340 check ci because gh is slow

.github/actions/project-setup/action.yml

@@ -10,6 +10,10 @@ inputs:
     description: "Additional Rust target to install"
     required: false
     default: ""
+  python_version:
+    description: "Python version to use"
+    required: false
+    default: "3.13"
 
 runs:
   using: "composite"
@@ -23,7 +27,7 @@ runs:
     # Python
     - uses: actions/setup-python@v5
       with:
-        python-version: "3.13"
+        python-version: ${{ inputs.python_version }}
 
     # ----- OS-Specific System Dependencies -----
 

.github/workflows/ci-python.yml

@@ -0,0 +1,43 @@
+name: ci
+
+on:
+  push:
+    branches: [main]
+    paths:
+      - encoderfile-py/**
+  pull_request:
+    paths:
+      - encoderfile-py/**
+  workflow_dispatch:
+
+jobs:
+  test_python:
+    runs-on: ubuntu-latest
+
+    steps:
+      - uses: actions/checkout@v4
+
+      - name: Project setup
+        uses: ./.github/actions/project-setup
+        with:
+          # use MSRV
+          python_version: "3.10"
+
+      - name: Install Maturin
+        run: |
+          cargo binstall maturin --force --disable-telemetry
+          ls -la /home/runner/.cargo/bin/
+
+      - name: Build rs (debug)
+        run: cargo build --locked
+
+      - name: Maturin develop
+        run: |
+          ls -la /home/runner/.cargo/bin/
+          just build-py
+
+      - name: Test
+        run: just test-py
+
+      - name: Stubtest
+        run: just stubtest

.gitignore

@@ -203,9 +203,9 @@ cython_debug/
 .abstra/
 
 # Visual Studio Code
-#   Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore 
+#   Visual Studio Code specific template is maintained in a separate VisualStudioCode.gitignore
 #   that can be found at https://github.com/github/gitignore/blob/main/Global/VisualStudioCode.gitignore
-#   and can be added to the global gitignore or merged into this file. However, if you prefer, 
+#   and can be added to the global gitignore or merged into this file. However, if you prefer,
 #   you could uncomment the following to ignore the entire vscode folder
 # .vscode/
 

Justfile

@@ -1,3 +1,4 @@
+export PATH := "/home/runner/.cargo/bin" + env_var("PATH")
 # Setup
 setup:
     @echo "creating .venv"
@@ -18,11 +19,11 @@ check:
     @cargo hack check --each-feature
 
 coverage:
-	@cargo llvm-cov \
-		--workspace \
-		--all-features \
-		--lcov \
-		--output-path lcov.info
+    @cargo llvm-cov \
+    	--workspace \
+    	--all-features \
+    	--lcov \
+    	--output-path lcov.info
 
 pre-commit:
     @uv run --dev pre-commit run --all-files
@@ -51,6 +52,12 @@ lint-py:
     # ruff check
     @uv run ruff check
 
+test-py:
+    @uv run \
+        --dev \
+        --directory encoderfile-py \
+        -m pytest
+
 stubtest:
     @uv run \
         --dev \
@@ -71,19 +78,20 @@ generate-docs:
         --all-features
 
 licenses:
-	@echo "Generating licenses..."
-	@cargo about generate about.hbs > THIRDPARTY.md
+    @echo "Generating licenses..."
+    @cargo about generate about.hbs > THIRDPARTY.md
 
 # Utilities
+
 target_max_mb := "2000"
 
 clean:
     #!/usr/bin/env bash
     if [ -d target ]; then
         target_size_mb=$(du -sm target | cut -f1)
         echo "target/ size: ${target_size_mb} MB"
-        if [ "${target_size_mb}" -gt "{{target_max_mb}}" ]; then
-            echo "target/ exceeds {{target_max_mb}} MB — running cargo clean..."
+        if [ "${target_size_mb}" -gt "{{ target_max_mb }}" ]; then
+            echo "target/ exceeds {{ target_max_mb }} MB — running cargo clean..."
             cargo clean
         else
             echo "target/ size within limits — skipping clean."

encoderfile-py/python/encoderfile/__main__.py

@@ -1,6 +1,7 @@
-from encoderfile import run_cli
 import sys
 
+from encoderfile import run_cli  # type: ignore
+
 if __name__ == "__main__":
     args = ["encoderfile"] + sys.argv[1:]
 

encoderfile-py/python/tests/assets/test_config.yml

@@ -1,13 +1,13 @@
 encoderfile:
   name: my-model-2
-  path: models/token_classification
+  path: ../models/token_classification
   model_type: token_classification
   output_path: /tmp/test-model.encoderfile
-  base_binary_path: ./target/debug/encoderfile-runtime
+  base_binary_path: ../target/debug/encoderfile-runtime
   transform: |
     --- Applies a softmax across token classification logits.
     --- Each token classification is normalized independently.
-    --- 
+    ---
     --- Args:
     ---   arr (Tensor): A tensor of shape [batch_size, n_tokens, n_labels].
     ---                 The softmax is applied along the third axis (n_labels).

encoderfile-py/python/tests/assets/test_config_lua.yml

@@ -1,16 +1,16 @@
 encoderfile:
   name: my-model-2-lua
-  path: models/token_classification
+  path: ../models/token_classification
   model_type: token_classification
   output_path: /tmp/test-model-lua.encoderfile
-  base_binary_path: ./target/debug/encoderfile-runtime
+  base_binary_path: ../target/debug/encoderfile-runtime
   lua_libs:
     - table
     - math
   transform: |
     --- Applies a softmax across token classification logits.
     --- Each token classification is normalized independently.
-    --- 
+    ---
     --- Args:
     ---   arr (Tensor): A tensor of shape [batch_size, n_tokens, n_labels].
     ---                 The softmax is applied along the third axis (n_labels).