prepare v0.6.1

.github/workflows/ci.yml

@@ -17,6 +17,7 @@ jobs:
           - "3.11"
           - "3.12"
           - "3.13"
+          - "3.14"
 
     steps:
     - name: Checkout the repository

.github/workflows/docs.yml

@@ -5,40 +5,40 @@ on:
     branches: [master]
 
 permissions:
-  contents: write
+  contents: read
+  pages: write
+  id-token: write
 
 jobs:
   docs:
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
     runs-on: ubuntu-latest
 
-    strategy:
-      matrix:
-        python-version:
-          - "3.13"
-
     steps:
-    - uses: actions/checkout@v4
-
-    - name: Install uv and set the python version
-      uses: astral-sh/setup-uv@v5
-      with:
-        python-version: ${{ matrix.python-version }}
-
-    - run: echo "cache_id=$(date --utc '+%V')" >> $GITHUB_ENV 
-
-    - uses: actions/cache@v4
-      env:
-        GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-      with:
-        key: mkdocs-material-${{ env.cache_id }}
-        path: .cache 
-        restore-keys: |
-          mkdocs-material-
-
-    - name: Install dependencies
-      run: uv sync --locked --only-group docs
-
-    - name: Run docs build
-      env:
-        GH_TOKEN: ${{ secrets.GITHUB_TOKEN }}
-      run: bash ./scripts/docs_build.sh
+      - name: Configure GitHub Pages
+        uses: actions/configure-pages@v6
+
+      - name: Checkout the repository
+        uses: actions/checkout@v7
+
+      - name: Install uv and set Python
+        uses: astral-sh/setup-uv@v5
+        with:
+          python-version: "3.14"
+
+      - name: Install documentation dependencies
+        run: uv sync --locked --no-default-groups --group docs
+
+      - name: Build documentation
+        run: bash ./scripts/docs_build.sh
+
+      - name: Upload GitHub Pages artifact
+        uses: actions/upload-pages-artifact@v5
+        with:
+          path: site
+
+      - name: Deploy GitHub Pages
+        id: deployment
+        uses: actions/deploy-pages@v5

.github/workflows/publish.yml

@@ -20,7 +20,7 @@ jobs:
       - name: Set up Python
         uses: actions/setup-python@v6.1.0
         with:
-          python-version: '3.13'
+          python-version: '3.14'
 
       - name: Install build tools
         run: |

AGENTS.md

@@ -0,0 +1,37 @@
+# AGENTS.md
+
+## Project shape
+
+- This is a single Python package, not a monorepo. Runtime code is under `src/dataguard`; tests are under `tests`; documentation is under `docs`; documentation examples are sourced from `notebooks`.
+- The distribution is named `peh-dataguard`, but the import name is `dataguard`. The public exports are `Validator`, `Filter`, and `ErrorCollector` from `src/dataguard/__init__.py`.
+- Python `>=3.10` is supported; CI tests Python `3.10` through `3.14`. Ruff targets Python 3.10, so do not use syntax or APIs newer than the minimum unless the project metadata is changed.
+- The main flow is configuration mapping -> recursive Pydantic expression/schema models -> Polars expression compilation -> Pandera validation for `Validator` or direct Polars filtering for `Filter`. The relevant wiring is in `config/config_reader.py`, `core/check/`, `core/filter/`, and `core/models/`.
+- `CaseCheckExpression` is recursive: `conjunction` and `disjunction` require at least two expressions, while `condition` requires exactly two. Built-in check names are mapped to Polars methods; custom callable checks are supported for validation, not filtering.
+- `Validator` normally collects Polars/Pandera/configuration errors in the shared `ErrorCollector`; pass `collect_exceptions=False` when an exception should propagate. `Filter.apply` preserves `DataFrame` versus `LazyFrame` input type.
+
+## Toolchain and commands
+
+- Use `uv` and commit `uv.lock` changes generated by `uv lock`; do not edit the lockfile by hand. The direct dependency bounds intentionally block next-minor releases while allowing patch updates.
+- Reproduce the CI environment with `uv sync --locked --all-extras --dev`. The docs environment is separate: `uv sync --locked --no-default-groups --group docs`.
+- Run the full test and coverage suite with `make test` (it runs `uv run pytest -s -vv --cov=src/` and then `coverage html`). It modifies the tracked `.coverage` file; restore it before committing unless the coverage data is an intentional change.
+- Run a focused test with `make test-set SET='tests/tests_filter/test_filter.py::test_filter_apply_dataframe'`, or use `uv run pytest` with the desired node id. Tests are local/unit-style and use Hypothesis; no external service is configured.
+- Run lint with `make lint`. It runs Ruff checks twice; it does not check formatting. For a non-mutating formatting check use `uv run ruff format --check .`. `make format` runs Ruff autofixes and formatting, so it changes files.
+- `bash scripts/pre_commit.sh` is the repository wrapper for `format -> lint -> test`; it is mutating because it runs the formatter first. There is no configured mypy/pyright type-check step.
+- Build the sdist and wheel with `uv build`. Keep `dist/`, `site/`, `htmlcov/`, caches, and virtual environments out of commits.
+
+## State and testing gotchas
+
+- `ErrorCollector` is cached/shared; clear it between tests that inspect collected reports. The validator test fixture demonstrates this cleanup.
+- `parse_checks()` destructively pops `name`, `error_level`, and `error_msg` from each check mapping. Do not reuse an input mapping after parsing if its original contents matter.
+- `expression_mapper` is global and `map_command()` mutates parsed expression objects. Filter tests reset this mapping; preserve that isolation when adding tests or changing command registration.
+- `CaseCheckExpression` and `SimpleCheckExpression` are the backend-neutral expression layer. Keep built-in lowering shared between validation and filtering rather than duplicating command logic.
+
+## Documentation and release operations
+
+- Zensical, not the `mkdocs` CLI, builds the site from the MkDocs-compatible `mkdocs.yml`. Use `make serve` for a local preview and `bash scripts/docs_build.sh` for the strict clean build; the script uses `uv run --no-default-groups --group docs` and builds `site/`.
+- The Markdown guides embed code with `--8<--` snippets from `notebooks/`. Edit the notebook source and rebuild rather than copying code into the generated guide. `docs/changelog.md` similarly includes the root `CHANGELOG.md`.
+- API documentation is generated from source docstrings through `mkdocstrings` (`paths: [../src]`), so update the relevant source docstrings when changing public APIs.
+- The `DOCS` workflow runs only for pushes to `master` and deploys the built `site/` directory through GitHub Pages Actions. GitHub Pages must be configured to use GitHub Actions as its source.
+- The publish workflow builds on branch pushes, publishes to TestPyPI on tag pushes, and publishes to PyPI when a GitHub release is published. A normal branch push alone does not publish the package.
+- For a release, update `project.version` in `pyproject.toml`, regenerate `uv.lock`, run `uv build`, and tag the matching commit; do not create a release with mismatched project and lock metadata.
+- `presentation.md` is a standalone 16:9 Marp deck; no Makefile or CI target renders it, so validate it separately if it is changed.

