[MBA-2202] Snowflake DDL migration for source, staging and data product tables

[hrabbani]

Summary

Adds a Snowflake translation of all 12 Teradata tables under a new ddl/snowflake/ directory. The Teradata scripts (ddl/00_source_tables.sql, ddl/01_staging_tables.sql, ddl/02_data_product_tables.sql) are left untouched as the migration reference; nothing else in the repo is modified.

Layout: Teradata databases become schemas inside one database per environment, with <ENV> supplied as a SnowSQL variable (snowsql -f ... -D env=DEV):

CORE_BANKING_DB    -> RETAIL_BANKING_&{env}.CORE_BANKING
TXN_PROCESSING_DB  -> RETAIL_BANKING_&{env}.TXN_PROCESSING
ETL_STAGING_DB     -> RETAIL_BANKING_&{env}.ETL_STAGING
DATA_PRODUCTS_DB   -> RETAIL_BANKING_&{env}.DATA_PRODUCTS

Every statement is CREATE ... IF NOT EXISTS, so the scripts are idempotent and re-runnable in DEV/UAT/PROD without dropping data (CREATE OR REPLACE was deliberately avoided on tables that hold operational/certified data).

Construct mapping (full tables in ddl/snowflake/README.md):

Teradata	Snowflake
MULTISET, NO FALLBACK, PRIMARY INDEX, COLLECT STATISTICS, CHARACTER SET LATIN NOT CASESPECIFIC, FORMAT 'YYYY-MM-DD'	removed (no Snowflake equivalent / handled automatically)
UNIQUE PRIMARY INDEX (TRANSACTION_TYPE_CD)	unenforced PRIMARY KEY constraint, to keep the uniqueness contract in the catalogue
PARTITION BY RANGE_N(TRANSACTION_DATE ... EACH INTERVAL '1' MONTH)	CLUSTER BY (TRANSACTION_DATE) on TRANSACTIONS
PARTITION BY COLUMN(REPORTING_PERIOD)	CLUSTER BY (REPORTING_PERIOD) on TRANSACTION_ANALYTICS
TIMESTAMP(6) / DECIMAL(p,s) / CHAR(n) / DATE	TIMESTAMP_NTZ(6) / NUMBER(p,s) / VARCHAR(n) / DATE
BIGINT / INTEGER / SMALLINT	NUMBER(19,0) / NUMBER(10,0) / NUMBER(5,0) (preserves Teradata integer ranges rather than defaulting to NUMBER(38,0))

No other table gets a clustering key: without a range/period access pattern, automatic micro-partitioning is sufficient and reclustering has ongoing cost.

Validation

• All 4 scripts parse cleanly with sqlglot in the snowflake dialect (after resolving the &{env} substitution).
• Automated column-parity check: all 12 tables present, with identical column names and order to the Teradata DDL.
• Keyword scan: no MULTISET, NO FALLBACK, PRIMARY INDEX, COLLECT STATISTICS, CASESPECIFIC or RANGE_N in executable SQL (they appear only in explanatory comments describing what was replaced).
• Not executed against a live Snowflake account — no credentials available in this session.

Assumptions

• Warehouses, roles and grants are provisioned by TICKET-02 and are intentionally not created here; this ticket only creates the database/schema containers the DDL depends on.
• ETL_STAGING_DB.ETL_RUN_LOG (referenced by the BTEQ scripts) has no Teradata DDL in the affected files, so it is out of scope for this ticket and left to the ETL migration ticket.
• Migrated queries must make case-insensitive comparisons explicit (ILIKE / UPPER()), since NOT CASESPECIFIC semantics do not carry over.

Devin-Org: engineering

Link to Devin session: https://app.devin.ai/sessions/cc38310e2ae84f23a9d0b38a2a858aba
Requested by: @hrabbani

---

Devin Review

Status	Commit
⚪ Not started	—

Run Devin Review

💡 Connect your GitHub account to enable automatic code reviews.

[devin-ai-integration]

🤖 Devin AI Engineer

I'll be helping with this pull request! Here's what you should know:

✅ I will automatically:

• Address comments on this PR. Add '(aside)' to your comment to have me ignore it.
• Look at CI failures and help fix them

Note: I can only respond to comments from users who have write access to this repository.

⚙️ Control Options:

• Disable automatic comment, CI, and merge conflict monitoring