Add Ingestion Module for GCS/S3 Resources to Snowflake

[cyclux]

Load jaffle-shop Parquet files from cloud storage into Snowflake.

API:

from data import load_from_gcs, load_from_s3

results = load_from_gcs(session, schema_name="RAW")
results = load_from_s3(session, bucket="s3://bucket/path/", schema_name="RAW")

Files:

• data/ingestion.py
• data/sql/ingestion/*.sql

Behavior:

1. GCS: Download → internal stage → COPY INTO
2. S3: External stage → COPY INTO
3. Schema inferred from Parquet (INFER_SCHEMA)
4. Idempotent (safe to re-run)

SQL Templates:

File	Purpose
create_parquet_file_format.sql	Parquet format definition
create_internal_stage.sql	Internal stage for GCS downloads
create_stage_s3_public.sql	External S3 stage
create_table_from_parquet.sql	Table creation with INFER_SCHEMA
copy_into_table.sql	COPY INTO with MATCH_BY_COLUMN_NAME