feat: regex and dynamic-metadata rewrite (#457)

Based on scikit-build/dynamic-metadata#1 (with
some updates). Fixes
#435.

This doesn't start using `tool.dynamic-metadata` yet, to give us room to
change that. Plus we have to back-compat support
`tool.scikit-build.metadata`, so might as well keep that as the only way
to set this for now.

Signed-off-by: Henry Schreiner <henryschreineriii@gmail.com>

Author: henryiii

docs/api/scikit_build_core.metadata.rst

@@ -17,6 +17,14 @@ scikit\_build\_core.metadata.fancy\_pypi\_readme module
    :undoc-members:
    :show-inheritance:
 
+scikit\_build\_core.metadata.regex module
+-----------------------------------------
+
+.. automodule:: scikit_build_core.metadata.regex
+   :members:
+   :undoc-members:
+   :show-inheritance:
+
 scikit\_build\_core.metadata.setuptools\_scm module
 ---------------------------------------------------
 

docs/configuration.md

@@ -445,6 +445,31 @@ metadata.readme.provider = "scikit_build_core.metadata.fancy_pypi_readme"
 
 :::
 
+:::{tab} Regex
+
+If you want to pull a string-valued expression (usually version) from an
+existing file, you can the integrated `regex` plugin to pull the information.
+
+```toml
+name = "mypackage"
+dynamic = ["version"]
+
+[tool.scikit-build.metadata.version]
+provider = "scikit_build_core.metadata.regex"
+input = "src/mypackage/__init__.py"
+```
+
+You can set a custom regex with `regex=`; use `(?p<value>...)` to capture the
+value you want to use. By default when targeting version, you get a reasonable
+regex for python files,
+`'(?i)^(__version__|VERSION) *= *([\'"])v?(?P<value>.+?)\2'`.
+
+```{versionadded} 0.5
+
+```
+
+:::
+
 ## Editable installs
 
 Experimental support for editable installs is provided, with some caveats and

src/scikit_build_core/metadata/fancy_pypi_readme.py

@@ -4,21 +4,21 @@
 
 from .._compat import tomllib
 
-__all__ = ["dynamic_metadata"]
+__all__ = ["dynamic_metadata", "get_requires_for_dynamic_metadata"]
 
 
 def __dir__() -> list[str]:
     return __all__
 
 
 def dynamic_metadata(
-    fields: frozenset[str],
+    field: str,
     settings: dict[str, list[str] | str] | None = None,
-) -> dict[str, str | dict[str, str | None]]:
+) -> str | dict[str, str]:
     from hatch_fancy_pypi_readme._builder import build_text
     from hatch_fancy_pypi_readme._config import load_and_validate_config
 
-    if fields != {"readme"}:
+    if field != "readme":
         msg = "Only the 'readme' field is supported"
         raise ValueError(msg)
 
@@ -35,13 +35,11 @@ def dynamic_metadata(
 
     # Version 22.3 does not have fragment support
     return {
-        "readme": {
-            "content-type": config.content_type,
-            "text": build_text(config.fragments, config.substitutions)
-            if hasattr(config, "substitutions")
-            # pylint: disable-next=no-value-for-parameter
-            else build_text(config.fragments),  # type: ignore[call-arg]
-        }
+        "content-type": config.content_type,
+        "text": build_text(config.fragments, config.substitutions)
+        if hasattr(config, "substitutions")
+        # pylint: disable-next=no-value-for-parameter
+        else build_text(config.fragments),  # type: ignore[call-arg]
     }
 
 

