rumdl 0.2.40

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
//! Rule MD082: Flag headings with no content before the next heading.
//!
//! A heading immediately followed by another heading, with no rendered body in
//! between, is an empty section. It usually signals a document that needs
//! restructuring: a parent heading with nothing under it before its first
//! child, or sibling headings with no body. This rule (opt-in) flags the
//! heading whose section is empty.
//!
//! Detection only: a fix would have to invent placeholder prose, so there is no
//! auto-fix. The `level` knob sets the minimum heading level that must have a
//! body. With the default `level = 1` every heading is checked, including
//! `# Title` straight into `## Section`. Set `level = 2` to exempt H1 while
//! still requiring content under H2 and deeper.
//!
//! What does not count as a section body: blank lines, HTML comments,
//! reference-link definitions (`[x]: url`), and lone thematic breaks (`---`).
//! Everything else that renders counts: paragraphs, lists, code blocks, tables,
//! blockquotes, and raw HTML. A `{#id}` attribute list on the line immediately
//! after a heading is its anchor (the parser folds it into the heading), so it
//! is treated as part of the heading rather than as the section body; an
//! attribute list anywhere else renders as ordinary text and counts as content.

use crate::lint_context::{HeadingStyle, LintContext};
use crate::rule::{FixCapability, LintError, LintResult, LintWarning, Rule, RuleCategory, Severity};
use crate::rule_config_serde::RuleConfig;
use serde::{Deserialize, Serialize};

fn default_level() -> u8 {
    1
}

/// Configuration for MD082 (No empty sections).
#[derive(Debug, Clone, Serialize, Deserialize, PartialEq)]
#[serde(rename_all = "kebab-case")]
pub struct MD082Config {
    /// Minimum heading level (1-6) that must be followed by content. A heading
    /// whose level is at least this value is flagged when it is immediately
    /// followed by another heading with no body in between. Default 1 checks
    /// every heading; set to 2 to exempt H1 (so `# Title` straight into
    /// `## Section` is allowed) while still requiring content under H2 and below.
    #[serde(default = "default_level")]
    pub level: u8,
}

impl Default for MD082Config {
    fn default() -> Self {
        Self { level: default_level() }
    }
}

impl RuleConfig for MD082Config {
    const RULE_NAME: &'static str = "MD082";
}

/// Position of a heading in the document, captured for adjacency analysis.
struct HeadingPos {
    /// 0-indexed line of the heading (the text line for a setext heading).
    index: usize,
    /// Heading level (1-6).
    level: u8,
    /// Whether the heading uses setext underlining (occupies two source lines).
    is_setext: bool,
    /// Whether the heading's anchor id came from a folded next-line `{#id}`
    /// attribute list rather than inline `{#id}` syntax in the heading text.
    /// Only a folded attribute list is part of the heading; an inline-id heading
    /// followed by a `{#id}` line leaves that line as ordinary content.
    id_from_next_line: bool,
    /// Heading text, for the diagnostic message.
    text: String,
}

#[derive(Debug, Clone, Default)]
pub struct MD082NoEmptySections {
    config: MD082Config,
}

impl MD082NoEmptySections {
    pub fn new() -> Self {
        Self::default()
    }

    pub fn from_config_struct(config: MD082Config) -> Self {
        Self { config }
    }

    /// Whether the line at `idx` (0-indexed) is a real section body line.
    /// Blank lines, HTML comments, reference definitions, and lone thematic
    /// breaks do not count as content. A heading's folded `{#id}` anchor line is
    /// handled in `check` by advancing the scan past it, not here, because an
    /// attribute list that is NOT folded renders as ordinary text and is content.
    fn is_content_line(&self, ctx: &LintContext, idx: usize) -> bool {
        let Some(li) = ctx.lines.get(idx) else {
            return false;
        };
        if li.is_blank || li.in_html_comment {
            return false;
        }
        // A lone thematic break is not a section body, but a `---` inside a
        // blockquote or list is that container's content (it renders), so only a
        // top-level thematic break is excluded.
        if li.is_horizontal_rule && li.blockquote.is_none() && !li.in_list_block {
            return false;
        }
        // Reference definitions: probe the first non-whitespace byte so an
        // indented definition is still recognised by the byte-range lookup.
        if ctx.is_in_reference_def(li.byte_offset + li.indent) {
            return false;
        }
        true
    }

    fn warn_empty_section(&self, ctx: &LintContext, heading: &HeadingPos) -> LintWarning {
        let line_content = ctx.lines.get(heading.index).map_or("", |l| l.content(ctx.content));
        let end_column = line_content.chars().count() + 1;
        LintWarning {
            rule_name: Some(self.name().to_string()),
            severity: Severity::Warning,
            line: heading.index + 1,
            column: 1,
            end_line: heading.index + 1,
            end_column,
            message: format!("Heading '{}' has no content before the next heading", heading.text),
            fix: None,
        }
    }
}

