mant-core 0.6.1

Structured manual and Markdown document engine used by ManT
Documentation
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
//! Identifies addressable semantic entries after source-specific lowering.
//!
//! Both libmandoc macro sets and Markdown produce definition lists. This
//! pass assigns one canonical option identity without leaking source macros
//! into the stable document contract.

use std::{
    collections::{HashSet, VecDeque},
    mem,
};

use mant_ast::{
    Block, DefinitionCase, DefinitionIdentity, DefinitionItem, DefinitionRole, Inline, LayoutHint,
    Section, SourceSpan,
};

use crate::inline::{DEFAULT_INLINE_TERM_MAX_WIDTH, plain_text, terms_fit_inline};

/// Annotate reliably recognizable command-line options and return every
/// inline anchor that the navigation resolver must retain.
pub(crate) fn identify_definitions(
    blocks: &mut Vec<Block>,
    sections: &mut [Section],
    reserved_targets: &HashSet<String>,
) -> HashSet<String> {
    let mut used = HashSet::new();
    collect_section_ids(sections, &mut used);
    let mut retained = HashSet::new();
    identify_blocks(blocks, &mut used, reserved_targets, &mut retained);
    for section in sections {
        identify_blocks(
            &mut section.blocks,
            &mut used,
            reserved_targets,
            &mut retained,
        );
        identify_sections(
            &mut section.children,
            &mut used,
            reserved_targets,
            &mut retained,
        );
    }
    retained
}

/// Return identified definition items in the single source order shared by
/// projection and search ownership, including the source of their containing
/// definition list.
pub(crate) fn definition_entries(blocks: &[Block]) -> Vec<(&DefinitionItem, Option<SourceSpan>)> {
    let mut entries = Vec::new();
    collect_definition_entries(blocks, &mut entries);
    entries
}

fn collect_definition_entries<'a>(
    blocks: &'a [Block],
    output: &mut Vec<(&'a DefinitionItem, Option<SourceSpan>)>,
) {
    for block in blocks {
        match block {
            Block::List { items, .. } => {
                for item in items {
                    collect_definition_entries(&item.blocks, output);
                }
            }
            Block::DefinitionList { items, source, .. } => {
                for item in items {
                    if item.identity.is_some() {
                        output.push((item, *source));
                    }
                    collect_definition_entries(&item.description, output);
                }
            }
            Block::Table { rows, .. } => {
                for row in rows {
                    for cell in &row.cells {
                        collect_definition_entries(&cell.blocks, output);
                    }
                }
            }
            Block::Paragraph { .. }
            | Block::Preformatted { .. }
            | Block::Equation { .. }
            | Block::VerticalSpace { .. }
            | Block::ThematicBreak { .. }
            | Block::Unsupported { .. } => {}
        }
    }
}

fn collect_section_ids(sections: &[Section], output: &mut HashSet<String>) {
    for section in sections {
        output.insert(section.id.clone());
        collect_section_ids(&section.children, output);
    }
}

fn identify_sections(
    sections: &mut [Section],
    used: &mut HashSet<String>,
    reserved: &HashSet<String>,
    retained: &mut HashSet<String>,
) {
    for section in sections {
        identify_blocks(&mut section.blocks, used, reserved, retained);
        identify_sections(&mut section.children, used, reserved, retained);
    }
}

fn identify_blocks(
    blocks: &mut Vec<Block>,
    used: &mut HashSet<String>,
    reserved: &HashSet<String>,
    retained: &mut HashSet<String>,
) {
    normalize_hanging_definitions(blocks);
    for block in blocks {
        match block {
            Block::List { items, .. } => {
                for item in items {
                    identify_blocks(&mut item.blocks, used, reserved, retained);
                }
            }
            Block::DefinitionList { items, .. } => {
                for item in items {
                    identify_item(item, used, reserved, retained);
                    identify_blocks(&mut item.description, used, reserved, retained);
                }
            }
            Block::Table { rows, .. } => {
                for row in rows {
                    for cell in &mut row.cells {
                        identify_blocks(&mut cell.blocks, used, reserved, retained);
                    }
                }
            }
            Block::Paragraph { .. }
            | Block::Preformatted { .. }
            | Block::Equation { .. }
            | Block::VerticalSpace { .. }
            | Block::ThematicBreak { .. }
            | Block::Unsupported { .. } => {}
        }
    }
}

