tasks/offline-compute/PySpark/pyspark_014/init/init_db.py

#!/usr/bin/env python3
"""pyspark_014 database initialization: create tables + load seed data"""
from pyspark.sql import SparkSession

spark = SparkSession.builder \
    .appName('dataclaw_eval_init_pyspark_014') \
    .enableHiveSupport() \
    .config('spark.sql.warehouse.dir', '/tmp/hive_warehouse') \
    .getOrCreate()

spark.sql('CREATE DATABASE IF NOT EXISTS internal_platform_db')

# Create input table
spark.sql('''
CREATE TABLE IF NOT EXISTS internal_platform_db.exp_hours_input_pyspark_102 (
    `fdate` STRING,
    `exp_group_id` STRING,
    `trigger_list` STRING
) PARTITIONED BY (`p_hour` STRING)
STORED AS ORC
''')

# Load seed data
spark.sql('''
INSERT INTO TABLE internal_platform_db.exp_hours_input_pyspark_102 PARTITION (p_hour="p_2026051300")
VALUES
    ("2026051300", "exp_001", "trigger_a,trigger_b"),
    ("2026051300", "exp_002", "trigger_c")
''')

spark.sql('''
INSERT INTO TABLE internal_platform_db.exp_hours_input_pyspark_102 PARTITION (p_hour="p_2026051223")
VALUES
    ("2026051223", "exp_001", "trigger_d"),
    ("2026051223", "exp_002", "trigger_e"),
    ("2026051223", "exp_003", "trigger_f")
''')

spark.sql('''
INSERT INTO TABLE internal_platform_db.exp_hours_input_pyspark_102 PARTITION (p_hour="p_2026051222")
VALUES
    ("2026051222", "exp_003", "trigger_g")
''')

print('Database initialization complete')
spark.stop()