dom-content-extraction 0.4.4

Rust implementation of Content extraction via text density paper
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

use ego_tree::NodeId;
use scraper::{Html, Selector};
use std::sync::LazyLock;

/// Selector for `<body>` tag
pub static BODY_SELECTOR: LazyLock<Selector> = LazyLock::new(|| {
    Selector::parse("body").expect("Can't be (parsing body selector)")
});

#[derive(Debug, Clone, Default)]
pub struct NodeMetrics {
    pub char_count: u32,
    pub tag_count: u32,
    pub link_char_count: u32,
    pub link_tag_count: u32,
}

pub trait TreeBuilder {
    fn build_metrics(&self, node_id: NodeId) -> NodeMetrics;
    fn get_children(&self, node_id: NodeId) -> Vec<NodeId>;
    fn get_parent(&self, node_id: NodeId) -> Option<NodeId>;
}

pub struct HtmlTreeBuilder<'a> {
    document: &'a Html,
}

impl<'a> HtmlTreeBuilder<'a> {
    pub fn new(document: &'a Html) -> Self {
        Self { document }
    }
}

impl TreeBuilder for HtmlTreeBuilder<'_> {
    fn build_metrics(&self, node_id: NodeId) -> NodeMetrics {
        let node = self.document.tree.get(node_id).unwrap();

        let mut metrics = NodeMetrics {
            char_count: 0,
            tag_count: 0,
            link_char_count: 0,
            link_tag_count: 0,
        };

        match node.value() {
            scraper::Node::Text(text) => {
                // NOTE: old method calculation
                // metrics.char_count = text.trim().len() as u32;
                let clean_text = text.trim();
                if !crate::utils::is_non_content_text(clean_text) {
                    metrics.char_count =
                        crate::unicode::count_graphemes(clean_text);
                }
            }
            scraper::Node::Element(elem) => {
                metrics.tag_count = 1;
                if elem.name() == "a"
                    || elem.name() == "button"
                    || elem.name() == "select"
                {
                    metrics.link_tag_count = 1;
                }
            }
            _ => {}
        }

        metrics
    }

    fn get_children(&self, node_id: NodeId) -> Vec<NodeId> {
        self.document
            .tree
            .get(node_id)
            .map(|node| {
                node.children()
                    .filter(|child| match child.value() {
                        scraper::Node::Element(elem) => {
                            !crate::utils::should_skip_element(elem)
                        }
                        scraper::Node::Text(text) => {
                            !crate::utils::is_non_content_text(text)
                        }
                        scraper::Node::Comment(_) | scraper::Node::Document => {
                            false
                        }
                        _ => true,
                    })
                    .map(|child| child.id())
                    .collect()
            })
            .unwrap_or_default()
    }

    fn get_parent(&self, node_id: NodeId) -> Option<NodeId> {
        self.document
            .tree
            .get(node_id)
            .and_then(|node| node.parent())
            .map(|parent| parent.id())
    }
}

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

    // Add metrics from another (child) node
    pub fn combine(&mut self, other: &NodeMetrics) {
        self.char_count += other.char_count;
        self.tag_count += other.tag_count;
        self.link_char_count += other.link_char_count;
        self.link_tag_count += other.link_tag_count;
    }

    // Calculate simple density (chars/tags ratio)
    pub fn calculate_simple_density(&self) -> f32 {
        if self.tag_count == 0 {
            0.0
        } else {
            self.char_count as f32 / self.tag_count as f32
        }
    }
}

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

    const TEST_HTML: &str = r#"
        <html>
        <body>
            <div class="content">
                Some text here
                <a href="\#">A link</a>
                <p>More content</p>
                <button>Click me</button>
                <script>console.log('skip');</script>
                <style>.skip{}</style>
            </div>
            <div class="sidebar">
                <select>
                    <option>Option 1</option>
                </select>
            </div>
        </body>
        </html>
    "#;

    #[test]
    fn test_body_selector_initialization() {
        // This will force the LazyLock to initialize
        let _ = &*BODY_SELECTOR;
    }

    #[test]
    fn test_node_metrics() {
        let document = Html::parse_document(TEST_HTML);
        let builder = HtmlTreeBuilder::new(&document);

        // Get content div
        let content_div = document
            .select(&scraper::Selector::parse("div.content").unwrap())
            .next()
            .unwrap();

        let metrics = builder.build_metrics(content_div.id());
        // notice that we do not count children nodes metrics
        assert_eq!(metrics.char_count, 0); // div itself has no direct text
        assert_eq!(metrics.tag_count, 1); // div counts as one tag
        assert_eq!(metrics.link_tag_count, 0); // div is not a link
    }

    #[test]
    fn test_get_children_filters() {
        let document = Html::parse_document(TEST_HTML);
        let builder = HtmlTreeBuilder::new(&document);

        let body = document
            .select(&scraper::Selector::parse("body").unwrap())
            .next()
            .unwrap();

        let children = builder.get_children(body.id());

        // Should get both divs but skip script and style
        assert_eq!(children.len(), 2);
    }

    #[test]
    fn test_link_metrics() {
        let document = Html::parse_document(TEST_HTML);
        let builder = HtmlTreeBuilder::new(&document);

        // Test link element
        let link = document
            .select(&scraper::Selector::parse("a").unwrap())
            .next()
            .unwrap();

        let metrics = builder.build_metrics(link.id());
        assert_eq!(metrics.link_tag_count, 1);

        // Test button
        let button = document
            .select(&scraper::Selector::parse("button").unwrap())
            .next()
            .unwrap();

        let metrics = builder.build_metrics(button.id());
        assert_eq!(metrics.link_tag_count, 1);

        // Test select
        let select = document
            .select(&scraper::Selector::parse("select").unwrap())
            .next()
            .unwrap();

        let metrics = builder.build_metrics(select.id());
        assert_eq!(metrics.link_tag_count, 1);
    }

    #[test]
    fn test_text_metrics() {
        let document = Html::parse_document(TEST_HTML);
        let builder = HtmlTreeBuilder::new(&document);

        // Find text node
        let text_node = document
            .select(&scraper::Selector::parse(".content").unwrap())
            .next()
            .unwrap()
            .first_child()
            .unwrap();

        let metrics = builder.build_metrics(text_node.id());
        assert_eq!(metrics.char_count, 14); // "Some text here"
        assert_eq!(metrics.tag_count, 0);
        assert_eq!(metrics.link_tag_count, 0);
    }

    #[test]
    fn test_get_parent() {
        let document = Html::parse_document(TEST_HTML);
        let builder = HtmlTreeBuilder::new(&document);

        let content_div = document
            .select(&scraper::Selector::parse("div.content").unwrap())
            .next()
            .unwrap();

        let parent = builder.get_parent(content_div.id());
        assert!(parent.is_some());

        // Body should be parent
        let body = document
            .select(&scraper::Selector::parse("body").unwrap())
            .next()
            .unwrap();

        assert_eq!(parent.unwrap(), body.id());
    }
}