fix(migrate-ts): store long index names as Postgres keeps them in every port by dmealing · Pull Request #581 · metaobjectsdev/metaobjects

dmealing · 2026-11-14T12:40:28Z

What Changed

migrate-ts now cuts index names over Postgres's 63-byte limit to the stored name, for declared index.lookup / identity.secondary names, @unique names and derived foreign-key indexes (<table>_<columns>_idx). The diff matches foreign keys by the stored name, so a converged schema proposes nothing on the next run. SQLite and D1 keep the whole name; MySQL is unchanged. Generated Drizzle, EF Core and Exposed mappings declare the same cut name on Postgres, while the names artifacts keep the declared name. The shared fixtures/foreign-key-index-conformance/ corpus gates the foreign-key rule and the declared.<dialect>.json names for TypeScript, C# and Kotlin.
The Java verify goal now runs the same model collision refusals as generate (refuseModelCollisions), so a model that generate refuses no longer verifies clean. ADR-0060 records verify.handEdits as a Node CLI setting by design, since only TypeScript merges hand edits forward.
Adds the release/<major>.<minor> maintenance-branch rule for patch releases (scripts/lib/release-branch.mjs, used by release.mjs and finish-release.mjs). CHANGELOG, docs/CONFORMANCE.md, the port pages and the migrations page are corrected to match the code.

Also in this PR

Forward-ports PR fix(csharp): make frozen metadata cache thread-safe; allow patch releases from release branches #578 (merged into release/1.2) to main: scripts/release.mjs and scripts/finish-release.mjs accept a release/<major>.<minor> maintenance branch and refuse a version older than the newest release, with its docs/RELEASING.md section and tests. docs/RELEASING.md also gains a checklist gate: before a minor is cut from main, every commit on the previous release/<major>.<minor> branch must be an ancestor of main.
The site-payload gate (scripts/ci-local.sh) runs bun test with --timeout 60000, so the site payload test no longer times out under load.

Issues

Closes #545
Closes #548

Refs #574: the migrations page now states what the code does. When meta migrate drops and recreates a Postgres view, it grants the view's existing privileges again; those are the only GRANT statements it writes, and it declares no privileges of its own. This PR does not close #574.

The raw evidence logs from validation are left out of this body; the scenario table below summarises them. Validation took two no-mistakes runs: the second added the generator-side cut of long declared index names.

Validation

Risk Assessment

✅ Low: The declared-index-name cut is applied consistently in migrate-ts, the Drizzle template, the C# EF context and the Kotlin Exposed generator, and a shared corpus with per-port runners pins it. I found no concrete failing input.

Testing

I ran the real meta CLI against a disposable Postgres 16 container with a model declaring a 74-byte index.lookup and a 78-byte identity.secondary. meta migrate --apply created both indexes under 63-byte names, and meta verify --db then showed no schema drift. The Drizzle tables generated for the same model declared exactly those 63-byte names on Postgres and the whole names on SQLite. The C# and Kotlin corpus tests, the Java verify-mojo refusal test, the full migrate-ts and codegen-ts bun suites (5074 pass, 0 fail) and scripts/ci-local.sh --only gates --only java-fast --only csharp all passed. The C# integration-test lane was cut off by my 560 s timeout (exit 124) and did not finish, so its result is unknown. The configured baseline (ts-fast, ts-unit) had already passed. The container, scratch project and temp files are removed, and the worktree is clean.

