aboutsummaryrefslogtreecommitdiffstats
path: root/src/tokenizer.rs
blob: 19e3392cd40a44d93b188d5d42d83cc70e67c600 (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
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
//! The tokenizer glues states from the state machine together.
//!
//! It facilitates everything needed to turn codes into tokens and events with
//! a state machine.
//! It also enables logic needed for parsing markdown, such as an [`attempt`][]
//! to parse something, which can succeed or, when unsuccessful, revert the
//! attempt.
//! Similarly, a [`check`][] exists, which does the same as an `attempt` but
//! reverts even if successful.
//!
//! [`attempt`]: Tokenizer::attempt
//! [`check`]: Tokenizer::check

use crate::parser::ParseState;
use crate::token::{Token, VOID_TOKENS};
use crate::util::edit_map::EditMap;

/// Embedded content type.
#[derive(Debug, Clone, PartialEq)]
pub enum ContentType {
    /// Represents [text content][crate::content::text].
    Text,
    /// Represents [string content][crate::content::string].
    String,
}

/// Enum representing a character code.
#[derive(Debug, Clone, Copy, PartialEq)]
pub enum Code {
    /// End of the input stream (called eof).
    None,
    /// Used to make parsing line endings easier as it represents both
    /// `Code::Char('\r')` and `Code::Char('\n')` combined.
    CarriageReturnLineFeed,
    /// the expansion of a tab (`Code::Char('\t')`), depending on where the tab
    /// ocurred, it’s followed by 0 to 3 (both inclusive) `Code::VirtualSpace`s.
    VirtualSpace,
    /// The most frequent variant of this enum is `Code::Char(char)`, which just
    /// represents a char, but micromark adds meaning to certain other values.
    Char(char),
}

/// A location in the document (`line`/`column`/`offset`).
///
/// The interface for the location in the document comes from unist `Point`:
/// <https://github.com/syntax-tree/unist#point>.
#[derive(Debug, Clone)]
pub struct Point {
    /// 1-indexed line number.
    pub line: usize,
    /// 1-indexed column number.
    /// This is increases up to a tab stop for tabs.
    /// Some editors count tabs as 1 character, so this position is not always
    /// the same as editors.
    pub column: usize,
    /// 0-indexed position in the document.
    pub offset: usize,
    /// Index into `codes`.
    pub index: usize,
}

/// Possible event types.
#[derive(Debug, PartialEq, Clone)]
pub enum EventType {
    /// The start of something.
    Enter,
    /// The end of something.
    Exit,
}

/// A link to another event.
#[derive(Debug, Clone)]
pub struct Link {
    pub previous: Option<usize>,
    pub next: Option<usize>,
    pub content_type: ContentType,
}

/// Something semantic happening somewhere.
#[derive(Debug, Clone)]
pub struct Event {
    pub event_type: EventType,
    pub token_type: Token,
    pub point: Point,
    pub link: Option<Link>,
}

/// The essence of the state machine are functions: `StateFn`.
/// It’s responsible for dealing with that single passed [`Code`][].
/// It yields a [`State`][].
pub type StateFn = dyn FnOnce(&mut Tokenizer) -> State;

/// Callback that can be registered and is called when the tokenizer is done.
///
/// Resolvers are supposed to change the list of events, because parsing is
/// sometimes messy, and they help expose a cleaner interface of events to
/// the compiler and other users.
pub type Resolver = dyn FnOnce(&mut Tokenizer);

/// The result of a state.
pub enum State {
    /// There is a future state: a boxed [`StateFn`][] to pass the next code to.
    Fn(Box<StateFn>),
    /// The state is successful.
    Ok,
    /// The state is not successful.
    Nok,
}

/// Loose label starts we found.
#[derive(Debug)]
pub struct LabelStart {
    /// Indices of where the label starts and ends in `events`.
    pub start: (usize, usize),
    /// A boolean used internally to figure out if a label start link can’t be
    /// used (because links in links are incorrect).
    pub inactive: bool,
    /// A boolean used internally to figure out if a label is balanced: they’re
    /// not media, it’s just balanced braces.
    pub balanced: bool,
}

