lexer.c 20.9 KB
Newer Older
Damien's avatar
Damien committed
1
2
3
4
5
6
7
8
9
10
11
12
/* lexer.c -- simple tokeniser for Python implementation
 */

#include <stdint.h>
#include <stdio.h>
#include <assert.h>

#include "misc.h"
#include "lexer.h"

#define TAB_SIZE (8)

13
14
15
// TODO seems that CPython allows NULL byte in the input stream
// don't know if that's intentional or not, but we don't allow it

16
struct _mp_lexer_t {
17
18
    const char *name;           // name of source
    void *stream_data;          // data for stream
19
20
    mp_lexer_stream_next_char_t stream_next_char;   // stream callback to get next char
    mp_lexer_stream_close_t stream_close;           // stream callback to free
Damien's avatar
Damien committed
21

22
    unichar chr0, chr1, chr2;   // current cached characters from source
Damien's avatar
Damien committed
23
24
25
26

    uint line;                  // source line
    uint column;                // source column

27
28
    int emit_dent;              // non-zero when there are INDENT/DEDENT tokens to emit
    int nested_bracket_level;   // >0 when there are nested brackets over multiple lines
Damien's avatar
Damien committed
29
30
31
32
33

    uint alloc_indent_level;
    uint num_indent_level;
    uint16_t *indent_level;

34
    vstr_t vstr;
35
    mp_token_t tok_cur;
Damien's avatar
Damien committed
36
37
};

38
// TODO replace with a call to a standard function
39
bool str_strn_equal(const char *str, const char *strn, int len) {
Damien's avatar
Damien committed
40
41
    uint i = 0;

42
    while (i < len && *str == *strn) {
Damien's avatar
Damien committed
43
44
        ++i;
        ++str;
45
        ++strn;
Damien's avatar
Damien committed
46
47
    }

48
    return i == len && *str == 0;
Damien's avatar
Damien committed
49
50
}

51
void mp_token_show(const mp_token_t *tok) {
52
    printf("(%s:%d:%d) kind:%d str:%p len:%d", tok->src_name, tok->src_line, tok->src_column, tok->kind, tok->str, tok->len);
Damien's avatar
Damien committed
53
54
55
56
57
    if (tok->str != NULL && tok->len > 0) {
        const char *i = tok->str;
        const char *j = i + tok->len;
        printf(" ");
        while (i < j) {
58
59
60
            unichar c = utf8_get_char(i);
            i = utf8_next_char(i);
            if (unichar_isprint(c)) {
Damien's avatar
Damien committed
61
62
63
64
65
66
67
68
69
                printf("%c", c);
            } else {
                printf("?");
            }
        }
    }
    printf("\n");
}

70
71
#define CUR_CHAR(lex) ((lex)->chr0)

72
73
static bool is_end(mp_lexer_t *lex) {
    return lex->chr0 == MP_LEXER_CHAR_EOF;
Damien's avatar
Damien committed
74
75
}

76
static bool is_physical_newline(mp_lexer_t *lex) {
Damien's avatar
Damien committed
77
78
79
    return lex->chr0 == '\n' || lex->chr0 == '\r';
}

80
static bool is_char(mp_lexer_t *lex, char c) {
Damien's avatar
Damien committed
81
82
83
    return lex->chr0 == c;
}

84
static bool is_char_or(mp_lexer_t *lex, char c1, char c2) {
Damien's avatar
Damien committed
85
86
87
    return lex->chr0 == c1 || lex->chr0 == c2;
}

88
static bool is_char_or3(mp_lexer_t *lex, char c1, char c2, char c3) {
Damien's avatar
Damien committed
89
90
91
92
    return lex->chr0 == c1 || lex->chr0 == c2 || lex->chr0 == c3;
}

/*
93
static bool is_char_following(mp_lexer_t *lex, char c) {
Damien's avatar
Damien committed
94
95
96
97
    return lex->chr1 == c;
}
*/

98
static bool is_char_following_or(mp_lexer_t *lex, char c1, char c2) {
Damien's avatar
Damien committed
99
100
101
    return lex->chr1 == c1 || lex->chr1 == c2;
}