impl Rule for MD082NoEmptySections {
    fn name(&self) -> &'static str {
        "MD082"
    }

    fn description(&self) -> &'static str {
        "Headings should have content before the next heading"
    }

    fn category(&self) -> RuleCategory {
        RuleCategory::Heading
    }

    fn should_skip(&self, ctx: &LintContext) -> bool {
        !ctx.has_valid_headings()
    }

    fn check(&self, ctx: &LintContext) -> LintResult {
        let headings: Vec<HeadingPos> = ctx
            .valid_headings()
            .map(|h| HeadingPos {
                index: h.line_num - 1,
                level: h.heading.level,
                is_setext: matches!(h.heading.style, HeadingStyle::Setext1 | HeadingStyle::Setext2),
                // The id was folded from the next line when the heading has an id
                // but its own text carries no inline `{#id}`.
                id_from_next_line: h.heading.custom_id.is_some()
                    && crate::utils::header_id_utils::extract_header_id(&h.heading.raw_text)
                        .1
                        .is_none(),
                text: h.heading.text.clone(),
            })
            .collect();

        if headings.len() < 2 {
            return Ok(Vec::new());
        }

        let mut warnings = Vec::new();
        for pair in headings.windows(2) {
            let cur = &pair[0];
            let next = &pair[1];

            if cur.level < self.config.level {
                continue;
            }

            // The section body begins after the heading construct. A setext
            // heading occupies two source lines (text + underline); an ATX
            // heading occupies one. Skipping the underline avoids counting it
            // as content.
            let content_start = if cur.is_setext { cur.index + 2 } else { cur.index + 1 };

            // rumdl folds a `{#id}` attribute list on the line immediately after
            // the heading into the heading's anchor (only when the heading has no
            // inline id). That folded line is part of the heading, not the body,
            // so skip it. Match by id so an unrelated attribute list - or one
            // after a heading that already had an inline id - still counts as
            // content, matching how the parser renders it.
            let mut scan_start = content_start;
            if cur.id_from_next_line
                && let Some(li) = ctx.lines.get(content_start)
                && crate::utils::header_id_utils::is_standalone_attr_list(li.content(ctx.content))
            {
                scan_start = content_start + 1;
            }

            let has_content = (scan_start..next.index).any(|idx| self.is_content_line(ctx, idx));
            if !has_content {
                warnings.push(self.warn_empty_section(ctx, cur));
            }
        }

        Ok(warnings)
    }

    fn fix_capability(&self) -> FixCapability {
        FixCapability::Unfixable
    }

    fn fix(&self, ctx: &LintContext) -> Result<String, LintError> {
        // Detection only: inventing a section body would be guesswork, so
        // fixing is a no-op that returns the content unchanged.
        Ok(ctx.content.to_string())
    }

    fn as_any(&self) -> &dyn std::any::Any {
        self
    }

    crate::impl_rule_config_methods!(MD082Config);
}

#[cfg(test)]
mod tests {
    use super::*;
    use crate::config::MarkdownFlavor;
    use crate::rule::LintWarning;

    fn check(content: &str, config: MD082Config) -> Vec<LintWarning> {
        let rule = MD082NoEmptySections::from_config_struct(config);
        let ctx = LintContext::new(content, MarkdownFlavor::Standard, None);
        rule.check(&ctx).unwrap()
    }

    fn check_default(content: &str) -> Vec<LintWarning> {
        check(content, MD082Config::default())
    }

    #[test]
    fn default_level_is_one() {
        assert_eq!(MD082Config::default().level, 1);
    }

    #[test]
    fn flags_atx_heading_immediately_followed_by_heading() {
        let w = check_default("# A\n## B\n\nBody text\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
        assert!(w[0].message.contains('A'), "got: {}", w[0].message);
    }

    #[test]
    fn accepts_heading_with_paragraph_body() {
        let w = check_default("# A\n\nSome text\n\n## B\n\nMore text\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn flags_nested_empty_section_from_issue() {
        // The issue's second example: H1 has a body, but the H2 runs straight
        // into an H3 with nothing between, so the H2 section is empty.
        let content =
            "# Level 1 heading\n\nLevel 1 content\n\n## Empty Section\n### Level 3 heading\n\nLevel 3 content\n";
        let w = check_default(content);
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 5);
        assert!(w[0].message.contains("Empty Section"));
    }

    #[test]
    fn default_level_flags_h1_into_h2() {
        let w = check_default("# Title\n## Section\n\nBody\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn level_2_exempts_h1_but_flags_h2() {
        let config = MD082Config { level: 2 };
        // H1 -> H2 with no body: exempt at level 2.
        assert!(check("# Title\n## Section\n\nBody\n", config.clone()).is_empty());
        // H2 -> H3 with no body: still flagged at level 2.
        let w = check("# Title\n\nIntro\n\n## A\n### B\n\nBody\n", config);
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 5);
    }

