SQLGlot C Extensions

30.4.2 · active · verified Sun Apr 12

sqlglotc provides performance-enhancing extensions for the sqlglot library, compiled using mypyc. SQLGlot is a robust, no-dependency Python library for parsing, transpiling, optimizing, and executing SQL across numerous dialects like DuckDB, Presto, Spark, Snowflake, and BigQuery. It aims to accurately process diverse SQL inputs and generate syntactically and semantically correct SQL in target dialects. The library is actively maintained with a rapid release cadence, currently at version 30.4.2.

Warnings

breaking Minor version updates of `sqlglot` (and thus `sqlglotc`) can introduce backwards-incompatible changes in SQL parsing or transpilation logic, especially concerning specific dialect behaviors.
Fix: Always pin `sqlglot` (and `sqlglotc`) to specific versions in production. Test upgrades thoroughly before deploying. Consult the `sqlglot` release notes for details on changes between versions.
gotcha Failing to explicitly specify the source (`read`) and target (`write`) SQL dialects when using `sqlglot.transpile` or `sqlglot.parse_one` can lead to incorrect parsing or output. `sqlglot` defaults to its 'SQLGlot dialect' if not specified.
Fix: Always provide the `read` and `write` (or `dialect` for `parse_one`) arguments to ensure correct interpretation and generation of SQL for your specific dialect.
gotcha If pre-built wheels for `sqlglotc` are not available for your specific platform or Python version, installation will attempt to build from source. This requires `mypyc` and a compatible C compiler (e.g., GCC or Clang) to be present on your system.
Fix: Ensure you have a C compiler and the `mypyc` package installed (e.g., `pip install mypyc`) if encountering build errors during `sqlglotc` installation.
gotcha `sqlglot` primarily acts as a parser, transpiler, and optimizer, not a full SQL validator. While it detects many syntax errors, it may not catch all semantic or dialect-specific validation issues.
Fix: Use `sqlglot` for transformation and analysis, but do not rely on it as a comprehensive SQL validator for all possible error conditions. Supplement with actual database execution or dedicated linting/validation tools.

Install

pip install sqlglotc Install sqlglotc directly
pip install "sqlglot[c]" Install sqlglot with C extensions

Imports

transpile
```
from sqlglot import transpile
```
sqlglotc enhances sqlglot's performance transparently; users typically interact with sqlglot's API directly.
parse_one
```
from sqlglot import parse_one
```
sqlglotc enhances sqlglot's performance transparently; users typically interact with sqlglot's API directly.

Quickstart

This quickstart demonstrates how to use the core `sqlglot` functionalities, `transpile` and `parse_one`, which are transparently accelerated by `sqlglotc` if installed. It shows converting a simple SQL query between dialects and parsing a query into an abstract syntax tree.

import sqlglot

# Example: Transpile SQL from DuckDB to Hive dialect
duckdb_sql = "SELECT EPOCH_MS(1618088028295)"
hive_sql = sqlglot.transpile(duckdb_sql, read='duckdb', write='hive')[0]

print(f"Original DuckDB SQL: {duckdb_sql}")
print(f"Transpiled Hive SQL: {hive_sql}")

# Example: Parse a SQL query into an expression tree
expression = sqlglot.parse_one("SELECT id, name FROM users WHERE age > 18")
print(f"Parsed Expression: {expression}")

view raw JSON →