docs: setup mkdocs

MkDocs is a popular tool in the Python ecosystem to generate
documentation. It has support for parsing Python docstrings in order to
generate API documentation quite easily.

Signed-off-by: JP-Ellis <josh@jpellis.me>

.github/workflows/docs.yml

@@ -0,0 +1,55 @@
+name: Generate and Publish Docs
+
+on:
+  push:
+
+env:
+  STABLE_PYTHON_VERSION: "3.11"
+  PYTEST_ADDOPTS: --color=yes
+
+jobs:
+  build:
+    name: Build docs
+
+    runs-on: ubuntu-latest
+
+    steps:
+      - name: Checkout code
+        uses: actions/checkout@v4
+        with:
+          fetch-depth: 0
+
+      - name: Set up Python
+        uses: actions/setup-python@v4
+        with:
+          python-version: ${{ env.STABLE_PYTHON_VERSION }}
+
+      - name: Install Hatch
+        run: pip install --upgrade hatch
+
+      - name: Build docs
+        run: |
+          hatch run mkdocs build
+
+      - name: Upload artifact
+        uses: actions/upload-pages-artifact@v2
+        with:
+          path: site
+
+  publish:
+    name: Publish docs
+
+    needs: build
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      pages: write
+      id-token: write
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
+
+    steps:
+      - name: Deploy to GitHub Pages
+        id: deployment
+        uses: actions/deploy-pages@v2

.pre-commit-config.yaml

@@ -23,6 +23,10 @@ repos:
       - id: check-toml
       - id: check-xml
       - id: check-yaml
+        exclude: |
+          (?x)^(
+            mkdocs.yml
+          )$
 
   - repo: https://gitlab.com/bmares/check-json5
     rev: v1.0.0

docs/SUMMARY.md

@@ -0,0 +1,5 @@
+-   [Home](README.md)
+    -   [Changelog](CHANGELOG.md)
+    -   [Contributing](CONTRIBUTING.md)
+-   [Pact](pact/)
+-   [Examples](examples/)

docs/scripts/markdown.py

@@ -0,0 +1,62 @@
+"""
+Script to merge Markdown documentation from the main codebase into the docs.
+
+This script is run by mkdocs-gen-files when the documentation is built and
+imports Markdown documentation from the main codebase so that it can be included
+in the documentation site. For example, a Markdown file located at
+`some/path/foo.md` will be treated as if it was located at
+`docs/some/path/foo.md` without the need for symlinks or copying the file.
+
+If the destination file already exists (either because it is a real file, or was
+otherwise already generated), the script will raise a RuntimeError.
+"""
+
+import subprocess
+import sys
+from pathlib import Path
+
+import mkdocs_gen_files
+from mkdocs_gen_files.editor import FilesEditor
+
+EDITOR = FilesEditor.current()
+
+# These paths are relative to the project root, *not* the current file.
+SRC_ROOT = "."
+DOCS_DEST = "."
+
+# List of all files version controlled files in the SRC_ROOT
+ALL_FILES = sorted(
+    map(
+        Path,
+        subprocess.check_output(["git", "ls-files", SRC_ROOT])  # noqa: S603, S607
+        .decode("utf-8")
+        .splitlines(),
+    ),
+)
+
+
+for source_path in filter(lambda p: p.suffix == ".md", ALL_FILES):
+    if source_path.parts[0] == "docs":
+        continue
+    dest_path = Path(DOCS_DEST, source_path)
+
+    if str(dest_path) in EDITOR.files:
+        print(  # noqa: T201
+            f"Unable to copy {source_path} to {dest_path} because the file already"
+            " exists at the destination.",
+            file=sys.stderr,
+        )
+        msg = f"File {dest_path} already exists."
+        raise RuntimeError(msg)
+
+    with Path(source_path).open("r", encoding="utf-8") as fi, mkdocs_gen_files.open(
+        dest_path,
+        "w",
+        encoding="utf-8",
+    ) as fd:
+        fd.write(fi.read())
+
+    mkdocs_gen_files.set_edit_path(
+        dest_path,
+        f"https://github.com/pact-foundation/pact-python/edit/develop/{source_path}",
+    )

