- BigQuery
- the dataset for the view
- GCP Service Account
- permission:
- role/BigQuery.dataEditor
- role/BigQuery.jobUser
- export: json key file
- permission:
- GitHub Actions
secrets.GCP_PROJECT_ID
- Project ID (not project number) of the Google Cloud project
secrets.GCP_SA_KEY
- Set json key (preferred to compress in single-line)
<1> checkout new git-branch
<2> add the following files to manage description about {test_dataset}.{test_table}
<2-1> ./queries/{test_dataset}/{test_table}.sql
SELECT
1 AS one,
2 AS two,
"3" AS three,
<2-2> ./tables/{test_dataset}/{test_table}.json
[
{
"name": "one",
"type": "INTEGER",
"mode": "NULLABLE",
"description": "this is one as integer"
},
{
"mode": "NULLABLE",
"name": "two",
"type": "INTEGER",
"description": "this is two as integer"
},
{
"mode": "NULLABLE",
"name": "three",
"type": "STRING",
"description": "this is three as string"
}
]
*you can copy schema.json for the table that already exists by the following command at GCP-console-shell
$ bq show --project_id={project_id} --schema --format=prettyjson {dataset_name}.{table_name}
*More details about schema definitions are available in official docments.
<3> create Pull Request
<4> merge to main branch
<5> BQ view will be updated after merge