aboutsummaryrefslogtreecommitdiffhomepage
path: root/src/parser.cpp
blob: 6b19f349653a91396ae6cf84cad03e4d363d68cc (plain)
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
// The fish parser. Contains functions for parsing and evaluating code.
#include "config.h"  // IWYU pragma: keep

#include <assert.h>
#include <stdbool.h>
#include <stdio.h>
#include <wchar.h>
#include <algorithm>
#include <memory>

#include "common.h"
#include "env.h"
#include "event.h"
#include "expand.h"
#include "fallback.h"  // IWYU pragma: keep
#include "function.h"
#include "intern.h"
#include "parse_constants.h"
#include "parse_execution.h"
#include "parse_tree.h"
#include "parse_util.h"
#include "parser.h"
#include "proc.h"
#include "reader.h"
#include "sanity.h"
#include "wutil.h"  // IWYU pragma: keep

class io_chain_t;

/// Error for evaluating in illegal scope.
#define INVALID_SCOPE_ERR_MSG _(L"Tried to evaluate commands using invalid block type '%ls'")

/// While block description.
#define WHILE_BLOCK N_(L"'while' block")

/// For block description.
#define FOR_BLOCK N_(L"'for' block")

/// Breakpoint block.
#define BREAKPOINT_BLOCK N_(L"block created by breakpoint")

/// If block description.
#define IF_BLOCK N_(L"'if' conditional block")

/// Function definition block description.
#define FUNCTION_DEF_BLOCK N_(L"function definition block")

/// Function invocation block description.
#define FUNCTION_CALL_BLOCK N_(L"function invocation block")

/// Function invocation block description.
#define FUNCTION_CALL_NO_SHADOW_BLOCK N_(L"function invocation block with no variable shadowing")

/// Switch block description.
#define SWITCH_BLOCK N_(L"'switch' block")

/// Fake block description.
#define FAKE_BLOCK N_(L"unexecutable block")

/// Top block description.
#define TOP_BLOCK N_(L"global root block")

/// Command substitution block description.
#define SUBST_BLOCK N_(L"command substitution block")

/// Begin block description.
#define BEGIN_BLOCK N_(L"'begin' unconditional block")

/// Source block description.
#define SOURCE_BLOCK N_(L"block created by the . builtin")

/// Source block description.
#define EVENT_BLOCK N_(L"event handler block")

/// Unknown block description.
#define UNKNOWN_BLOCK N_(L"unknown/invalid block")

/// Data structure to describe a block type, like while blocks, command substitution blocks, etc.
struct block_lookup_entry {
    // The block type id. The legal values are defined in parser.h.
    block_type_t type;
    // The name of the builtin that creates this type of block, if any.
    const wchar_t *name;
    // A description of this block type.
    const wchar_t *desc;
};

/// List of all legal block types.
static const struct block_lookup_entry block_lookup[] = {
    {WHILE, L"while", WHILE_BLOCK},
    {FOR, L"for", FOR_BLOCK},
    {IF, L"if", IF_BLOCK},
    {FUNCTION_DEF, L"function", FUNCTION_DEF_BLOCK},
    {FUNCTION_CALL, 0, FUNCTION_CALL_BLOCK},
    {FUNCTION_CALL_NO_SHADOW, 0, FUNCTION_CALL_NO_SHADOW_BLOCK},
    {SWITCH, L"switch", SWITCH_BLOCK},
    {FAKE, 0, FAKE_BLOCK},
    {TOP, 0, TOP_BLOCK},
    {SUBST, 0, SUBST_BLOCK},
    {BEGIN, L"begin", BEGIN_BLOCK},
    {SOURCE, L".", SOURCE_BLOCK},
    {EVENT, 0, EVENT_BLOCK},
    {BREAKPOINT, L"breakpoint", BREAKPOINT_BLOCK},
    {(block_type_t)0, 0, 0}};

// Given a file path, return something nicer. Currently we just "unexpand" tildes.
static wcstring user_presentable_path(const wcstring &path) {
    return replace_home_directory_with_tilde(path);
}

parser_t::parser_t() : cancellation_requested(false), is_within_fish_initialization(false) {}