102
static bool is_char_following_following_or(mp_lexer_t *lex, char c1, char c2) {
Damien's avatar
Damien committed
103
104
105
    return lex->chr2 == c1 || lex->chr2 == c2;
}

106
static bool is_char_and(mp_lexer_t *lex, char c1, char c2) {
Damien's avatar
Damien committed
107
108
109
    return lex->chr0 == c1 && lex->chr1 == c2;
}

110
static bool is_whitespace(mp_lexer_t *lex) {
111
    return unichar_isspace(lex->chr0);
Damien's avatar
Damien committed
112
113
}

114
static bool is_letter(mp_lexer_t *lex) {
115
    return unichar_isalpha(lex->chr0);
Damien's avatar
Damien committed
116
117
}

118
static bool is_digit(mp_lexer_t *lex) {
119
    return unichar_isdigit(lex->chr0);
Damien's avatar
Damien committed
120
121
}

122
static bool is_following_digit(mp_lexer_t *lex) {
123
    return unichar_isdigit(lex->chr1);
Damien's avatar
Damien committed
124
125
126
}

// TODO UNICODE include unicode characters in definition of identifiers
127
static bool is_head_of_identifier(mp_lexer_t *lex) {
Damien's avatar
Damien committed
128
129
130
131
    return is_letter(lex) || lex->chr0 == '_';
}

// TODO UNICODE include unicode characters in definition of identifiers
132
static bool is_tail_of_identifier(mp_lexer_t *lex) {
Damien's avatar
Damien committed
133
134
135
    return is_head_of_identifier(lex) || is_digit(lex);
}

136
137
static void next_char(mp_lexer_t *lex) {
    if (lex->chr0 == MP_LEXER_CHAR_EOF) {
Damien's avatar
Damien committed
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
        return;
    }

    int advance = 1;

    if (lex->chr0 == '\n') {
        // LF is a new line
        ++lex->line;
        lex->column = 1;
    } else if (lex->chr0 == '\r') {
        // CR is a new line
        ++lex->line;
        lex->column = 1;
        if (lex->chr1 == '\n') {
            // CR LF is a single new line
            advance = 2;
        }
    } else if (lex->chr0 == '\t') {
        // a tab
        lex->column = (((lex->column - 1 + TAB_SIZE) / TAB_SIZE) * TAB_SIZE) + 1;
    } else {
        // a character worth one column
        ++lex->column;
    }

    for (; advance > 0; advance--) {
        lex->chr0 = lex->chr1;
        lex->chr1 = lex->chr2;
166
        lex->chr2 = lex->stream_next_char(lex->stream_data);
167
        if (lex->chr2 == MP_LEXER_CHAR_EOF) {
Damien's avatar
Damien committed
168
            // EOF
169
            if (lex->chr1 != MP_LEXER_CHAR_EOF && lex->chr1 != '\n' && lex->chr1 != '\r') {
Damien's avatar
Damien committed
170
171
172
173
174
175
                lex->chr2 = '\n'; // insert newline at end of file
            }
        }
    }
}

176
void indent_push(mp_lexer_t *lex, uint indent) {
Damien's avatar
Damien committed
177
    if (lex->num_indent_level >= lex->alloc_indent_level) {
178
        lex->indent_level = m_renew(uint16_t, lex->indent_level, lex->alloc_indent_level, lex->alloc_indent_level * 2);
Damien's avatar
Damien committed
179
180
181
182
183
        lex->alloc_indent_level *= 2;
    }
    lex->indent_level[lex->num_indent_level++] = indent;
}

184
uint indent_top(mp_lexer_t *lex) {
Damien's avatar
Damien committed
185
186
187
    return lex->indent_level[lex->num_indent_level - 1];
}

188
void indent_pop(mp_lexer_t *lex) {
Damien's avatar
Damien committed
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
    lex->num_indent_level -= 1;
}

