LocalStack Snowflake Extension

This LocalStack extension provides basic emulation of the Snowflake API.

⚠️ Please note that this extension is experimental and still under development.

Prerequisites

• LocalStack Pro
• Docker
• Python

Installation

Before installing the extension, make sure you're logged into LocalStack. If not, log in using the following command:

localstack login

You can then install this extension using the following command:

localstack extensions install localstack-extension-snowflake

Usage

Once the extension is installed, configure your Snowflake client connector to point to the API endpoint https://snowflake.localhost.localstack.cloud. For example, when using the Snowflake Python connector:

client = snowflake.connector.connect(
    user="test",
    password="test",
    account="test",
    host="snowflake.localhost.localstack.cloud",
)
client.cursor().execute("...")

Change Log

• 0.1.20: Initial simple support for table streams; add support for SHOW DATABASES, SHOW VIEWS; enhance parity for Arrow results of TIMESTAMP_NTZ values; more refactoring into QueryProcessor classes; fix identifier uppercasing for ALTER TABLE queries; fix extraction of DB name from CREATE SCHEMA queries
• 0.1.19: Return SELECT results in arrow format for pandas compatibility; add add_months function; fix UDFs with raw expressions; upgrade to Postgres v15; distinguish internal/external VARIANTs; print query processors in CI logs; enable lazy installation of plv8 extension; enhance parity around CREATE VIEW queries; extend analytics; return row_count for UPDATE queries; add statementTypeId
• 0.1.18: Add support for various array and aggregation functions; enhance FILE FORMAT operations; fix CTAS (Create Table AS) queries; support INFER_SCHEMA(..) for getting schema from parquet files; better handling of upper/lowercase identifiers
• 0.1.17: Support creation/deletion of stages; add IS_ARRAY function; remove DuckDB based DB engine; refactor codebase to use QueryProcessor interface; enhance proper handling of column names for table aliases
• 0.1.16: Add support for SHOW PROCEDURES and SHOW IMPORTED KEYS; add basic support for session parameters
• 0.1.15: Fix result type conversation for GET_PATH(..) util function
• 0.1.14: Enhance parity around SHOW TABLES/OBJECTS/COLUMNS queries; add more array util functions; fix STRING_AGG for different DISTINCT/GROUP combinations
• 0.1.13: Support some CURRENT_* functions; enhance LISTAGG for distinct values; add test for JS UDFs with exports
• 0.1.12: Cast params for string_agg/listagg; more parity fixes for upper/lowercase names
• 0.1.11: Enhance parity for array aggregation functions; enhance parity around subqueries and timestamp timezones; add logic to keep track of case-sensitive db/table identifiers
• 0.1.10: Add query transforms for CLUSTER BY; add SF_S3_ENDPOINT config; more parity fixes
• 0.1.9: Add support for Python UDFs; enhance parity around CREATE OR REPLACE FUNCTION queries; add analytics setup
• 0.1.8: Add SF_LOG config to enable request/response trace logging
• 0.1.7: Add initial support for Snowflake JavaScript UDFs; enhance parity around responses for DB/table creation; enhancements for Snowflake streaming logic
• 0.1.6: Introduce session state to retain DB/schema across queries; support async queries and result_scan(..)
• 0.1.5: Enhance parity around DESCRIBE TABLE results; support MIN_BY/MAX_BY aggregate functions
• 0.1.4: Add logic to parse and replace DB references in queries
• 0.1.3: Add DBEngine abstraction, experimental support for duckdb; enhance support for JSON queries
• 0.1.2: Add logic to ingest a CSV file from a Snowflake stage into a table
• 0.1.1: Initial support for Kafka connector and snowpipe/streaming APIs
• 0.1.0: Initial release of the extension

License

(c) 2023 LocalStack