/// A pointer to the principal parser (which is a static local).
static parser_t *s_principal_parser = NULL;

parser_t &parser_t::principal_parser(void) {
    ASSERT_IS_NOT_FORKED_CHILD();
    ASSERT_IS_MAIN_THREAD();
    static parser_t parser;
    if (!s_principal_parser) {
        s_principal_parser = &parser;
    }
    return parser;
}

void parser_t::set_is_within_fish_initialization(bool flag) {
    is_within_fish_initialization = flag;
}

void parser_t::skip_all_blocks(void) {
    // Tell all blocks to skip.
    if (s_principal_parser) {
        s_principal_parser->cancellation_requested = true;

        // write(2, "Cancelling blocks\n", strlen("Cancelling blocks\n"));
        for (size_t i = 0; i < s_principal_parser->block_count(); i++) {
            s_principal_parser->block_at_index(i)->skip = true;
        }
    }
}

void parser_t::push_block(block_t *new_current) {
    const enum block_type_t type = new_current->type();
    new_current->src_lineno = parser_t::get_lineno();

    const wchar_t *filename = parser_t::current_filename();
    if (filename != NULL) {
        new_current->src_filename = intern(filename);
    }

    const block_t *old_current = this->current_block();
    if (old_current && old_current->skip) {
        new_current->skip = true;
    }

    // New blocks should be skipped if the outer block is skipped, except TOP ans SUBST block, which
    // open up new environments. Fake blocks should always be skipped. Rather complicated... :-(
    new_current->skip = old_current ? old_current->skip : 0;

    // Type TOP and SUBST are never skipped.
    if (type == TOP || type == SUBST) {
        new_current->skip = 0;
    }

    // Fake blocks and function definition blocks are never executed.
    if (type == FAKE || type == FUNCTION_DEF) {
        new_current->skip = 1;
    }

    new_current->job = 0;
    new_current->loop_status = LOOP_NORMAL;

    this->block_stack.push_back(new_current);

    // Types TOP and SUBST are not considered blocks for the purposes of `status -b`.
    if (type != TOP && type != SUBST) {
        is_block = 1;
    }

    if ((new_current->type() != FUNCTION_DEF) && (new_current->type() != FAKE) &&
        (new_current->type() != TOP)) {
        env_push(type == FUNCTION_CALL);
        new_current->wants_pop_env = true;
    }
}

void parser_t::pop_block() {
    if (block_stack.empty()) {
        debug(1, L"function %s called on empty block stack.", __func__);
        bugreport();
        return;
    }

    block_t *old = block_stack.back();
    block_stack.pop_back();

    if (old->wants_pop_env) env_pop();

    delete old;

    // Figure out if `status -b` should consider us to be in a block now.
    int new_is_block = 0;
    for (std::vector<block_t *>::const_iterator it = block_stack.begin(), end = block_stack.end();
         it != end; ++it) {
        const enum block_type_t type = (*it)->type();
        if (type != TOP && type != SUBST) {
            new_is_block = 1;
            break;
        }
    }
    is_block = new_is_block;
}

void parser_t::pop_block(const block_t *expected) {
    assert(expected == this->current_block());
    this->pop_block();
}

const wchar_t *parser_t::get_block_desc(int block) const {
    for (size_t i = 0; block_lookup[i].desc; i++) {
        if (block_lookup[i].type == block) {
            return _(block_lookup[i].desc);
        }
    }
    return _(UNKNOWN_BLOCK);
}

#if 0
// TODO: Lint says this isn't used (which is true). Should this be removed?
wcstring parser_t::block_stack_description() const {
    wcstring result;
    size_t idx = this->block_count();
    size_t spaces = 0;
    while (idx--) {
        if (spaces > 0) {
            result.push_back(L'\n');
        }
        for (size_t j = 0; j < spaces; j++) {
            result.push_back(L' ');
        }
        result.append(this->block_at_index(idx)->description());
        spaces++;
    }
    return result;
}
#endif

const block_t *parser_t::block_at_index(size_t idx) const {
    // Zero corresponds to the last element in our vector.
    size_t count = block_stack.size();
    return idx < count ? block_stack.at(count - idx - 1) : NULL;
}