docs/scripts/other.py

@@ -0,0 +1,117 @@
+"""
+Create placeholder files for all other files in the codebase.
+
+This script is run by mkdocs-gen-files when the documentation is built and
+creates placeholder files for all other files in the codebase. This is done so
+that the documentation site can link to all files in the codebase, even if they
+aren't part of the documentation proper.
+
+If the files are binary, they are copied as-is (e.g. for images), otherwise a
+HTML redirect is created.
+
+If the destination file already exists (either because it is a real file, or was
+otherwise already generated), the script will ignore the current file and
+continue silently.
+"""
+
+import subprocess
+from pathlib import Path
+from typing import TYPE_CHECKING
+
+import mkdocs_gen_files
+from mkdocs_gen_files.editor import FilesEditor
+
+if TYPE_CHECKING:
+    import io
+
+EDITOR = FilesEditor.current()
+
+# These paths are relative to the project root, *not* the current file.
+SRC_ROOT = "."
+DOCS_DEST = "."
+
+# List of all files version controlled files in the SRC_ROOT
+ALL_FILES = sorted(
+    map(
+        Path,
+        subprocess.check_output(["git", "ls-files", SRC_ROOT])  # noqa: S603, S607
+        .decode("utf-8")
+        .splitlines(),
+    ),
+)
+
+
+def is_binary(buffer: bytes) -> bool:
+    """
+    Determine whether the given buffer is binary or not.
+
+    The check is done by attempting to decode the buffer as UTF-8. If this
+    succeeds, the buffer is not binary. If it fails, the buffer is binary.
+
+    The entire buffer will be checked, therefore if checking whether a file is
+    binary, only the start of the file should be passed.
+
+    Args:
+        buffer:
+            The buffer to check.
+
+    Returns:
+        True if the buffer is binary, False otherwise.
+    """
+    try:
+        buffer.decode("utf-8")
+    except UnicodeDecodeError:
+        return True
+    else:
+        return False
+
+
+for source_path in ALL_FILES:
+    if not source_path.is_file():
+        continue
+    if source_path.parts[0] in ["docs"]:
+        continue
+
+    dest_path = Path(DOCS_DEST, source_path)
+
+    if str(dest_path) in EDITOR.files:
+        continue
+
+    fi: "io.IOBase"
+    with Path(source_path).open("rb") as fi:
+        buf = fi.read(2048)
+
+    if is_binary(buf):
+        if source_path.stat().st_size < 16 * 2**20:
+            # Copy the file only if it's less than 16MB.
+            with Path(source_path).open("rb") as fi, mkdocs_gen_files.open(
+                dest_path,
+                "wb",
+            ) as fd:
+                fd.write(fi.read())
+        else:
+            # File is too big, create a redirect.
+            url = (
+                "https://github.com"
+                "/pact-foundation/pact-python"
+                "/raw"
+                "/develop"
+                f"/{source_path}"
+            )
+            with mkdocs_gen_files.open(dest_path, "w", encoding="utf-8") as fd:
+                fd.write(f'<meta http-equiv="refresh" content="0; url={url}">')
+                fd.write(f"# Redirecting to {url}...")
+                fd.write(f"[Click here if you are not redirected]({url})")
+
+            mkdocs_gen_files.set_edit_path(
+                dest_path,
+                f"https://github.com/pact-foundation/pact-python/edit/develop/{source_path}",
+            )
+
+    else:
+        with Path(source_path).open("r", encoding="utf-8") as fi, mkdocs_gen_files.open(
+            dest_path,
+            "w",
+            encoding="utf-8",
+        ) as fd:
+            fd.write(fi.read())

docs/scripts/python.py

@@ -0,0 +1,70 @@
+"""
+Script used by mkdocs-gen-files to generate documentation for Pact Python.
+
+The script is run by mkdocs-gen-files when the documentation is built in order
+to generate documentation from Python docstrings.
+"""
+import subprocess
+from pathlib import Path
+from typing import Union
+
+import mkdocs_gen_files
+
+
+def process_python(src: str, dest: Union[str, None] = None) -> None:
+    """
+    Process the Python files in the given directory.
+
+    The source directory is relative to the root of the repository, and only
+    Python files which are version controlled are processed. The generated
+    documentation may optionally written to a different directory.
+    """
+    dest = dest or src
+
+    # List of all files version controlled files in the SRC_ROOT
+    files = sorted(
+        map(
+            Path,
+            subprocess.check_output(["git", "ls-files", src])  # noqa: S603, S607
+            .decode("utf-8")
+            .splitlines(),
+        ),
+    )
+    files = filter(lambda p: p.suffix == ".py", files)
+    files = sorted(files)
+
+    for source_path in files:
+        module_path = source_path.relative_to(src).with_suffix("")
+        doc_path = source_path.relative_to(src).with_suffix(".md")
+        full_doc_path = Path(dest, doc_path)
+
+        parts = [src, *module_path.parts]
+
+        # Skip __main__ modules
+        if parts[-1] == "__main__":
+            continue
+
+        # The __init__ modules are implicit in the directory structure.
+        if parts[-1] == "__init__":
+            parts = parts[:-1]
+            full_doc_path = full_doc_path.parent / "README.md"
+
+        if full_doc_path.exists():
+            with mkdocs_gen_files.open(full_doc_path, "a", encoding="utf-8") as fd:
+                python_identifier = ".".join(parts)
+                print("# " + parts[-1], file=fd)
+                print("::: " + python_identifier, file=fd)
+        else:
+            with mkdocs_gen_files.open(full_doc_path, "w", encoding="utf-8") as fd:
+                python_identifier = ".".join(parts)
+                print("# " + parts[-1], file=fd)
+                print("::: " + python_identifier, file=fd)
+
+        mkdocs_gen_files.set_edit_path(
+            full_doc_path,
+            f"https://github.com/pact-foundation/pact-python/edit/master/pact/{module_path}.py",
+        )
+
+
+process_python("pact")
+process_python("examples")

docs/scripts/ruff.toml

@@ -0,0 +1,4 @@
+extend = "../../pyproject.toml"
+ignore = [
+  "INP001",  # Forbid implicit namespaces
+]