/// Media we found.
#[derive(Debug)]
pub struct Media {
    /// Indices of where the media’s label start starts and ends in `events`.
    pub start: (usize, usize),
    /// Indices of where the media’s label end starts and ends in `events`.
    pub end: (usize, usize),
    /// Identifier
    pub id: String,
}

/// Supported containers.
#[derive(Debug, PartialEq)]
pub enum Container {
    BlockQuote,
    ListItem,
}

/// Info used to tokenize the current container.
///
/// This info is shared between the initial construct and its continuation.
/// It’s only used for list items.
#[derive(Debug)]
pub struct ContainerState {
    /// Kind.
    pub kind: Container,
    /// Whether the first line was blank.
    pub blank_initial: bool,
    /// The size of the initial construct.
    pub size: usize,
}

/// The internal state of a tokenizer, not to be confused with states from the
/// state machine, this instead is all the information about where we currently
/// are and what’s going on.
#[derive(Debug, Clone)]
struct InternalState {
    /// Length of `events`. We only add to events, so reverting will just pop stuff off.
    events_len: usize,
    /// Length of the stack. It’s not allowed to decrease the stack in a check or an attempt.
    stack_len: usize,
    /// Previous code.
    previous: Code,
    /// Current code.
    current: Code,
    /// Current relative and absolute position in the file.
    point: Point,
}

/// A tokenizer itself.
#[allow(clippy::struct_excessive_bools)]
pub struct Tokenizer<'a> {
    /// Jump between line endings.
    column_start: Vec<usize>,
    // First line.
    line_start: usize,
    /// Track whether a character is expected to be consumed, and whether it’s
    /// actually consumed
    ///
    /// Tracked to make sure everything’s valid.
    consumed: bool,
    /// Track whether this tokenizer is done.
    drained: bool,
    /// Current character code.
    pub current: Code,
    /// Previous character code.
    pub previous: Code,
    /// Current relative and absolute place in the file.
    pub point: Point,
    /// Semantic labels of one or more codes in `codes`.
    pub events: Vec<Event>,
    /// Hierarchy of semantic labels.
    ///
    /// Tracked to make sure everything’s valid.
    pub stack: Vec<Token>,
    /// To do.
    pub map: EditMap,
    /// List of attached resolvers, which will be called when done feeding,
    /// to clean events.
    resolvers: Vec<Box<Resolver>>,
    /// List of names associated with attached resolvers.
    resolver_ids: Vec<String>,
    /// Shared parsing state across tokenizers.
    pub parse_state: &'a ParseState<'a>,
    /// To do.
    pub index: usize,
    /// Stack of label (start) that could form images and links.
    ///
    /// Used when tokenizing [text content][crate::content::text].
    pub label_start_stack: Vec<LabelStart>,
    /// Stack of label (start) that cannot form images and links.
    ///
    /// Used when tokenizing [text content][crate::content::text].
    pub label_start_list_loose: Vec<LabelStart>,
    /// Stack of images and links.
    ///
    /// Used when tokenizing [text content][crate::content::text].
    pub media_list: Vec<Media>,
    /// Current container state.
    pub container: Option<ContainerState>,
    /// Whether we would be interrupting something.
    ///
    /// Used when tokenizing [flow content][crate::content::flow].
    pub interrupt: bool,
    /// Whether containers cannot “pierce” into the current construct.
    ///
    /// Used when tokenizing [document content][crate::content::document].
    pub concrete: bool,
    /// Whether this line is lazy.
    ///
    /// The previous line was a paragraph, and this line’s containers did not
    /// match.
    pub lazy: bool,
}

