rumdl 0.1.84

A fast Markdown linter written in Rust (Ru(st) MarkDown Linter)
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
//! Label generation and CommonMark normalization for MD054 reference-style fixes.
//!
//! Reference-link labels in CommonMark are matched case-insensitively after collapsing
//! internal whitespace runs to a single space and trimming leading/trailing whitespace
//! (see CommonMark §4.7). When MD054 converts an inline link to a reference-style link,
//! it must:
//!
//! 1. Derive a candidate label slug from the link text (or URL when text is empty).
//! 2. Avoid colliding with any reference definition that already exists for a *different*
//!    URL — same URL means we can reuse the existing label.
//! 3. Avoid colliding with another label generated earlier in the same fix pass.
//!
//! The slug rule used here is the human-readable form documented in the MD054 issue:
//! lowercase, runs of non-alphanumeric characters replaced by `-`, leading and trailing
//! `-` trimmed. CJK and other Unicode letters are kept as alphanumeric so non-ASCII
//! text produces sensible labels rather than empty strings.

use std::collections::{HashMap, HashSet};

/// Normalize a label for matching, mirroring CommonMark §4.7.
///
/// - Unicode case-fold (lowercased).
/// - Internal whitespace runs collapsed to a single space.
/// - Leading and trailing whitespace trimmed.
pub(super) fn normalize_label(label: &str) -> String {
    let mut out = String::with_capacity(label.len());
    let mut in_space = true; // skip leading whitespace
    for ch in label.chars() {
        if ch.is_whitespace() {
            if !in_space {
                out.push(' ');
                in_space = true;
            }
        } else {
            for lower in ch.to_lowercase() {
                out.push(lower);
            }
            in_space = false;
        }
    }
    if out.ends_with(' ') {
        out.pop();
    }
    out
}

/// Build a slug suitable for use as a reference label.
///
/// `text` is typically a link's display text. The result is lowercased, with runs
/// of non-alphanumeric characters collapsed to a single `-`, and trimmed of `-`.
/// Returns an empty string when the input collapses to nothing — callers should
/// treat that as a signal to fall back to a different source (e.g. the URL).
pub(super) fn slugify(text: &str) -> String {
    let mut out = String::with_capacity(text.len());
    let mut prev_dash = true; // skip leading dashes
    for ch in text.chars() {
        if ch.is_alphanumeric() {
            for lower in ch.to_lowercase() {
                out.push(lower);
            }
            prev_dash = false;
        } else if !prev_dash {
            out.push('-');
            prev_dash = true;
        }
    }
    if out.ends_with('-') {
        out.pop();
    }
    out
}

/// Outcome of a label lookup: either a fresh label that the caller must define,
/// or an existing one that already has a reference definition.
#[derive(Debug, Clone, PartialEq, Eq)]
pub(super) struct LabelChoice {
    pub label: String,
    /// True iff the label belongs to a definition that doesn't yet exist —
    /// the caller must emit `[label]: url ...` for it.
    pub is_new: bool,
}

/// Composite key for label reuse: a `(url, title)` pair. Two links to the same
/// URL but with *different* titles must produce distinct reference definitions
/// — one shared ref def could only carry one title, silently dropping the
/// other. Treating the title as part of the identity preserves both.
type DefKey = (String, Option<String>);

/// Resolved-target metadata for an existing label, kept on `by_label` so we can
/// distinguish "same URL, same title" (safe reuse) from "same label, different
/// destination" (collision).
#[derive(Debug, Clone)]
struct ExistingTarget {
    url: String,
    title: Option<String>,
}

/// Tracks already-used reference labels and assigns new ones, sharing labels for
/// duplicate `(url, title)` destinations and disambiguating slug collisions
/// with `-2`, `-3`, ... suffixes.
pub(super) struct LabelGenerator {
    /// Map from normalized label to its resolved destination.
    /// Used to detect collisions with existing reference definitions.
    by_label: HashMap<String, ExistingTarget>,
    /// Map from `(url, title)` to the label assigned to it. Two distinct
    /// destinations always get distinct labels, even when the URL matches.
    by_def: HashMap<DefKey, String>,
    /// `(url, title)` pairs that came from existing reference definitions in
    /// the document. Reuses of these do NOT need a new ref def emitted.
    existing_defs: HashSet<DefKey>,
}

