Save benchmarks with pixi task

Author: eholum

.github/workflows/benchmarks.yml

@@ -55,11 +55,10 @@ jobs:
           manifest-path: pixi.toml
           # TODO: Remove this when we do not have self-hosted runners in the public repo
           pixi-bin-path: ~/actions-runner/.pixi/bin/pixi
-      - name: Build and test
+      - name: Build and benchmark
         run: |
           pixi run install_all
-      - name: Run python benchmarks
-        run: pytest benchmarks/ --benchmark-json pytest_benchmarks.json
+          pixi run test_benchmarks
       - name: Store benchmark results
         uses: benchmark-action/github-action-benchmark@v1
         with:

.gitignore

@@ -7,3 +7,6 @@
 # pixi environments
 .pixi
 *.egg-info
+
+# Benchmarking results
+pytest_benchmarks.json

pixi.toml

@@ -180,7 +180,7 @@ test_cpp = { depends-on = [
   { task = "test", args = ["roboplan_toppra"] },
 ] }
 test_all = { depends-on = ["test_py", "test_cpp"] }
-test_benchmarks = { cmd = "pytest benchmarks/" }
+test_benchmarks = { cmd = "pytest benchmarks/ --benchmark-json pytest_benchmarks.json" }
 
 [dependencies]