fix(sql): accept the OUTER keyword in LEFT/RIGHT/FULL OUTER JOIN#96
Merged
Conversation
Closes #94. ANSI SQL allows the optional `OUTER` keyword after LEFT / RIGHT / FULL — `LEFT OUTER JOIN` and `LEFT JOIN` are the same. The previous regexes on each JoinType (`Expr("LEFT")` etc.) matched only the bare keyword, so the standard `FULL OUTER JOIN` / `LEFT OUTER JOIN` / `RIGHT OUTER JOIN` forms failed to parse, breaking any BI tool or ORM that emits the explicit OUTER syntax. Fix at the model level: each of `LeftJoin`, `RightJoin`, `FullJoin` now overrides `words` to list the longer `<keyword>\s+OUTER` form first and the bare keyword second. `TokenRegex.regex` builds `(?i)(LEFT\s+OUTER|LEFT)\b`, and regex alternation prefers the longest-first match — so both surface forms parse into the same AST node. This is intentionally a model-side fix rather than a parser-side fix so the regex is reused everywhere `LeftJoin.regex` / `RightJoin.regex` / `FullJoin.regex` is referenced. The `sql` value used for AST round-trip rendering stays on the canonical short form, so `Update.sql` / `Join.sql` keep emitting `LEFT JOIN` rather than `LEFT OUTER JOIN`. Three new ParserSpec tests assert that the short and OUTER forms produce identical AST nodes (`LeftJoin`, `RightJoin`, `FullJoin`). Full ParserSpec — 192 tests — still green. `+ sql/compile` clean under 2.12 and 2.13.
…ialized views
Closes softclient4es-extensions#19's user-facing documentation gap.
The MV engine now rejects RIGHT JOIN and FULL OUTER JOIN with an
actionable error message; this commit documents the constraint so
users hit it once, in the docs, instead of at CREATE MATERIALIZED
VIEW time.
Both copies are updated to stay in sync:
- documentation/sql/materialized_views.md (this repo)
- softclient4es-web/src/content/docs/sql/materialized-views.mdx
(published site)
Each gets:
- A new row in the Limitations table flagging RIGHT JOIN and
FULL OUTER JOIN as unsupported.
- A new "Supported JOIN types" subsection explaining the underlying
architectural reason — the ingest pipeline is driven by writes to
the main (left-hand) FROM table and can only enrich into the
main-table document via EnrichProcessor — and the workaround
(rewrite RIGHT JOIN with swapped table order; FULL OUTER JOIN is
simply not expressible in this model).
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
ANSI SQL allows the optional
OUTERkeyword after LEFT / RIGHT / FULL—
LEFT OUTER JOINandLEFT JOINare the same. The previous regexeson each JoinType (
Expr("LEFT")etc.) matched only the bare keyword,so the standard
FULL OUTER JOIN/LEFT OUTER JOIN/RIGHT OUTER JOINforms failed to parse, breaking any BI tool or ORMthat emits the explicit OUTER syntax.
Fix at the model level: each of
LeftJoin,RightJoin,FullJoinnow overrides
wordsto list the longer<keyword>\s+OUTERformfirst and the bare keyword second.
TokenRegex.regexbuilds(?i)(LEFT\s+OUTER|LEFT)\b, and regex alternation prefers thelongest-first match — so both surface forms parse into the same
AST node.
This is intentionally a model-side fix rather than a parser-side fix
so the regex is reused everywhere
LeftJoin.regex/RightJoin.regex/
FullJoin.regexis referenced. Thesqlvalue used for ASTround-trip rendering stays on the canonical short form, so
Update.sql/Join.sqlkeep emittingLEFT JOINrather thanLEFT OUTER JOIN.Closes #94.