src/scikit_build_core/metadata/regex.py

@@ -0,0 +1,48 @@
+from __future__ import annotations
+
+import re
+from collections.abc import Mapping
+from pathlib import Path
+from typing import Any
+
+__all__ = ["dynamic_metadata"]
+
+
+def __dir__() -> list[str]:
+    return __all__
+
+
+def dynamic_metadata(
+    field: str,
+    settings: Mapping[str, Any],
+) -> str:
+    # Input validation
+    if field not in {"version", "description", "requires-python"}:
+        msg = "Only string fields supported by this plugin"
+        raise RuntimeError(msg)
+    if settings.keys() > {"input", "regex"}:
+        msg = "Only 'input' and 'regex' settings allowed by this plugin"
+        raise RuntimeError(msg)
+    if "input" not in settings:
+        msg = "Must contain the 'input' setting to perform a regex on"
+        raise RuntimeError(msg)
+    if field != "version" and "regex" not in settings:
+        msg = "Must contain the 'regex' setting if not getting version"
+        raise RuntimeError(msg)
+    if not all(isinstance(x, str) for x in settings.values()):
+        msg = "Must set 'input' and/or 'regex' to strings"
+        raise RuntimeError(msg)
+
+    input_filename = settings["input"]
+    regex = settings.get(
+        "regex", r'(?i)^(__version__|VERSION) *= *([\'"])v?(?P<value>.+?)\2'
+    )
+
+    with Path(input_filename).open(encoding="utf-8") as f:
+        match = re.search(regex, f.read(), re.MULTILINE)
+
+    if not match:
+        msg = f"Couldn't find {regex!r} in {input_filename}"
+        raise RuntimeError(msg)
+
+    return match.group("value")

src/scikit_build_core/metadata/setuptools_scm.py

@@ -1,20 +1,20 @@
 from __future__ import annotations
 
-__all__ = ["dynamic_metadata"]
+__all__ = ["dynamic_metadata", "get_requires_for_dynamic_metadata"]
 
 
 def __dir__() -> list[str]:
     return __all__
 
 
 def dynamic_metadata(
-    fields: frozenset[str],
+    field: str,
     settings: dict[str, object] | None = None,
-) -> dict[str, str | dict[str, str | None]]:
+) -> str:
     # this is a classic implementation, waiting for the release of
     # vcs-versioning and an improved public interface
 
-    if fields != {"version"}:
+    if field != "version":
         msg = "Only the 'version' field is supported"
         raise ValueError(msg)
 
@@ -27,7 +27,7 @@ def dynamic_metadata(
     config = Configuration.from_file("pyproject.toml")
     version: str = _get_version(config)
 
-    return {"version": version}
+    return version
 
 
 def get_requires_for_dynamic_metadata(

src/scikit_build_core/settings/_load_provider.py

@@ -2,13 +2,13 @@
 
 import importlib
 import sys
-from collections.abc import Iterable
+from collections.abc import Generator, Iterable, Mapping
 from pathlib import Path
-from typing import Any
+from typing import Any, Union
 
 from .._compat.typing import Protocol
 
-__all__ = ["load_provider"]
+__all__ = ["load_provider", "load_dynamic_metadata"]
 
 
 def __dir__() -> list[str]:
@@ -27,10 +27,31 @@ def get_requires_for_dynamic_metadata(self, settings: dict[str, Any]) -> list[st
         ...
 
 
+class DynamicMetadataWheelProtocol(DynamicMetadataProtocol, Protocol):
+    def dynamic_wheel(
+        self, field: str, settings: Mapping[str, Any] | None = None
+    ) -> bool:
+        ...
+
+
+class DynamicMetadataRequirementsWheelProtocol(
+    DynamicMetadataRequirementsProtocol, DynamicMetadataWheelProtocol, Protocol
+):
+    ...
+
+
+DMProtocols = Union[
+    DynamicMetadataProtocol,
+    DynamicMetadataRequirementsProtocol,
+    DynamicMetadataWheelProtocol,
+    DynamicMetadataRequirementsWheelProtocol,
+]
+
+
 def load_provider(
     provider: str,
     provider_path: str | None = None,
-) -> DynamicMetadataProtocol | DynamicMetadataRequirementsProtocol:
+) -> DMProtocols:
     if provider_path is None:
         return importlib.import_module(provider)
 
@@ -43,3 +64,16 @@ def load_provider(
         return importlib.import_module(provider)
     finally:
         sys.path.pop(0)
+
+
+def load_dynamic_metadata(
+    metadata: Mapping[str, Mapping[str, str]]
+) -> Generator[tuple[str, DMProtocols | None, dict[str, str]], None, None]:
+    for field, orig_config in metadata.items():
+        if "provider" in orig_config:
+            config = dict(orig_config)
+            provider = config.pop("provider")
+            provider_path = config.pop("provider-path", None)
+            yield field, load_provider(provider, provider_path), config
+        else:
+            yield field, None, dict(orig_config)

src/scikit_build_core/settings/metadata.py

@@ -1,12 +1,13 @@
 from __future__ import annotations
 
+from collections.abc import Mapping
 from typing import Any
 
 from packaging.version import Version
 from pyproject_metadata import StandardMetadata
 
 from ..settings.skbuild_model import ScikitBuildSettings
-from ._load_provider import load_provider
+from ._load_provider import load_dynamic_metadata
 
