Performant and lightweight Node.js library to convert PDF file to images. Designed with high focus on performance and developer experience.
| Test/Feature | pdf-to-image-generator | pdf-to-png-converter | pdf-to-img |
|---|---|---|---|
| 2-pages-pdf-with-forms | 381ms | 91% slower | 115% slower |
| large-pdf-with-old-characters | 13629ms | 7% slower | 31% slower |
| rich-pdf-with-images-form-text | 14658ms | 3% slower | 22% slower |
| jpeg support | ✔️ | ❌ | ❌ |
| webp support | ✔️ | ❌ | ❌ |
| pausing flow | ✔️ | ❌ | ❌ |
| resume flow | ✔️ | ❌ | ❌ |
| stop flow | ✔️ | ❌ | ❌ |
| get number pages | ✔️ | ❌ | ❌ |
| get text content | ✔️ | ❌ | ❌ |
| progression event | ✔️ | ❌ | ❌ |
| class based API* | ✔️ | ❌ | ❌ |
*A class-based API offers more flexibility, allowing for reusable instances and caching, which improves efficiency across multiple conversion processes.
Installation:
npm install pdf-to-image-generator
Find the complete documentation here.
// The path of your local PDF
const filePath = path.join(__dirname, '../my_document.pdf');
// Conversion options
const options: PDFToIMGOptions = {
// the name of the folder where files will be rendered
outputFolderName: 'upload',
// controls scaling
viewportScale: 2,
};
// Load the pdf file
const pdf = await new PDFToImage().load(filePath);
// Get the number of pages of the document
const numPages = pdf.document.numPages;
// Get the text content of the document
const textContent = await pdf.getTextContent();
setTimeout(() => {
// Simulating a user pausing the conversion flows
pdf.pause();
setTimeout(async () => {
// Resume conversion flows
await pdf.resume();
}, 500);
}, 500);
// Listen to a progression event
pdf.on('progress', (data) => {
console.log(
`Page: ${data.currentPage}. Total: ${data.totalPages}. Progression: ${data.progress}%`
);
});
// Notification of end of conversion
pdf.on('end', (data) => {
console.log('End of conversion.', data);
});
// Convert your PDF to image
const pdfConversion = await pdf.convert(options);
// Other conversion flows with the same pdf but different options...
// Remove generated images on disk
await pdf.removeGeneratedImagesOnDisk();
Focus on developer experience. Properties and methods are exhaustively documented and typed. Use your IDE's autocompletion or hover on properties and methods for help.
If you use my library in a Next.js project, please add the following code in your project:
import { GlobalWorkerOptions } from 'pdfjs-dist';
import pdfjsWorker from 'pdfjs-dist/build/pdf.worker.min.mjs';
GlobalWorkerOptions.workerSrc = new URL(pdfjsWorker, import.meta.url).toString();
// OR
const nextConfig = {
experimental: {
serverComponentsExternalPackages: ['pdf-to-image-generator'],
},
...
}
Next.js optimizes dependencies by importing only the strictly necessary elements (aggressive tree shaking), which results in the removal of the worker from pdfjs-dist, a dependency used in my library. The second lines raises a Typescript error, which you can ignore. I'm not personnaly using Next.js, so if you find a better solution, please share it in the Discussion on Github.
Feel free! Open a ticket.
V0.0.3: API changes. Exposing the text content in a separate method. Width and height have been removed from the returned object of conversion.
V0.0.5: Exposing the name of the file in the getTextContent method.
V0.0.6: API changes, giving more flexibility and offering a cleaner way of options initialisation. Adding pausing, resume and stop methods to give more control over the conversion flow. Adding progression and end events.
V0.0.7: Removal of possible duplicate page index in the option provided by the user and adding verification.
V0.0.8: Addition of a method to remove generated images on disk. Addition of a getter for accessing the loaded pdf document and all its methods. Minor changes to the returned object of conversion.
V0.0.9: Bug path correction.
V1.0.0: Performance improvement and adding child process documentation.
V1.0.1: Bug correction, setting jpeg as default type, changing behavior if empty array of pages is passed (no action is taken) and adding tests.
V1.0.2: Removing node canvas dependency, adapting options.
V1.0.4: Improving DX for multiple conversions with option includeBufferContent.
V1.0.4: Removing unnecessary images writing, changing buffer content to asynchronous behaviour.
All libraries are permanently supported. Discover them here.