cvxtract 0.2.0

LLM-powered structured extraction from CVs/resumes — PDF, DOCX, HTML, TXT input; typed Rust structs output.
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

use super::{Document, DocumentElement, DocumentLoader, DocumentMetadata, FileType, Result};
use std::fs;
use std::path::Path;

/// Simple text file loader
pub struct TextLoader;

impl TextLoader {
    pub fn new() -> Self {
        Self
    }
}

impl DocumentLoader for TextLoader {
    fn load_from_path(&self, path: &Path) -> Result<Document> {
        let content = fs::read_to_string(path)?;
        let metadata = fs::metadata(path)?;

        let doc_metadata = DocumentMetadata {
            filename: path.file_name().map(|s| s.to_string_lossy().to_string()),
            file_size: Some(metadata.len()),
            file_type: FileType::Text,
            page_count: None, // Text files don't have pages
        };

        // Basic text processing - split into paragraphs as chunks
        let elements = content
            .split("\n\n")
            .filter(|s| !s.trim().is_empty())
            .enumerate()
            .map(|(i, paragraph)| DocumentElement {
                element_type: "paragraph".to_string(),
                text: paragraph.trim().to_string(),
                metadata: Some(format!("paragraph_{i}")),
            })
            .collect();

        Ok(Document {
            content,
            metadata: doc_metadata,
            elements,
        })
    }

    fn load_from_bytes(&self, data: &[u8], filename: Option<&str>) -> Result<Document> {
        let content = String::from_utf8_lossy(data).to_string();

        let doc_metadata = DocumentMetadata {
            filename: filename.map(|s| s.to_string()),
            file_size: Some(data.len() as u64),
            file_type: FileType::Text,
            page_count: None,
        };

        // Basic text processing - split into paragraphs as chunks
        let elements = content
            .split("\n\n")
            .filter(|s| !s.trim().is_empty())
            .enumerate()
            .map(|(i, paragraph)| DocumentElement {
                element_type: "paragraph".to_string(),
                text: paragraph.trim().to_string(),
                metadata: Some(format!("paragraph_{i}")),
            })
            .collect();

        Ok(Document {
            content,
            metadata: doc_metadata,
            elements,
        })
    }
}