Repository navigation
feat(spec)!: return a statement's result as a rows type - #184
Open
colinhacks wants to merge 2 commits into
Open
colinhacks wants to merge 2 commits into
colinhacks wants to merge 2 commits into
Conversation
|
The latest updates on your projects. Learn more about Vercel for GitHub.
|
Contributor
There was a problem hiding this comment.
ℹ️ No critical issues — three small rough edges inline.
Reviewed changes
Reviewed the move of a statement descriptor's columns and cardinality into one result of the new rows kind, everywhere the format is written, read and documented.
- Format —
RowsType(cardinality,columns, optionalkeyandordered: true) is appended toType, andStatementDescriptor.resultrefers to it; the schema, the generatedspec/index.tsand the published0.1.0.jsonagree. The version stays0.1.0, and that holds up: thereleasebranch has nospec/typedefs.schema.json, so no released build has served this version. - Engine —
Kind::Rowsis handled inrank,name,collect_names,renamed,short,check_atandjson_schema.Analysis::descriptornow carries thekeyandorderedthe analysis already computed, andreylo analyze's line for a failed statement uses the new shape. - Run time —
descriptor()andrefused()insrc/rtd.tswrite the same shape.withBasesandcolumnKindhandlerows. The.reylo/cache still holdsAnalysis, withcolumnsandcardinalityat the top, solab/compareand the playground, which read that cache, are unaffected. - Tests and docs — every reader of
reylo analyzeoutput is moved toresult.*, includinginstall-check.ts. A new Postgres check showskeyandorderedreachingreylo/rtd. The example and field table inrtd.mdxare updated.
claude-opus-5-5 ｜ 𝕏
Contributor
There was a problem hiding this comment.
✅ No new issues found. The new commit fixes all three points from the last review.
Reviewed changes
This run covered 64cc401, which tightens how Rust types
8000
a statement's result and adds checks on its key.
- Typed
result:StatementDescriptor::resultis nowResultRows, which serde reads and writes as aType. Deserialization refuses any kind other thanrows, as the schema does. The JSON is unchanged, and a new core test checks both directions. - Key rule check:
check_atforKind::Rowsnow refuses akeythat names a column the result doesn't have. It also refuses akeyon rows whose cardinality isn'tmany. This matches what the engine emits: Postgres'sresult_keyreturnsNoneunless the result ismanyrows, and the other engines writekey: None. - Docs: the
resultrow inrtd.mdxnow says whatkeyandorderedmean.
claude-opus-5-5 ｜ 𝕏
A result of another kind now fails to deserialize, as it fails the schema, and the rule check holds a key to its columns and cardinality. The rtd page says what key and ordered mean.
colinhacks
force-pushed
the
typedefs-rows
branch
from
November 5, 2026 18:47
64cc401 to
6b19b18
Compare
Contributor
There was a problem hiding this comment.
✅ No new issues found. The rebase onto
maincarries the change over cleanly.
Reviewed changes
Since 64cc401, the branch was rebased onto main after #191, which renamed reylo analyze to reylo compile. The PR's own changes now sit on the renamed files.
- Renamed paths: the
resultshape now lives insrc/compile.ts(descriptor()andrefused()),engine/cli/src/compile.rs,tests/compile-cli.test.ts,tests/compile.test.ts,tests/cli.test.ts,tests/connect.test.tsand.github/scripts/install-check.ts. Every reader there usesresult.columnsandresult.cardinality. I found no remaining reader of the old top-level fields insrc/, the skill or the docs. - Core: the
Kind::Unionarm inrenamednow matches theUnionTypefrommain, which no longer has a second field.ResultRowsand thekeychecks from the previous commit are unchanged. - Docs: the example in
08c-api.mdxuses therowsshape. The page onmainhas no field table, sokeyandorderedare explained only in the JSON Schema it links to. - Format text: the
StatementDescriptordescription in the schema,spec/index.tsand the published0.1.0.jsonnow namesreylo compile.
claude-opus-5-5 ｜ 𝕏
This branch has not been deployed
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
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.
A statement descriptor's
columnsandcardinalitybecome oneresult, of a new kind for tabular data,rows, which also carries the two facts Reylo infers about a result and the format didn't hold: its key and whether its order is total.{ "$schema": "https://reylo.dev/typedefs/0.1.0.json", "dialect": { "name": "postgres", "version": "18.6" }, "sql": "select id, email from users order by id", "kind": "select", "params": [], "result": { "kind": "rows", "cardinality": "many", "columns": [ { "name": "id", "type": { "kind": "integer", "bits": 64, "name": "int8" } }, { "name": "email", "type": { "kind": "text", "size": { "max": 255 }, "name": "varchar" } } ], "key": ["id"], "ordered": true }, "diagnostics": [] }RowsTypeiskind: "rows"withcardinality,columns, and where they're knownkeyandordered. It is appended toType, since a relation nested in a value (SQL'sMULTISET, an Oracle nested table) is the same shape; no engine writes a nested one yet. A statement that returns one value is still a row of one column, which keeps the column's name, as.run()does.spec/typedefs.schema.json, thennub run typedefsforspec/index.tsand the published0.1.0.json. The version stays 0.1.0: no release has published it.Kind::RowsandStatementDescriptor::resultinengine/core/src/types.rs;Analysis::descriptorandreylo analyze's line for a failed statement write it. What the engine answers the loader is unchanged, so no declaration and no.run()result changes: only whatreylo analyzeandreylo/rtdreturn.descriptor()andrefused()insrc/rtd.tswrite the same shape, andtests/rtd.test.tsstill holds them toreylo analyzefield by field, with a new check thatkeyandorderedreach it on Postgres.install-check.ts, which reads a descriptor.Verified
cargo build --workspace --all-targetswith no warning; every crate's tests.nub run typecheck, andnub run testagainst Postgres 18 and MySQL 8.4 on the release binary: 352 passed, none skipped.nub run buildinsite/.