Vendor Python packages (take 2)

This PR vendors all Python packages using the [vendoring](https://pypi.org/project/vendoring/) library.

Vendoring allows users to install packages that rely on versions of libraries that are in conflict with Cog's dependencies (e.g. Pydantic, see replicate#1562, replicate#1384, replicate#1186, replicate#1586, replicate#1336, replicate#785).

This PR is broken up into two commits: First one adds all the files in the `_vendor` directory, second one is the actually changed code.

Closes replicate#409

Signed-off-by: andreasjansson <[email protected]>

CONTRIBUTING.md

@@ -207,6 +207,26 @@ To publish a prerelease version, append a [SemVer prerelease identifer](https://
     git tag -a v0.1.0-alpha -m "Prerelease v0.1.0"
     git push --tags
 
+## Vendoring Python packages
+
+We vendor the few Python libraries we depend on to avoid dependency hell. We use [vendoring](https://pypi.org/project/vendoring/), the same tool used by pip.
+
+The vendored packages are defined in `python/cog/_vendor/vendor.txt`. If you add/change anything in there, run this to update the vendored libraries:
+
+    pip install vendoring
+    vendoring sync -v .
+
+If you run into issues you may need to manually add a patch. Patches are stored in `python/tools/vendoring/patches`.
+
+To create a new patch, copy the file you want to patch to `$FILENAME.new`, and edit that file as needed. For example:
+
+    cp python/cog/_vendor/curio/__main__.py python/cog/_vendor/curio/__main__.py.new
+
+Then `diff` the new and old version to a patch file in `python/tools/vendoring/patches/$PACKAGE.patch` and re-run vendoring. For example:
+
+    diff -u ./python/cog/_vendor/curio/__main__.py ./python/cog/_vendor/curio/__main__.py.new >> python/tools/vendoring/patches/curio.patch
+    vendoring sync -v .
+
 ## Troubleshooting
 
 ### `cog command not found`

pyproject.toml

@@ -12,16 +12,7 @@ urls."Source" = "https://github.com/replicate/cog"
 
 requires-python = ">=3.7"
 dependencies = [
-  # intentionally loose. perhaps these should be vendored to not collide with user code?
-  "attrs>=20.1,<24",
-  "fastapi>=0.75.2,<0.99.0",
-  "pydantic>=1.9,<2",
-  "PyYAML",
-  "requests>=2,<3",
-  "structlog>=20,<25",
   'typing-compat; python_version < "3.8"',
-  "typing_extensions>=4.4.0",
-  "uvicorn[standard]>=0.12,<1",
 ]
 
 optional-dependencies = { "dev" = [
@@ -111,3 +102,36 @@ extend-exclude = [
   "S607", # Starting a process with a partial executable path"
   "ANN",
 ]
+
+[tool.black]
+exclude = '(\.eggs|\.git|\.hg|\.mypy|_cache|\.nox|\.tox|\.venv|\.svn|_build|buck-out|build|dist|_vendor)'
+
+[tool.vendoring]
+destination = "python/cog/_vendor/"
+requirements = "python/cog/_vendor/vendor.txt"
+namespace = "cog._vendor"
+
+protected-files = ["__init__.py", "README.rst", "vendor.txt"]
+patches-dir = "python/tools/vendoring/patches"
+
+[tool.vendoring.transformations]
+substitute = []
+drop = [
+  # pyyaml falls back to plain python
+  "_yaml.*.so",
+
+  # sniffio ships with tests
+  "sniffio/_tests",
+
+  # h11 ships with tests
+  "h11/tests",
+
+  # pydantic hypothesis plugin, fall back to python
+  "pydantic/_hypothesis_plugin.*.so",
+]
+
+[tool.vendoring.typing-stubs]
+
+[tool.vendoring.license.directories]
+
+[tool.vendoring.license.fallback-urls]

python/cog/__init__.py

@@ -1,4 +1,4 @@
-from pydantic import BaseModel
+from ._vendor.pydantic import BaseModel
 
 from .predictor import BasePredictor
 from .types import ConcatenateIterator, File, Input, Path, Secret

python/cog/files.py

@@ -4,7 +4,7 @@
 import os
 from urllib.parse import urlparse
 
-import requests
+from ._vendor import requests
 
 
 def upload_file(fh: io.IOBase, output_file_prefix: str = None) -> str:

python/cog/json.py

@@ -4,7 +4,7 @@
 from types import GeneratorType
 from typing import Any, Callable
 