// some tricky operator encoding:
//     <op>  = begin with <op>, if this opchar matches then begin here
//     e<op> = end with <op>, if this opchar matches then end
//     E<op> = mandatory end with <op>, this opchar must match, then end
//     c<op> = continue with <op>, if this opchar matches then continue matching
// this means if the start of two ops are the same then they are equal til the last char

static const char *tok_enc =
    "()[]{},:;@~" // singles
    "<e=c<e="     // < <= << <<=
    ">e=c>e="     // > >= >> >>=
    "*e=c*e="     // * *= ** **=
    "+e="         // + +=
    "-e=e>"       // - -= ->
    "&e="         // & &=
    "|e="         // | |=
    "/e=c/e="     // / /= // //=
    "%e="         // % %=
    "^e="         // ^ ^=
    "=e="         // = ==
    "!E="         // !=
    ".c.E.";      // . ...

// TODO static assert that number of tokens is less than 256 so we can safely make this table with byte sized entries
static const uint8_t tok_enc_kind[] = {
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
    MP_TOKEN_DEL_PAREN_OPEN, MP_TOKEN_DEL_PAREN_CLOSE,
    MP_TOKEN_DEL_BRACKET_OPEN, MP_TOKEN_DEL_BRACKET_CLOSE,
    MP_TOKEN_DEL_BRACE_OPEN, MP_TOKEN_DEL_BRACE_CLOSE,
    MP_TOKEN_DEL_COMMA, MP_TOKEN_DEL_COLON, MP_TOKEN_DEL_SEMICOLON, MP_TOKEN_DEL_AT, MP_TOKEN_OP_TILDE,

    MP_TOKEN_OP_LESS, MP_TOKEN_OP_LESS_EQUAL, MP_TOKEN_OP_DBL_LESS, MP_TOKEN_DEL_DBL_LESS_EQUAL,
    MP_TOKEN_OP_MORE, MP_TOKEN_OP_MORE_EQUAL, MP_TOKEN_OP_DBL_MORE, MP_TOKEN_DEL_DBL_MORE_EQUAL,
    MP_TOKEN_OP_STAR, MP_TOKEN_DEL_STAR_EQUAL, MP_TOKEN_OP_DBL_STAR, MP_TOKEN_DEL_DBL_STAR_EQUAL,
    MP_TOKEN_OP_PLUS, MP_TOKEN_DEL_PLUS_EQUAL,
    MP_TOKEN_OP_MINUS, MP_TOKEN_DEL_MINUS_EQUAL, MP_TOKEN_DEL_MINUS_MORE,
    MP_TOKEN_OP_AMPERSAND, MP_TOKEN_DEL_AMPERSAND_EQUAL,
    MP_TOKEN_OP_PIPE, MP_TOKEN_DEL_PIPE_EQUAL,
    MP_TOKEN_OP_SLASH, MP_TOKEN_DEL_SLASH_EQUAL, MP_TOKEN_OP_DBL_SLASH, MP_TOKEN_DEL_DBL_SLASH_EQUAL,
    MP_TOKEN_OP_PERCENT, MP_TOKEN_DEL_PERCENT_EQUAL,
    MP_TOKEN_OP_CARET, MP_TOKEN_DEL_CARET_EQUAL,
    MP_TOKEN_DEL_EQUAL, MP_TOKEN_OP_DBL_EQUAL,
    MP_TOKEN_OP_NOT_EQUAL,
Damien George's avatar
Damien George committed
234
    MP_TOKEN_DEL_PERIOD, MP_TOKEN_ELLIPSIS,
Damien's avatar
Damien committed
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
};

// must have the same order as enum in lexer.h
static const char *tok_kw[] = {
    "False",
    "None",
    "True",
    "and",
    "as",
    "assert",
    "break",
    "class",
    "continue",
    "def",
    "del",
    "elif",
    "else",
    "except",
    "finally",
    "for",
    "from",
    "global",
    "if",
    "import",
    "in",
    "is",
    "lambda",
    "nonlocal",
    "not",
    "or",
    "pass",
    "raise",
    "return",
    "try",
    "while",
    "with",
    "yield",
    NULL,
};