impl<'a> Tokenizer<'a> {
    /// Create a new tokenizer.
    pub fn new(point: Point, parse_state: &'a ParseState) -> Tokenizer<'a> {
        Tokenizer {
            previous: Code::None,
            current: Code::None,
            // To do: reserve size when feeding?
            column_start: vec![],
            line_start: point.line,
            consumed: true,
            drained: false,
            point,
            stack: vec![],
            events: vec![],
            parse_state,
            index: 0,
            map: EditMap::new(),
            label_start_stack: vec![],
            label_start_list_loose: vec![],
            media_list: vec![],
            container: None,
            interrupt: false,
            concrete: false,
            lazy: false,
            // Assume about 10 resolvers.
            resolvers: Vec::with_capacity(10),
            resolver_ids: Vec::with_capacity(10),
        }
    }

    /// Register a resolver.
    pub fn register_resolver(&mut self, id: String, resolver: Box<Resolver>) {
        if !self.resolver_ids.contains(&id) {
            self.resolver_ids.push(id);
            self.resolvers.push(resolver);
        }
    }

    /// Register a resolver, before others.
    pub fn register_resolver_before(&mut self, id: String, resolver: Box<Resolver>) {
        if !self.resolver_ids.contains(&id) {
            self.resolver_ids.push(id);
            self.resolvers.insert(0, resolver);
        }
    }

    /// Prepare for a next code to get consumed.
    pub fn expect(&mut self, code: Code) {
        assert!(self.consumed, "expected previous character to be consumed");
        self.consumed = false;
        self.current = code;
    }

    /// Define a jump between two places.
    pub fn define_skip(&mut self, point: &Point) {
        define_skip_impl(self, point.line, point.index);
    }

    /// Define the current place as a jump between two places.
    pub fn define_skip_current(&mut self) {
        define_skip_impl(self, self.point.line, self.point.index);
    }

    /// Increment the current positional info if we’re right after a line
    /// ending, which has a skip defined.
    fn account_for_potential_skip(&mut self) {
        let at = self.point.line - self.line_start;

        if self.point.column == 1 && at != self.column_start.len() {
            self.move_to(self.column_start[at]);
        }
    }

    /// Consume the current character.
    /// Each [`StateFn`][] is expected to call this to signal that this code is
    /// used, or call a next `StateFn`.
    pub fn consume(&mut self) {
        log::debug!("consume: `{:?}` ({:?})", self.current, self.point);
        assert!(!self.consumed, "expected code to not have been consumed: this might be because `x(code)` instead of `x` was returned");

        self.move_to(self.index + 1);

        self.previous = self.current;
        // Mark as consumed.
        self.consumed = true;
    }

    /// To do.
    pub fn move_to(&mut self, to: usize) {
        while self.index < to {
            let code = &self.parse_state.codes[self.index];
            self.point.index += 1;
            self.index += 1;

            match code {
                Code::CarriageReturnLineFeed | Code::Char('\n' | '\r') => {
                    self.point.line += 1;
                    self.point.column = 1;
                    self.point.offset += if *code == Code::CarriageReturnLineFeed {
                        2
                    } else {
                        1
                    };

                    if self.point.line - self.line_start + 1 > self.column_start.len() {
                        self.column_start.push(self.point.index);
                    }

                    self.account_for_potential_skip();
                    log::debug!("position: after eol: `{:?}`", self.point);
                }
                Code::VirtualSpace => {
                    // Empty.
                }
                _ => {
                    self.point.column += 1;
                    self.point.offset += 1;
                }
            }
        }
    }

    /// Mark the start of a semantic label.
    pub fn enter(&mut self, token_type: Token) {
        self.enter_with_link(token_type, None);
    }

    pub fn enter_with_content(&mut self, token_type: Token, content_type_opt: Option<ContentType>) {
        self.enter_with_link(
            token_type,
            content_type_opt.map(|content_type| Link {
                content_type,
                previous: None,
                next: None,
            }),
        );
    }

    pub fn enter_with_link(&mut self, token_type: Token, link: Option<Link>) {
        log::debug!("enter: `{:?}` ({:?})", token_type, self.point);
        self.events.push(Event {
            event_type: EventType::Enter,
            token_type: token_type.clone(),
            point: self.point.clone(),
            link,
        });
        self.stack.push(token_type);
    }