block_t *parser_t::block_at_index(size_t idx) {
    size_t count = block_stack.size();
    return idx < count ? block_stack.at(count - idx - 1) : NULL;
}

block_t *const parser_t::current_block() { return block_stack.empty() ? NULL : block_stack.back(); }

void parser_t::forbid_function(const wcstring &function) { forbidden_function.push_back(function); }

void parser_t::allow_function() { forbidden_function.pop_back(); }

/// Print profiling information to the specified stream.
static void print_profile(const std::vector<profile_item_t *> &items, FILE *out) {
    for (size_t pos = 0; pos < items.size(); pos++) {
        const profile_item_t *me, *prev;
        size_t i;
        int my_time;

        me = items.at(pos);
        if (!me->skipped) {
            my_time = me->parse + me->exec;

            for (i = pos + 1; i < items.size(); i++) {
                prev = items.at(i);
                if (prev->skipped) {
                    continue;
                }

                if (prev->level <= me->level) {
                    break;
                }

                if (prev->level > me->level + 1) {
                    continue;
                }

                my_time -= prev->parse;
                my_time -= prev->exec;
            }

            if (me->cmd.size() > 0) {
                if (fwprintf(out, L"%d\t%d\t", my_time, me->parse + me->exec) < 0) {
                    wperror(L"fwprintf");
                    return;
                }

                for (i = 0; i < me->level; i++) {
                    if (fwprintf(out, L"-") < 0) {
                        wperror(L"fwprintf");
                        return;
                    }
                }
                if (fwprintf(out, L"> %ls\n", me->cmd.c_str()) < 0) {
                    wperror(L"fwprintf");
                    return;
                }
            }
        }
    }
}

void parser_t::emit_profiling(const char *path) const {
    // Save profiling information. OK to not use CLO_EXEC here because this is called while fish is
    // dying (and hence will not fork).
    FILE *f = fopen(path, "w");
    if (!f) {
        debug(1, _(L"Could not write profiling information to file '%s'"), path);
    } else {
        if (fwprintf(f, _(L"Time\tSum\tCommand\n"), profile_items.size()) < 0) {
            wperror(L"fwprintf");
        } else {
            print_profile(profile_items, f);
        }

        if (fclose(f)) {
            wperror(L"fclose");
        }
    }
}

void parser_t::expand_argument_list(const wcstring &arg_list_src, expand_flags_t eflags,
                                    std::vector<completion_t> *output_arg_list) {
    assert(output_arg_list != NULL);

    // Parse the string as an argument list.
    parse_node_tree_t tree;
    if (!parse_tree_from_string(arg_list_src, parse_flag_none, &tree, NULL /* errors */,
                                symbol_freestanding_argument_list)) {
        // Failed to parse. Here we expect to have reported any errors in test_args.
        return;
    }

    // Get the root argument list.
    assert(!tree.empty());
    const parse_node_t *arg_list = &tree.at(0);
    assert(arg_list->type == symbol_freestanding_argument_list);

    // Extract arguments from it.
    while (arg_list != NULL) {
        const parse_node_t *arg_node =
            tree.next_node_in_node_list(*arg_list, symbol_argument, &arg_list);
        if (arg_node != NULL) {
            const wcstring arg_src = arg_node->get_source(arg_list_src);
            if (expand_string(arg_src, output_arg_list, eflags, NULL) == EXPAND_ERROR) {
                break;  // failed to expand a string
            }
        }
    }
}

wcstring parser_t::stack_trace() const {
    wcstring trace;
    this->stack_trace_internal(0, &trace);
    return trace;
}

void parser_t::stack_trace_internal(size_t block_idx, wcstring *buff) const {
    // Check if we should end the recursion.
    if (block_idx >= this->block_count()) return;

    const block_t *b = this->block_at_index(block_idx);

    if (b->type() == EVENT) {
        // This is an event handler.
        const event_block_t *eb = static_cast<const event_block_t *>(b);
        wcstring description = event_get_desc(eb->event);
        append_format(*buff, _(L"in event handler: %ls\n"), description.c_str());
        buff->append(L"\n");