Live validation: ✅ go - 6 of 6 scenarios driven live against the product

Scenario	Result	Live	Evidence
Live Postgres: `meta migrate --apply` of a 74-byte index.lookup and 78-byte identity.secondary creates them at 63 bytes, and `meta verify --db` reports no drift	✅ pass	live	live-postgres-vs-drizzle.txt; `meta verify --db` returned schemaDrift changes=[] findings=[]
Generated Drizzle table declares the cut 63-byte name on Postgres and the whole name on SQLite, matching the live database	✅ pass	live	live-postgres-vs-drizzle.txt: the Drizzle postgres names equal the pg_indexes names byte for byte; the sqlite names are 74 and 78 bytes
C# EF Core context declares the same cut names as the shared declared.<dialect>.json corpus	✅ pass	live	csharp-fkindex.log: 71 passed, 0 failed
Kotlin Exposed table declares the same cut names as the shared corpus (SQLite whole, Postgres cut)	✅ pass	live	kotlin-fkindex.log and surefire reports: 5 + 4 tests, 0 failures
Adversarial: Java/Kotlin verify mojo refuses a layout collision (package acme::Sales vs acme::sales) the way generate does	✅ pass	live	mojo-verify.log: MetaDataVerifyModelRefusalTest 2 pass
Regression: full migrate-ts and codegen-ts suites, gates, and csharp, java and kotlin conformance	✅ pass	live	ts-migrate-codegen.log: 5074 pass, 0 fail; ci-local log: gates and csharp, java and kotlin conformance green

Pipeline

Updates from git push no-mistakes

✅ **Intent** - passed

✅ No issues found.

✅ **Rebase** - passed

✅ No issues found.

✅ **Review** - passed

✅ No issues found.

✅ **Test** - passed

✅ No issues found.

Live validation: ✅ go - 6 of 6 scenarios driven live against the product

Scenario	Result	Live	Evidence
Live Postgres: `meta migrate --apply` of a 74-byte index.lookup and 78-byte identity.secondary creates them at 63 bytes, and `meta verify --db` reports no drift	✅ pass	live	live-postgres-vs-drizzle.txt; `meta verify --db` returned schemaDrift changes=[] findings=[]
Generated Drizzle table declares the cut 63-byte name on Postgres and the whole name on SQLite, matching the live database	✅ pass	live	live-postgres-vs-drizzle.txt: the Drizzle postgres names equal the pg_indexes names byte for byte; the sqlite names are 74 and 78 bytes
C# EF Core context declares the same cut names as the shared declared.<dialect>.json corpus	✅ pass	live	csharp-fkindex.log: 71 passed, 0 failed
Kotlin Exposed table declares the same cut names as the shared corpus (SQLite whole, Postgres cut)	✅ pass	live	kotlin-fkindex.log and surefire reports: 5 + 4 tests, 0 failures
Adversarial: Java/Kotlin verify mojo refuses a layout collision (package acme::Sales vs acme::sales) the way generate does	✅ pass	live	mojo-verify.log: MetaDataVerifyModelRefusalTest 2 pass
Regression: full migrate-ts and codegen-ts suites, gates, and csharp, java and kotlin conformance	✅ pass	live	ts-migrate-codegen.log: 5074 pass, 0 fail; ci-local log: gates and csharp, java and kotlin conformance green

scripts/ci-local.sh --only ts-fast --only ts-unit --strict-toolchains
meta migrate --db <postgres> --slug init --apply on a model with 74-byte index.lookup and 78-byte identity.secondary names; psql pg_indexes showed both at 63 bytes
meta verify --db <postgres> on the same database: schemaDrift changes=[] and findings=[]
Drizzle entityFile generated for the same model, postgres and sqlite dialects, compared with the live index names
bun test migrate-ts foreign-key-index-conformance, long-declared-index-name and codegen-ts declared-index-name-cut (12 pass)
dotnet test MetaObjects.Codegen.Tests --filter ForeignKeyIndex|EjectedGeneratorsAlone (71 pass)
mvn -o -pl codegen-kotlin -am test -Dtest=KotlinForeignKeyIndexConformanceTest,KotlinRouteNamingTest (9 pass)
mvn -o -pl maven-plugin -am test -Dtest=MetaDataVerifyModelRefusalTest (2 pass)
bun test packages/migrate-ts packages/codegen-ts in server/typescript (5074 pass, 0 fail)
scripts/ci-local.sh --only gates --only java-fast --only csharp --strict-toolchains: gates, csharp, java and kotlin conformance green; cut off in the csharp integration lane by my timeout, no failure recorded
baseline already run: scripts/ci-local.sh --only ts-fast --only ts-unit --strict-toolchains

✅ **Document** - passed

✅ No issues found.

✅ **Lint** - passed

✅ No issues found.

✅ **Push** - passed

✅ No issues found.