-from pydantic import BaseModel
+from ._vendor.pydantic import BaseModel
 
 from .types import Path
 

python/cog/logging.py

@@ -1,8 +1,8 @@
 import logging
 import os
 
-import structlog
-from structlog.typing import EventDict
+from ._vendor import structlog
+from ._vendor.structlog.typing import EventDict
 
 
 def replace_level_with_severity(

python/cog/predictor.py

@@ -21,7 +21,7 @@
 )
 from unittest.mock import patch
 
-import structlog
+from ._vendor import structlog
 
 import cog.code_xforms as code_xforms
 
@@ -30,12 +30,12 @@
 except ImportError:  # Python < 3.8
     from typing_compat import get_args, get_origin  # type: ignore
 
-import yaml
-from pydantic import BaseModel, Field, create_model
-from pydantic.fields import FieldInfo
+from ._vendor import yaml
+from ._vendor.pydantic import BaseModel, Field, create_model
+from ._vendor.pydantic.fields import FieldInfo
 
 # Added in Python 3.9. Can be from typing if we drop support for <3.9
-from typing_extensions import Annotated
+from ._vendor.typing_extensions import Annotated
 
 from .errors import ConfigDoesNotExist, PredictorNotSet
 from .types import (

python/cog/schema.py

@@ -7,7 +7,7 @@
 from enum import Enum
 from types import ModuleType
 
-import pydantic
+from ._vendor import pydantic
 
 BUNDLED_SCHEMA_PATH = ".cog/schema.py"
 

python/cog/server/eventtypes.py

@@ -1,6 +1,6 @@
 from typing import Any, Dict
 
-from attrs import define, field, validators
+from .._vendor.attrs import define, field, validators
 
 
 # From worker parent process

python/cog/server/http.py

@@ -25,15 +25,15 @@
 if TYPE_CHECKING:
     from typing import ParamSpec
 
-import attrs
-import structlog
-import uvicorn
-from fastapi import Body, FastAPI, Header, HTTPException, Path, Response
-from fastapi.encoders import jsonable_encoder
-from fastapi.exceptions import RequestValidationError
-from fastapi.responses import JSONResponse
-from pydantic import ValidationError
-from pydantic.error_wrappers import ErrorWrapper
+from .._vendor import attrs
+from .._vendor import structlog
+from .._vendor import uvicorn
+from .._vendor.fastapi import Body, FastAPI, Header, HTTPException, Path, Response
+from .._vendor.fastapi.encoders import jsonable_encoder
+from .._vendor.fastapi.exceptions import RequestValidationError
+from .._vendor.fastapi.responses import JSONResponse
+from .._vendor.pydantic import ValidationError
+from .._vendor.pydantic.error_wrappers import ErrorWrapper
 
 from .. import schema
 from ..errors import PredictorNotSet

python/cog/server/runner.py

@@ -7,12 +7,12 @@
 from multiprocessing.pool import AsyncResult, ThreadPool
 from typing import Any, Callable, Optional, Tuple, Union, cast
 
