Skip to main content

Type stubs for google-api-python-client

Project description

Type stubs for google-api-python-client

PyPI version

This package provides type stubs for the google-api-python-client library. It allows you to type check usage of the library with e.g. mypy and will also improve autocomplete in many editors.

This is in no way affiliated with Google.

The stubs were generated automatically based on Google's Discovery Documents, that are bundled with google-api-python-client as of v2.

If you find incorrect annotations, please create an issue.

Releases can be somewhat infrequent. If you need a new release, create an issue and I'll probably get around to it faster.

Installation

$ pip install google-api-python-client-stubs

The stubs should be automatically picked up by your editor or typechecker.

Caveats

Performance

The stubs contain a separate overload of googleapiclient.discovery.build for each service and version (see discovery.pyi). This can lead to slow type inference for this function. Mypy will generally be pretty fast after the first run, but you might experience slow autocomplete in your editor. If you're experiencing this problem you can bypass type inference with explicit annotations, e.g. sheets_service: SheetsResource = build("sheets", "v4") instead of sheets_service = build("sheets", "v4"). See the next section for some caveats to this approach.

Explicit annotations

The google-api-python-client-stubs is quite dynamic. All resources are just instances of a single class with dynamically attached methods and the requests and responses are just dictionaries. The way this is annotated is with classes and TypedDicts that don't exist in the actual library, only in the stub files. This means you cannot use any of these types at runtime, it will cause your code to crash. If you rely solely on type inference, this is not an issue, but as soon as you want to explicitly annotate a variable, argument or return type in your code with one of these types, you must follow these rules:

  1. Either put from __future__ import annotations at the top of your file or surround the annotations with quotes. This ensures that Python doesn't attempt to evaluate the types at runtime. Note that the import is only supported in Python 3.7 and above.
  2. Only import the types inside an if typing.TYPE_CHECKING block. This ensures that the imports are only evaluated during type checking. Note that any type not available at runtime is located under the googleapiclient._apis package. For example, SheetsResource should be imported from googleapiclient._apis.sheets.v4. Note that nested resources have a nested class structure, e.g. the return type of sheets_service.spreadsheets().values() is SheetsResource.SpreadsheetsResource.ValuesResource. If autocompleting import paths doesn't work you can find the correct path by using Mypy's reveal_type on the thing you want to explicitly annotate. For TypedDicts this will also give you useful info about the structure of the dictionary.

Recursive types

The stubs previously didn't include recursive type definitions due to a lack of type checker support, but this is now included. Note that if you're using Mypy, v0.990 or higher is required for this to work.

Stubs for non-API-specific parts

There are detailed stubs for the API services, but other parts of the library have only been annotated with stubgen, so they're mostly typed as Any. I believe these parts are mostly used internally by the library itself, so for most users this should be fine. Contributions to improve these stubs are welcome, though.

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

google_api_python_client_stubs-1.25.0.tar.gz (2.1 MB view hashes)

Uploaded Source

Built Distribution

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page