        // Stop recursing at event handler. No reason to believe that any other code is relevant.
        //
        // It might make sense in the future to continue printing the stack trace of the code that
        // invoked the event, if this is a programmatic event, but we can't currently detect that.
        return;
    }

    if (b->type() == FUNCTION_CALL || b->type() == FUNCTION_CALL_NO_SHADOW || b->type() == SOURCE ||
        b->type() == SUBST) {
        // These types of blocks should be printed.
        int i;

        switch (b->type()) {
            case SOURCE: {
                const source_block_t *sb = static_cast<const source_block_t *>(b);
                const wchar_t *source_dest = sb->source_file;
                append_format(*buff, _(L"from sourcing file %ls\n"),
                              user_presentable_path(source_dest).c_str());
                break;
            }
            case FUNCTION_CALL:
            case FUNCTION_CALL_NO_SHADOW: {
                const function_block_t *fb = static_cast<const function_block_t *>(b);
                append_format(*buff, _(L"in function '%ls'\n"), fb->name.c_str());
                break;
            }
            case SUBST: {
                append_format(*buff, _(L"in command substitution\n"));
                break;
            }
            default: {
                break;  // can't get here
            }
        }

        const wchar_t *file = b->src_filename;

        if (file) {
            append_format(*buff, _(L"\tcalled on line %d of file %ls\n"), b->src_lineno,
                          user_presentable_path(file).c_str());
        } else if (is_within_fish_initialization) {
            append_format(*buff, _(L"\tcalled during startup\n"));
        } else {
            append_format(*buff, _(L"\tcalled on standard input\n"));
        }

        if (b->type() == FUNCTION_CALL) {
            const function_block_t *fb = static_cast<const function_block_t *>(b);
            const process_t *const process = fb->process;
            if (process->argv(1)) {
                wcstring tmp;

                for (i = 1; process->argv(i); i++) {
                    if (i > 1) tmp.push_back(L' ');
                    tmp.append(process->argv(i));
                }
                append_format(*buff, _(L"\twith parameter list '%ls'\n"), tmp.c_str());
            }
        }

        append_format(*buff, L"\n");
    }

    // Recursively print the next block.
    parser_t::stack_trace_internal(block_idx + 1, buff);
}

/// Returns the name of the currently evaluated function if we are currently evaluating a function,
/// null otherwise. This is tested by moving down the block-scope-stack, checking every block if it
/// is of type FUNCTION_CALL.
const wchar_t *parser_t::is_function() const {
    // PCA: Have to make this a string somehow.
    ASSERT_IS_MAIN_THREAD();

    const wchar_t *result = NULL;
    for (size_t block_idx = 0; block_idx < this->block_count(); block_idx++) {
        const block_t *b = this->block_at_index(block_idx);
        if (b->type() == FUNCTION_CALL || b->type() == FUNCTION_CALL_NO_SHADOW) {
            const function_block_t *fb = static_cast<const function_block_t *>(b);
            result = fb->name.c_str();
            break;
        } else if (b->type() == SOURCE) {
            // If a function sources a file, obviously that function's offset doesn't contribute.
            break;
        }
    }
    return result;
}

int parser_t::get_lineno() const {
    int lineno = -1;
    if (!execution_contexts.empty()) {
        lineno = execution_contexts.back()->get_current_line_number();

        // If we are executing a function, we have to add in its offset.
        const wchar_t *function_name = is_function();
        if (function_name != NULL) {
            lineno += function_get_definition_offset(function_name);
        }
    }
    return lineno;
}

const wchar_t *parser_t::current_filename() const {
    ASSERT_IS_MAIN_THREAD();

    for (size_t i = 0; i < this->block_count(); i++) {
        const block_t *b = this->block_at_index(i);
        if (b->type() == FUNCTION_CALL || b->type() == FUNCTION_CALL_NO_SHADOW) {
            const function_block_t *fb = static_cast<const function_block_t *>(b);
            return function_get_definition_file(fb->name);
        } else if (b->type() == SOURCE) {
            const source_block_t *sb = static_cast<const source_block_t *>(b);
            return sb->source_file;
        }
    }