275
static void mp_lexer_next_token_into(mp_lexer_t *lex, mp_token_t *tok, bool first_token) {
276
    // skip white space and comments
Damien's avatar
Damien committed
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
    bool had_physical_newline = false;
    while (!is_end(lex)) {
        if (is_physical_newline(lex)) {
            had_physical_newline = true;
            next_char(lex);
        } else if (is_whitespace(lex)) {
            next_char(lex);
        } else if (is_char(lex, '#')) {
            next_char(lex);
            while (!is_end(lex) && !is_physical_newline(lex)) {
                next_char(lex);
            }
            // had_physical_newline will be set on next loop
        } else if (is_char(lex, '\\')) {
            // backslash (outside string literals) must appear just before a physical newline
            next_char(lex);
            if (!is_physical_newline(lex)) {
294
295
296
297
298
299
300
301
302
                // SyntaxError: unexpected character after line continuation character
                tok->src_name = lex->name;
                tok->src_line = lex->line;
                tok->src_column = lex->column;
                tok->kind = MP_TOKEN_BAD_LINE_CONTINUATION;
                vstr_reset(&lex->vstr);
                tok->str = vstr_str(&lex->vstr);
                tok->len = 0;
                return;
Damien's avatar
Damien committed
303
304
305
306
307
308
309
310
            } else {
                next_char(lex);
            }
        } else {
            break;
        }
    }

311
    // set token source information
Damien's avatar
Damien committed
312
313
314
315
    tok->src_name = lex->name;
    tok->src_line = lex->line;
    tok->src_column = lex->column;

316
317
318
319
320
321
322
323
    // start new token text
    vstr_reset(&lex->vstr);

    if (first_token && lex->line == 1 && lex->column != 1) {
        // check that the first token is in the first column
        // if first token is not on first line, we get a physical newline and
        // this check is done as part of normal indent/dedent checking below
        // (done to get equivalence with CPython)
324
        tok->kind = MP_TOKEN_INDENT;
325
326

    } else if (lex->emit_dent < 0) {
327
        tok->kind = MP_TOKEN_DEDENT;
Damien's avatar
Damien committed
328
329
330
        lex->emit_dent += 1;

    } else if (lex->emit_dent > 0) {
331
        tok->kind = MP_TOKEN_INDENT;
Damien's avatar
Damien committed
332
333
        lex->emit_dent -= 1;

334
    } else if (had_physical_newline && lex->nested_bracket_level == 0) {
335
        tok->kind = MP_TOKEN_NEWLINE;
Damien's avatar
Damien committed
336
337
338
339
340
341
342
343
344
345
346
347
348

        uint num_spaces = lex->column - 1;
        lex->emit_dent = 0;
        if (num_spaces == indent_top(lex)) {
        } else if (num_spaces > indent_top(lex)) {
            indent_push(lex, num_spaces);
            lex->emit_dent += 1;
        } else {
            while (num_spaces < indent_top(lex)) {
                indent_pop(lex);
                lex->emit_dent -= 1;
            }
            if (num_spaces != indent_top(lex)) {
349
                tok->kind = MP_TOKEN_DEDENT_MISMATCH;
Damien's avatar
Damien committed
350
351
352
353
354
            }
        }

    } else if (is_end(lex)) {
        if (indent_top(lex) > 0) {
355
            tok->kind = MP_TOKEN_NEWLINE;
Damien's avatar
Damien committed
356
357
358
359
360
361
            lex->emit_dent = 0;
            while (indent_top(lex) > 0) {
                indent_pop(lex);
                lex->emit_dent -= 1;
            }
        } else {
362
            tok->kind = MP_TOKEN_END;
Damien's avatar
Damien committed
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
        }

    } else if (is_char_or(lex, '\'', '\"')
               || (is_char_or3(lex, 'r', 'u', 'b') && is_char_following_or(lex, '\'', '\"'))
               || ((is_char_and(lex, 'r', 'b') || is_char_and(lex, 'b', 'r')) && is_char_following_following_or(lex, '\'', '\"'))) {
        // a string or bytes literal