CHANGELOG.md

@@ -0,0 +1,35 @@
+# Changelog
+
+All notable changes to DataGuard are documented in this file. The format is
+based on [Keep a Changelog](https://keepachangelog.com/en/1.1.0/), and this
+project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+## [0.6.1] - 2026-09-25
+
+This maintenance release updates packaging, supported Python versions,
+dependencies, CI, and documentation without changing the public runtime API.
+
+### Added
+
+- Added end-user installation instructions for `uv` and `pip` to the
+  documentation and README.
+- Added this changelog and exposed it in the documentation navigation.
+
+### Changed
+
+- Bumped the package version to `0.6.1`.
+- Added Python 3.14 to the CI, documentation, and publishing workflows and
+  updated the contributor setup.
+- Migrated the documentation from MkDocs Material to Zensical, including the
+  GitHub Pages deployment workflow.
+- Declared direct Polars and Pydantic runtime dependencies and added
+  next-minor upper bounds for direct runtime, development, build, and
+  documentation dependencies.
+- Regenerated `uv.lock` with the validated dependency graph.
+- Updated the README with installation, development, documentation, and
+  release-history links.
+
+[Unreleased]: https://github.com/eu-parc/dataguard/compare/v0.6.1...HEAD
+[0.6.1]: https://github.com/eu-parc/dataguard/compare/v0.6.0...v0.6.1

README.md

@@ -1,16 +1,19 @@
 # DataGuard
 
 <p align="center">
-   <a href="https://github.com/eu-parc/dataguard/actions?query=workflow%3ACI" targe>
+   <a href="https://github.com/eu-parc/dataguard/actions?query=workflow%3ACI" target="_blank" rel="noopener">
     <img src="https://github.com/eu-parc/dataguard/actions/workflows/ci.yml/badge.svg" alt="CI">
    </a>
-   <a href="https://github.com/eu-parc/dataguard/actions?query=workflow%3ADOCS" targe>
-    <img src="https://github.com/eu-parc/dataguard/actions/workflows/docs.yml/badge.svg" alt="CI">
+   <a href="https://github.com/eu-parc/dataguard/actions?query=workflow%3ADOCS" target="_blank" rel="noopener">
+    <img src="https://github.com/eu-parc/dataguard/actions/workflows/docs.yml/badge.svg" alt="Documentation build">
    </a>
-<p/>
+</p>
 
-Source: https://github.com/eu-parc/dataguard  
-Documentation: https://eu-parc.github.io/dataguard/
+- Source: [GitHub repository](https://github.com/eu-parc/dataguard)
+- Documentation: [DataGuard documentation](https://eu-parc.github.io/dataguard/)
+- Changelog: [CHANGELOG.md](https://github.com/eu-parc/dataguard/blob/master/CHANGELOG.md)
+
+DataGuard supports Python 3.10 and later.
 