    // We query a global array for the current file name, but only do that if we are the principal
    // parser.
    if (this == &principal_parser()) {
        return reader_current_filename();
    }
    return NULL;
}

wcstring parser_t::current_line() {
    if (execution_contexts.empty()) {
        return wcstring();
    }
    const parse_execution_context_t *context = execution_contexts.back();
    assert(context != NULL);

    int source_offset = context->get_current_source_offset();
    if (source_offset < 0) {
        return wcstring();
    }

    const int lineno = this->get_lineno();
    const wchar_t *file = this->current_filename();

    wcstring prefix;

    // If we are not going to print a stack trace, at least print the line number and filename.
    if (!shell_is_interactive() || is_function()) {
        if (file) {
            append_format(prefix, _(L"%ls (line %d): "), user_presentable_path(file).c_str(),
                          lineno);
        } else if (is_within_fish_initialization) {
            append_format(prefix, L"%ls: ", _(L"Startup"), lineno);
        } else {
            append_format(prefix, L"%ls: ", _(L"Standard input"), lineno);
        }
    }

    bool is_interactive = shell_is_interactive();
    bool skip_caret = is_interactive && !is_function();

    // Use an error with empty text.
    assert(source_offset >= 0);
    parse_error_t empty_error = {};
    empty_error.source_start = source_offset;

    wcstring line_info =
        empty_error.describe_with_prefix(context->get_source(), prefix, is_interactive, skip_caret);
    if (!line_info.empty()) {
        line_info.push_back(L'\n');
    }

    line_info.append(this->stack_trace());
    return line_info;
}

void parser_t::job_add(job_t *job) {
    assert(job != NULL);
    assert(job->first_process != NULL);
    this->my_job_list.push_front(job);
}

bool parser_t::job_remove(job_t *j) {
    job_list_t::iterator iter = std::find(my_job_list.begin(), my_job_list.end(), j);
    if (iter != my_job_list.end()) {
        my_job_list.erase(iter);
        return true;
    } else {
        debug(1, _(L"Job inconsistency"));
        sanity_lose();
        return false;
    }
}

void parser_t::job_promote(job_t *job) {
    job_list_t::iterator loc = std::find(my_job_list.begin(), my_job_list.end(), job);
    assert(loc != my_job_list.end());

    // Move the job to the beginning.
    my_job_list.splice(my_job_list.begin(), my_job_list, loc);
}

job_t *parser_t::job_get(job_id_t id) {
    job_iterator_t jobs(my_job_list);
    job_t *job;
    while ((job = jobs.next())) {
        if (id <= 0 || job->job_id == id) return job;
    }
    return NULL;
}

job_t *parser_t::job_get_from_pid(int pid) {
    job_iterator_t jobs;
    job_t *job;
    while ((job = jobs.next())) {
        if (job->pgid == pid) return job;
    }
    return 0;
}

profile_item_t *parser_t::create_profile_item() {
    profile_item_t *result = NULL;
    if (g_profiling_active) {
        result = new profile_item_t();
        profile_items.push_back(result);
    }
    return result;
}

int parser_t::eval(const wcstring &cmd, const io_chain_t &io, enum block_type_t block_type) {
    // Parse the source into a tree, if we can.
    parse_node_tree_t tree;
    parse_error_list_t error_list;
    if (!parse_tree_from_string(cmd, parse_flag_none, &tree, &error_list)) {
        // Get a backtrace. This includes the message.
        wcstring backtrace_and_desc;
        this->get_backtrace(cmd, error_list, &backtrace_and_desc);

        // Print it.
        fprintf(stderr, "%ls", backtrace_and_desc.c_str());

        return 1;
    }
    return this->eval_acquiring_tree(cmd, io, block_type, moved_ref<parse_node_tree_t>(tree));
}

int parser_t::eval_acquiring_tree(const wcstring &cmd, const io_chain_t &io,
                                  enum block_type_t block_type, moved_ref<parse_node_tree_t> tree) {
    CHECK_BLOCK(1);
    assert(block_type == TOP || block_type == SUBST);

    if (tree.val.empty()) {
        return 0;
    }

    // Determine the initial eval level. If this is the first context, it's -1; otherwise it's the
    // eval level of the top context. This is sort of wonky because we're stitching together a
    // global notion of eval level from these separate objects. A better approach would be some
    // profile object that all contexts share, and that tracks the eval levels on its own.
    int exec_eval_level =
        (execution_contexts.empty() ? -1 : execution_contexts.back()->current_eval_level());

