This library is in Alpha. We will make an effort to support the library, but we reserve the right to make incompatible changes when necessary.
Version 0.9 is not compatible with previous versions. See MIGRATING for additional details on how to migrate to the latest version.
Add this line to your application's Gemfile:
gem 'google-api-client', '0.9.pre1'
And then execute:
$ bundle
Or install it yourself as:
$ gem install google-api-client
To use an API, include the corresponding generated file and instantiate the service. For example to use the Drive API:
require 'google/apis/drive_v2'
Drive = Google::Apis::DriveV2 # Alias the module
drive = Drive::DriveService.new
drive.authorization = authorization # See Googleauth or Signet libraries
# Search for files in Drive (first page only)
files = drive.list_files(q: "title contains 'finances'")
files.items.each do |file|
puts file.title
end
# Upload a file
metadata = Drive::File.new(title: 'My document')
metadata = drive.insert_file(metadata, upload_source: 'test.txt', content_type: 'text/plain')
# Download a file
drive.get_file(metadata.id, download_dest: '/tmp/myfile.txt')
Methods that allow media operations have additional parameters to specify the upload source or download destination.
For uploads, the upload_source
parameter can be specified with either a path to a file, an IO
stream, or StringIO
instance.
For downloads, the download_dest
parameter can also be either a path to a file, an IO
stream, or StringIO
instance.
Both uploads & downloads are resumable. If an error occurs during transmission the request will be automatically retried from the last received byte.
Retries are disabled by default, but enabling retries is strongly encouraged. The number of retries can be configured
via Google::Apis::RequestOptions
. Any number greater than 0 will enable retries.
To enable retries for all services:
Google::Apis::RequestOptions.default.retries = 5
With retries enabled globally, retries can be disabled for specific calls by including a retry value of 0 in the request options:
drive.insert_file(metadata, upload_source: 'test.txt', content_type: 'text/plain', options: { retries: 0 })
When retries are enabled, if a server or rate limit error occurs during a request it is automatically retried with an exponentially increasing delay on subsequent retries. If a request can not be retried or if the maximum number of retries is exceeded, an exception is thrown.
A block an be specified when making calls. If present, the block will be called with the result or error, rather than returning the result from the call or raising the error. Example:
# Search for files in Drive (first page only)
drive.list_files(q: "title contains 'finances'") do |res, err|
if err
# Handle error
else
# Handle response
end
end
This calling style is required when making batch requests as responses are not available until the entire batch is complete.
Multiple requests can be batched together into a single HTTP request to reduce overhead. Batched calls are executed in parallel and the responses processed once all results are available
# Fetch a bunch of files by ID
ids = ['file_id_1', 'file_id_2', 'file_id_3', 'file_id_4']
drive.batch do |drive|
ids.each do |id|
drive.get_file(id) do |res, err|
# Handle response
end
end
end
Media operations -- uploads & downloads -- can not be included in batch with other requests.
However, some APIs support batch uploads. To upload multiple files in a batch, use the batch_upload
method instead.
Batch uploads should only be used when uploading multiple small files. For large files, upload files individually to
take advantage of the libraries built-in reusmable upload support.
While the API will always return instances of schema classes, plain hashes are accepted in method calls for convenience. Hash keys must be symbols matching the attribute names on the corresponding object the hash is meant to replace. For example:
file = {id: '123', title: 'My document', labels: { starred: true }}
file = drive.insert_file(file) # Returns a Drive::File instance
is equivalent to:
file = Drive::File.new(id: '123', title: 'My document')
file.labels = Drive::File::Labels.new(starred: true)
file = drive.update_file(file) # Returns a Drive::File instance
OAuth 2 is used to authorize applications. This library uses both Signet and Google Auth Library for Ruby for OAuth 2 support.
The Google Auth Library for Ruby provides an implementation of [application default credentials] for Ruby. It offers a simple way to get authorization credentials for use in calling Google APIs, best suited for cases when the call needs to have the same identity and authorization level for the application independent of the user. This is the recommended approach to authorize calls to Cloud APIs, particularly when you're building an application that uses Google Compute Engine.
For per-user authorization, use Signet to obtain user authorization.
Authorization can be specified for the entire client, for an individual service instance, or on a per-request basis.
Set authorization for all service:
Google::Apis::RequestOptions.default.authorization = authorization
# Services instantiated after this will inherit the authorization
On a per-service level:
drive = Google::Apis::DriveV2::DriveService.new
drive.authorization = authorization
# All requests made with this service will use the same authorization
Per-request:
drive.get_file('123', options: { authorization: authorization })
For Cloud Endpoints or other APIs not included in the gem, ruby code can be generated from the discovery document.
To generate from a local discovery file:
$ generate-api gen <outdir> --file=<path>
A URL can also be specified:
$ generate-api gen <outdir> --url=<url>
- ETag support (if-not-modified)
- Auto-paging results (maybe)
- Caching
- Model validations
This library is licensed under Apache 2.0. Full license text is available in LICENSE.txt.
See CONTRIBUTING.
Please report bugs at the project on Github. Don't hesitate to ask questions about the client or APIs on StackOverflow.