/// Turn renderer-neutral hanging-indent runs into semantic definitions.
///
/// Some man(7) generators use `.PP` followed by `.RS` instead of `.TP` for
/// option entries. Native parsers correctly retain that layout, but
/// neither representation is a definition list on its own. Recognising the
/// shared visible shape here keeps option identity independent of the source
/// macro set or source parser used by the query pipeline.
fn normalize_hanging_definitions(blocks: &mut Vec<Block>) {
    let mut pending: VecDeque<Block> = mem::take(blocks).into();
    let mut normalized = Vec::with_capacity(pending.len());

    while let Some(block) = pending.pop_front() {
        let Some(term_indent) = option_term_indent(&block) else {
            normalized.push(block);
            continue;
        };

        let mut description = Vec::new();
        while let Some(next) = pending.front() {
            if option_term_indent(next) == Some(term_indent) {
                break;
            }
            if matches!(next, Block::VerticalSpace { .. }) {
                if description.is_empty() {
                    break;
                }
                description.push(pending.pop_front().expect("front exists"));
                continue;
            }
            if block_indent(next) <= term_indent {
                break;
            }
            description.push(pending.pop_front().expect("front exists"));
        }

        if description.is_empty() {
            normalized.push(block);
            continue;
        }

        let Block::Paragraph {
            children,
            layout,
            source,
        } = block
        else {
            unreachable!("option_term_indent only accepts paragraphs");
        };
        let description_origin = term_indent.saturating_add(4);
        for child in &mut description {
            shift_block_indent(child, description_origin);
        }
        let terms = vec![children];
        normalized.push(Block::DefinitionList {
            items: vec![DefinitionItem {
                identity: None,
                inline_term: terms_fit_inline(&terms, DEFAULT_INLINE_TERM_MAX_WIDTH),
                terms,
                description,
                spacing_before_lines: Some(layout.spacing_before_lines),
            }],
            compact: true,
            layout: LayoutHint {
                indent_columns: term_indent,
                spacing_before_lines: 0,
            },
            source,
        });
    }

    *blocks = normalized;
}

fn option_term_indent(block: &Block) -> Option<u16> {
    let Block::Paragraph {
        children, layout, ..
    } = block
    else {
        return None;
    };
    let text = plain_text(children);
    let trimmed = text.trim_start();
    (trimmed.starts_with('-')
        && !option_names_from_terms(std::slice::from_ref(children)).is_empty())
    .then_some(layout.indent_columns)
}

fn block_indent(block: &Block) -> u16 {
    match block {
        Block::Paragraph { layout, .. }
        | Block::Preformatted { layout, .. }
        | Block::List { layout, .. }
        | Block::DefinitionList { layout, .. }
        | Block::Table { layout, .. }
        | Block::Equation { layout, .. }
        | Block::Unsupported { layout, .. } => layout.indent_columns,
        Block::VerticalSpace { .. } | Block::ThematicBreak { .. } => 0,
    }
}

fn shift_block_indent(block: &mut Block, origin: u16) {
    let layout = match block {
        Block::Paragraph { layout, .. }
        | Block::Preformatted { layout, .. }
        | Block::List { layout, .. }
        | Block::DefinitionList { layout, .. }
        | Block::Table { layout, .. }
        | Block::Equation { layout, .. }
        | Block::Unsupported { layout, .. } => layout,
        Block::VerticalSpace { .. } | Block::ThematicBreak { .. } => return,
    };
    layout.indent_columns = layout.indent_columns.saturating_sub(origin);
}

