react-native-pdf-text-search

Full-text search inside PDF files for React Native. Returns, for every match, the page, the matched text, and normalized rectangles you can draw as highlights.

| Platform | Engine | |----------|--------| | iOS | PDFKit | | Android | PdfBox-Android |

Status

Extracted from a production React Native app, where these modules have been shipping since 2024. The API is new, so it may still change before 1.0.

Install

npm install react-native-pdf-text-search
cd ios && pod install

Contains native code, so it needs a custom dev client — it does not run in Expo Go.

Usage

import { searchPDF } from 'react-native-pdf-text-search';

const { matches } = await searchPDF('/path/to/file.pdf', 'sustainability');

// matches[0] =
// {
//   page: 3,
//   text: 'sustainability',
//   rects: [{ x: 0.12, y: 0.34, width: 0.21, height: 0.02 }]
// }

Options

await searchPDF(path, 'term', {
  caseSensitive: false, // default false
  wholeWord: false,     // default false
  startPage: 1,         // 1-based, inclusive
  endPage: 10,          // 1-based, inclusive
});

Drawing the highlights

rects are normalized 0..1 against the page, origin top-left, one rect per line a match spans. Multiply by the size of the rendered page:

<View style={{ width, height }}>
  <Pdf source={{ uri }} style={{ flex: 1 }} page={page} />
  <View pointerEvents="none" style={StyleSheet.absoluteFill}>
    {matches
      .filter((m) => m.page === page)
      .flatMap((m, i) =>
        m.rects.map((r, j) => (
          <View
            key={`${i}-${j}`}
            style={{
              position: 'absolute',
              left: r.x * width,
              top: r.y * height,
              width: r.width * width,
              height: r.height * height,
              backgroundColor: 'rgba(255,235,59,0.4)',
            }}
          />
        )),
      )}
  </View>
</View>

Because the values are relative, they stay correct at any zoom level or screen size.

Errors

The promise rejects with:

| Code | Meaning | |------|---------| | E_FILE_NOT_FOUND | no file at the given path | | E_OPEN_FAILED | the file is not a readable PDF | | E_SEARCH_FAILED | any other native failure |

Limitations

• Only PDFs with a text layer. Scanned documents without OCR return no matches.
• Matches that wrap across lines yield several rects; there is no single bounding box.
• wholeWord uses alphanumeric boundaries, which suits Latin scripts best.
• Large documents are searched page by page on a background thread; a very large PDF still takes a few seconds.
• height differs slightly between platforms: iOS reports the selection line height (PDFKit includes leading), Android the glyph height — roughly half. x, y and width agree to four decimals. Pad the rect vertically if you need identical visuals.

License

MIT