impl LabelGenerator {
    /// Seed the generator with the document's existing reference definitions.
    ///
    /// Each item is `(label, url, title)`. CommonMark says the first definition
    /// for a label wins; later definitions with the same normalized label are
    /// shadowed and won't actually resolve, so they're skipped here too — adding
    /// them to `by_def` would cause us to reuse a label that doesn't actually
    /// resolve to the requested destination.
    pub(super) fn from_existing<I, L, U, T>(existing: I) -> Self
    where
        I: IntoIterator<Item = (L, U, Option<T>)>,
        L: AsRef<str>,
        U: AsRef<str>,
        T: AsRef<str>,
    {
        let mut by_label: HashMap<String, ExistingTarget> = HashMap::new();
        let mut by_def: HashMap<DefKey, String> = HashMap::new();
        let mut existing_defs: HashSet<DefKey> = HashSet::new();
        for (label, url, title) in existing {
            let label = label.as_ref();
            let url = url.as_ref().to_string();
            let title = title.as_ref().map(|t| t.as_ref().to_string());
            let normalized = normalize_label(label);
            // CommonMark §4.7: first definition wins. Skip shadowed ones —
            // their label doesn't actually resolve, so we must NOT add them
            // to `by_def` either, otherwise we'd suggest reusing a label
            // that points elsewhere.
            if by_label.contains_key(&normalized) {
                continue;
            }
            by_label.insert(
                normalized,
                ExistingTarget {
                    url: url.clone(),
                    title: title.clone(),
                },
            );
            let key: DefKey = (url, title);
            by_def.entry(key.clone()).or_insert_with(|| label.to_string());
            existing_defs.insert(key);
        }
        Self {
            by_label,
            by_def,
            existing_defs,
        }
    }

    /// Reserve `label` *exactly* (no `-N` suffixing) for `(url, title)`.
    ///
    /// Returns `None` when the normalized label is already taken by a *different*
    /// destination — the caller cannot safely use this label, since
    /// collapsed/shortcut references can't be disambiguated with a suffix
    /// without changing the link's visible text.
    ///
    /// Returns `Some(LabelChoice)` when the label is free or already maps to the
    /// same destination. `is_new` is `true` when the caller must emit a fresh
    /// ref def.
    pub(super) fn reserve_exact(&mut self, label: &str, url: &str, title: Option<&str>) -> Option<LabelChoice> {
        let normalized = normalize_label(label);
        let key: DefKey = (url.to_string(), title.map(str::to_string));
        match self.by_label.get(&normalized) {
            Some(existing) if existing.url == url && existing.title.as_deref() == title => {
                let is_new = !self.existing_defs.contains(&key);
                self.by_def.entry(key).or_insert_with(|| label.to_string());
                Some(LabelChoice {
                    label: label.to_string(),
                    is_new,
                })
            }
            Some(_) => None,
            None => {
                self.by_label.insert(
                    normalized,
                    ExistingTarget {
                        url: url.to_string(),
                        title: title.map(str::to_string),
                    },
                );
                self.by_def.entry(key).or_insert_with(|| label.to_string());
                Some(LabelChoice {
                    label: label.to_string(),
                    is_new: true,
                })
            }
        }
    }

    /// Get the label to use for `(text, url, title)`.
    ///
    /// - If a label is already assigned to this `(url, title)` pair (from
    ///   existing refs or earlier in this pass), reuse it. `is_new` is `false`
    ///   when that label came from a pre-existing reference definition in the
    ///   document.
    /// - Otherwise, slugify `text` (falling back to `url` if the slug is empty),
    ///   then suffix `-2`, `-3`, ... until the candidate doesn't collide with a
    ///   different destination. `is_new` is `true` for the freshly-assigned
    ///   label.
    pub(super) fn label_for(&mut self, text: &str, url: &str, title: Option<&str>) -> LabelChoice {
        let key: DefKey = (url.to_string(), title.map(str::to_string));
        if let Some(existing) = self.by_def.get(&key) {
            let is_new = !self.existing_defs.contains(&key);
            return LabelChoice {
                label: existing.clone(),
                is_new,
            };
        }

        let mut base = slugify(text);
        if base.is_empty() {
            base = slugify(url);
        }
        if base.is_empty() {
            base = "ref".to_string();
        }

