Support of Google Cloud Storage for cloud storage

[jasonkit]

Motivation and context

Add Google Cloud Storage support for cloud storage.
Also update Credentials class to include key_file_path and
add KEY_FILE_PATH as a new credential type which is required for using
google cloud storage.
It is expected to set the key_file_path to the Google cloud's service account key json file.

The specific_attributes for Google Cloud Storage includes:

• prefix: key prefix for filtering blobs that going to fetch during initialize_content()
• location: The location of bucket if we are going to create a bucket
• project: The Google Cloud project id the bucket if we are going to create a bucket

All of the above attributes are optional.

How has this been tested?

Tested manually.

Tested sending HTTP POST to /api/v1/cloudstorages with following body

{
  "provider_type": "GOOGLE_CLOUD_STORAGE",
  "resource": "my_bucket",
  "display_name": "My Bucket",
  "credentials_type": "KEY_FILE_PATH",
  "key_file_path": "/credentials/gcs.json",
  "specific_attributes": "prefix=test"
}

and confirmed the cloudstorages record could be created on DB

To make this call works, we need to ensure /credentials/gcs.json exists on cvat server
and this service account key json file should has permission to access gs://my_bucket

Also tested creating new task data with using this newly created cloud storage
by sending HTTP POST to /api/v1/tasks/:task_id/data with following body

{
  "chunk_size": 1,
  "image_quality": 70,
  "server_files": [
     "test/img1.jpg",
     "test/img2.jpg",
     "test/manifest.jsonl"
  ],
  "cloud_storage_id": 1,
  "use_zip_chunks": true,
  "use_cache": true
}

and confirmed the task could be created, and the all the images could be loaded.

Checklist

• I submit my changes into the develop branch
• I have added description of my changes into CHANGELOG file
• I have updated the documentation accordingly
• I have added tests to cover my changes
• I have linked related issues (read github docs)
• I have increased versions of npm packages if it is necessary (cvat-canvas,
  cvat-core, cvat-data and cvat-ui)

License

• I submit my code changes under the same MIT License that covers the project.
  Feel free to contact the maintainers if that's a concern.
• I have updated the license header for each file (see an example below)

# Copyright (C) 2021 Intel Corporation
#
# SPDX-License-Identifier: MIT

[nmanovic]

@jasonkit , thanks for the contribution. Please give us some time to review the PR. We are in the process of adding UI for the feature.

@Marishka17 , please review the PR next week.

[Marishka17]

@jasonkit , Hi. Thanks for the contribution, please give me some time to test the functionality of the PR.

[Marishka17]

Good job, I have only one small comment: i don`t see the migration file, could you please add it?

[Marishka17]

Will it work if location == None?

[jasonkit]

Yes, location is indeed an optional argument for create_bucket, if location is None, it will default to us

ref: https://googleapis.dev/python/storage/latest/client.html#google.cloud.storage.client.Client.create_bucket

[jasonkit]

@Marishka17
I have added a commit for the auto generated migration file for change in engine_cloud_provider table

[bsekachev]

Thank you for the contribution!