_error(Parser *p)
{
    if (PyErr_Occurred()) {
        return -1;
    }

    const char *msg = NULL;
    PyObject* errtype = PyExc_SyntaxError;
    Py_ssize_t col_offset = -1;
    switch (p->tok->done) {
        case E_TOKEN:
            msg = "invalid token";
            break;
        case E_EOF:
            if (p->tok->level) {
                raise_unclosed_parentheses_error(p);
            } else {
                RAISE_SYNTAX_ERROR("unexpected EOF while parsing");
            }
            return -1;
        case E_DEDENT:
            RAISE_INDENTATION_ERROR("unindent does not match any outer indentation level");
            return -1;
        case E_INTR:
            if (!PyErr_Occurred()) {
                PyErr_SetNone(PyExc_KeyboardInterrupt);
            }
            return -1;
        case E_NOMEM:
            PyErr_NoMemory();
            return -1;
        case E_TABSPACE:
            errtype = PyExc_TabError;
            msg = "inconsistent use of tabs and spaces in indentation";
            break;
        case E_TOODEEP:
            errtype = PyExc_IndentationError;
            msg = "too many levels of indentation";
            break;
        case E_LINECONT: {
            col_offset = p->tok->cur - p->tok->buf - 1;
            msg = "unexpected character after line continuation character";
            break;
        }
        case E_COLUMNOVERFLOW:
            PyErr_SetString(PyExc_OverflowError,
                    "Parser column offset overflow - source line is too big");
            return -1;
        default:
            msg = "unknown parsing error";
    }

    RAISE_ERROR_KNOWN_LOCATION(p, errtype, p->tok->lineno,
                               col_offset >= 0 ? col_offset : 0,
                               p->tok->lineno, -1, msg);
    return -1;
}

int
_Pypegen_raise_decode_error(Parser *p)
{
    assert(PyErr_Occurred());
    const char *errtype = NULL;
    if (PyErr_ExceptionMatches(PyExc_UnicodeError)) {
        errtype = "unicode error";
    }
    else if (PyErr_ExceptionMatches(PyExc_ValueError)) {
        errtype = "value error";
    }
    if (errtype) {
        PyObject *type;
        PyObject *value;
        PyObject *tback;
        PyObject *errstr;
        PyErr_Fetch(&type, &value, &tback);
        errstr = PyObject_Str(value);
        if (errstr) {
            RAISE_SYNTAX_ERROR("(%s) %U", errtype, errstr);
            Py_DECREF(errstr);
        }
        else {
            PyErr_Clear();
            RAISE_SYNTAX_ERROR("(%s) unknown error", errtype);
        }
        Py_XDECREF(type);
        Py_XDECREF(value);
        Py_XDECREF(tback);
    }

    return -1;
}

static int
_PyPegen_tokenize_full_source_to_check_for_errors(Parser *p) {
    // Tokenize the whole input to see if there are any tokenization
    // errors such as mistmatching parentheses. These will get priority
    // over generic syntax errors only if the line number of the error is
    // before the one that we had for the generic error.

    // We don't want to tokenize to the end for interactive input
    if (p->tok->prompt != NULL) {
        return 0;
    }

    PyObject *type, *value, *traceback;
    PyErr_Fetch(&type, &value, &traceback);

    Token *current_token = p->known_err_token != NULL ? p->known_err_token : p->tokens[p->fill - 1];
    Py_ssize_t current_err_line = current_token->lineno;

    int ret = 0;

    for (;;) {
        const char *start;
        const char *end;
        switch (_PyTokenizer_Get(p->tok, &start, &end)) {
            case ERRORTOKEN:
                if (PyErr_Occurred()) {
                    ret = -1;
                    goto exit;
                }
                if (p->tok->level != 0) {
                    int error_lineno = p->tok->parenlinenostack[p->tok->level-1];
                    if (current_err_line > error_lineno) {
                        raise_unclosed_parentheses_error(p);
                        ret = -1;
                        goto exit;
                    }
                }
                break;
            case ENDMARKER:
                break;
            default:
                continue;
        }
        break;
    }


exit:
    if (PyErr_Occurred()) {
        Py_XDECREF(value);
        Py_XDECREF(type);
        Py_XDECREF(traceback);
    } else {
        PyErr_Restore(type, value, traceback);
    }
    return ret;
}

// PARSER ERRORS

void *
_PyPegen_raise_error(Parser *p, PyObject *errtype, const char *errmsg, ...)
{
    // Bail out if we already have an error set.
    if (p->error_indicator && PyErr_Occurred()) {
        return NULL;
    }
    if (p->fill == 0) {
        va_list va;
        va_start(va, errmsg);
        _PyPegen_raise_error_known_location(p, errtype, 0, 0, 0, -1, errmsg, va);
        va_end(va);
        return NULL;
    }