    /// Mark the end of a semantic label.
    pub fn exit(&mut self, token_type: Token) {
        let current_token = self.stack.pop().expect("cannot close w/o open tokens");

        assert_eq!(
            current_token, token_type,
            "expected exit token to match current token"
        );

        let previous = self.events.last().expect("cannot close w/o open event");
        let mut point = self.point.clone();

        assert!(
            current_token != previous.token_type || previous.point.index != point.index,
            "expected non-empty token"
        );

        if VOID_TOKENS.iter().any(|d| d == &token_type) {
            assert!(
                current_token == previous.token_type,
                "expected token to be void (`{:?}`), instead of including `{:?}`",
                current_token,
                previous.token_type
            );
        }

        // A bit weird, but if we exit right after a line ending, we *don’t* want to consider
        // potential skips.
        if matches!(
            self.previous,
            Code::CarriageReturnLineFeed | Code::Char('\n' | '\r')
        ) {
            point.column = 1;
            point.offset = previous.point.offset
                + if self.previous == Code::CarriageReturnLineFeed {
                    2
                } else {
                    1
                };
            point.index = previous.point.index + 1;
        }

        log::debug!("exit: `{:?}` ({:?})", token_type, point);
        self.events.push(Event {
            event_type: EventType::Exit,
            token_type,
            point,
            link: None,
        });
    }

    /// Capture the internal state.
    fn capture(&mut self) -> InternalState {
        InternalState {
            previous: self.previous,
            current: self.current,
            point: self.point.clone(),
            events_len: self.events.len(),
            stack_len: self.stack.len(),
        }
    }

    /// Apply the internal state.
    fn free(&mut self, previous: InternalState) {
        self.previous = previous.previous;
        self.current = previous.current;
        self.point = previous.point;
        assert!(
            self.events.len() >= previous.events_len,
            "expected to restore less events than before"
        );
        self.events.truncate(previous.events_len);
        assert!(
            self.stack.len() >= previous.stack_len,
            "expected to restore less stack items than before"
        );
        self.stack.truncate(previous.stack_len);
    }

    /// Parse with `state_fn` and its future states, switching to `ok` when
    /// successful, and passing [`State::Nok`][] back up if it occurs.
    ///
    /// This function does not capture the current state, in case of
    /// `State::Nok`, as it is assumed that this `go` is itself wrapped in
    /// another `attempt`.
    #[allow(clippy::unused_self)]
    pub fn go(
        &mut self,
        state_fn: impl FnOnce(&mut Tokenizer) -> State + 'static,
        after: impl FnOnce(&mut Tokenizer) -> State + 'static,
    ) -> Box<StateFn> {
        attempt_impl(
            state_fn,
            None,
            self.index,
            |end: usize, tokenizer: &mut Tokenizer, state| {
                if matches!(state, State::Ok) {
                    tokenizer.index = end;
                    tokenizer.consumed = true;
                    State::Fn(Box::new(after))
                } else {
                    state
                }
            },
        )
    }

    /// Like `go`, but this lets you *hijack* back to some other state after a
    /// certain code.
    #[allow(clippy::unused_self)]
    pub fn go_until(
        &mut self,
        state_fn: impl FnOnce(&mut Tokenizer) -> State + 'static,
        until: impl Fn(Code) -> bool + 'static,
        done: impl FnOnce(State) -> Box<StateFn> + 'static,
    ) -> Box<StateFn> {
        attempt_impl(
            state_fn,
            Some(Box::new(until)),
            self.index,
            |end: usize, tokenizer: &mut Tokenizer, state| {
                tokenizer.index = end;
                tokenizer.consumed = true;
                State::Fn(done(state))
            },
        )
    }

