gcp_cloud_function_sample

import time
from typing import List
import os

os.environ["GOOGLE_APPLICATION_CREDENTIALS"]= "credentials.json"

def deidentify_cloud_storage(
    project = ["YOUR_PROJECT_ID"],
    input_gcs_bucket = ["INPUT_DATA_BUCKET_NAME"],
    output_gcs_bucket = ["SINK_DATA_BUCKET_NAME"],
    info_types = ['PERSON_NAME', 'STREET_ADDRESS', 'LOCATION'],
    timeout: int = 1000,
) -> None:
    
    # Instantiate a client.
    dlp = google.cloud.dlp_v2.DlpServiceClient()

    # Construct the configuration dictionary.
    # Specify the type of info the inspection will look for.
    # See https://cloud.google.com/dlp/docs/infotypes-reference for complete list of info types.
    inspect_config = {"info_types": [{"name": info_type} for info_type in info_types]}

    # Construct cloud_storage_options dictionary with the bucket's URL.
    storage_config = {
        "cloud_storage_options": {"file_set": {"url": f"gs://{input_gcs_bucket}"}}
    }

    # Specify the big query table to store the transformation details.
    # big_query_table = {
    #     "project_id": project,
    #     "dataset_id": dataset_id,
    #     "table_id": table_id,
    # }

    # Convert the project id into a full resource id.
    parent = f"projects/{project}/locations/global"

    # Construct Transformation Configuration with de-identify Templates used
    # for transformation.
    # transformation_config = {
    #     "deidentify_template": f"{parent}/deidentifyTemplates/{deid_template_id}",
    #     "structured_deidentify_template": f"{parent}/deidentifyTemplates/{structured_deid_template_id}",
    #     "image_redact_template": f"{parent}/deidentifyTemplates/{image_redact_template_id}",
    # }

    # Tell the API where to send notification when the job is completed.
    actions = [
        {
            "deidentify": {
                "cloud_storage_output": f"gs://{output_gcs_bucket}",
                "file_types_to_transform": ["TEXT_FILE"],
            }
        }
    ]

    # Construct the job definition.
    inspect_job = {
        "inspect_config": inspect_config,
        "storage_config": storage_config,
        "actions": actions,
    }

    # Call the API.
    response = dlp.create_dlp_job(
        request={
            "parent": parent,
            "inspect_job": inspect_job,
        }
    )

    job_name = response.name
    print(f"Inspection Job started : {job_name}")

    # Waiting for the job to get completed.
    job = dlp.get_dlp_job(request={"name": job_name})
    # Since the sleep time is kept as 30s, number of calls would be timeout/30.
    no_of_attempts = timeout // 30
    while no_of_attempts != 0:
        # Check if the job has completed.
        if job.state == google.cloud.dlp_v2.DlpJob.JobState.DONE:
            break
        if job.state == google.cloud.dlp_v2.DlpJob.JobState.FAILED:
            print("Job Failed, Please check the configuration.")
            break

        # Sleep for a short duration before checking the job status again.
        time.sleep(30)
        no_of_attempts -= 1

        # Get DLP job status.
        job = dlp.get_dlp_job(request={"name": job_name})

    if job.state != google.cloud.dlp_v2.DlpJob.JobState.DONE:
        print(f"Job did not complete within {timeout} minutes.")
        return

    # Print out the results.
    print(f"Job name: {job.name}")
    result = job.inspect_details.result
    print(f"Processed Bytes: {result.processed_bytes}")
    if result.info_type_stats:
        for stats in result.info_type_stats:
            print(f"Info type: {stats.info_type.name}")
            print(f"Count: {stats.count}")
    else:
        print("No findings.")