        let mut candidate = base.clone();
        let mut suffix = 2u32;
        loop {
            let normalized = normalize_label(&candidate);
            match self.by_label.get(&normalized) {
                Some(existing) if existing.url == url && existing.title.as_deref() == title => {
                    // Same destination — collapse onto the existing label. If
                    // the label came from a pre-existing definition, no new ref
                    // def is needed.
                    let is_new = !self.existing_defs.contains(&key);
                    self.by_def.insert(key, candidate.clone());
                    return LabelChoice {
                        label: candidate,
                        is_new,
                    };
                }
                Some(_) => {
                    candidate = format!("{base}-{suffix}");
                    suffix += 1;
                }
                None => {
                    self.by_label.insert(
                        normalized,
                        ExistingTarget {
                            url: url.to_string(),
                            title: title.map(str::to_string),
                        },
                    );
                    self.by_def.insert(key, candidate.clone());
                    return LabelChoice {
                        label: candidate,
                        is_new: true,
                    };
                }
            }
        }
    }
}

#[cfg(test)]
mod tests {
    use super::*;

    #[test]
    fn slugify_basic() {
        assert_eq!(slugify("Hello World"), "hello-world");
        assert_eq!(slugify("documentation"), "documentation");
        assert_eq!(slugify("API Reference Guide"), "api-reference-guide");
    }

    #[test]
    fn slugify_punctuation_collapses_to_dash() {
        assert_eq!(slugify("foo!bar"), "foo-bar");
        assert_eq!(slugify("a.b.c"), "a-b-c");
        assert_eq!(slugify("!!!hello!!!"), "hello");
    }

    #[test]
    fn slugify_unicode_kept_as_alphanumeric() {
        assert_eq!(slugify("café"), "café");
        assert_eq!(slugify("日本語"), "日本語");
        assert_eq!(slugify("hello 日本"), "hello-日本");
    }

    #[test]
    fn slugify_empty_for_punctuation_only() {
        assert_eq!(slugify(""), "");
        assert_eq!(slugify("!!!"), "");
        assert_eq!(slugify("---"), "");
    }

    #[test]
    fn normalize_label_case_and_whitespace() {
        assert_eq!(normalize_label("Hello   World"), "hello world");
        assert_eq!(normalize_label("\tfoo\nbar  "), "foo bar");
        assert_eq!(normalize_label("ALREADY-LOW"), "already-low");
    }

    fn no_existing() -> LabelGenerator {
        LabelGenerator::from_existing(std::iter::empty::<(&str, &str, Option<&str>)>())
    }

    fn with_existing(defs: Vec<(&str, &str, Option<&str>)>) -> LabelGenerator {
        LabelGenerator::from_existing(defs)
    }

    #[test]
    fn generator_reuses_label_for_same_url_and_title() {
        let mut g = no_existing();
        let l1 = g.label_for("docs", "https://example.com/x", None);
        let l2 = g.label_for("documentation", "https://example.com/x", None);
        assert_eq!(l1.label, l2.label);
        // First call assigns a new label; second call reuses an in-pass label,
        // which is still considered "new" because no pre-existing ref def covers it.
        assert!(l1.is_new);
        assert!(l2.is_new);
    }

    #[test]
    fn generator_disambiguates_same_url_different_titles() {
        // Same URL but different titles must produce distinct labels and
        // distinct ref defs — sharing a single def would silently drop one
        // of the titles.
        let mut g = no_existing();
        let a = g.label_for("first", "https://example.com", Some("Title A"));
        let b = g.label_for("second", "https://example.com", Some("Title B"));
        assert_ne!(a.label, b.label, "different titles must get different labels");
        assert!(a.is_new);
        assert!(b.is_new);
        // A third call with a third title gets yet another label.
        let c = g.label_for("third", "https://example.com", Some("Title C"));
        assert_ne!(c.label, a.label);
        assert_ne!(c.label, b.label);
    }

    #[test]
    fn generator_treats_no_title_distinctly_from_empty_or_present_title() {
        let mut g = no_existing();
        let with_title = g.label_for("a", "https://example.com", Some("T"));
        let without_title = g.label_for("a", "https://example.com", None);
        assert_ne!(
            with_title.label, without_title.label,
            "presence vs absence of title must produce distinct labels"
        );
    }

    #[test]
    fn generator_disambiguates_collision() {
        let mut g = no_existing();
        let a = g.label_for("docs", "https://a.example.com", None);
        let b = g.label_for("docs", "https://b.example.com", None);
        assert_eq!(a.label, "docs");
        assert_eq!(b.label, "docs-2");
        let c = g.label_for("docs", "https://c.example.com", None);
        assert_eq!(c.label, "docs-3");
    }

