readabilityrs 0.1.4

A Rust port of Mozilla's Readability library for extracting article content from web pages
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

//! Quick readability check without full parsing.
//!
//! This module provides the [`is_probably_readerable`] function, which performs
//! a fast pre-flight check to determine if a document is likely to have extractable
//! article content without doing a full parse.
//!
//! ## Use Case
//!
//! Use this function to quickly filter out pages that are unlikely to contain article
//! content, saving the cost of a full parse:
//!
//! ```rust
//! use readabilityrs::{is_probably_readerable, Readability};
//!
//! let html = "<html>...</html>";
//!
//! // Quick check first
//! if is_probably_readerable(html, None) {
//!     // Do full parse
//!     let readability = Readability::new(html, None, None).unwrap();
//!     if let Some(article) = readability.parse() {
//!         println!("Article extracted!");
//!     }
//! } else {
//!     println!("Not an article page, skipping parse");
//! }
//! ```
//!
//! ## Performance
//!
//! This check is significantly faster than a full parse because it only looks
//! for basic content signals without doing deep analysis or scoring.

use crate::constants::REGEXPS;
use crate::dom_utils;
use scraper::{ElementRef, Html, Selector};

/// Options for the readability pre-flight check.
///
/// Controls the thresholds used by [`is_probably_readerable`] to determine
/// if a document is likely to be parseable.
///
/// ## Example
///
/// ```rust
/// use readabilityrs::{is_probably_readerable, ReaderableOptions};
///
/// let html = "<html>...</html>";
///
/// let options = ReaderableOptions {
///     min_content_length: 200,
///     min_score: 30.0,
/// };
///
/// let is_readerable = is_probably_readerable(html, Some(options));
/// ```
#[derive(Debug, Clone)]
pub struct ReaderableOptions {
    /// Minimum content length to consider a paragraph.
    ///
    /// Paragraphs shorter than this are ignored when calculating the
    /// readability score.
    ///
    /// Default: `140`
    pub min_content_length: usize,

    /// Minimum score threshold to consider a page readerable.
    ///
    /// The score is calculated based on the length and number of content
    /// paragraphs found in the document.
    ///
    /// Default: `20.0`
    pub min_score: f64,
}

impl Default for ReaderableOptions {
    fn default() -> Self {
        Self {
            min_content_length: 140,
            min_score: 20.0,
        }
    }
}

/// Quick check to determine if a document is likely to be readerable.
///
/// This function performs a fast analysis to predict whether full article extraction
/// is likely to succeed, without doing the expensive full parse. It looks for basic
/// content signals like paragraphs with sufficient text.
///
/// ## Arguments
///
/// * `html` - The HTML document to check
/// * `options` - Optional custom thresholds (uses defaults if `None`)
///
/// ## Returns
///
/// `true` if the document likely contains extractable article content, `false` otherwise.
///
/// ## Example
///
/// ```rust
/// use readabilityrs::is_probably_readerable;
///
/// let article_html = r#"
///     <html><body>
///         <article>
///             <p>This is a substantial paragraph with enough content to indicate that this page
///             likely contains article text that can be extracted by the readability algorithm.
///             The paragraph needs to be long enough to pass the minimum content length threshold.</p>
///         </article>
///     </body></html>
/// "#;
///
/// assert!(is_probably_readerable(article_html, None));
///
/// let non_article_html = "<html><body><p>Short</p></body></html>";
/// assert!(!is_probably_readerable(non_article_html, None));
/// ```
///
/// ## With Custom Options
///
/// ```rust
/// use readabilityrs::{is_probably_readerable, ReaderableOptions};
///
/// let html = "<html>...</html>";
/// let options = ReaderableOptions {
///     min_content_length: 200,
///     min_score: 30.0,
/// };
///
/// if is_probably_readerable(html, Some(options)) {
///     println!("Likely readerable with stricter thresholds");
/// }
/// ```
///
/// ## Algorithm
///
/// This mirrors Mozilla's `isProbablyReaderable`. The candidate set is every
/// `<p>`, `<pre>` and `<article>` element, plus any `<div>` with a direct `<br>`
/// child, which is how prose laid out with line breaks instead of paragraphs
/// gets counted.
///
/// A candidate is skipped when it is hidden, when its class or id marks it as an
/// unlikely candidate (comment thread, sidebar, footer) without also matching the
/// "maybe a candidate" pattern, or when it is a `<p>` directly inside an `<li>`.
/// Surviving candidates shorter than `min_content_length` contribute nothing;
/// the rest add `sqrt(len - min_content_length)`. The function returns `true`
/// as soon as the running score passes `min_score`.
///
/// ## Performance
///
/// This is far cheaper than a full parse, which makes it useful for batch processing
/// large numbers of URLs, pre-filtering in crawlers or scrapers, and quick content
/// classification.
pub fn is_probably_readerable(html: &str, options: Option<ReaderableOptions>) -> bool {
    let options = options.unwrap_or_default();
    let document = Html::parse_document(html);

