Rewrote protobuf generation scripts in Python (#12527)

scripts/sync_protobuf/_utils.py

@@ -0,0 +1,67 @@
+from __future__ import annotations
+
+import subprocess
+import sys
+from http.client import HTTPResponse
+from pathlib import Path
+from typing import TYPE_CHECKING, Iterable
+from urllib.request import urlopen
+from zipfile import ZipFile
+
+import tomlkit
+from mypy_protobuf.main import (  # type: ignore[import-untyped]  # pyright: ignore[reportMissingTypeStubs]
+    __version__ as mypy_protobuf__version__,
+)
+
+if TYPE_CHECKING:
+    from _typeshed import StrOrBytesPath, StrPath
+
+REPO_ROOT = Path(__file__).absolute().parent.parent.parent
+MYPY_PROTOBUF_VERSION = mypy_protobuf__version__
+
+
+def download_file(url: str, destination: StrPath) -> None:
+    print(f"Downloading '{url}' to '{destination}'")
+    resp: HTTPResponse
+    with urlopen(url) as resp:
+        if resp.getcode() != 200:
+            raise RuntimeError(f"Error downloading {url}")
+        with open(destination, "wb") as file:
+            file.write(resp.read())
+
+
+def extract_archive(archive_path: StrPath, destination: StrPath) -> None:
+    print(f"Extracting '{archive_path}' to '{destination}'")
+    with ZipFile(archive_path) as file_in:
+        file_in.extractall(destination)
+
+
+def update_metadata(metadata_folder: StrPath, new_extra_description: str) -> None:
+    metadata_path = Path(metadata_folder) / "METADATA.toml"
+    with open(metadata_path) as file:
+        metadata = tomlkit.load(file)
+    metadata["extra_description"] = new_extra_description
+    with open(metadata_path, "w") as file:
+        # tomlkit.dump has partially unknown IO type
+        tomlkit.dump(metadata, file)  # pyright: ignore[reportUnknownMemberType]
+    print(f"Updated {metadata_path}")
+
+
+def run_protoc(
+    proto_paths: Iterable[StrPath], mypy_out: StrPath, proto_globs: Iterable[str], cwd: StrOrBytesPath | None = None
+) -> str:
+    """TODO: Describe parameters and return"""
+    protoc_version = (
+        subprocess.run([sys.executable, "-m", "grpc_tools.protoc", "--version"], capture_output=True).stdout.decode().strip()
+    )
+    print()
+    print(protoc_version)
+    protoc_args = [
+        *[f"--proto_path={proto_path}" for proto_path in proto_paths],
+        "--mypy_out",
+        f"relax_strict_optional_primitives:{mypy_out}",
+        *proto_globs,
+    ]
+    print("Running: protoc\n    " + "\n    ".join(protoc_args) + "\n")
+    subprocess.run((sys.executable, "-m", "grpc_tools.protoc", *protoc_args), cwd=cwd, check=True)
+    return protoc_version

