tasks/offline-compute/PrestoSQL/prestosql_002_en/task.md

metadata

id: offline-compute_PrestoSQL_prestosql_002
name: Ad ASA Attribution Migration
category: offline-compute/PrestoSQL
timeout_seconds: 900
modality: pure-text
engine: prestosql

Prompt

Task Objective: From the ad attribution raw logs, filter valid records of two attribution types (ad_attribution and huawei), aggregate the log count and deduplicated user count by version and media dimensions, and write the results to the output table.

Input Table:

• internal_platform_db.t_sh_event_log_v2_06000061157_prestosql_002 (ad attribution raw logs)
  • ftime STRING — time
  • extinfo STRING — extended information
  • ad_version STRING — ad version (ad_attribution/huawei)
  • guid STRING — user identifier
  • ch STRING — channel
  • first_date STRING — first date
  • last_date STRING — last date
  • click_date STRING — click date
  • attribution STRING — attribution flag
  • conversion_date STRING — conversion date
  • conversion_type STRING — conversion type
  • creativeset_id STRING — creative set ID
  • org_id STRING — organization ID
  • campaign_id STRING — campaign ID
  • adgroup_id STRING — ad group ID
  • keyword_id STRING — keyword ID
  • country_oregion STRING — country/region
  • enter_ag_time STRING — ad group entry time
  • installed_finish_time STRING — installation completion time
  • start_download_time STRING — download start time
  • sub_channel STRING — sub-channel
  • callback STRING — callback information
  • task_id STRING — task ID
  • sub_task_id STRING — sub-task ID
  • rta_id STRING — RTA identifier
  • device_id STRING — device_id identifier
  • ext_info1 STRING — extended information 1
  • ext_info2 STRING — extended information 2
  • caid STRING — CAID
  • claim_type STRING — claim type
  • ai_assistant_device_id STRING — YB device_id
  • ds STRING — partition field

Filter Conditions:

• ds = '2026060919'
• Valid record condition: (ad_version = 'ad_attribution' AND attribution = 'true') OR (ad_version = 'phonebrandh' AND callback != '')

Output Requirements:

• Target table: internal_platform_db.t_my_ads_tracking_log_metrics_cand_prestosql_002
• Output fields and order: ds BIGINT, ad_version STRING, media_id INT, log_num BIGINT, log_num_qimei BIGINT
• ds: Cast the original table's ds field to BIGINT
• media_id: CASE ad_version WHEN 'ad_attribution' THEN 39 WHEN 'phonebrandh' THEN 28 END
• log_num: COUNT(*) log count
• log_num_qimei: COUNT(DISTINCT device_id) deduplicated user count
• Group by ds, ad_version
• If the target table does not exist, first create the table, then write the data
• Use Presto/Trino SQL syntax; do not use Hive/Spark SQL dialects

Environment and Execution Notes:

• Presto is running, connected via the Hive catalog
• Execute SQL: presto-cli --catalog hive --schema internal_platform_db -f /tmp_workspace/result.sql
• After writing result.sql, you must execute it yourself to verify that it runs successfully and produces correct data