Add C# language bindings for payjoin-ffi by chavic · Pull Request #1318 · payjoin/rust-payjoin

chavic · 2026-02-09T08:43:55Z

This work in progress adds C# bindings and implements tests following the pattern in the other languages

coveralls · 2026-02-09T08:50:04Z

Pull Request Test Coverage Report for Build 22152746445

Details

0 of 0 changed or added relevant lines in 0 files are covered.
No unchanged relevant lines lost coverage.
Overall coverage remained the same at 83.169%

Totals
Change from base Build 22148988412:	0.0%
Covered Lines:	10219
Relevant Lines:	12287

💛 - Coveralls

DanGould · 2026-02-09T13:58:32Z

BOOYEAH

I like the direction this is going.

DanGould · 2026-02-10T06:40:56Z

payjoin-ffi/csharp/Payjoin.Tests.csproj

+<Project Sdk="Microsoft.NET.Sdk">
+
+  <PropertyGroup>
+    <TargetFramework>net8.0</TargetFramework>


Is net8.0 the most common target? I'm very curious what the actual needs of a BTCpayServer plugin or Strike are, because I've dealt with incompatibility issue here, albeit years ago.

it can: https://blog.btcpayserver.org/btcpay-server-1-12-0/#upgrade-to-net-8. Strike is likely ahead because they don't need to support a bunch of open source stuff.

I see, any idea which specific runtime they're on?

I did a quick check beforehand to target the most common runtime right now...

Gonna look loser into the Strike repos specifically

strike very likely on net8.0. probably private repos. i'll ask their team

Confirmed net8.0 for them and BTCpayServer, and both likely upgrade to net10.0 very soon.

ValeraFinebits · 2026-02-11T21:49:38Z

I've been working on an experimental C# bindings prototype in my fork (ValeraFinebits#1) while building a BTCPayServer Payjoin plugin, mainly to explore integration possibilities.

This PR aligns closely with what I implemented there, and in my local testing everything works as expected. I'll migrate my plugin to these upstream bindings and share any issues or improvement ideas as I go.

DanGould · 2026-02-13T03:15:43Z

Hi @ValeraFinebits I saw your PR open and close. Thanks bigly for writing us updates here. Very eager to see what that plugin can do and challenges hooking it up to BTCPay via the plugin framework.

DanGould

What's the plan with the dependency on your personal fork? i'm OK with that with giant, blaring ⚠️, even better would be in the payjoin project, but ultimately we want to upstream whatever fix was necessary. The process of making the PR also serves as documenting what changed. I don't know what changed at all in this PR other than the fact that it's forked. On this note, I'd much rather pin a commit hash than a branch so we have exact dependency that can't drift.
README references chavic/external-types-support but the script installs chavic/csharp-handle-fix. One of them is stale. Prefer commit & to doc in tested code > README in part for this reason
Missing async persister tests (compared to other bindings) Python and Dart both test JsonReceiverSessionPersisterAsync / JsonSenderSessionPersisterAsync. C# only tests the sync path. Is async support not generated yet,
or just untested?
No Windows CI
C# is disproportionately a Windows ecosystem language. BTCPayServer runs on Windows. The matrix only has ubuntu-latest and macos-latest. Is this a known
limitation or an oversight? I do see this changing, even NicolasDorier BTCPay Emperor has been tweeting about giving up windows, but C# specifically might deserve a windows flow.
Why brew install llvm on macOS?
The other bindings (Python, Dart) don't need this. Is it required for uniffi-bindgen-cs compilation, or the native lib? Should be documented.

payjoin-ffi/csharp/scripts/generate_bindings.sh

DanGould · 2026-02-16T13:49:14Z

payjoin-ffi/csharp/Payjoin.Tests.csproj

+    <TargetFramework>net8.0</TargetFramework>
+    <ImplicitUsings>enable</ImplicitUsings>
+    <Nullable>enable</Nullable>
+    <AllowUnsafeBlocks>true</AllowUnsafeBlocks>


Do we need unsafe? known lifetime gotchas? Ideally we'd NOT do this.

Yes, AllowUnsafeBlocks is currently required for the generated interop code (pointer usage in generated C#). I added an inline comment in the project file documenting this and noting that disabling /unsafe fails with CS0227. I'll revisit this to see if there's a way to avoid it

DanGould

We probably need nuspec too. We've got the Payjoin handle https://www.nuget.org/packages/Payjoin/

chavic · 2026-02-16T14:31:28Z

What's the plan with the dependency on your personal fork? i'm OK with that with giant, blaring ⚠️, even better would be in the payjoin project, but ultimately we want to upstream whatever fix was necessary. The process of making the PR also serves as documenting what changed. I don't know what changed at all in this PR other than the fact that it's forked. On this note, I'd much rather pin a commit hash than a branch so we have exact dependency that can't drift.