scripts/sync_protobuf/google_protobuf.py

@@ -0,0 +1,94 @@
+"""
+Generates the protobuf stubs for the given protobuf version using mypy-protobuf.
+Generally, new minor versions are a good time to update the stubs.
+"""
+
+from __future__ import annotations
+
+import json
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+from typing import Any
+
+from _utils import MYPY_PROTOBUF_VERSION, REPO_ROOT, download_file, extract_archive, run_protoc, update_metadata
+
+# Whenever you update PACKAGE_VERSION here, version should be updated
+# in stubs/protobuf/METADATA.toml and vice-versa.
+PACKAGE_VERSION = "27.1"
+
+STUBS_FOLDER = REPO_ROOT / "stubs" / "protobuf"
+ARCHIVE_FILENAME = f"protobuf-{PACKAGE_VERSION}.zip"
+ARCHIVE_URL = f"https://github.com/protocolbuffers/protobuf/releases/download/v{PACKAGE_VERSION}/{ARCHIVE_FILENAME}"
+EXTRACTED_PACKAGE_DIR = f"protobuf-{PACKAGE_VERSION}"
+
+VERSION_PATTERN = re.compile(r'def game_version\(\):\n    return "(.+?)"')
+PROTO_FILE_PATTERN = re.compile(r'"//:(.*)_proto"')
+
+
+def extract_python_version(file_path: Path) -> str:
+    """Extract the Python version from https://github.com/protocolbuffers/protobuf/blob/main/version.json"""
+    with open(file_path) as file:
+        data: dict[str, Any] = json.load(file)
+    # The root key will be the protobuf source code version
+    version = next(iter(data.values()))["languages"]["python"]
+    assert isinstance(version, str)
+    return version
+
+
+def extract_proto_file_paths(temp_dir: Path) -> list[str]:
+    """
+    Roughly reproduce the subset of .proto files on the public interface
+    as described in py_proto_library calls in
+    https://github.com/protocolbuffers/protobuf/blob/main/python/dist/BUILD.bazel
+    """
+    with open(temp_dir / EXTRACTED_PACKAGE_DIR / "python" / "dist" / "BUILD.bazel") as file:
+        matched_lines = filter(None, (re.search(PROTO_FILE_PATTERN, line) for line in file))
+        proto_files = [
+            EXTRACTED_PACKAGE_DIR + "/src/google/protobuf/" + match.group(1).replace("compiler_", "compiler/") + ".proto"
+            for match in matched_lines
+        ]
+    return proto_files
+
+
+def main() -> None:
+    temp_dir = Path(tempfile.mkdtemp())
+    # Fetch s2clientprotocol (which contains all the .proto files)
+    archive_path = temp_dir / ARCHIVE_FILENAME
+    download_file(ARCHIVE_URL, archive_path)
+    extract_archive(archive_path, temp_dir)
+
+    # Remove existing pyi
+    for old_stub in STUBS_FOLDER.rglob("*_pb2.pyi"):
+        old_stub.unlink()
+
+    PROTOC_VERSION = run_protoc(
+        proto_paths=(f"{EXTRACTED_PACKAGE_DIR}/src",),
+        mypy_out=STUBS_FOLDER,
+        proto_globs=extract_proto_file_paths(temp_dir),
+        cwd=temp_dir,
+    )
+
+    PYTHON_PROTOBUF_VERSION = extract_python_version(temp_dir / EXTRACTED_PACKAGE_DIR / "version.json")
+
+    # Cleanup after ourselves, this is a temp dir, but it can still grow fast if run multiple times
+    shutil.rmtree(temp_dir)
+
+    update_metadata(
+        STUBS_FOLDER,
+        f"""Partially generated using \
+[mypy-protobuf=={MYPY_PROTOBUF_VERSION}](https://github.com/nipunn1313/mypy-protobuf/tree/v{MYPY_PROTOBUF_VERSION}) \
+and {PROTOC_VERSION} on \
+[protobuf v{PACKAGE_VERSION}](https://github.com/protocolbuffers/protobuf/releases/tag/v{PACKAGE_VERSION}) \
+(python `protobuf=={PYTHON_PROTOBUF_VERSION}`).""",
+    )
+
+    # Run pre-commit to cleanup the stubs
+    subprocess.run((sys.executable, "-m", "pre_commit", "run", "--files", *STUBS_FOLDER.rglob("*_pb2.pyi")))
+
+
+if __name__ == "__main__":
+    main()

