feat: Add simple Pixi.toml generation without conflict resolution

PIXI_PLAN.md

@@ -0,0 +1,111 @@
+# Pixi Integration - Simple Implementation Plan
+
+## Core Philosophy
+**Let UniDep translate, let Pixi resolve**
+
+UniDep should act as a simple translator from `requirements.yaml`/`pyproject.toml` to `pixi.toml` format.
+Pixi handles all dependency resolution, conflict management, and lock file generation.
+
+## Current Problem with `pixi` Branch
+- **Over-engineering**: Pre-resolves conflicts that Pixi can handle
+- **Origin tracking**: Complex system to track where each dependency came from
+- **Unnecessary complexity**: ~500+ lines of code for what should be ~100 lines
+
+## New Simple Architecture
+
+### Phase 1: Basic Pixi.toml Generation ✅
+- [x] Create minimal `_pixi.py` module (~100 lines)
+- [ ] Parse requirements WITHOUT resolution
+- [ ] Create features with literal dependencies
+- [ ] Generate pixi.toml with proper structure
+- [ ] Add `--pixi` flag to merge command
+
+### Phase 2: Pixi Lock Command
+- [ ] Add `pixi-lock` subcommand to CLI
+- [ ] Simple wrapper around `pixi lock` command
+- [ ] Support platform selection
+- [ ] Add basic tests
+
+### Phase 3: Monorepo Support (Optional)
+- [ ] Generate sub-lock files if needed
+- [ ] But let Pixi handle the complexity
+
+## Implementation Details
+
+### 1. Simple Pixi.toml Structure
+```python
+def generate_pixi_toml(requirements_files, output_file):
+    pixi_data = {
+        "project": {
+            "name": "myenv",
+            "channels": channels,
+            "platforms": platforms,
+        },
+        "dependencies": {},
+        "pypi-dependencies": {},
+    }
+
+    # For monorepo: create features
+    if len(requirements_files) > 1:
+        pixi_data["feature"] = {}
+        pixi_data["environments"] = {}
+
+        for req_file in requirements_files:
+            feature_name = req_file.parent.stem
+            deps = parse_single_file(req_file)  # NO RESOLUTION!
+
+            pixi_data["feature"][feature_name] = {
+                "dependencies": deps.conda,  # Literal copy
+                "pypi-dependencies": deps.pip,  # Literal copy
+            }
+
+        # Create environments
+        all_features = list(pixi_data["feature"].keys())
+        pixi_data["environments"]["default"] = all_features
+        for feat in all_features:
+            pixi_data["environments"][feat.replace("_", "-")] = [feat]
+```
+
+### 2. Key Simplifications
+- **NO conflict resolution** - Pixi handles this
+- **NO origin tracking** - Not needed
+- **NO version pinning combination** - Pixi does this
+- **NO platform resolution** - Use Pixi's native platform support
+
+### 3. Testing Strategy
+- Test pixi.toml generation (structure validation)
+- Test CLI integration
+- Test with example monorepo
+- Let Pixi handle the actual resolution testing
+
+## Files to Create/Modify
+
+### New Files
+1. `unidep/_pixi.py` - Simple pixi.toml generation (~100 lines)
+2. `tests/test_pixi.py` - Basic tests (~50 lines)
+
+### Modified Files
+1. `unidep/_cli.py` - Add `--pixi` flag and `pixi-lock` command
+2. `README.md` - Document new Pixi support
+
+## Success Criteria
+- [ ] Generate valid pixi.toml files
+- [ ] Pass all tests
+- [ ] Work with monorepo example
+- [ ] Total implementation < 200 lines (vs 500+ in old branch)
+
+## Timeline
+- **Hour 1-2**: Basic pixi.toml generation ✅
+- **Hour 3-4**: CLI integration and testing
+- **Hour 5-6**: Documentation and polish
+
+## Testing Checkpoints
+After each major change:
+1. Run tests: `pytest tests/test_pixi.py -xvs`
+2. Test with monorepo: `unidep merge --pixi tests/simple_monorepo`
+3. Validate pixi.toml: `pixi list`
+
+## Commit Strategy
+- Commit after each working phase
+- Clear commit messages
+- Test before each commit

README.md

@@ -550,7 +550,8 @@ positional arguments:
                         Subcommands
     merge               Combine multiple (or a single) `requirements.yaml` or
                         `pyproject.toml` files into a single Conda installable
-                        `environment.yaml` file.
+                        `environment.yaml` file or Pixi installable
+                        `pixi.toml` file.
     install             Automatically install all dependencies from one or
                         more `requirements.yaml` or `pyproject.toml` files.
                         This command first installs dependencies with Conda,
@@ -603,30 +604,33 @@ See `unidep merge -h` for more information:
 <!-- ⚠️ This content is auto-generated by `markdown-code-runner`. -->
 ```bash
 usage: unidep merge [-h] [-o OUTPUT] [-n NAME] [--stdout]
-                    [--selector {sel,comment}] [-d DIRECTORY] [--depth DEPTH]
-                    [-v]
+                    [--selector {sel,comment}] [--pixi] [-d DIRECTORY]
+                    [--depth DEPTH] [-v]
                     [-p {linux-64,linux-aarch64,linux-ppc64le,osx-64,osx-arm64,win-64}]
                     [--skip-dependency SKIP_DEPENDENCY]
                     [--ignore-pin IGNORE_PIN] [--overwrite-pin OVERWRITE_PIN]
 
 Combine multiple (or a single) `requirements.yaml` or `pyproject.toml` files
-into a single Conda installable `environment.yaml` file. Example usage:
-`unidep merge --directory . --depth 1 --output environment.yaml` to search for
-`requirements.yaml` or `pyproject.toml` files in the current directory and its
-subdirectories and create `environment.yaml`. These are the defaults, so you
-can also just run `unidep merge`.
+into a single Conda installable `environment.yaml` file or Pixi installable
+`pixi.toml` file. Example usage: `unidep merge --directory . --depth 1
+--output environment.yaml` to search for `requirements.yaml` or
+`pyproject.toml` files in the current directory and its subdirectories and
+create `environment.yaml`. These are the defaults, so you can also just run
+`unidep merge`.
 
 options:
   -h, --help            show this help message and exit
   -o, --output OUTPUT   Output file for the conda environment, by default
-                        `environment.yaml`
+                        `environment.yaml` or `pixi.toml` if `--pixi` is used
   -n, --name NAME       Name of the conda environment, by default `myenv`
   --stdout              Output to stdout instead of a file
   --selector {sel,comment}
                         The selector to use for the environment markers, if
                         `sel` then `- numpy # [linux]` becomes `sel(linux):
                         numpy`, if `comment` then it remains `- numpy #
                         [linux]`, by default `sel`
+  --pixi                Generate a `pixi.toml` file instead of
+                        `environment.yaml`
   -d, --directory DIRECTORY
                         Base directory to scan for `requirements.yaml` or
                         `pyproject.toml` file(s), by default `.`