    let Ok(candidate_selector) = Selector::parse("p, pre, article, div") else {
        return false;
    };

    let mut score = 0.0;

    for node in document.select(&candidate_selector) {
        if !is_scored_candidate(node) {
            continue;
        }

        let text_len = node.text().collect::<String>().trim().len();
        if text_len < options.min_content_length {
            continue;
        }

        score += ((text_len - options.min_content_length) as f64).sqrt();

        if score > options.min_score {
            return true;
        }
    }

    false
}

/// Whether a node from the candidate selector should contribute to the score.
///
/// A `<div>` only qualifies through a direct `<br>` child, which is how Mozilla
/// picks up prose laid out with line breaks instead of paragraphs. The selector
/// matches every div so the tree is walked once, and the rest are rejected here.
fn is_scored_candidate(node: ElementRef) -> bool {
    let tag = node.value().name();

    if tag.eq_ignore_ascii_case("div") && !has_direct_br_child(node) {
        return false;
    }

    // Mozilla excludes list-item paragraphs: they are almost always navigation
    // or link lists rather than prose.
    if tag.eq_ignore_ascii_case("p") && parent_is_list_item(node) {
        return false;
    }

    if !dom_utils::is_probably_visible(node) {
        return false;
    }

    let class = node.value().attr("class").unwrap_or("");
    let id = node.value().attr("id").unwrap_or("");
    let match_string = format!("{class} {id}");

    if REGEXPS.unlikely_candidates.is_match(&match_string)
        && !REGEXPS.ok_maybe_its_a_candidate.is_match(&match_string)
    {
        return false;
    }

    true
}

fn has_direct_br_child(node: ElementRef) -> bool {
    node.children()
        .filter_map(ElementRef::wrap)
        .any(|child| child.value().name().eq_ignore_ascii_case("br"))
}

fn parent_is_list_item(node: ElementRef) -> bool {
    node.parent()
        .and_then(ElementRef::wrap)
        .is_some_and(|parent| parent.value().name().eq_ignore_ascii_case("li"))
}

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

    #[test]
    fn test_is_probably_readerable() {
        let html = r#"
            <html>
                <body>
                    <article>
                        <p>This is a long enough paragraph that should make the content readerable.
                        It has sufficient content to pass the minimum threshold check. Adding more text here to ensure
                        we definitely exceed the 140 character minimum requirement for each paragraph element.</p>
                        <p>Another paragraph with more content to increase the score. This paragraph also needs to be
                        long enough to contribute to the overall readability score calculation and help us pass the test.</p>
                    </article>
                </body>
            </html>
        "#;

        assert!(is_probably_readerable(html, None));
    }

    #[test]
    fn test_not_readerable() {
        let html = r#"
            <html>
                <body>
                    <p>Short</p>
                </body>
            </html>
        "#;