    // Append to the execution context stack.
    parse_execution_context_t *ctx =
        new parse_execution_context_t(tree, cmd, this, exec_eval_level);
    execution_contexts.push_back(ctx);

    // Execute the first node.
    this->eval_block_node(0, io, block_type);

    // Clean up the execution context stack.
    assert(!execution_contexts.empty() && execution_contexts.back() == ctx);
    execution_contexts.pop_back();
    delete ctx;

    return 0;
}

int parser_t::eval_block_node(node_offset_t node_idx, const io_chain_t &io,
                              enum block_type_t block_type) {
    // Paranoia. It's a little frightening that we're given only a node_idx and we interpret this in
    // the topmost execution context's tree. What happens if two trees were to be interleaved?
    // Fortunately that cannot happen (yet); in the future we probably want some sort of reference
    // counted trees.
    parse_execution_context_t *ctx = execution_contexts.back();
    assert(ctx != NULL);

    CHECK_BLOCK(1);

    // Handle cancellation requests. If our block stack is currently empty, then we already did
    // successfully cancel (or there was nothing to cancel); clear the flag. If our block stack is
    // not empty, we are still in the process of cancelling; refuse to evaluate anything.
    if (this->cancellation_requested) {
        if (!block_stack.empty()) {
            return 1;
        }
        this->cancellation_requested = false;
    }

    // Only certain blocks are allowed.
    if ((block_type != TOP) && (block_type != SUBST)) {
        debug(1, INVALID_SCOPE_ERR_MSG, parser_t::get_block_desc(block_type));
        bugreport();
        return 1;
    }

    job_reap(0);  // not sure why we reap jobs here

    /* Start it up */
    const block_t *const start_current_block = current_block();
    block_t *scope_block = new scope_block_t(block_type);
    this->push_block(scope_block);
    int result = ctx->eval_node_at_offset(node_idx, scope_block, io);

    // Clean up the block stack.
    this->pop_block();
    while (start_current_block != current_block()) {
        if (current_block() == NULL) {
            debug(0, _(L"End of block mismatch. Program terminating."));
            bugreport();
            FATAL_EXIT();
            break;
        }
        this->pop_block();
    }

    job_reap(0);  // reap again

    return result;
}

bool parser_t::detect_errors_in_argument_list(const wcstring &arg_list_src, wcstring *out,
                                              const wchar_t *prefix) {
    bool errored = false;
    parse_error_list_t errors;

    // Use empty string for the prefix if it's NULL.
    if (prefix == NULL) {
        prefix = L"";
    }

    // Parse the string as an argument list.
    parse_node_tree_t tree;
    if (!parse_tree_from_string(arg_list_src, parse_flag_none, &tree, &errors,
                                symbol_freestanding_argument_list)) {
        // Failed to parse.
        errored = true;
    }

    if (!errored) {
        // Get the root argument list.
        assert(!tree.empty());
        const parse_node_t *arg_list = &tree.at(0);
        assert(arg_list->type == symbol_freestanding_argument_list);

        // Extract arguments from it.
        while (arg_list != NULL && !errored) {
            const parse_node_t *arg_node =
                tree.next_node_in_node_list(*arg_list, symbol_argument, &arg_list);
            if (arg_node != NULL) {
                const wcstring arg_src = arg_node->get_source(arg_list_src);
                if (parse_util_detect_errors_in_argument(*arg_node, arg_src, &errors)) {
                    errored = true;
                }
            }
        }
    }

    if (!errors.empty() && out != NULL) {
        out->assign(errors.at(0).describe_with_prefix(
            arg_list_src, prefix, false /* not interactive */, false /* don't skip caret */));
    }
    return errored;
}

void parser_t::get_backtrace(const wcstring &src, const parse_error_list_t &errors,
                             wcstring *output) const {
    assert(output != NULL);
    if (!errors.empty()) {
        const parse_error_t &err = errors.at(0);

        const bool is_interactive = shell_is_interactive();