…t generate refuses - migrate.foreignKeyIndexes: a <table>_<columns>_idx name over 63 bytes is cut to the 63 bytes Postgres stores (SQLite and D1 keep it whole), in migrate-ts, the C# EF context and the Kotlin Exposed tables, so a converged schema proposes nothing. The diff also keys a Postgres foreign key by its stored name, so a long <table>_<column>_fk is not proposed as a rename every run. - New shared corpus fixtures/foreign-key-index-conformance/, passed by the TypeScript, C# and Kotlin rules; two Postgres migration scenarios with a long table and column name. - mvn metaobjects:verify runs the generate goal's model refusals (output package clash, collection-name clash), with a test; Kotlin gets its own collection-name clash test. - docs/CONFORMANCE.md: the clash is per-port unit tests, not the segments corpus; the Drizzle identity start is string-tested only. - ADR-0060 records verify.handEdits as Node-CLI only by design. - The migration docs say what the code does with view grants (#574).

…e a minor on the previous one being merged Forward-ports #578 from release/1.2: release.mjs and finish-release.mjs accept a maintenance branch for a patch on its own line, refuse a version that would move npm's latest backwards, and RELEASING documents the path. A minor or major cut is now refused while the previous release/<M>.<m> holds a commit that is not on main, and RELEASING Step 0 carries the same check.

# Conflicts: # CHANGELOG.md # docs/CONFORMANCE.md

… rules; site gate timeout; ts-poet names - docs/features/api-contract.md owns api.pagination, api.unknownFields and the validation issue list; language-differences, the five port pages and the CHANGELOG point to it instead of restating them. - The site-payload gate runs its tests with a 60 s per-test timeout, so a busy machine does not read as a red release gate. - CHANGELOG names def, literalOf, arrayOf, asOneline and Import.fromMaybeString among the ts-poet names the vendored printer does not carry. - site-payload.json counts the new shared foreign-key-index corpus.

…stores it meta migrate creates and compares a declared index.lookup / identity.secondary name over 63 bytes under the name Postgres stores (its first 63 bytes). The Drizzle tables, the EF Core context and the Exposed tables now declare that cut name on Postgres (SQLite and D1 keep the whole name; MySQL is exempt), so the mapping names the index the database holds. The names artifacts keep the declared, dialect-neutral name. The shared corpus fixtures/foreign-key-index-conformance/ gains a long index.lookup and identity.secondary and declared.<dialect>.json, written by migrate-ts and read by the TypeScript, C# and Kotlin generator tests. C# ForeignKeyIndexes joins the eject engine surface, as the schema contract CheckConstraints already is.

dmealing added 9 commits November 13, 2026 22:11

Merge remote-tracking branch 'origin/main' into fm/mo-main-followups

a7783fc

# Conflicts: # CHANGELOG.md # docs/CONFORMANCE.md

no-mistakes(review): Restore changelog bullets; cut long index names

8d59242

no-mistakes(review): Rename case-only layout-collision test resources

a5ef614

no-mistakes(document): docs: document the declared-index name cut

01600bb

no-mistakes(document): docs: stored index names and verify-goal refusals

925b508

dmealing changed the title ~~fix: cut long foreign-key index names to what Postgres stores; verify refuses what generate refuses~~ fix(migrate-ts): store long index names as Postgres keeps them in every port Nov 14, 2026

dmealing merged commit 059563c into main Nov 14, 2026
9 checks passed

dmealing deleted the fm/mo-main-followups branch November 14, 2026 15:14

Navigation Menu

Repository navigation

fix(migrate-ts): store long index names as Postgres keeps them in every port - #581

fix(migrate-ts): store long index names as Postgres keeps them in every port#581
dmealing merged 9 commits into
mainfrom
fm/mo-main-followups

dmealing commented Nov 14, 2026 •

edited

Loading

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

dmealing commented Nov 14, 2026 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

What Changed

Also in this PR

Issues

Risk Assessment

Testing

Pipeline

Uh oh!

Uh oh!

Reviewers

Uh oh!

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

dmealing commented Nov 14, 2026 •

edited

Loading