GoogleSQLite is a Go project that runs GoogleSQL — the SQL dialect used by BigQuery and Cloud Spanner — on top of a SQLite backend. It is exposed as a database/sql driver, so any database/sql consumer gets a local GoogleSQL execution engine with no external services to run.

GoogleSQL parsing and analysis is provided by goccy/go-googlesql, and query execution runs on ncruces/go-sqlite3. Both are implemented in pure Go, so GoogleSQLite is pure Go as well — no cgo is required.

Because GoogleSQL is the dialect behind BigQuery and Cloud Spanner, this project gives strong support to use cases such as locally emulating those services.

import (
	"database/sql"

	_ "github.com/goccy/googlesqlite"
)

db, err := sql.Open("googlesqlite", ":memory:")
if err != nil {
	log.Fatal(err)
}
defer db.Close()

rows, err := db.Query(`SELECT FORMAT('%t', DATE '2026-05-14') AS today`)
// ...

The driver registers under googlesqlite. DSN is the underlying SQLite path (:memory:, a file path, or file:foo.db?cache=shared).

Try GoogleSQLite in your browser, with nothing to install:

https://goccy.github.io/googlesqlite/

The Playground is the whole engine — GoogleSQL parsing, analysis, and the SQLite backend — compiled to WebAssembly and run entirely client-side. Highlights:

• No server, no install. Every query is parsed, analyzed, and executed inside the browser tab; nothing is sent anywhere.
• Responsive UI. The engine runs in a Web Worker, so editing and running queries never blocks the page.
• Persistent database. The SQLite database lives in the Origin Private File System (OPFS), so your tables and data survive page reloads with no explicit save step.
• A real SQL editor, with query history and result export.

It is the quickest way to explore the supported GoogleSQL syntax and functions without setting up a Go environment.

Every function and type is backed by a declarative spec under docs/specs/ and at least one upstream-sourced test case under testdata/specs/. The full support matrix — per-function and per-type status with a link to each spec — is generated from spec frontmatter into docs/specs/INDEX.md.

Specs marked partial are documented gaps with a concrete infrastructure dependency; see each spec's frontmatter notes: field for what is missing.

make build            # build the bin/googlesqlite CLI
make test             # run unit and spec tests
make lint             # golangci-lint via tools/go.mod
make spec/upstream-sync   # refresh docs/third_party/googlesql-docs from upstream

• docs/specs/ — canonical, normalized spec markdown (one file per function/type).
• testdata/specs/ — declarative YAML test cases referenced by spec frontmatter. The repo-root TestSpec runner executes each case against the driver as a Go subtest.
• docs/third_party/googlesql-docs/ — Apache-2.0 snapshot of upstream google/googlesql docs/. Refreshed by make spec/upstream-sync.
• cmd/specctl/ — operational CLI (normalize, upstream-sync, check, ...).
• tools/go.mod — pinned developer tools (golangci-lint).

This is a personal project. It is developed by referencing the open-source GoogleSQL repository, but it receives no support of any kind from Google — no sponsorship, no contributions, no promotion.

For example, Google has had a request for a BigQuery emulator open on its Issue Tracker (issue 129248927) for seven years without taking any action. Unable to watch that any longer, in 2022 I built goccy/bigquery-emulator — the only BigQuery emulator in the world — and have maintained it ever since.

This project is used by that bigquery-emulator. I do not use BigQuery in my own job, however. I simply happen to have the skills to build this, noticed how many people are struggling without it, and so I spend my personal time and money on it.

Because of that, keeping this project alive needs your help. Emulating BigQuery locally brings many benefits to development and can significantly cut development costs. Could you return a small part of those savings to me? I believe doing so leads to a better future for both me and your company.

See CLAUDE.md for the project's working rules.