        // Determine if we want to try to print a caret to point at the source error. The
        // err.source_start <= src.size() check is due to the nasty way that slices work, which is
        // by rewriting the source.
        size_t which_line = 0;
        bool skip_caret = true;
        if (err.source_start != SOURCE_LOCATION_UNKNOWN && err.source_start <= src.size()) {
            // Determine which line we're on.
            which_line = 1 + std::count(src.begin(), src.begin() + err.source_start, L'\n');

            // Don't include the caret if we're interactive, this is the first line of text, and our
            // source is at its beginning, because then it's obvious.
            skip_caret = (is_interactive && which_line == 1 && err.source_start == 0);
        }

        wcstring prefix;
        const wchar_t *filename = this->current_filename();
        if (filename) {
            if (which_line > 0) {
                prefix = format_string(_(L"%ls (line %lu): "),
                                       user_presentable_path(filename).c_str(), which_line);
            } else {
                prefix = format_string(_(L"%ls: "), user_presentable_path(filename).c_str());
            }
        } else {
            prefix = L"fish: ";
        }

        const wcstring description =
            err.describe_with_prefix(src, prefix, is_interactive, skip_caret);
        if (!description.empty()) {
            output->append(description);
            output->push_back(L'\n');
        }
        output->append(this->stack_trace());
    }
}

block_t::block_t(block_type_t t)
    : block_type(t),
      skip(),
      tok_pos(),
      node_offset(NODE_OFFSET_INVALID),
      loop_status(LOOP_NORMAL),
      job(),
      src_filename(),
      src_lineno(),
      wants_pop_env(false),
      event_blocks() {}

block_t::~block_t() {}

wcstring block_t::description() const {
    wcstring result;
    switch (this->type()) {
        case WHILE: {
            result.append(L"while");
            break;
        }
        case FOR: {
            result.append(L"for");
            break;
        }
        case IF: {
            result.append(L"if");
            break;
        }
        case FUNCTION_DEF: {
            result.append(L"function_def");
            break;
        }
        case FUNCTION_CALL: {
            result.append(L"function_call");
            break;
        }
        case FUNCTION_CALL_NO_SHADOW: {
            result.append(L"function_call_no_shadow");
            break;
        }
        case SWITCH: {
            result.append(L"switch");
            break;
        }
        case FAKE: {
            result.append(L"fake");
            break;
        }
        case SUBST: {
            result.append(L"substitution");
            break;
        }
        case TOP: {
            result.append(L"top");
            break;
        }
        case BEGIN: {
            result.append(L"begin");
            break;
        }
        case SOURCE: {
            result.append(L"source");
            break;
        }
        case EVENT: {
            result.append(L"event");
            break;
        }
        case BREAKPOINT: {
            result.append(L"breakpoint");
            break;
        }
        default: {
            append_format(result, L"unknown type %ld", (long)this->type());
            break;
        }
    }

    if (this->src_lineno >= 0) {
        append_format(result, L" (line %d)", this->src_lineno);
    }
    if (this->src_filename != NULL) {
        append_format(result, L" (file %ls)", this->src_filename);
    }
    return result;
}

// Various block constructors.

if_block_t::if_block_t() : block_t(IF) {}

event_block_t::event_block_t(const event_t &evt) : block_t(EVENT), event(evt) {}

function_block_t::function_block_t(const process_t *p, const wcstring &n, bool shadows)
    : block_t(shadows ? FUNCTION_CALL : FUNCTION_CALL_NO_SHADOW), process(p), name(n) {}

source_block_t::source_block_t(const wchar_t *src) : block_t(SOURCE), source_file(src) {}

for_block_t::for_block_t() : block_t(FOR) {}

while_block_t::while_block_t() : block_t(WHILE) {}

switch_block_t::switch_block_t() : block_t(SWITCH) {}

fake_block_t::fake_block_t() : block_t(FAKE) {}

scope_block_t::scope_block_t(block_type_t type) : block_t(type) {
    assert(type == BEGIN || type == TOP || type == SUBST);
}

breakpoint_block_t::breakpoint_block_t() : block_t(BREAKPOINT) {}