Create get_packaged_python_script.py

Author: cthoyt

src/fair_mappings_schema/get_packaged_python_script.py

@@ -0,0 +1,118 @@
+"""Extract metadata from Python script."""
+
+import re
+from typing import Any
+
+from fair_mappings_schema.datamodel.fair_mappings_schema_pydantic import (
+    MappingSpecification,
+    Person,
+    MappingSpecificationTypeEnum,
+)
+import click
+import requests
+from pystow.utils import model_dump_yaml
+import tomllib
+
+__all__ = ["get_python_script"]
+
+
+def get_python_script(script_url: str) -> MappingSpecification:
+    """Get a mapping specification from a package."""
+    owner, repo = _get_repository(script_url)
+    data = _get_pyproject_toml(owner, repo)
+    project = data["project"]
+    urls = data.get("urls", {})
+    documentation = urls.get("documentation") or urls.get("Documentation")
+    xxx = dict(
+        content_url=script_url,
+        # TODO add explicit way of saying it's code
+        type=MappingSpecificationTypeEnum.other,
+        mapping_method="Python script",
+        name=project["name"],
+        version=project["version"],
+        description=project.get("description"),
+        license=project.get("license"),
+        author=_get_person(project, "authors"),
+        documentation=documentation,
+    )
+    xxx.update(_get_script_data(script_url))
+    return MappingSpecification.model_validate(xxx)
+
+
+def _get_script_data(script_url: str) -> dict[str, Any]:
+    """Get a mapping specification from a package."""
+    res = requests.get(script_url, timeout=5)
+    res.raise_for_status()
+    dd = extract_script_toml(res.text)
+    return dd.get("tool", {}).get("fair-mappings")
+
+
+def _get_person(project: dict[str, Any], key: str) -> Person | None:
+    people = project.get(key)
+    if not people:
+        return None
+    person = people[0]
+    name = person["name"]
+    # TODO add email to person model
+    # TODO see if we can recognize ORCiD here more officially
+    orcid = person.get("orcid")
+    if orcid:
+        orcid = orcid.removeprefix("https://orcid.com")
+        orcid = orcid.removeprefix("http://orcid.com")
+    return Person(name=name, orcid=orcid)
+
+
+def _get_pyproject_toml(owner: str, repo: str, branch: str = "main"):
+    url = f"https://github.com/{owner}/{repo}/raw/refs/heads/{branch}/pyproject.toml"
+    res = requests.get(url, timeout=5)
+    data = tomllib.loads(res.text)
+    return data
+
+
+def _get_repository(url: str) -> tuple[str, str]:
+    """Get a mapping specification from a package."""
+    if url.startswith("https://github.com/"):
+        url = url.removeprefix("https://github.com/")
+    if url.startswith("http://github.com/"):
+        url = url.removeprefix("http://github.com/")
+
+    parts = url.split("/")
+    if len(parts) == 1:
+        raise ValueError(f"no repository given, just an owner: {url}")
+    owner, repo, *_ = parts
+    return owner, repo
+
+
+def extract_script_toml(source: str) -> dict[str, Any] | None:
+    """
+    Extract the raw TOML string from a `# /// script` ... `# ///` block.
+    Returns the TOML text, or None if no such block is found.
+    """
+    pattern = re.compile(
+        r"^#\s/// script\s*\n"  # opening marker
+        r"((?:#[^\n]*\n)*?)"  # captured comment lines
+        r"#\s///\s*$",  # closing marker
+        re.MULTILINE,
+    )
+    match = pattern.search(source)
+    if not match:
+        return None
+
+    # Strip the leading `# ` (or `#`) from every line
+    toml_lines = []
+    for line in match.group(1).splitlines():
+        # Remove exactly one leading `# ` or `#`
+        toml_lines.append(re.sub(r"^#( ?)", "", line))
+
+    return tomllib.loads("\n".join(toml_lines))
+
+
+def _main():
+    # TODO make function that fixes URL to be raw
+    better_url = "https://github.com/data-literacy-alliance/oerbservatory/raw/refs/heads/main/src/oerbservatory/sources/dalia.py"
+    model = get_python_script(better_url)
+    click.echo(model_dump_yaml(model, exclude_none=True, exclude={"author.type"}))
+
+
+if __name__ == "__main__":
+    _main()