        // parse type codes
        bool is_raw = false;
        bool is_bytes = false;
        if (is_char(lex, 'u')) {
            next_char(lex);
        } else if (is_char(lex, 'b')) {
            is_bytes = true;
            next_char(lex);
            if (is_char(lex, 'r')) {
                is_raw = true;
                next_char(lex);
            }
        } else if (is_char(lex, 'r')) {
            is_raw = true;
            next_char(lex);
            if (is_char(lex, 'b')) {
                is_bytes = true;
                next_char(lex);
            }
        }

        // set token kind
        if (is_bytes) {
393
            tok->kind = MP_TOKEN_BYTES;
Damien's avatar
Damien committed
394
        } else {
395
            tok->kind = MP_TOKEN_STRING;
Damien's avatar
Damien committed
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
        }

        // get first quoting character
        char quote_char = '\'';
        if (is_char(lex, '\"')) {
            quote_char = '\"';
        }
        next_char(lex);

        // work out if it's a single or triple quoted literal
        int num_quotes;
        if (is_char_and(lex, quote_char, quote_char)) {
            // triple quotes
            next_char(lex);
            next_char(lex);
            num_quotes = 3;
        } else {
            // single quotes
            num_quotes = 1;
        }

        // parse the literal
        int n_closing = 0;
        while (!is_end(lex) && (num_quotes > 1 || !is_char(lex, '\n')) && n_closing < num_quotes) {
            if (is_char(lex, quote_char)) {
                n_closing += 1;
422
                vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
423
424
425
426
            } else {
                n_closing = 0;
                if (!is_raw && is_char(lex, '\\')) {
                    next_char(lex);
427
428
                    unichar c = CUR_CHAR(lex);
                    switch (c) {
429
430
                        case MP_LEXER_CHAR_EOF: break; // TODO a proper error message?
                        case '\n': c = MP_LEXER_CHAR_EOF; break; // TODO check this works correctly (we are supposed to ignore it
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
                        case '\\': break;
                        case '\'': break;
                        case '"': break;
                        case 'a': c = 0x07; break;
                        case 'b': c = 0x08; break;
                        case 't': c = 0x09; break;
                        case 'n': c = 0x0a; break;
                        case 'v': c = 0x0b; break;
                        case 'f': c = 0x0c; break;
                        case 'r': c = 0x0d; break;
                        // TODO \ooo octal
                        case 'x': // TODO \xhh
                        case 'N': // TODO \N{name} only in strings
                        case 'u': // TODO \uxxxx only in strings
                        case 'U': // TODO \Uxxxxxxxx only in strings
                        default: break; // TODO error message
                    }
448
                    if (c != MP_LEXER_CHAR_EOF) {
449
450
451
452
                        vstr_add_char(&lex->vstr, c);
                    }
                } else {
                    vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
453
454
455
456
457
458
459
                }
            }
            next_char(lex);
        }

        // check we got the required end quotes
        if (n_closing < num_quotes) {
460
            tok->kind = MP_TOKEN_LONELY_STRING_OPEN;
Damien's avatar
Damien committed
461
462
        }

463
464
        // cut off the end quotes from the token text
        vstr_cut_tail(&lex->vstr, n_closing);
Damien's avatar
Damien committed
465
466

    } else if (is_head_of_identifier(lex)) {
467
        tok->kind = MP_TOKEN_NAME;
Damien's avatar
Damien committed
468

469
470
        // get first char
        vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
471
472
        next_char(lex);

473
        // get tail chars
Damien's avatar
Damien committed
474
        while (!is_end(lex) && is_tail_of_identifier(lex)) {
475
            vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
476
477
478
479
            next_char(lex);
        }

    } else if (is_digit(lex) || (is_char(lex, '.') && is_following_digit(lex))) {
480
        tok->kind = MP_TOKEN_NUMBER;
Damien's avatar
Damien committed
481

482
483
        // get first char
        vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
484
485
        next_char(lex);