    Token *t = p->known_err_token != NULL ? p->known_err_token : p->tokens[p->fill - 1];
    Py_ssize_t col_offset;
    Py_ssize_t end_col_offset = -1;
    if (t->col_offset == -1) {
        if (p->tok->cur == p->tok->buf) {
            col_offset = 0;
        } else {
            const char* start = p->tok->buf  ? p->tok->line_start : p->tok->buf;
            col_offset = Py_SAFE_DOWNCAST(p->tok->cur - start, intptr_t, int);
        }
    } else {
        col_offset = t->col_offset + 1;
    }

    if (t->end_col_offset != -1) {
        end_col_offset = t->end_col_offset + 1;
    }

    va_list va;
    va_start(va, errmsg);
    _PyPegen_raise_error_known_location(p, errtype, t->lineno, col_offset, t->end_lineno, end_col_offset, errmsg, va);
    va_end(va);

    return NULL;
}

static PyObject *
get_error_line_from_tokenizer_buffers(Parser *p, Py_ssize_t lineno)
{
    /* If the file descriptor is interactive, the source lines of the current
     * (multi-line) statement are stored in p->tok->interactive_src_start.
     * If not, we're parsing from a string, which means that the whole source
     * is stored in p->tok->str. */
    assert((p->tok->fp == NULL && p->tok->str != NULL) || p->tok->fp == stdin);

    char *cur_line = p->tok->fp_interactive ? p->tok->interactive_src_start : p->tok->str;
    if (cur_line == NULL) {
        assert(p->tok->fp_interactive);
        // We can reach this point if the tokenizer buffers for interactive source have not been
        // initialized because we failed to decode the original source with the given locale.
        return PyUnicode_FromStringAndSize("", 0);
    }

    Py_ssize_t relative_lineno = p->starting_lineno ? lineno - p->starting_lineno + 1 : lineno;
    const char* buf_end = p->tok->fp_interactive ? p->tok->interactive_src_end : p->tok->inp;

    if (buf_end < cur_line) {
        buf_end = cur_line + strlen(cur_line);
    }

    for (int i = 0; i < relative_lineno - 1; i++) {
        char *new_line = strchr(cur_line, '\n');
        // The assert is here for debug builds but the conditional that
        // follows is there so in release builds we do not crash at the cost
        // to report a potentially wrong line.
        assert(new_line != NULL && new_line + 1 < buf_end);
        if (new_line == NULL || new_line + 1 > buf_end) {
            break;
        }
        cur_line = new_line + 1;
    }

    char *next_newline;
    if ((next_newline = strchr(cur_line, '\n')) == NULL) { // This is the last line
        next_newline = cur_line + strlen(cur_line);
    }
    return PyUnicode_DecodeUTF8(cur_line, next_newline - cur_line, "replace");
}

void *
_PyPegen_raise_error_known_location(Parser *p, PyObject *errtype,
                                    Py_ssize_t lineno, Py_ssize_t col_offset,
                                    Py_ssize_t end_lineno, Py_ssize_t end_col_offset,
                                    const char *errmsg, va_list va)
{
    PyObject *value = NULL;
    PyObject *errstr = NULL;
    PyObject *error_line = NULL;
    PyObject *tmp = NULL;
    p->error_indicator = 1;

    if (end_lineno == CURRENT_POS) {
        end_lineno = p->tok->lineno;
    }
    if (end_col_offset == CURRENT_POS) {
        end_col_offset = p->tok->cur - p->tok->line_start;
    }

    if (p->start_rule == Py_fstring_input) {
        const char *fstring_msg = "f-string: ";
        Py_ssize_t len = strlen(fstring_msg) + strlen(errmsg);

        char *new_errmsg = PyMem_Malloc(len + 1); // Lengths of both strings plus NULL character
        if (!new_errmsg) {
            return (void *) PyErr_NoMemory();
        }