    #[test]
    fn flags_setext_heading_into_setext_heading() {
        // "Title" (setext H1) runs straight into "Section" (setext H2) with only
        // the underline between, so the H1 section is empty.
        let w = check_default("Title\n=====\nSection\n-------\ncontent\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn accepts_setext_heading_with_body() {
        let w = check_default("Title\n=====\n\nSome body\n\nSection\n-------\n\nMore\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn blank_lines_do_not_count_as_content() {
        let w = check_default("# A\n\n\n## B\n\ncontent\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn html_comment_does_not_count_as_content() {
        let w = check_default("# A\n\n<!-- a comment -->\n\n## B\n\ncontent\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn reference_definition_does_not_count_as_content() {
        let w = check_default("# A\n\n[ref]: https://example.com\n\n## B\n\ncontent\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn thematic_break_does_not_count_as_content() {
        // All three CommonMark thematic-break styles are treated as non-content.
        for marker in ["---", "***", "___"] {
            let input = format!("# A\n\n{marker}\n\n## B\n\ncontent\n");
            let w = check_default(&input);
            assert_eq!(w.len(), 1, "marker {marker:?}: got: {w:?}");
            assert_eq!(w[0].line, 1, "marker {marker:?}");
        }
    }

    #[test]
    fn code_block_counts_as_content() {
        let w = check_default("# A\n\n```\ncode\n```\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn list_counts_as_content() {
        let w = check_default("# A\n\n- item\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn raw_html_block_counts_as_content() {
        let w = check_default("# A\n\n<div>hello</div>\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn trailing_heading_at_eof_is_not_flagged() {
        // Only heading-into-heading is in scope; a final heading with no body is
        // not flagged.
        let w = check_default("# A\n\nbody\n\n## B\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn single_heading_is_not_flagged() {
        assert!(check_default("# Only heading\n\ncontent\n").is_empty());
    }

    #[test]
    fn document_without_headings_is_not_flagged() {
        assert!(check_default("Just some text\nand more text\n").is_empty());
    }

    #[test]
    fn invalid_heading_renders_as_content() {
        // `#nospace` (lowercase, no space after `#`) is not a CommonMark heading;
        // it renders as a paragraph, so it counts as a body and the section is
        // not empty. (Uppercase-first variants are classified as valid by the
        // heading detector's heuristic; use lowercase to ensure invalidity.)
        let w = check_default("# A\n\n#nospace is text\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn standalone_attr_list_does_not_count_as_content() {
        // `{#a}` on its own line is folded into the heading's id, not a body.
        let w = check_default("# A\n{#a}\n## B\n\ntext\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn setext_standalone_attr_list_does_not_count_as_content() {
        // For setext the attr list sits on the line after the underline.
        let w = check_default("Title\n=====\n{#a}\nSection\n-------\ntext\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }

    #[test]
    fn non_folded_attr_list_counts_as_content() {
        // An attribute list separated from the heading by a blank line is NOT
        // folded into the heading; the parser renders it as an ordinary
        // paragraph, so it is real content and the section is not empty.
        let w = check_default("## A\n\n{#stray}\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn setext_non_folded_attr_list_counts_as_content() {
        // Setext heading with a blank line between the underline and the attr
        // list: not folded, so the attr list counts as content.
        let w = check_default("Title\n=====\n\n{#a}\nSection\n-------\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn inline_id_heading_with_following_attr_list_counts_as_content() {
        // The heading already has an inline id, so the next-line attr list is
        // NOT folded; it renders as a paragraph and counts as content.
        let w = check_default("## A {#x}\n{#y}\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn inline_id_then_matching_attr_list_counts_as_content() {
        // The heading has an INLINE id `{#x}`, so the parser does not fold the
        // following `{#x}` line; it renders as a paragraph and counts as content
        // even though its id matches the heading's. Must not flag.
        let w = check_default("## A {#x}\n{#x}\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn setext_inline_id_then_matching_attr_list_counts_as_content() {
        // Same as above for a setext heading with an inline id.
        let w = check_default("Title {#x}\n======\n{#x}\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn blockquoted_thematic_break_counts_as_content() {
        // `> ---` renders as a blockquote containing a thematic break: visible
        // content, not a bare top-level break. Must not flag.
        let w = check_default("# A\n\n> ---\n\n## B\n\ntext\n");
        assert!(w.is_empty(), "got: {w:?}");
    }

    #[test]
    fn multiline_reference_definition_is_an_empty_section() {
        // A reference definition is invisible, even when its title sits on a
        // continuation line. A section whose only body is such a definition is
        // empty and is flagged.
        let w = check_default("# A\n\n[ref]: https://example.com\n  \"title\"\n\n## B\n\ntext\n");
        assert_eq!(w.len(), 1, "got: {w:?}");
        assert_eq!(w[0].line, 1);
    }
}