    /// Parse with `state_fn` and its future states, to check if it result in
    /// [`State::Ok`][] or [`State::Nok`][], revert on both cases, and then
    /// call `done` with whether it was successful or not.
    ///
    /// This captures the current state of the tokenizer, returns a wrapped
    /// state that captures all codes and feeds them to `state_fn` and its
    /// future states until it yields `State::Ok` or `State::Nok`.
    /// It then applies the captured state, calls `done`, and feeds all
    /// captured codes to its future states.
    pub fn check(
        &mut self,
        state_fn: impl FnOnce(&mut Tokenizer) -> State + 'static,
        done: impl FnOnce(bool) -> Box<StateFn> + 'static,
    ) -> Box<StateFn> {
        let previous = self.capture();
        let start = self.index;

        attempt_impl(
            state_fn,
            None,
            start,
            move |_: usize, tokenizer: &mut Tokenizer, state| {
                tokenizer.free(previous);
                tokenizer.index = start;
                tokenizer.consumed = true;
                State::Fn(done(matches!(state, State::Ok)))
            },
        )
    }

    /// Parse with `state_fn` and its future states, to check if it results in
    /// [`State::Ok`][] or [`State::Nok`][], revert on the case of
    /// `State::Nok`, and then call `done` with whether it was successful or
    /// not.
    ///
    /// This captures the current state of the tokenizer, returns a wrapped
    /// state that captures all codes and feeds them to `state_fn` and its
    /// future states until it yields `State::Ok`, at which point it calls
    /// `done` and yields its result.
    /// If instead `State::Nok` was yielded, the captured state is applied,
    /// `done` is called, and all captured codes are fed to its future states.
    pub fn attempt(
        &mut self,
        state_fn: impl FnOnce(&mut Tokenizer) -> State + 'static,
        done: impl FnOnce(bool) -> Box<StateFn> + 'static,
    ) -> Box<StateFn> {
        let previous = self.capture();

        attempt_impl(
            state_fn,
            None,
            self.index,
            |end: usize, tokenizer: &mut Tokenizer, state| {
                let ok = matches!(state, State::Ok);

                if !ok {
                    tokenizer.free(previous);
                }

                log::debug!("attempt: {:?}, at {:?}", ok, tokenizer.point);

                tokenizer.index = end;
                tokenizer.consumed = true;
                State::Fn(done(ok))
            },
        )
    }

    /// Just like [`attempt`][Tokenizer::attempt], but many.
    pub fn attempt_n(
        &mut self,
        mut state_fns: Vec<Box<StateFn>>,
        done: impl FnOnce(bool) -> Box<StateFn> + 'static,
    ) -> Box<StateFn> {
        if state_fns.is_empty() {
            done(false)
        } else {
            let state_fn = state_fns.remove(0);
            self.attempt(state_fn, move |ok| {
                if ok {
                    done(ok)
                } else {
                    Box::new(|t| t.attempt_n(state_fns, done)(t))
                }
            })
        }
    }

    /// Just like [`attempt`][Tokenizer::attempt], but for when you don’t care
    /// about `ok`.
    pub fn attempt_opt(
        &mut self,
        state_fn: impl FnOnce(&mut Tokenizer) -> State + 'static,
        after: impl FnOnce(&mut Tokenizer) -> State + 'static,
    ) -> Box<StateFn> {
        self.attempt(state_fn, |_ok| Box::new(after))
    }