README references chavic/external-types-support but the script installs chavic/csharp-handle-fix. One of them is stale. Prefer commit & to doc in tested code > README in part for this reason

Missing async persister tests (compared to other bindings) Python and Dart both test JsonReceiverSessionPersisterAsync / JsonSenderSessionPersisterAsync. C# only tests the sync path. Is async support not generated yet,
or just untested?

No Windows CI
C# is disproportionately a Windows ecosystem language. BTCPayServer runs on Windows. The matrix only has ubuntu-latest and macos-latest. Is this a known
limitation or an oversight? I do see this changing, even NicolasDorier BTCPay Emperor has been tweeting about giving up windows, but C# specifically might deserve a windows flow.

Why brew install llvm on macOS?
The other bindings (Python, Dart) don't need this. Is it required for uniffi-bindgen-cs compilation, or the native lib? Should be documented.

Going to make an upstream PR after this works. Could do a draft for documentation.

Addressing the rest in upcoming commits.

chavic · 2026-02-18T18:51:44Z

Addressed most of the issues mentioned by @DanGould I think this PR it's ready for a second look. @spacebear21 @ValeraFinebits

ValeraFinebits · 2026-02-18T20:31:05Z

Addressed most of the issues mentioned by @DanGould I think this PR it's ready for a second look. @spacebear21 @ValeraFinebits

Windows now works fine. 👍

DanGould · 2026-02-20T07:19:22Z

Stoked this is working on Windows now.

On very first blush I see your new commits are telling me what happened but not why. The rationale is supercritical. Don't forget! Why > What!

DanGould · 2026-02-20T07:34:17Z

The changes you made to uniffi-bindgen-cs look sane, so I'd like to see them as open PRs there right away so we can get feedback ASAP. They can be one or multiple PRs but I think the feedback there is the most valuable next step we could have.

DanGould

I think this is pretty much ready to land as an expreiment but I would want @spacebear21's review and I really want to see open PRs upstream so we know we're on the right path.

It would be really nice to have a csharp devshell. I think we could also remove dart from the top level flake and make a devshell so that it doesn't pollute the top level with 300mb of stuff not everyone needs as well which could be done at the same time as as a follow up

  csharpDevShell = pkgs.mkShell {                                                                                                                                    
    name = "csharp-dev";                                                                                                                                             
    buildInputs = with pkgs; [                                                                                                                                       
      dotnet-sdk_8                                                                                                                                                   
      rustToolchains.msrv                                                                                                                                            
    ];                                                                                                                                                               
    shellHook = ''                                                                                                                                                   
      cd payjoin-ffi/csharp                                                                                                                                          
    '';                                                                                                                                                              
  };                                                                                                                                                                 
                                                                                                                                                                     
  # then in devShells:                                                                                                                                               
  devShells = devShells // {                                                                                                                                         
    default = devShells.nightly;                                                                                                                                     
    csharp = csharpDevShell;                                                                                                                                         
    # python = pythonDevShell;  # from PR #864                                                                                                                       
  };

DanGould · 2026-02-20T07:36:10Z

ohttp-relay/src/lib.rs

Ok so I see now that without this commit, cargo build fails on Windows. Since the C# CI matrix includes windows-latest, this commit unblocks compilation.

This took me about 20 minutes of review to confirm and could have been alleviated by a simple commit message. Next time I will have to refuse to review commits without rationale as part of their message.

DanGould · 2026-02-20T07:38:57Z

payjoin-ffi/csharp/README.md

+# Run all tests
+dotnet test


Would be nice to mirror the same pattern as other repos with a contrib/test.sh

#!/usr/bin/env bash set -euo pipefail SCRIPT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")" && pwd)" cd "$SCRIPT_DIR/.." bash ./scripts/generate_bindings.sh dotnet test --logger "console;verbosity=minimal"

chavic · 2026-02-20T09:11:13Z

Stoked this is working on Windows now.

On very first blush I see your new commits are telling me what happened but not why. The rationale is supercritical. Don't forget! Why > What!