 __all__ = ["get_standard_metadata"]
 
@@ -17,37 +18,24 @@ def __dir__() -> list[str]:
 
 # If pyproject-metadata eventually supports updates, this can be simplified
 def get_standard_metadata(
-    pyproject_dict: dict[str, Any],
+    pyproject_dict: Mapping[str, Any],
     settings: ScikitBuildSettings,
 ) -> StandardMetadata:
+    new_pyproject_dict = dict(pyproject_dict)
     # Handle any dynamic metadata
-    calls: dict[frozenset[tuple[str, Any]], set[str]] = {}
-    for field, raw_settings in settings.metadata.items():
-        if field not in pyproject_dict.get("project", {}).get("dynamic", []):
-            msg = f"{field} is not in project.dynamic"
-            raise KeyError(msg)
-        if "provider" not in raw_settings:
+    for field, provider, config in load_dynamic_metadata(settings.metadata):
+        if provider is None:
             msg = f"{field} is missing provider"
             raise KeyError(msg)
-        calls.setdefault(frozenset(raw_settings.items()), set()).add(field)
-
-    for call, fields in calls.items():
-        args = dict(call)
-        provider = args.pop("provider")
-        provider_path = args.pop("provider-path", None)
-        computed = load_provider(provider, provider_path).dynamic_metadata(
-            frozenset(fields), args
-        )
-        if set(computed) != fields:
-            msg = f"{provider} did not return requested fields"
+        if field not in pyproject_dict.get("project", {}).get("dynamic", []):
+            msg = f"{field} is not in project.dynamic"
             raise KeyError(msg)
-        pyproject_dict["project"].update(computed)
-        for field in fields:
-            pyproject_dict["project"]["dynamic"].remove(field)
+        new_pyproject_dict["project"][field] = provider.dynamic_metadata(field, config)
+        new_pyproject_dict["project"]["dynamic"].remove(field)
 
-    metadata = StandardMetadata.from_pyproject(pyproject_dict)
+    metadata = StandardMetadata.from_pyproject(new_pyproject_dict)
     # pyproject-metadata normalizes the name - see https://github.com/FFY00/python-pyproject-metadata/pull/65
     # For scikit-build-core 0.5+, we keep the un-normalized name, and normalize it when using it for filenames
     if settings.minimum_version is None or settings.minimum_version >= Version("0.5"):
-        metadata.name = pyproject_dict["project"]["name"]
+        metadata.name = new_pyproject_dict["project"]["name"]
     return metadata

tests/packages/custom_cmake/CMakeLists.txt

@@ -3,7 +3,7 @@ cmake_minimum_required(VERSION 3.15...3.26)
 project(
   ${SKBUILD_PROJECT_NAME}
   LANGUAGES
-  VERSION ${SKBUILD_PROJECT_VERSION})
+  VERSION 2.3.4)
 
 find_package(ExamplePkg REQUIRED)
 

tests/packages/custom_cmake/pyproject.toml

@@ -4,4 +4,13 @@ build-backend = "scikit_build_core.build"
 
 [project]
 name = "custom_modules"
-version = "0.0.1"
+dynamic = ["version"]
+
+[tool.scikit-build]
+wheel.packages = []
+wheel.license-files = []
+
+[tool.scikit-build.metadata.version]
+provider = "scikit_build_core.metadata.regex"
+input = "CMakeLists.txt"
+regex = 'project\([^)]+ VERSION (?P<value>[0-9.]+)'

tests/packages/dynamic_metadata/plugins/local/version/nested/__init__.py

@@ -8,15 +8,15 @@ def __dir__() -> list[str]:
 
 
 def dynamic_metadata(
-    fields: frozenset[str],
+    field: str,
     settings: dict[str, object] | None = None,
-) -> dict[str, str | dict[str, str | None]]:
-    if fields != {"version"}:
+) -> str:
+    if field != "version":
         msg = "Only the 'version' field is supported"
         raise ValueError(msg)
 
     if settings:
         msg = "No inline configuration is supported"
         raise ValueError(msg)
 
-    return {"version": "3.2.1"}
+    return "3.2.1"