docuparse 0.0.1

Document extraction and parsing — PDF text, rendering, and OCR
Documentation
use pdfium_render::prelude::PdfPageIndex;

use crate::{
    error::{PdfError, TextError},
    PdfDocument,
};

impl<'p> PdfDocument<'p> {
    /// Extract native text layer from a page.
    /// Returns None if the page has no text layer (scanned).
    pub fn extract_text_layer(&self, page_index: u32) -> Result<Option<String>, PdfError> {
        if page_index >= self.page_count() {
            return Err(TextError::ExtractionFailed { page: page_index }.into());
        }

        let page = self
            .pages()
            .get(page_index as PdfPageIndex)
            .map_err(|_| TextError::ExtractionFailed { page: page_index })?;

        let text = page
            .text()
            .map_err(|_| TextError::ExtractionFailed { page: page_index })?
            .all();

        if text.trim().is_empty() {
            Ok(None)
        } else {
            Ok(Some(text))
        }
    }

    /// Returns (page_index, Option<text>) for all pages.
    /// None means the page needs OCR.
    pub fn extract_all_text_layers(
        &self,
    ) -> impl Iterator<Item = (u32, Result<Option<String>, PdfError>)> + '_ {
        (0..self.page_count()).map(|i| (i, self.extract_text_layer(i)))
    }
}