Will do

spacebear21

utACK b859525 (didn't test locally but CI looks good on all platforms)

+1 on writing explanatory commit descriptions for why each change is made. Other than that the commits organization was nice and easy to follow.

Some thoughts on follow-ups:

Add integration tests for the end-to-end payjoin flow like we have in the other downstream languages.
Although I started the trend for dart to include 3rd party bindings generation in payjoin-ffi by hijacking uniffi-bindgen.rs, I think it will be more manageable as we add more languages to remove those optional dependencies/features from payjoin-ffi, and instead silo each language's generator inside that language's directory e.g. by using a cli generator like we do for javascript currently. Notably, adding these optional dependencies still inflates the main Cargo.lock files which can lead to dependency conflicts, MSRV issues, etc.

ValeraFinebits · 2026-02-21T12:41:05Z

utACK b859525 (didn't test locally but CI looks good on all platforms)

+1 on writing explanatory commit descriptions for why each change is made. Other than that the commits organization was nice and easy to follow.

Some thoughts on follow-ups:

Add integration tests for the end-to-end payjoin flow like we have in the other downstream languages.

Although I started the trend for dart to include 3rd party bindings generation in payjoin-ffi by hijacking uniffi-bindgen.rs, I think it will be more manageable as we add more languages to remove those optional dependencies/features from payjoin-ffi, and instead silo each language's generator inside that language's directory e.g. by using a cli generator like we do for javascript currently. Notably, adding these optional dependencies still inflates the main Cargo.lock files which can lead to dependency conflicts, MSRV issues, etc.

Integration tests have been added here:
#1350

chavic force-pushed the chavic/c-sharp-bindings branch 2 times, most recently from b4799da to ae74db7 Compare February 9, 2026 12:14

DanGould reviewed Feb 10, 2026

View reviewed changes

chavic force-pushed the chavic/c-sharp-bindings branch from ae74db7 to 4e75d20 Compare February 12, 2026 15:34

chavic force-pushed the chavic/c-sharp-bindings branch 2 times, most recently from ef26b10 to 769f0a9 Compare February 15, 2026 13:22

chavic requested a review from spacebear21 February 15, 2026 13:41

chavic self-assigned this Feb 15, 2026

chavic requested a review from DanGould February 15, 2026 13:41

chavic marked this pull request as ready for review February 15, 2026 13:43

DanGould reviewed Feb 16, 2026

View reviewed changes

chavic force-pushed the chavic/c-sharp-bindings branch 2 times, most recently from 6646f19 to bf7ea11 Compare February 17, 2026 20:14

chavic added 5 commits February 18, 2026 18:16

Gate unix socket support by platform in ohttp-relay

7af8451

Pin C# generator revision and use cargo-managed uniffi-bindgen

b891d5d

Add C# bindings scaffold, generation scripts, and tests

3b01da6

Add C# GitHub Actions workflow for Linux, macOS, and Windows

a2d30b4

Add draft C# nuspec

b859525

chavic force-pushed the chavic/c-sharp-bindings branch 2 times, most recently from eb9f897 to b859525 Compare February 18, 2026 18:38

DanGould approved these changes Feb 20, 2026

View reviewed changes

spacebear21 approved these changes Feb 20, 2026

View reviewed changes

spacebear21 merged commit 1a7effa into payjoin:master Feb 20, 2026
23 checks passed

chavic mentioned this pull request Feb 23, 2026

Standup Input: Week of 2026-02-23 #1352

Closed

Conversation

chavic commented Feb 9, 2026

Uh oh!

coveralls commented Feb 9, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Pull Request Test Coverage Report for Build 22152746445

Details

💛 - Coveralls

Uh oh!

DanGould commented Feb 9, 2026

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

ValeraFinebits commented Feb 11, 2026

Uh oh!

DanGould commented Feb 13, 2026

Uh oh!

DanGould left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

DanGould left a comment

Choose a reason for hiding this comment

Uh oh!

chavic commented Feb 16, 2026

Uh oh!

chavic commented Feb 18, 2026

Uh oh!

ValeraFinebits commented Feb 18, 2026

Uh oh!

DanGould commented Feb 20, 2026

Uh oh!

DanGould commented Feb 20, 2026

Uh oh!

DanGould left a comment

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

chavic commented Feb 20, 2026

Uh oh!

spacebear21 left a comment

Choose a reason for hiding this comment

Uh oh!

Uh oh!

ValeraFinebits commented Feb 21, 2026

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

5 participants

coveralls commented Feb 9, 2026 •

edited

Loading