        // Copy both strings into new buffer
        memcpy(new_errmsg, fstring_msg, strlen(fstring_msg));
        memcpy(new_errmsg + strlen(fstring_msg), errmsg, strlen(errmsg));
        new_errmsg[len] = 0;
        errmsg = new_errmsg;
    }
    errstr = PyUnicode_FromFormatV(errmsg, va);
    if (!errstr) {
        goto error;
    }

    if (p->tok->fp_interactive && p->tok->interactive_src_start != NULL) {
        error_line = get_error_line_from_tokenizer_buffers(p, lineno);
    }
    else if (p->start_rule == Py_file_input) {
        error_line = _PyErr_ProgramDecodedTextObject(p->tok->filename,
                                                     (int) lineno, p->tok->encoding);
    }

    if (!error_line) {
        /* PyErr_ProgramTextObject was not called or returned NULL. If it was not called,
           then we need to find the error line from some other source, because
           p->start_rule != Py_file_input. If it returned NULL, then it either unexpectedly
           failed or we're parsing from a string or the REPL. There's a third edge case where
           we're actually parsing from a file, which has an E_EOF SyntaxError and in that case
           `PyErr_ProgramTextObject` fails because lineno points to last_file_line + 1, which
           does not physically exist */
        assert(p->tok->fp == NULL || p->tok->fp == stdin || p->tok->done == E_EOF);

        if (p->tok->lineno <= lineno && p->tok->inp > p->tok->buf) {
            Py_ssize_t size = p->tok->inp - p->tok->buf;
            error_line = PyUnicode_DecodeUTF8(p->tok->buf, size, "replace");
        }
        else if (p->tok->fp == NULL || p->tok->fp == stdin) {
            error_line = get_error_line_from_tokenizer_buffers(p, lineno);
        }
        else {
            error_line = PyUnicode_FromStringAndSize("", 0);
        }
        if (!error_line) {
            goto error;
        }
    }

    if (p->start_rule == Py_fstring_input) {
        col_offset -= p->starting_col_offset;
        end_col_offset -= p->starting_col_offset;
    }

    Py_ssize_t col_number = col_offset;
    Py_ssize_t end_col_number = end_col_offset;

    if (p->tok->encoding != NULL) {
        col_number = _PyPegen_byte_offset_to_character_offset(error_line, col_offset);
        if (col_number < 0) {
            goto error;
        }
        if (end_col_number > 0) {
            Py_ssize_t end_col_offset = _PyPegen_byte_offset_to_character_offset(error_line, end_col_number);
            if (end_col_offset < 0) {
                goto error;
            } else {
                end_col_number = end_col_offset;
            }
        }
    }
    tmp = Py_BuildValue("(OnnNnn)", p->tok->filename, lineno, col_number, error_line, end_lineno, end_col_number);
    if (!tmp) {
        goto error;
    }
    value = PyTuple_Pack(2, errstr, tmp);
    Py_DECREF(tmp);
    if (!value) {
        goto error;
    }
    PyErr_SetObject(errtype, value);

    Py_DECREF(errstr);
    Py_DECREF(value);
    if (p->start_rule == Py_fstring_input) {
        PyMem_Free((void *)errmsg);
    }
    return NULL;

error:
    Py_XDECREF(errstr);
    Py_XDECREF(error_line);
    if (p->start_rule == Py_fstring_input) {
        PyMem_Free((void *)errmsg);
    }
    return NULL;
}

void
_Pypegen_set_syntax_error(Parser* p, Token* last_token) {
    // Existing sintax error
    if (PyErr_Occurred()) {
        // Prioritize tokenizer errors to custom syntax errors raised
        // on the second phase only if the errors come from the parser.
        int is_tok_ok = (p->tok->done == E_DONE || p->tok->done == E_OK);
        if (is_tok_ok && PyErr_ExceptionMatches(PyExc_SyntaxError)) {
            _PyPegen_tokenize_full_source_to_check_for_errors(p);
        }
        // Propagate the existing syntax error.
        return;
    }
    // Initialization error
    if (p->fill == 0) {
        RAISE_SYNTAX_ERROR("error at start before reading any input");
    }
    // Parser encountered EOF (End of File) unexpectedtly
    if (last_token->type == ERRORTOKEN && p->tok->done == E_EOF) {
        if (p->tok->level) {
            raise_unclosed_parentheses_error(p);
        } else {
            RAISE_SYNTAX_ERROR("unexpected EOF while parsing");
        }
        return;
    }
    // Indentation error in the tokenizer
    if (last_token->type == INDENT || last_token->type == DEDENT) {
        RAISE_INDENTATION_ERROR(last_token->type == INDENT ? "unexpected indent" : "unexpected unindent");
        return;
    }
    // Unknown error (generic case)

    // Use the last token we found on the first pass to avoid reporting
    // incorrect locations for generic syntax errors just because we reached
    // further away when trying to find specific syntax errors in the second
    // pass.
    RAISE_SYNTAX_ERROR_KNOWN_LOCATION(last_token, "invalid syntax");
    // _PyPegen_tokenize_full_source_to_check_for_errors will override the existing
    // generic SyntaxError we just raised if errors are found.
    _PyPegen_tokenize_full_source_to_check_for_errors(p);
}