        assert!(!is_probably_readerable(html, None));
    }

    /// Three paragraphs of real prose, long enough that the page is readerable
    /// unless a filter excludes them. `class_attr` lands on the paragraphs
    /// themselves, because the unlikely-candidate check reads each scored node's
    /// own class and id and never walks ancestors.
    fn prose_paragraphs(class_attr: &str) -> String {
        (1..=3)
            .map(|i| {
                format!(
                    "<p class=\"{class_attr}\">Paragraph {i} carries well over the hundred and \
                     forty character minimum that a candidate needs before it contributes to the \
                     score at all, so three of them comfortably clear the threshold.</p>"
                )
            })
            .collect()
    }

    #[test]
    fn test_unlikely_candidate_is_excluded() {
        let html = format!("<html><body>{}</body></html>", prose_paragraphs("comment"));

        assert!(!is_probably_readerable(&html, None));
    }

    #[test]
    fn test_ok_maybe_candidate_overrides_unlikely() {
        let html = format!(
            "<html><body>{}</body></html>",
            prose_paragraphs("comment main-content")
        );

        assert!(is_probably_readerable(&html, None));
    }

    /// The check is per node, matching Mozilla: a container marked unlikely does
    /// not disqualify unmarked paragraphs inside it. Worth pinning, since the
    /// opposite is the intuitive assumption.
    #[test]
    fn test_unlikely_wrapper_does_not_exclude_its_children() {
        let html = format!(
            "<html><body><div class=\"comment\">{}</div></body></html>",
            prose_paragraphs("")
        );

        assert!(is_probably_readerable(&html, None));
    }

    #[test]
    fn test_hidden_content_is_excluded() {
        let html = format!(
            "<html><body><div style=\"display:none\">{}</div></body></html>",
            prose_paragraphs("")
        );

        assert!(!is_probably_readerable(&html, None));
    }

    /// Prose separated by `<br>` rather than wrapped in paragraphs is only
    /// reachable through the `div > br` rule, so this fails without it.
    #[test]
    fn test_div_with_br_child_is_counted() {
        let line = "A line of prose long enough on its own to matter, padded out past the \
                    hundred and forty character minimum that a candidate needs before it \
                    contributes anything to the score.";
        // The div is a single candidate, so its whole text yields one sqrt term;
        // it needs to clear min_score of 20 on its own, i.e. over 540 characters.
        let body = [line; 6].join("<br>");
        let html = format!("<html><body><div>{body}</div></body></html>");

        assert!(is_probably_readerable(&html, None));
    }

    /// A bare `<div>` of prose with neither `<br>` nor `<p>` children is not a
    /// candidate at all. Without the `div > br` rule every div would score, so
    /// this is what keeps the rule honest.
    #[test]
    fn test_bare_div_of_text_is_not_a_candidate() {
        let filler = "Plenty of prose sitting directly in a div with no line breaks and no \
                      paragraph children whatsoever. "
            .repeat(10);
        let html = format!("<html><body><div>{filler}</div></body></html>");

        assert!(!is_probably_readerable(&html, None));
    }

    #[test]
    fn test_div_without_br_is_not_counted_itself() {
        // The div is skipped, but its paragraphs are scored in their own right.
        let html = format!(
            "<html><body><div>{}</div></body></html>",
            prose_paragraphs("")
        );

        assert!(is_probably_readerable(&html, None));
    }

    #[test]
    fn test_paragraphs_inside_list_items_are_excluded() {
        // Each entry is long enough that a single one would clear min_score, so
        // the page is only unreadable because the list-item rule drops them.
        let items: String = (1..=3)
            .map(|i| {
                let filler = "This entry is deliberately verbose so that its length alone would \
                              carry the page over the score threshold. "
                    .repeat(6);
                format!("<li><p>Entry {i}. {filler}</p></li>")
            })
            .collect();
        let html = format!("<html><body><ul>{items}</ul></body></html>");

        assert!(!is_probably_readerable(&html, None));
    }
}