SYSTEM Cited by 1 source

GoogleSQL¶

GoogleSQL is the SQL dialect shared across Google's data systems — BigQuery, F1, Spanner, and other internal Google query engines. Rather than each engine defining its own dialect, GoogleSQL is a single language spec consumed by multiple planners. Dialect-level features (new operators, new syntax, new semantics) therefore propagate across the whole Google query-engine fleet without per-engine rollout.

Why it matters on the wiki¶

Production precedent for extending — not replacing — an entrenched query language to fix its known design problems. The 2024-08-24 paper Pipe Syntax in SQL describes native |> pipe-syntax support added to GoogleSQL, positioned as the canonical wiki production instance of concepts/language-extension-vs-replacement and patterns/pipe-syntax-query-language.
Cross-engine dialect governance. A single dialect spec consumed by BigQuery + F1 + Spanner lets one grammar change ship across the whole stack. This is the same architectural lever as single-source service definitions or universal resource provisioning — one authoritative spec, many downstream executors.

Known features¶

Standard SQL surface. GoogleSQL is ANSI-SQL-compatible at the core: SELECT … FROM … WHERE … GROUP BY … HAVING … ORDER BY with the usual clause-order-vs-evaluation-order inversion that the paper flags as a learnability problem.
Pipe syntax extension (|>). Additive extension layered on top of classic SQL, not a replacement — see patterns/pipe-syntax-query-language and the worked discussion on the source page. Every pipe stage is a named existing operator (WHERE, SELECT, AGGREGATE, ORDER BY, JOIN, EXTEND, ...) applied in data-flow order.
Shared across engines. BigQuery, F1, Spanner are the named hosts; additional Google data systems consume GoogleSQL but are not enumerated in the abstract.

Adjacent systems¶

BigQuery — Google's serverless analytics data warehouse; primary public-facing GoogleSQL surface.
F1 — Google's distributed SQL database, historically the backbone of AdWords.
Spanner — Google's globally-distributed transactional database; exposes GoogleSQL as one of its query surfaces.

These engines have no wiki pages yet; they are named here for orientation and may get dedicated pages as future Google sources cover them directly.

Caveats¶

No public dialect spec ingested. The wiki currently cites GoogleSQL via the 2024-08-24 pipe-syntax paper only — a language-design paper, not a dialect reference. Feature coverage here is therefore narrow.
Propagation to other engines. Whether pipe syntax (or other GoogleSQL extensions) propagates to ANSI SQL or to non-Google engines (Postgres, MySQL, Snowflake, DuckDB, ...) is an open downstream question unanswered by the paper.

Seen in¶

sources/2024-08-24-google-pipe-syntax-in-sql — 2024-08-24 Google Research paper describing pipe syntax as a native GoogleSQL extension; dialect-level change propagating across BigQuery, F1, Spanner.

concepts/language-extension-vs-replacement — the design stance GoogleSQL's pipe-syntax extension embodies.
patterns/pipe-syntax-query-language — the query-language shape.
systems/postgresql, systems/mysql — SQL engines with their own dialects; structural counterexamples where one dialect governs one engine family rather than a whole query-engine fleet.