    /// Feed a list of `codes` into `start`.
    ///
    /// This is set up to support repeatedly calling `feed`, and thus streaming
    /// markdown into the state machine, and normally pauses after feeding.
    pub fn push(
        &mut self,
        min: usize,
        max: usize,
        start: impl FnOnce(&mut Tokenizer) -> State + 'static,
        drain: bool,
    ) -> State {
        assert!(!self.drained, "cannot feed after drain");

        let mut result = feed_impl(self, min, max, start);

        if drain {
            let func = match result {
                State::Fn(func) => func,
                _ => unreachable!("expected next state"),
            };

            result = flush_impl(self, max, func);

            self.drained = true;

            while !self.resolvers.is_empty() {
                let resolver = self.resolvers.remove(0);
                resolver(self);
            }

            self.map.consume(&mut self.events);
        }

        result
    }

    /// Flush the tokenizer.
    pub fn flush(&mut self, start: impl FnOnce(&mut Tokenizer) -> State + 'static) -> State {
        flush_impl(self, self.index, start)
    }
}

/// Internal utility to wrap states to also capture codes.
///
/// Recurses into itself.
/// Used in [`Tokenizer::attempt`][Tokenizer::attempt] and  [`Tokenizer::check`][Tokenizer::check].
fn attempt_impl(
    state: impl FnOnce(&mut Tokenizer) -> State + 'static,
    pause: Option<Box<dyn Fn(Code) -> bool + 'static>>,
    start: usize,
    done: impl FnOnce(usize, &mut Tokenizer, State) -> State + 'static,
) -> Box<StateFn> {
    Box::new(move |tokenizer| {
        if let Some(ref func) = pause {
            if tokenizer.index > start && func(tokenizer.previous) {
                return done(tokenizer.index, tokenizer, State::Fn(Box::new(state)));
            }
        }

        let state = state(tokenizer);

        match state {
            State::Ok => {
                assert!(
                    tokenizer.index >= start,
                    "`end` must not be smaller than `start`"
                );
                done(tokenizer.index, tokenizer, state)
            }
            State::Nok => done(start, tokenizer, state),
            State::Fn(func) => State::Fn(attempt_impl(func, pause, start, done)),
        }
    })
}

/// Feed a list of `codes` into `start`.
fn feed_impl(
    tokenizer: &mut Tokenizer,
    min: usize,
    max: usize,
    start: impl FnOnce(&mut Tokenizer) -> State + 'static,
) -> State {
    let mut state = State::Fn(Box::new(start));

    tokenizer.index = min;
    tokenizer.consumed = true;

    while tokenizer.index < max {
        let code = tokenizer.parse_state.codes[tokenizer.index];

        match state {
            State::Ok | State::Nok => {
                break;
            }
            State::Fn(func) => {
                log::debug!("main: passing: `{:?}` ({:?})", code, tokenizer.index);
                tokenizer.expect(code);
                state = func(tokenizer);
            }
        }
    }

    state
}

/// Flush `start`: pass `eof`s to it until done.
fn flush_impl(
    tokenizer: &mut Tokenizer,
    max: usize,
    start: impl FnOnce(&mut Tokenizer) -> State + 'static,
) -> State {
    let mut state = State::Fn(Box::new(start));
    tokenizer.consumed = true;

    loop {
        match state {
            State::Ok | State::Nok => break,
            State::Fn(func) => {
                let code = if tokenizer.index < max {
                    tokenizer.parse_state.codes[tokenizer.index]
                } else {
                    Code::None
                };
                log::debug!("main: flushing {:?}", code);
                tokenizer.expect(code);
                state = func(tokenizer);
            }
        }
    }

    match state {
        State::Ok => {}
        _ => unreachable!("expected final state to be `State::Ok`"),
    }

    state
}

/// Define a jump between two places.
///
/// This defines how much columns, offsets, and the `index` are increased when
/// consuming a line ending.
fn define_skip_impl(tokenizer: &mut Tokenizer, line: usize, index: usize) {
    log::debug!("position: define skip: {:?} -> ({:?})", line, index);
    let at = line - tokenizer.line_start;

    if at == tokenizer.column_start.len() {
        tokenizer.column_start.push(index);
    } else {
        tokenizer.column_start[at] = index;
    }

    tokenizer.account_for_potential_skip();
}