486
        // get tail chars
Damien's avatar
Damien committed
487
488
        while (!is_end(lex)) {
            if (is_char_or(lex, 'e', 'E')) {
489
                vstr_add_char(&lex->vstr, 'e');
Damien's avatar
Damien committed
490
491
                next_char(lex);
                if (is_char(lex, '+') || is_char(lex, '-')) {
492
                    vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
493
494
495
                    next_char(lex);
                }
            } else if (is_letter(lex) || is_digit(lex) || is_char_or(lex, '_', '.')) {
496
                vstr_add_char(&lex->vstr, CUR_CHAR(lex));
Damien's avatar
Damien committed
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
                next_char(lex);
            } else {
                break;
            }
        }

    } else {
        // search for encoded delimiter or operator

        const char *t = tok_enc;
        uint tok_enc_index = 0;
        for (; *t != 0 && !is_char(lex, *t); t += 1) {
            if (*t == 'e' || *t == 'c') {
                t += 1;
            } else if (*t == 'E') {
                tok_enc_index -= 1;
                t += 1;
            }
            tok_enc_index += 1;
        }

        next_char(lex);

        if (*t == 0) {
            // didn't match any delimiter or operator characters
522
            tok->kind = MP_TOKEN_INVALID;
Damien's avatar
Damien committed
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546

        } else {
            // matched a delimiter or operator character

            // get the maximum characters for a valid token
            t += 1;
            uint t_index = tok_enc_index;
            for (;;) {
                for (; *t == 'e'; t += 1) {
                    t += 1;
                    t_index += 1;
                    if (is_char(lex, *t)) {
                        next_char(lex);
                        tok_enc_index = t_index;
                        break;
                    }
                }

                if (*t == 'E') {
                    t += 1;
                    if (is_char(lex, *t)) {
                        next_char(lex);
                        tok_enc_index = t_index;
                    } else {
547
                        tok->kind = MP_TOKEN_INVALID;
Damien's avatar
Damien committed
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
                    }
                    break;
                }

                if (*t == 'c') {
                    t += 1;
                    t_index += 1;
                    if (is_char(lex, *t)) {
                        next_char(lex);
                        tok_enc_index = t_index;
                        t += 1;
                    } else {
                        break;
                    }
                } else {
                    break;
                }
            }

            // set token kind
            tok->kind = tok_enc_kind[tok_enc_index];

            // compute bracket level for implicit line joining
571
            if (tok->kind == MP_TOKEN_DEL_PAREN_OPEN || tok->kind == MP_TOKEN_DEL_BRACKET_OPEN || tok->kind == MP_TOKEN_DEL_BRACE_OPEN) {
Damien's avatar
Damien committed
572
                lex->nested_bracket_level += 1;
573
            } else if (tok->kind == MP_TOKEN_DEL_PAREN_CLOSE || tok->kind == MP_TOKEN_DEL_BRACKET_CLOSE || tok->kind == MP_TOKEN_DEL_BRACE_CLOSE) {
Damien's avatar
Damien committed
574
575
576
577
578
                lex->nested_bracket_level -= 1;
            }
        }
    }

579
580
581
    // point token text to vstr buffer
    tok->str = vstr_str(&lex->vstr);
    tok->len = vstr_len(&lex->vstr);
Damien's avatar
Damien committed
582

583
    // check for keywords
584
    if (tok->kind == MP_TOKEN_NAME) {
Damien's avatar
Damien committed
585
        for (int i = 0; tok_kw[i] != NULL; i++) {
586
            if (str_strn_equal(tok_kw[i], tok->str, tok->len)) {
587
                tok->kind = MP_TOKEN_KW_FALSE + i;
Damien's avatar
Damien committed
588
589
590
591
592
593
                break;
            }
        }
    }
}

594
595
mp_lexer_t *mp_lexer_new(const char *src_name, void *stream_data, mp_lexer_stream_next_char_t stream_next_char, mp_lexer_stream_close_t stream_close) {
    mp_lexer_t *lex = m_new(mp_lexer_t, 1);
Damien's avatar
Damien committed
596