scripts/sync_protobuf/s2clientprotocol.py

@@ -0,0 +1,72 @@
+"""
+Generates the protobuf stubs for the given s2clientprotocol version using mypy-protobuf.
+Generally, new minor versions are a good time to update the stubs.
+"""
+
+from __future__ import annotations
+
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+from _utils import MYPY_PROTOBUF_VERSION, REPO_ROOT, download_file, extract_archive, run_protoc, update_metadata
+
+# Whenever you update PACKAGE_VERSION here, version should be updated
+# in stubs/s2clientprotocol/METADATA.toml and vice-versa.
+PACKAGE_VERSION = "c04df4adbe274858a4eb8417175ee32ad02fd609"
+
+STUBS_FOLDER = REPO_ROOT / "stubs" / "s2clientprotocol"
+ARCHIVE_FILENAME = f"{PACKAGE_VERSION}.zip"
+ARCHIVE_URL = f"https://github.com/Blizzard/s2client-proto/archive/{ARCHIVE_FILENAME}"
+EXTRACTED_PACKAGE_DIR = f"s2client-proto-{PACKAGE_VERSION}"
+
+VERSION_PATTERN = re.compile(r'def game_version\(\):\n    return "(.+?)"')
+
+
+def extract_python_version(file_path: Path) -> str:
+    """Extract Python version from s2clientprotocol's build file"""
+    match = re.search(VERSION_PATTERN, file_path.read_text())
+    assert match
+    return match.group(1)
+
+
+def main() -> None:
+    temp_dir = Path(tempfile.mkdtemp())
+    # Fetch s2clientprotocol (which contains all the .proto files)
+    archive_path = temp_dir / ARCHIVE_FILENAME
+    download_file(ARCHIVE_URL, archive_path)
+    extract_archive(archive_path, temp_dir)
+
+    # Remove existing pyi
+    for old_stub in STUBS_FOLDER.rglob("*_pb2.pyi"):
+        old_stub.unlink()
+
+    PROTOC_VERSION = run_protoc(
+        proto_paths=(EXTRACTED_PACKAGE_DIR,),
+        mypy_out=STUBS_FOLDER,
+        proto_globs=(f"{EXTRACTED_PACKAGE_DIR}/s2clientprotocol/*.proto",),
+        cwd=temp_dir,
+    )
+
+    PYTHON_S2_CLIENT_PROTO_VERSION = extract_python_version(temp_dir / EXTRACTED_PACKAGE_DIR / "s2clientprotocol" / "build.py")
+
+    # Cleanup after ourselves, this is a temp dir, but it can still grow fast if run multiple times
+    shutil.rmtree(temp_dir)
+
+    update_metadata(
+        STUBS_FOLDER,
+        f"""Partially generated using \
+[mypy-protobuf=={MYPY_PROTOBUF_VERSION}](https://github.com/nipunn1313/mypy-protobuf/tree/v{MYPY_PROTOBUF_VERSION}) \
+and {PROTOC_VERSION} on \
+[s2client-proto {PYTHON_S2_CLIENT_PROTO_VERSION}](https://github.com/Blizzard/s2client-proto/tree/{PACKAGE_VERSION}).""",
+    )
+
+    # Run pre-commit to cleanup the stubs
+    subprocess.run((sys.executable, "-m", "pre_commit", "run", "--files", *STUBS_FOLDER.rglob("*_pb2.pyi")))
+
+
+if __name__ == "__main__":
+    main()