    #[test]
    fn generator_respects_existing_labels() {
        let mut g = with_existing(vec![("docs", "https://existing.com/docs", None)]);
        // Same URL/title → reuse existing label, and signal that no new ref def is needed.
        let same = g.label_for("documentation", "https://existing.com/docs", None);
        assert_eq!(same.label, "docs");
        assert!(!same.is_new, "reusing pre-existing ref def must report is_new=false");
        // Different URL → must avoid the slug "docs"; this one IS new.
        let diff = g.label_for("docs", "https://other.com/docs", None);
        assert_eq!(diff.label, "docs-2");
        assert!(diff.is_new);
    }

    #[test]
    fn generator_falls_back_to_url_when_text_empty() {
        let mut g = no_existing();
        let choice = g.label_for("", "https://example.com/page", None);
        assert_eq!(choice.label, "https-example-com-page");
        assert!(choice.is_new);
    }

    #[test]
    fn generator_falls_back_to_ref_when_both_empty() {
        let mut g = no_existing();
        let first = g.label_for("", "", None);
        assert_eq!(first.label, "ref");
        assert!(first.is_new);
        let again = g.label_for("", "", None);
        assert_eq!(again.label, "ref");
        assert!(again.is_new);
    }

    #[test]
    fn generator_normalizes_when_checking_collision() {
        // Existing definition uses different case/whitespace.
        let mut g = with_existing(vec![("Hello   World", "https://existing.com", None)]);
        // Different URL with text that slugifies to "hello-world" — must disambiguate.
        let choice = g.label_for("Hello World", "https://other.com", None);
        // "hello-world" normalized differs from "hello world", so no collision.
        // (Slug uses dashes, normalize_label uses spaces — they're distinct keys.)
        assert_eq!(choice.label, "hello-world");
        assert!(choice.is_new);
        // But a slug that *exactly* matches the existing label must be disambiguated.
        let mut g2 = with_existing(vec![("hello-world", "https://existing.com", None)]);
        let choice2 = g2.label_for("Hello World", "https://other.com", None);
        assert_eq!(choice2.label, "hello-world-2");
        assert!(choice2.is_new);
    }

    #[test]
    fn generator_marks_existing_url_match_as_not_new() {
        // Existing ref def for the URL — even when the inline link's text would
        // normally produce a different slug, the planner reuses the existing
        // label and must NOT emit a duplicate definition.
        let mut g = with_existing(vec![("site", "https://example.com", None)]);
        let choice = g.label_for("docs", "https://example.com", None);
        assert_eq!(choice.label, "site");
        assert!(!choice.is_new);
    }

    #[test]
    fn generator_skips_shadowed_existing_defs() {
        // CommonMark §4.7: when two ref defs share a normalized label, only
        // the first resolves. The second is shadowed and its URL doesn't
        // actually map to any label in the parsed document. The generator
        // must NOT seed `by_def` with the shadowed entry — otherwise a
        // future `label_for` call for that URL would propose the shadowed
        // label, which doesn't resolve there.
        let mut g = with_existing(vec![
            ("docs", "https://first.com", None),
            ("DOCS", "https://second.com", None),
        ]);
        // Looking up the *shadowed* URL must not return the shadowed label;
        // the generator should produce a fresh, non-colliding slug instead.
        let choice = g.label_for("anchor", "https://second.com", None);
        assert_ne!(choice.label, "docs");
        assert_ne!(choice.label, "DOCS");
        assert!(choice.is_new);
    }

    #[test]
    fn reserve_exact_treats_title_as_part_of_identity() {
        // Pre-existing def carries title T1 — that's its identity.
        let mut g = with_existing(vec![("anchor", "https://example.com", Some("T1"))]);
        // Same label, same URL, *different* title → unsafe (label already
        // resolved to a different destination).
        assert!(g.reserve_exact("anchor", "https://example.com", Some("T2")).is_none());
        // Same label, same URL, same title → safe reuse against an existing
        // def, so `is_new` is false (no new ref def needed).
        let reuse = g
            .reserve_exact("anchor", "https://example.com", Some("T1"))
            .expect("same destination must reuse");
        assert_eq!(reuse.label, "anchor");
        assert!(!reuse.is_new);
    }
}