How to create path using execution date in Airflow?

问题

I have the following Airflow dag:

    start_task = DummyOperator(task_id='start_task', dag=dag)

    gcs_export_uri_template = 'adstest/2018/08/31/*'
    update_bigquery = GoogleCloudStorageToBigQueryOperator(
        dag=dag,
        task_id='load_ads_to_BigQuery',
        bucket=GCS_BUCKET_ID,
        destination_project_dataset_table=table_name_template,
        source_format='CSV',
        source_objects=[gcs_export_uri_template],
        schema_fields=dc(),
        create_disposition='CREATE_IF_NEEDED',
        write_disposition='WRITE_APPEND',
        skip_leading_rows = 1,
        google_cloud_storage_conn_id=CONNECTION_ID,
        bigquery_conn_id=CONNECTION_ID
    )

start_task >> update_bigquery

This dag load data from adstest/2018/08/31/* to BigQuery and it works great.

I want to modify the Dag to run over dates based on execution date:

Execution date
Execution date - 1 days
Execution date - 2 days

Example if Execution date is 2018-09-02 I want the DAG to go to:

Execution date : adstest/2018/09/02/*
Execution date - 1 days : adstest/2018/09/01/*
Execution date - 2 days : adstest/2018/08/31/*

How can I do that?

Edit: This is my updated code:

for i in range(5, 0, -1):
    gcs_export_uri_template = ['''adstest/{{ macros.ds_format(macros.ds_add(ds, -{0}), '%Y-%m-%d', '%Y/%m/%d') }}/*'''.format(i)]
    update_bigquery = GoogleCloudStorageToBigQueryOperator(
        dag=dag,
        task_id='load_ads_to_BigQuery-{}'.format(i),
        bucket=GCS_BUCKET_ID,
        destination_project_dataset_table=table_name_template,
        source_format='CSV',
        source_objects=gcs_export_uri_template,
        schema_fields=dc(),
        create_disposition='CREATE_IF_NEEDED',
        write_disposition='WRITE_APPEND',
        skip_leading_rows=1,
        google_cloud_storage_conn_id=CONNECTION_ID,
        bigquery_conn_id=CONNECTION_ID
    )
    start_task >> update_bigquery

Edit 2:

My code:

for i in range(5, 0, -1):
    gcs_export_uri_template = ['''adstest/{{ macros.ds_format(macros.ds_add(ds, -params.i), '%Y-%m-%d', '%Y/%m/%d') }}/*'''.format(i)]

    update_bigquery = GoogleCloudStorageToBigQueryOperator(
        dag=dag,
        task_id='load_ads_to_BigQuery-{}'.format(i),
        bucket=GCS_BUCKET_ID,
        destination_project_dataset_table=table_name_template,
        source_format='CSV',
        source_objects=gcs_export_uri_template,
        schema_fields=dc(),
        params={'i': i},
        create_disposition='CREATE_IF_NEEDED',
        write_disposition='WRITE_APPEND',
        skip_leading_rows=1,
        google_cloud_storage_conn_id=CONNECTION_ID,
        bigquery_conn_id=CONNECTION_ID
    )

Template:

The code gives this error:

"Source URI must not contain the ',' character: gs://adstest/{ macros.ds_format(macros.ds_add(ds, -params.i), '%Y-%m-%d', '%Y/%m/%d') }/*">

回答1:

You can use Airflow Macros to achieve this as follows:

gcs_export_uri_template=[
    "adstest/{{ macros.ds_format(ds, '%Y-%m-%d', '%Y/%m/%d') }}/*",
    "adstest/{{ macros.ds_format(prev_ds, '%Y-%m-%d', '%Y/%m/%d') }}/*",
    "adstest/{{ macros.ds_format(macros.ds_add(ds, -2), '%Y-%m-%d', '%Y/%m/%d') }}/*"
]

update_bigquery = GoogleCloudStorageToBigQueryOperator(
    dag=dag,
    task_id='load_ads_to_BigQuery',
    bucket=GCS_BUCKET_ID,
    destination_project_dataset_table=table_name_template,
    source_format='CSV',
    source_objects=gcs_export_uri_template,
    schema_fields=dc(),
    create_disposition='CREATE_IF_NEEDED',
    write_disposition='WRITE_APPEND',
    skip_leading_rows = 1,
    google_cloud_storage_conn_id=CONNECTION_ID,
    bigquery_conn_id=CONNECTION_ID
)

When you run the above code, you can check in the Web UI, the rendered parameter:

---

For EDITED Comment:

You will need to pass the value of the loop variable i in params parameter and use it in the string as params.i as follows:

for i in range(5, 0, -1):
    gcs_export_uri_template = ["adstest/{{ macros.ds_format(macros.ds_add(ds, -params.i), '%Y-%m-%d', '%Y/%m/%d') }}/*"]
    update_bigquery = GoogleCloudStorageToBigQueryOperator(
        dag=dag,
        task_id='load_ads_to_BigQuery-{}'.format(i),
        bucket=GCS_BUCKET_ID,
        destination_project_dataset_table=table_name_template,
        source_format='CSV',
        source_objects=gcs_export_uri_template,
        schema_fields=dc(),
        params={'i': i},
        create_disposition='CREATE_IF_NEEDED',
        write_disposition='WRITE_APPEND',
        skip_leading_rows=1,
        google_cloud_storage_conn_id=CONNECTION_ID,
        bigquery_conn_id=CONNECTION_ID
    )
    start_task >> update_bigquery

来源：https://stackoverflow.com/questions/52137002/how-to-create-path-using-execution-date-in-airflow