scripts/sync_protobuf/tensorflow.py

@@ -0,0 +1,137 @@
+"""
+Generates the protobuf stubs for the given tensorflow version using mypy-protobuf.
+Generally, new minor versions are a good time to update the stubs.
+"""
+
+from __future__ import annotations
+
+import os
+import re
+import shutil
+import subprocess
+import sys
+import tempfile
+from pathlib import Path
+
+from _utils import MYPY_PROTOBUF_VERSION, REPO_ROOT, download_file, extract_archive, run_protoc, update_metadata
+
+# Whenever you update PACKAGE_VERSION here, version should be updated
+# in stubs/tensorflow/METADATA.toml and vice-versa.
+PACKAGE_VERSION = "2.17.0"
+
+STUBS_FOLDER = REPO_ROOT / "stubs" / "tensorflow"
+ARCHIVE_FILENAME = f"v{PACKAGE_VERSION}.zip"
+ARCHIVE_URL = f"https://github.com/tensorflow/tensorflow/archive/refs/tags/{ARCHIVE_FILENAME}"
+EXTRACTED_PACKAGE_DIR = f"tensorflow-{PACKAGE_VERSION}"
+
+PROTOS_TO_REMOVE = (
+    "compiler/xla/autotune_results_pb2.pyi",
+    "compiler/xla/autotuning_pb2.pyi",
+    "compiler/xla/service/buffer_assignment_pb2.pyi",
+    "compiler/xla/service/hlo_execution_profile_data_pb2.pyi",
+    "core/protobuf/autotuning_pb2.pyi",
+    "core/protobuf/conv_autotuning_pb2.pyi",
+    "core/protobuf/critical_section_pb2.pyi",
+    "core/protobuf/eager_service_pb2.pyi",
+    "core/protobuf/master_pb2.pyi",
+    "core/protobuf/master_service_pb2.pyi",
+    "core/protobuf/replay_log_pb2.pyi",
+    "core/protobuf/tpu/compile_metadata_pb2.pyi",
+    "core/protobuf/worker_pb2.pyi",
+    "core/protobuf/worker_service_pb2.pyi",
+    "core/util/example_proto_fast_parsing_test_pb2.pyi",
+)
+"""
+These protos exist in a folder with protos used in python,
+but are not included in the python wheel.
+They are likely only used for other language builds.
+stubtest was used to identify them by looking for ModuleNotFoundError.
+(comment out ".*_pb2.*" from the allowlist)
+"""
+
+TSL_IMPORT_PATTERN = re.compile(r"(\[|\s)tsl\.")
+XLA_IMPORT_PATTERN = re.compile(r"(\[|\s)xla\.")
+
+
+def post_creation() -> None:
+    """Move third-party and fix imports"""
+    # Can't use shutil.move because it can't merge existing directories.
+    print()
+    print(f"Moving '{STUBS_FOLDER}/tsl' to '{STUBS_FOLDER}/tensorflow/tsl'")
+    shutil.copytree(f"{STUBS_FOLDER}/tsl", f"{STUBS_FOLDER}/tensorflow/tsl", dirs_exist_ok=True)
+    shutil.rmtree(f"{STUBS_FOLDER}/tsl")
+
+    print(f"Moving '{STUBS_FOLDER}/xla' to '{STUBS_FOLDER}/tensorflow/compiler/xla'")
+    shutil.copytree(f"{STUBS_FOLDER}/xla", f"{STUBS_FOLDER}/tensorflow/compiler/xla", dirs_exist_ok=True)
+    shutil.rmtree(f"{STUBS_FOLDER}/xla")
+
+    for path in STUBS_FOLDER.rglob("*_pb2.pyi"):
+        print(f"Fixing imports in '{path}'")
+        with open(path) as file:
+            filedata = file.read()
+
+        # Replace the target string
+        filedata = re.sub(TSL_IMPORT_PATTERN, "\\1tensorflow.tsl.", filedata)
+        filedata = re.sub(XLA_IMPORT_PATTERN, "\\1tensorflow.compiler.xla.", filedata)
+
+        # Write the file out again
+        with open(path, "w") as file:
+            file.write(filedata)
+
+    print()
+    for to_remove in PROTOS_TO_REMOVE:
+        file_path = STUBS_FOLDER / "tensorflow" / to_remove
+        os.remove(file_path)
+        print(f"Removed '{file_path}'")
+
+
+def main() -> None:
+    temp_dir = Path(tempfile.mkdtemp())
+    # Fetch tensorflow (which contains all the .proto files)
+    archive_path = temp_dir / ARCHIVE_FILENAME
+    download_file(ARCHIVE_URL, archive_path)
+    extract_archive(archive_path, temp_dir)
+
+    # Remove existing pyi
+    for old_stub in STUBS_FOLDER.rglob("*_pb2.pyi"):
+        old_stub.unlink()
+
+    PROTOC_VERSION = run_protoc(
+        proto_paths=(
+            f"{EXTRACTED_PACKAGE_DIR}/third_party/xla/third_party/tsl",
+            f"{EXTRACTED_PACKAGE_DIR}/third_party/xla",
+            f"{EXTRACTED_PACKAGE_DIR}",
+        ),
+        mypy_out=STUBS_FOLDER,
+        proto_globs=(
+            f"{EXTRACTED_PACKAGE_DIR}/third_party/xla/xla/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/third_party/xla/xla/service/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/core/example/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/core/framework/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/core/protobuf/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/core/protobuf/tpu/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/core/util/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/tensorflow/python/keras/protobuf/*.proto",
+            f"{EXTRACTED_PACKAGE_DIR}/third_party/xla/third_party/tsl/tsl/protobuf/*.proto",
+        ),
+        cwd=temp_dir,
+    )
+
+    # Cleanup after ourselves, this is a temp dir, but it can still grow fast if run multiple times
+    shutil.rmtree(temp_dir)
+
+    post_creation()
+
+    update_metadata(
+        STUBS_FOLDER,
+        f"""Partially generated using \
+[mypy-protobuf=={MYPY_PROTOBUF_VERSION}](https://github.com/nipunn1313/mypy-protobuf/tree/v{MYPY_PROTOBUF_VERSION}) \
+and {PROTOC_VERSION} on `tensorflow=={PACKAGE_VERSION}`.""",
+    )
+
+    # Run pre-commit to cleanup the stubs
+    subprocess.run((sys.executable, "-m", "pre_commit", "run", "--files", *STUBS_FOLDER.rglob("*_pb2.pyi")))
+
+
+if __name__ == "__main__":
+    main()