Improve the update_schemastore script (#11353)

scripts/pyproject.toml

@@ -26,5 +26,11 @@ select = [
   "RUF",
 ]
 
+ignore = [
+  # only relevant if you run a script with `python -0`,
+  # which seems unlikely for any of the scripts in this directory
+  "B011"
+]
+
 [tool.ruff.lint.isort]
 required-imports = ["from __future__ import annotations"]

scripts/update_schemastore.py

@@ -7,13 +7,13 @@ to schemastore in the CLI.
 
 from __future__ import annotations
 
+import enum
 import json
 from pathlib import Path
 from subprocess import check_call, check_output
 from tempfile import TemporaryDirectory
+from typing import NamedTuple, assert_never
 
-schemastore_fork = "git@github.com:astral-sh/schemastore.git"
-schemastore_upstream = "git@github.com:SchemaStore/schemastore.git"
 ruff_repo = "https://github.com/astral-sh/ruff"
 root = Path(
     check_output(["git", "rev-parse", "--show-toplevel"], text=True).strip(),
@@ -21,35 +21,64 @@ root = Path(
 ruff_json = Path("schemas/json/ruff.json")
 
 
-def update_schemastore(schemastore: Path) -> None:
+class SchemastoreRepos(NamedTuple):
-    if not schemastore.is_dir():
+    fork: str
-        check_call(["git", "clone", schemastore_fork, schemastore])
+    upstream: str
+
+
+class GitProtocol(enum.Enum):
+    SSH = "ssh"
+    HTTPS = "https"
+
+    def schemastore_repos(self) -> SchemastoreRepos:
+        match self:
+            case GitProtocol.SSH:
+                return SchemastoreRepos(
+                    fork="git@github.com:astral-sh/schemastore.git",
+                    upstream="git@github.com:SchemaStore/schemastore.git",
+                )
+            case GitProtocol.HTTPS:
+                return SchemastoreRepos(
+                    fork="https://github.com/astral-sh/schemastore.git",
+                    upstream="https://github.com/SchemaStore/schemastore.git",
+                )
+            case _:
+                assert_never(self)
+
+
+def update_schemastore(
+    schemastore_path: Path, schemastore_repos: SchemastoreRepos
+) -> None:
+    if not schemastore_path.is_dir():
+        check_call(
+            ["git", "clone", schemastore_repos.fork, schemastore_path, "--depth=1"]
+        )
         check_call(
             [
                 "git",
                 "remote",
                 "add",
                 "upstream",
-                schemastore_upstream,
+                schemastore_repos.upstream,
             ],
-            cwd=schemastore,
+            cwd=schemastore_path,
         )
     # Create a new branch tagged with the current ruff commit up to date with the latest
     # upstream schemastore
-    check_call(["git", "fetch", "upstream"], cwd=schemastore)
+    check_call(["git", "fetch", "upstream"], cwd=schemastore_path)
     current_sha = check_output(["git", "rev-parse", "HEAD"], text=True).strip()
     branch = f"update-ruff-{current_sha}"
     check_call(
         ["git", "switch", "-c", branch],
-        cwd=schemastore,
+        cwd=schemastore_path,
     )
     check_call(
         ["git", "reset", "--hard", "upstream/master"],
-        cwd=schemastore,
+        cwd=schemastore_path,
     )
 
     # Run npm install
-    src = schemastore.joinpath("src")
+    src = schemastore_path.joinpath("src")
     check_call(["npm", "install"], cwd=src)
 
     # Update the schema and format appropriately
@@ -71,7 +100,7 @@ def update_schemastore(schemastore: Path) -> None:
 
     # Check if the schema has changed
     # https://stackoverflow.com/a/9393642/3549270
-    if check_output(["git", "status", "-s"], cwd=schemastore).strip():
+    if check_output(["git", "status", "-s"], cwd=schemastore_path).strip():
         # Schema has changed, commit and push
         commit_url = f"{ruff_repo}/commit/{current_sha}"
         commit_body = (
@@ -88,24 +117,43 @@ def update_schemastore(schemastore: Path) -> None:
                 "-m",
                 commit_body,
             ],
-            cwd=schemastore,
+            cwd=schemastore_path,
         )
         # This should show the link to create a PR
         check_call(
             ["git", "push", "--set-upstream", "origin", branch],
-            cwd=schemastore,
+            cwd=schemastore_path,
         )
     else:
         print("No changes")
 
 
+def determine_git_protocol(argv: list[str] | None = None) -> GitProtocol:
+    import argparse
+
+    parser = argparse.ArgumentParser(
+        description=__doc__, formatter_class=argparse.RawDescriptionHelpFormatter
+    )
+    parser.add_argument(
+        "--proto",
+        required=True,
+        choices=[proto.value for proto in GitProtocol],
+        help="Protocol to use for cloning git repos",
+    )
+    args = parser.parse_args(argv)
+    return GitProtocol(args.proto)
+
+
 def main() -> None:
+    schemastore_repos = determine_git_protocol().schemastore_repos()
     schemastore_existing = root.joinpath("schemastore")
     if schemastore_existing.is_dir():
-        update_schemastore(schemastore_existing)
+        update_schemastore(schemastore_existing, schemastore_repos)
     else:
         with TemporaryDirectory() as temp_dir:
-            update_schemastore(Path(temp_dir).joinpath("schemastore"))
+            update_schemastore(
+                Path(temp_dir).joinpath("schemastore"), schemastore_repos
+            )
 
 
 if __name__ == "__main__":