 ## Features
 
@@ -24,29 +27,67 @@ Documentation: https://eu-parc.github.io/dataguard/
 - **Detailed error reporting**: Collect and format validation errors with custom messages and severity levels
 - **Polars support**: Built on Polars for efficient data processing at scale
 
-## Installation  
-To install the library for development.
+## Installation
+
+The package is published on PyPI as `peh-dataguard` and imported as
+`dataguard`.
+
+### Install with uv
+
+For a `uv`-managed project, add DataGuard as a dependency:
+
+```bash
+uv add peh-dataguard
+```
+
+To install it into the current environment without changing a project file:
+
+```bash
+uv pip install peh-dataguard
+```
+
+### Install with pip
+
+Create and activate a virtual environment, then install the package:
+
+```bash
+python -m venv .venv
+source .venv/bin/activate
+python -m pip install --upgrade pip
+python -m pip install peh-dataguard
+```
+
+On Windows PowerShell, activate the environment with
+`.venv\Scripts\Activate.ps1` instead of `source`.
+
+Verify that the public API can be imported:
+
+```bash
+python -c "from dataguard import Filter, Validator; print('DataGuard is ready')"
+```
+
+For a `uv`-managed project, use `uv run python ...` instead of `python ...`.
 
-### Synchronize the project dependencies
-
-   ```bash
-   $ uv sync
-   ```
+## Development
 
-#### Tests
+For a source checkout, install the development dependencies with Python 3.14:
 
-   ```bash
-   $ make test
-   ```
+```bash
+git clone https://github.com/eu-parc/dataguard.git
+cd dataguard
+uv sync --python 3.14
+```
 
-#### Format
+Run the project checks with:
 
-   ```bash
-   $ make format
-   ```
+```bash
+make test
+make lint
+make format
+```
 
-#### lint
+To preview the Zensical documentation locally:
 
-   ```bash
-   $ make lint
-   ```
+```bash
+make serve
+```

docs/changelog.md

@@ -0,0 +1,3 @@
+# Changelog
+
+--8<-- "CHANGELOG.md:2"

docs/getting_started.md

@@ -1,6 +1,47 @@
 # Getting Started
 
-## Dataguard workflow
+## Installation
+
+The PyPI distribution is named `peh-dataguard`, while the Python import
+name is `dataguard`.
+
+### Install with `uv`
+
+For a `uv`-managed project, add DataGuard as a dependency:
+
+```bash
+uv add peh-dataguard
+```
+
+To install it into the current environment without changing a project file:
+
+```bash
+uv pip install peh-dataguard
+```
+
+### Install with `pip`
+
+Create and activate a virtual environment, then install the package:
+
+```bash
+python -m venv .venv
+source .venv/bin/activate
+python -m pip install --upgrade pip
+python -m pip install peh-dataguard
+```
+
+On Windows PowerShell, activate the environment with
+`.venv\Scripts\Activate.ps1` instead of `source`.
+
+### Verify the installation
+
+```bash
+python -c "from dataguard import Filter, Validator; print('DataGuard is ready')"
+```
+
+For a `uv`-managed project, use `uv run python ...` instead of `python ...`.
+
+## DataGuard workflow
 
 ```py title="getting_started.py" linenums="1" hl_lines="3-9 13 15"
 --8<-- "notebooks/getting_started.py:1:15"
@@ -15,7 +56,7 @@ While this example uses empty lists and an empty DataFrame for simplicity, it il
 
 ## Validating real-world constraints
 
-Consider an `age` column configuration that demonstrates Dataguard's data quality enforcement capabilities:
+Consider an `age` column configuration that demonstrates DataGuard's data quality enforcement capabilities:
 
 - **Type Safety**: Enforcing `integer` data type prevents string or float contamination
 - **Null Prevention**: `nullable: False` ensures no missing age values slip through

docs/index.md

@@ -1,6 +1,6 @@
-# Welcome to Dataguard
+# Welcome to DataGuard
 
-Dataguard is a Python package for performing data validation and filtering on dataframe objects.
+DataGuard is a Python package for performing data validation and filtering on dataframe objects.
 
 It provides a flexible and extensible framework for defining validation rules and filter expressions, ensuring that your data meets quality and consistency requirements.
 
@@ -10,3 +10,5 @@ For filtering data, please see [How to filter](how-to-filter.md)
 
 For API documentation, please see the [API Reference](api.md).
 
+For release history, see the [Changelog](changelog.md).
+

mkdocs.yml

@@ -2,8 +2,11 @@ site_name: DataGuard
 repo_url: https://github.com/eu-parc/dataguard
 repo_name: eu-parc/dataguard
 edit_uri: edit/master/docs/
+# Zensical intentionally reads this MkDocs-compatible configuration.
+# The classic variant preserves the existing Material-style presentation.
 theme:
   name: material
+  variant: classic
 
 plugins:
   - search
@@ -39,6 +42,7 @@ markdown_extensions:
 nav:
   - Home: index.md
   - Getting Started: getting_started.md
+  - Changelog: changelog.md
   - User Guide:
       - How to create checks: how-to-create_checks.md
       - How to filter: how-to-filter.md