Make scanned PDFs searchable.

use pdfluent::{Sdk, ocr::{OcrLayerOptions, HocrResult}};
use std::process::Command;

fn main() -> pdfluent::Result<()> {
    let sdk = Sdk::new()?;
    let doc = sdk.open("scanned_contract.pdf")?;

    // Detect which pages are image-only (no text layer)
    let scanned_pages: Vec<u32> = doc.pages()
        .filter(|p| p.is_image_only())
        .map(|p| p.index())
        .collect();

    println!("{} of {} pages are scanned", scanned_pages.len(), doc.page_count());

    let mut builder = doc.add_ocr_layer();

    for page_index in &scanned_pages {
        // Extract the page as a 300 DPI PNG for OCR
        let img = doc.render_page(*page_index, Default::default())?;
        img.save(format!("/tmp/page_{}.png", page_index))?;

        // Run Tesseract and get hOCR output
        Command::new("tesseract")
            .args([
                &format!("/tmp/page_{}.png", page_index),
                &format!("/tmp/page_{}", page_index),
                "-l", "eng", "hocr",
            ])
            .status()?;

        let hocr = std::fs::read_to_string(
            format!("/tmp/page_{}.hocr", page_index)
        )?;

        // Write invisible text overlay back into the page
        let result = HocrResult::parse(&hocr)?;
        builder.add_page(*page_index, result);
    }

    let opts = OcrLayerOptions::builder()
        .text_rendering_mode(pdfluent::ocr::TextRenderingMode::Invisible)
        .conform_to_pdfa2b(true)
        .build();

    let searchable = builder.finish(opts)?;
    searchable.save("scanned_contract_searchable.pdf")?;

    println!("Saved searchable PDF with {} OCR pages", scanned_pages.len());
    Ok(())
}