fn identify_item(
    item: &mut DefinitionItem,
    used: &mut HashSet<String>,
    reserved: &HashSet<String>,
    retained: &mut HashSet<String>,
) {
    let (role, case, names) = item.identity.as_ref().map_or_else(
        || {
            (
                DefinitionRole::Option,
                DefinitionCase::Sensitive,
                option_names(item),
            )
        },
        |identity| (identity.role, identity.case, identity.names.clone()),
    );
    if names.is_empty() {
        return;
    }

    let mut anchors = Vec::new();
    for term in &item.terms {
        collect_anchor_ids(term, &mut anchors);
    }
    retained.extend(anchors.iter().cloned());

    let existing = anchors.first().cloned();
    let preferred = existing.clone().unwrap_or_else(|| {
        format!(
            "{}-{}",
            role_id_prefix(role),
            role_name_slug(role, &names[0])
        )
    });
    // A copied libmandoc anchor may itself be an explicit `.Tg` destination,
    // so it is allowed to match the reserved set. Generated IDs are not.
    let id = if existing.is_some() && !used.contains(&preferred) {
        used.insert(preferred.clone());
        preferred
    } else {
        unique_id(&preferred, used, reserved)
    };
    if !anchors.iter().any(|anchor| anchor == &id)
        && let Some(term) = item.terms.first_mut()
    {
        term.insert(0, Inline::Anchor { id: id.clone() });
    }
    retained.insert(id.clone());
    item.identity = Some(DefinitionIdentity {
        id,
        role,
        case,
        names,
    });
}

fn role_name_slug(role: DefinitionRole, name: &str) -> String {
    if role == DefinitionRole::Variable {
        match name {
            "$?" => return "question-mark".to_owned(),
            "$$" => return "dollar-dollar".to_owned(),
            "$^" => return "caret".to_owned(),
            "$_" => return "underscore".to_owned(),
            _ => {}
        }
    }
    slug(name)
}

const fn role_id_prefix(role: DefinitionRole) -> &'static str {
    match role {
        DefinitionRole::Option => "option",
        DefinitionRole::Command => "command",
        DefinitionRole::EnvironmentVariable => "environment",
        DefinitionRole::Variable => "variable",
    }
}

fn option_names(item: &DefinitionItem) -> Vec<String> {
    option_names_from_terms(&item.terms)
}

pub(crate) fn option_names_from_terms(terms: &[Vec<Inline>]) -> Vec<String> {
    let mut names = Vec::new();
    for term in terms {
        let text = plain_text(term);
        for token in text.split(|character: char| {
            character.is_whitespace() || matches!(character, ',' | '|' | '/' | ';')
        }) {
            let token = token.trim_matches(|character: char| {
                matches!(
                    character,
                    '[' | ']' | '(' | ')' | '{' | '}' | '' | '' | '' | ''
                )
            });
            let Some(name) = option_prefix(token) else {
                continue;
            };
            if !names.iter().any(|existing| existing == name) {
                names.push(name.to_owned());
            }
        }
    }
    names
}

pub(crate) fn option_prefix(token: &str) -> Option<&str> {
    if !token.starts_with('-') || token == "-" {
        return None;
    }
    let end = token
        .char_indices()
        .skip(1)
        .take_while(|(_, character)| {
            character.is_ascii_alphanumeric() || matches!(character, '-' | '_' | '?')
        })
        .map(|(index, character)| index + character.len_utf8())
        .last()?;
    let candidate = &token[..end];
    let body = candidate.trim_start_matches('-');
    (!body.is_empty()
        && body
            .chars()
            .any(|character| character.is_ascii_alphanumeric() || character == '?'))
    .then_some(candidate)
}

