edgeparse_core/output/
text.rs

1//! Plain text output generator.
2
3use crate::models::content::ContentElement;
4use crate::models::document::PdfDocument;
5use crate::models::table::TableTokenRow;
6use crate::EdgePdfError;
7
8/// Generate plain text representation of a PdfDocument.
9///
10/// # Errors
11/// Returns `EdgePdfError::OutputError` on write failures.
12pub fn to_text(doc: &PdfDocument) -> Result<String, EdgePdfError> {
13    let mut output = String::new();
14
15    if doc.kids.is_empty() {
16        output.push_str("[No content extracted]\n");
17        return Ok(output);
18    }
19
20    for element in &doc.kids {
21        render_element(&mut output, element);
22    }
23
24    Ok(output)
25}
26
27/// Extract text from table token rows.
28fn token_rows_text(rows: &[TableTokenRow]) -> String {
29    rows.iter()
30        .flat_map(|row| row.iter())
31        .map(|token| token.base.value.as_str())
32        .collect::<Vec<_>>()
33        .join(" ")
34}
35
36fn render_element(out: &mut String, element: &ContentElement) {
37    match element {
38        ContentElement::Heading(h) => {
39            let text = h.base.base.value();
40            let trimmed = text.trim();
41            if !trimmed.is_empty() {
42                out.push_str(trimmed);
43                out.push_str("\n\n");
44            }
45        }
46        ContentElement::Paragraph(p) => {
47            let text = p.base.value();
48            let trimmed = clean_text(&text);
49            if !trimmed.is_empty() {
50                out.push_str(&trimmed);
51                out.push_str("\n\n");
52            }
53        }
54        ContentElement::List(list) => {
55            for item in &list.list_items {
56                let label = token_rows_text(&item.label.content);
57                let body = token_rows_text(&item.body.content);
58                let label_trimmed = label.trim();
59                let body_trimmed = body.trim();
60                if !label_trimmed.is_empty() || !body_trimmed.is_empty() {
61                    if !label_trimmed.is_empty() && !body_trimmed.is_empty() {
62                        out.push_str(&format!("  {} {}\n", label_trimmed, body_trimmed));
63                    } else if !body_trimmed.is_empty() {
64                        out.push_str(&format!("  {}\n", body_trimmed));
65                    } else {
66                        out.push_str(&format!("  {}\n", label_trimmed));
67                    }
68                }
69            }
70            out.push('\n');
71        }
72        ContentElement::Image(_) => {
73            out.push_str("[Image]\n\n");
74        }
75        ContentElement::HeaderFooter(_) => {
76            // Skip headers/footers in text by default
77        }
78        ContentElement::TextBlock(tb) => {
79            let text = tb.value();
80            let trimmed = clean_text(&text);
81            if !trimmed.is_empty() {
82                out.push_str(&trimmed);
83                out.push_str("\n\n");
84            }
85        }
86        ContentElement::TextLine(tl) => {
87            let text = tl.value();
88            let trimmed = text.trim();
89            if !trimmed.is_empty() {
90                out.push_str(trimmed);
91                out.push('\n');
92            }
93        }
94        ContentElement::TextChunk(tc) => {
95            out.push_str(&tc.value);
96        }
97        _ => {}
98    }
99}
100
101/// Clean paragraph text: trim whitespace, collapse multiple spaces.
102fn clean_text(text: &str) -> String {
103    let trimmed = text.trim();
104    if trimmed.is_empty() {
105        return String::new();
106    }
107    let mut result = String::with_capacity(trimmed.len());
108    let mut prev_space = false;
109    for ch in trimmed.chars() {
110        if ch == ' ' || ch == '\t' {
111            if !prev_space {
112                result.push(' ');
113                prev_space = true;
114            }
115        } else {
116            result.push(ch);
117            prev_space = false;
118        }
119    }
120    result
121}
122
123#[cfg(test)]
124mod tests {
125    use super::*;
126
127    #[test]
128    fn test_empty_doc() {
129        let doc = PdfDocument::new("test.pdf".to_string());
130        let text = to_text(&doc).unwrap();
131        assert!(text.contains("No content extracted"));
132    }
133}
edgeparse_core/output/text.rs

edgeparse_core/output/
text.rs
