Integrate Dangerzone in Rust app?

[NielDuysters]

Hi

I am working on a hobby-project Arti Chat which is a decentralized peer-to-peer chat messenger routed over Tor.

Recently I have implemented the ability to send images. Now I also want to support sharing PDF-files. My app is focused on security and I like Dangerzone to provide the safety for PDF-files.

But I am not sure about the best way to integrate Dangerzone in my app. My current plan is the following:

• The feature for PDF-files is only enabled if the user has Dangerzone installed.
• When receiving a PDF-file the bytes are written to a temporary file.
• My app automatically uses the Dangerzone CLI to convert the PDF file in a safe format.
• Path to the safe file is displayed in chat.

I guess this would work, but it is sub-optimal. Is there a better way to integrate Dangerzone in my app, by using it more like a kind of library instead of a separate process?

[almet]

Hi, and thanks for opening this. The timing is great: we have been talking about having a way to make dangerzone usable as a library, both for our usage and for external usage like the one you describe. It should happen in 2026 Q1.

The fact you mention rust is also of use: I'm not sure if and how we will make that possible, as for now most of the tooling is in python, but we have some proof of concept in rust that might prove useful.

All in all, the current status is that it's currently not possible to use dangerzone as a library, and the best way to accomplish this right now is to use the dangerzone-cli CLI tool.

Hope that helps!

[NielDuysters]

Thanks for your reply @almet , those plans sounds exciting. Is the PoC in Rust publicly available in a repo yet? Being familiar with Python, fluent in Rust and extremely interested in privacy / internet freedom I'd be happy to become a consistent contributor to that project.

If helpful, I could share a few thoughts on making Dangerzone accessible in Rust. As I see it, there are several options:

• A full rewrite in Rust as a separate library, decoupled from the Python codebase and maintained in parallel (powerful long-term, but resource-intensive and harder to keep behavior in sync).
• Keeping the core logic in Python, but extending it so it can act as an RPC server. Language-specific libraries (starting with Rust) would then communicate with it over IPC, preserving process isolation while avoiding tight language bindings.
• Use PyO3 (good for Rust, but RPC would also allow a clients in other languages). I have briefly used PyO3 in the past.