597
598
599
    lex->name = src_name; // TODO do we need to strdup this?
    lex->stream_data = stream_data;
    lex->stream_next_char = stream_next_char;
600
    lex->stream_close = stream_close;
Damien's avatar
Damien committed
601
602
603
604
605
606
607
608
    lex->line = 1;
    lex->column = 1;
    lex->emit_dent = 0;
    lex->nested_bracket_level = 0;
    lex->alloc_indent_level = 16;
    lex->num_indent_level = 1;
    lex->indent_level = m_new(uint16_t, lex->alloc_indent_level);
    lex->indent_level[0] = 0;
609
    vstr_init(&lex->vstr, 32);
Damien's avatar
Damien committed
610
611

    // preload characters
612
613
614
615
616
    lex->chr0 = stream_next_char(stream_data);
    lex->chr1 = stream_next_char(stream_data);
    lex->chr2 = stream_next_char(stream_data);

    // if input stream is 0, 1 or 2 characters long and doesn't end in a newline, then insert a newline at the end
617
    if (lex->chr0 == MP_LEXER_CHAR_EOF) {
618
        lex->chr0 = '\n';
619
    } else if (lex->chr1 == MP_LEXER_CHAR_EOF) {
Damien's avatar
Damien committed
620
        if (lex->chr0 != '\n' && lex->chr0 != '\r') {
621
            lex->chr1 = '\n';
Damien's avatar
Damien committed
622
        }
623
    } else if (lex->chr2 == MP_LEXER_CHAR_EOF) {
Damien's avatar
Damien committed
624
        if (lex->chr1 != '\n' && lex->chr1 != '\r') {
625
            lex->chr2 = '\n';
Damien's avatar
Damien committed
626
627
628
        }
    }

629
    // preload first token
630
    mp_lexer_next_token_into(lex, &lex->tok_cur, true);
Damien's avatar
Damien committed
631
632
633
634

    return lex;
}

635
void mp_lexer_free(mp_lexer_t *lex) {
636
    if (lex) {
637
638
        if (lex->stream_close) {
            lex->stream_close(lex->stream_data);
639
        }
640
        vstr_clear(&lex->vstr);
641
        m_del_obj(mp_lexer_t, lex);
Damien's avatar
Damien committed
642
643
644
    }
}

645
646
void mp_lexer_to_next(mp_lexer_t *lex) {
    mp_lexer_next_token_into(lex, &lex->tok_cur, false);
Damien's avatar
Damien committed
647
648
}

649
const mp_token_t *mp_lexer_cur(const mp_lexer_t *lex) {
Damien's avatar
Damien committed
650
651
652
    return &lex->tok_cur;
}

653
bool mp_lexer_is_kind(mp_lexer_t *lex, mp_token_kind_t kind) {
Damien's avatar
Damien committed
654
655
656
657
    return lex->tok_cur.kind == kind;
}

/*
658
659
bool mp_lexer_is_str(mp_lexer_t *lex, const char *str) {
    return mp_token_is_str(&lex->tok_cur, str);
Damien's avatar
Damien committed
660
661
}

662
663
664
bool mp_lexer_opt_kind(mp_lexer_t *lex, mp_token_kind_t kind) {
    if (mp_lexer_is_kind(lex, kind)) {
        mp_lexer_to_next(lex);
Damien's avatar
Damien committed
665
666
667
668
669
        return true;
    }
    return false;
}

670
671
672
bool mp_lexer_opt_str(mp_lexer_t *lex, const char *str) {
    if (mp_lexer_is_str(lex, str)) {
        mp_lexer_to_next(lex);
Damien's avatar
Damien committed
673
674
675
676
677
678
        return true;
    }
    return false;
}
*/

679
680
681
bool mp_lexer_show_error_pythonic_prefix(mp_lexer_t *lex) {
    printf("  File \"%s\", line %d column %d\n", lex->tok_cur.src_name, lex->tok_cur.src_line, lex->tok_cur.src_column);
    return false;
Damien's avatar
Damien committed
682
}
683

684
bool mp_lexer_show_error_pythonic(mp_lexer_t *lex, const char *msg) {
685
686
687
    printf("  File \"%s\", line %d column %d\n%s\n", lex->tok_cur.src_name, lex->tok_cur.src_line, lex->tok_cur.src_column, msg);
    return false;
}