fn collect_anchor_ids(nodes: &[Inline], output: &mut Vec<String>) {
    for node in nodes {
        match node {
            Inline::Anchor { id } => output.push(id.clone()),
            Inline::Strong { children }
            | Inline::Emphasis { children }
            | Inline::ExternalLink { children, .. }
            | Inline::EmailLink { children, .. }
            | Inline::ManualReference { children, .. }
            | Inline::SectionReference { children, .. } => collect_anchor_ids(children, output),
            Inline::Text { .. } | Inline::Code { .. } | Inline::LineBreak => {}
        }
    }
}

fn slug(value: &str) -> String {
    if value.trim_start_matches(['-', '/']) == "?" {
        return "help".to_owned();
    }
    let slug = value
        .trim_start_matches(['-', '/'])
        .chars()
        .flat_map(char::to_lowercase)
        .map(|character| {
            if character.is_alphanumeric() {
                character
            } else {
                '-'
            }
        })
        .collect::<String>()
        .split('-')
        .filter(|part| !part.is_empty())
        .collect::<Vec<_>>()
        .join("-");
    if slug.is_empty() {
        "entry".to_owned()
    } else {
        slug
    }
}

fn unique_id(base: &str, used: &mut HashSet<String>, reserved: &HashSet<String>) -> String {
    let mut candidate = base.to_owned();
    let mut suffix = 2;
    while used.contains(&candidate) || reserved.contains(&candidate) {
        candidate = format!("{base}-{suffix}");
        suffix += 1;
    }
    used.insert(candidate.clone());
    candidate
}

#[cfg(test)]
mod tests {
    use std::collections::HashSet;

    use mant_ast::{Block, DefinitionItem, Inline, LayoutHint, Section};

    use super::{identify_definitions, option_names};

    fn item(value: &str) -> DefinitionItem {
        DefinitionItem {
            identity: None,
            inline_term: false,
            terms: vec![vec![Inline::Text {
                value: value.into(),
            }]],
            description: Vec::new(),
            spacing_before_lines: None,
        }
    }

    #[test]
    fn extracts_aliases_without_argument_placeholders() {
        assert_eq!(
            option_names(&item("-g, --listed-incremental=FILE")),
            ["-g", "--listed-incremental"]
        );
        assert_eq!(option_names(&item("ordinary term")), Vec::<String>::new());
    }

    #[test]
    fn normalizes_hanging_option_layout_before_assigning_identity() {
        let paragraph = |value: &str, indent_columns| Block::Paragraph {
            children: vec![Inline::Text {
                value: value.to_owned(),
            }],
            layout: LayoutHint {
                indent_columns,
                spacing_before_lines: 1,
            },
            source: None,
        };
        let mut sections = vec![Section {
            id: "options".to_owned(),
            title: "OPTIONS".to_owned(),
            spacing_before_lines: 0,
            blocks: vec![
                paragraph("-v, --version", 0),
                paragraph("Print version information.", 4),
                paragraph("-C <path>", 0),
                paragraph("Run from path.", 4),
            ],
            children: Vec::new(),
            source: None,
        }];

        identify_definitions(&mut Vec::new(), &mut sections, &HashSet::new());

        assert_eq!(sections[0].blocks.len(), 2);
        let Block::DefinitionList { items, layout, .. } = &sections[0].blocks[0] else {
            panic!("hanging option should become a definition list");
        };
        assert_eq!(layout.indent_columns, 0);
        assert_eq!(
            items[0].identity.as_ref().expect("option identity").names,
            ["-v", "--version"]
        );
        assert!(matches!(
            &items[0].description[0],
            Block::Paragraph { layout, .. } if layout.indent_columns == 0
        ));
        let Block::DefinitionList { items, .. } = &sections[0].blocks[1] else {
            panic!("second option should remain independently addressable");
        };
        assert_eq!(
            items[0].identity.as_ref().expect("option identity").names,
            ["-C"]
        );
    }
}