-import requests
-import structlog
-from attrs import define
-from fastapi.encoders import jsonable_encoder
-from requests.adapters import HTTPAdapter
-from requests.packages.urllib3.util.retry import Retry  # type: ignore
+from .._vendor import requests
+from .._vendor import structlog
+from .._vendor.attrs import define
+from .._vendor.fastapi.encoders import jsonable_encoder
+from .._vendor.requests.adapters import HTTPAdapter
+from .._vendor.urllib3.util.retry import Retry  # type: ignore
 
 from .. import schema, types
 from ..files import put_file_to_signed_endpoint

python/cog/server/webhook.py

@@ -1,10 +1,10 @@
 import os
 from typing import Any, Callable, Set
 
-import requests
-import structlog
-from requests.adapters import HTTPAdapter
-from requests.packages.urllib3.util.retry import Retry  # type: ignore
+from .._vendor import requests
+from .._vendor import structlog
+from .._vendor.requests.adapters import HTTPAdapter
+from .._vendor.urllib3.util.retry import Retry  # type: ignore
 
 from ..schema import Status, WebhookEvent
 from .response_throttler import ResponseThrottler

python/cog/types.py

@@ -8,8 +8,8 @@
 import urllib.request
 from typing import Any, Dict, Iterator, List, Optional, TypeVar, Union
 
-import requests
-from pydantic import Field, SecretStr
+from ._vendor import requests
+from ._vendor.pydantic import Field, SecretStr
 
 FILENAME_ILLEGAL_CHARS = set("\u0000/")
 

python/tests/server/conftest.py

@@ -6,10 +6,10 @@
 from unittest import mock
 
 import pytest
-from attrs import define
+from cog._vendor.attrs import define
 from cog.command import ast_openapi_schema
 from cog.server.http import create_app
-from fastapi.testclient import TestClient
+from cog._vendor.fastapi.testclient import TestClient
 
 
 @define

python/tests/server/fixtures/complex_output.py

@@ -1,4 +1,4 @@
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 class Output(BaseModel):

python/tests/server/fixtures/input_unsupported_type.py

@@ -1,5 +1,5 @@
 from cog import BasePredictor
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 class Input(BaseModel):

python/tests/server/fixtures/openapi_custom_output_type.py

@@ -1,5 +1,5 @@
 from cog import BasePredictor
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 # Calling this `MyOutput` to test if cog renames it to `Output` in the schema

python/tests/server/fixtures/openapi_output_type.py

@@ -1,5 +1,5 @@
 from cog import BasePredictor
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 # An output object called `Output` needs to be special cased because pydantic tries to dedupe it with the internal `Output`

python/tests/server/fixtures/output_complex.py

@@ -1,7 +1,7 @@
 import io
 
 from cog import BasePredictor, File
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 class Output(BaseModel):

python/tests/server/fixtures/output_iterator_complex.py

@@ -1,7 +1,7 @@
 from typing import Iterator, List
 
 from cog import BasePredictor
-from pydantic import BaseModel
+from cog import BaseModel
 
 
 class Output(BaseModel):

python/tests/server/test_http_input.py

@@ -10,6 +10,8 @@
 
 from .conftest import uses_predictor
 
+responses.mock.target = "cog._vendor.requests.adapters.HTTPAdapter.send"
+
 
 @uses_predictor("input_none")
 def test_no_input(client, match):

python/tests/server/test_http_output.py

@@ -6,6 +6,8 @@
 
 from .conftest import uses_predictor, uses_predictor_with_client_options
 
+responses.mock.target = "cog._vendor.requests.adapters.HTTPAdapter.send"
+
 
 @uses_predictor("output_wrong_type")
 def test_return_wrong_type(client):

python/tests/server/test_webhook.py

@@ -1,4 +1,4 @@
-import requests
+from cog._vendor import requests
 import responses
 from cog.schema import WebhookEvent
 from cog.server.webhook import webhook_caller, webhook_caller_filtered

python/tests/server/test_worker.py

@@ -3,7 +3,7 @@
 from typing import Any, Optional
 
 import pytest
-from attrs import define
+from cog._vendor.attrs import define
 from cog.server.eventtypes import (
     Done,
     Heartbeat,

python/tests/test_json.py

@@ -5,7 +5,7 @@
 import numpy as np
 from cog.files import upload_file
 from cog.json import make_encodeable, upload_files
-from pydantic import BaseModel
+from cog._vendor.pydantic import BaseModel
 
 
 def test_make_encodeable_recursively_encodes_tuples():