GSoC 2026: Interested in contributing to Wayback PDF Changes

[welcomemeet]

Hello,

I am a Google Summer of Code 2026 applicant interested in the "Wayback PDF Changes" project.

I have previously contributed to the Internet Archive repository archive-pdf-tools where my pull requests were merged. I would like to continue contributing and prepare for this project.

I also tried accessing the Internet Archive interest form mentioned in the GSoC document, but it shows "Access Required".

Could you please guide me on:

1. How I can connect with the mentors for this project
2. Where I should start contributing in the Wayback diff system

Thank you for your time.

[welcomemeet]

Hello,

I am preparing a GSoC proposal for the Wayback PDF Changes idea.

While exploring the project, I built a small prototype that extracts text from two PDFs, detects differences, and visualizes them in a browser-based viewer.

Repository:
https://github.com/welcomemeet/wayback-pdf-diff-prototype

I would really appreciate feedback on whether this approach aligns with the current Wayback diff architecture and how PDF support could best integrate with the existing system.

Thank you!
Meet Darji

[vbanos]

The current HTML diff https://web.archive.org/web/diff/20220901000000/20230215235959/https://www.iskme.org
works using https://github.com/edgi-govdata-archiving/web-monitoring-diff.

We must create another PDF diff project that has the same JSON output as web-monitoring-diff.
Then, with a small WBM UI change, we call the PDF diff when the mimetype of the captures we compare is PDF.

You don't need to do any UI work. We only need a backend. The UI is already working fine as long as your program has the same output as web-monitoring-diff.
Your code should be independent from the web service. It should be like a class / package that is imported by the web service.
In the future we may integrate it with https://github.com/edgi-govdata-archiving/web-monitoring-diff or other apps.
It should be easy to do that.
Another requirement: your project should be a package that can be installed via pip install .
These 2 example projects showcase how the package should be like:
https://github.com/internetarchive/warcprox
https://github.com/internetarchive/brozzler

The target Python version is 3.10+.

[welcomemeet]

Thank you for the detailed explanation. That makes the requirements much clearer. I understand that the goal is to implement a backend PDF diff engine that produces the same JSON output format as web-monitoring-diff, without requiring any UI work. I will study the structure of web-monitoring-diff and the example projects you shared, and start working on a Python package that performs PDF diffing and returns compatible JSON output. Thanks again for the guidance. Meet Darji

…

On Thu, 12 Mar 2026 at 13:59, Vangelis Banos ***@***.***> wrote: *vbanos* left a comment (internetarchive/wayback-diff#195) <#195 (comment)> The current HTML diff https://web.archive.org/web/diff/20220901000000/20230215235959/https://www.iskme.org works using https://github.com/edgi-govdata-archiving/web-monitoring-diff. We must create another PDF diff project that has the same JSON output as web-monitoring-diff. Then, with a small WBM UI change, we call the PDF diff when the mimetype of the captures we compare is PDF. You don't need to do any UI work. We only need a backend. The UI is already working fine as long as your program has the same output as web-monitoring-diff. Your code should be independent from the web service. It should be like a class / package that is imported by the web service. In the future we may integrate it with https://github.com/edgi-govdata-archiving/web-monitoring-diff or other apps. It should be easy to do that. Another requirement: your project should be a package that can be installed via pip install . These 2 example projects showcase how the package should be like: https://github.com/internetarchive/warcprox https://github.com/internetarchive/brozzler The target Python version is 3.10+. — Reply to this email directly, view it on GitHub <#195 (comment)>, or unsubscribe <https://github.com/notifications/unsubscribe-auth/BE5R5RPZGKNBQ2YOO7EF4UT4QJYQBAVCNFSM6AAAAACWNZGWU2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DANBUHA4TKNRYGE> . You are receiving this because you authored the thread.Message ID: ***@***.***>

[welcomemeet]

I created a backend prototype for comparing two PDF documents and generating structured diff output. The prototype extracts text using PyMuPDF, compares content using Python’s difflib, and outputs structured JSON results. I also structured the project as a pip-installable Python package as suggested. Repository: https://github.com/welcomemeet/wayback-pdf-diff-prototype Next I will work on aligning the JSON output format with the structure used by web-monitoring-diff.

…

On Thu, 12 Mar 2026, 14:11 Meet Darji, ***@***.***> wrote: Thank you for the detailed explanation. That makes the requirements much clearer. I understand that the goal is to implement a backend PDF diff engine that produces the same JSON output format as web-monitoring-diff, without requiring any UI work. I will study the structure of web-monitoring-diff and the example projects you shared, and start working on a Python package that performs PDF diffing and returns compatible JSON output. Thanks again for the guidance. Meet Darji On Thu, 12 Mar 2026 at 13:59, Vangelis Banos ***@***.***> wrote: > *vbanos* left a comment (internetarchive/wayback-diff#195) > <#195 (comment)> > > The current HTML diff > https://web.archive.org/web/diff/20220901000000/20230215235959/https://www.iskme.org > works using https://github.com/edgi-govdata-archiving/web-monitoring-diff > . > > We must create another PDF diff project that has the same JSON output as > web-monitoring-diff. > Then, with a small WBM UI change, we call the PDF diff when the mimetype > of the captures we compare is PDF. > > You don't need to do any UI work. We only need a backend. The UI is > already working fine as long as your program has the same output as > web-monitoring-diff. > Your code should be independent from the web service. It should be like a > class / package that is imported by the web service. > In the future we may integrate it with > https://github.com/edgi-govdata-archiving/web-monitoring-diff or other > apps. > It should be easy to do that. > Another requirement: your project should be a package that can be > installed via pip install . > These 2 example projects showcase how the package should be like: > https://github.com/internetarchive/warcprox > https://github.com/internetarchive/brozzler > > The target Python version is 3.10+. > > — > Reply to this email directly, view it on GitHub > <#195 (comment)>, > or unsubscribe > <https://github.com/notifications/unsubscribe-auth/BE5R5RPZGKNBQ2YOO7EF4UT4QJYQBAVCNFSM6AAAAACWNZGWU2VHI2DSMVQWIX3LMV43OSLTON2WKQ3PNVWWK3TUHM2DANBUHA4TKNRYGE> > . > You are receiving this because you authored the thread.Message ID: > ***@***.***> >

[aam-naa]

Hi, I'm also applying to GSOC 2026 for the Wayback PDF Changes project, and after reading this thread I've come to the understanding that the requirement is a pip-installable Python package that is purely backend. I've built a small prototype that does text-based PDF diffing already using PyMuPDF and difflib.

In regards to the actual requirements of the project, I was curious to know if textual diffs are the only things that need to be detected or if diffs for visual elements like tables or graphs were also required. Thank you. @vbanos

[vbanos]

Thank you for your interest, everything must be supported, not just text.

[welcomemeet]

check my repo and give me suggestion for further process

https://github.com/welcomemeet/wayback-pdf-diff-prototype