TML-2916: un-namespaced PG models default to public namespace per ADR 223#838
Merged
Merged
Conversation
Contributor
|
No actionable comments were generated in the recent review. 🎉 ℹ️ Recent review info⚙️ Run configurationConfiguration used: Path: .coderabbit.yml Review profile: CHILL Plan: Pro Run ID: ⛔ Files ignored due to path filters (19)
📒 Files selected for processing (125)
✅ Files skipped from review due to trivial changes (73)
🚧 Files skipped from review as they are similar to previous changes (44)
📝 WalkthroughWalkthroughUn-namespaced models now resolve to ChangesConditional namespace defaults and regenerated artifacts
Estimated code review effort🎯 4 (Complex) | ⏱️ ~60 minutes Possibly related PRs
Suggested reviewers
Poem
✨ Finishing Touches📝 Generate docstrings
🧪 Generate unit tests (beta)
|
@prisma-next/extension-author-tools
@prisma-next/mongo-runtime
@prisma-next/family-mongo
@prisma-next/sql-runtime
@prisma-next/family-sql
@prisma-next/extension-arktype-json
@prisma-next/middleware-cache
@prisma-next/mongo
@prisma-next/extension-paradedb
@prisma-next/extension-pgvector
@prisma-next/extension-postgis
@prisma-next/postgres
@prisma-next/sql-orm-client
@prisma-next/sqlite
@prisma-next/extension-supabase
@prisma-next/target-mongo
@prisma-next/adapter-mongo
@prisma-next/driver-mongo
@prisma-next/contract
@prisma-next/utils
@prisma-next/config
@prisma-next/errors
@prisma-next/framework-components
@prisma-next/operations
@prisma-next/ts-render
@prisma-next/contract-authoring
@prisma-next/ids
@prisma-next/psl-parser
@prisma-next/psl-printer
@prisma-next/cli
@prisma-next/cli-telemetry
@prisma-next/emitter
@prisma-next/migration-tools
prisma-next
@prisma-next/vite-plugin-contract-emit
@prisma-next/mongo-codec
@prisma-next/mongo-contract
@prisma-next/mongo-value
@prisma-next/mongo-contract-psl
@prisma-next/mongo-contract-ts
@prisma-next/mongo-emitter
@prisma-next/mongo-schema-ir
@prisma-next/mongo-query-ast
@prisma-next/mongo-orm
@prisma-next/mongo-query-builder
@prisma-next/mongo-lowering
@prisma-next/mongo-wire
@prisma-next/sql-contract
@prisma-next/sql-errors
@prisma-next/sql-operations
@prisma-next/sql-schema-ir
@prisma-next/sql-contract-psl
@prisma-next/sql-contract-ts
@prisma-next/sql-contract-emitter
@prisma-next/sql-lane-query-builder
@prisma-next/sql-relational-core
@prisma-next/sql-builder
@prisma-next/target-postgres
@prisma-next/target-sqlite
@prisma-next/adapter-postgres
@prisma-next/adapter-sqlite
@prisma-next/driver-postgres
@prisma-next/driver-sqlite
commit: |
size-limit report 📦
|
wmadden
approved these changes
Jun 16, 2026
… 223
The contract authoring pipeline injected an empty `__unbound__` storage
namespace into every Postgres contract, in violation of ADR 223 — which
makes `target.defaultNamespaceId` ("public" for Postgres) the static,
target-owned default. Two sites are guarded on `defaultNamespaceId ===
UNBOUND_NAMESPACE_ID` so only SQLite and Mongo (whose default is the
unbound sentinel) keep the slot:
1. `build-contract.ts` — the `ensureUnboundNamespaceSlot` call site at
the end of `buildSqlContractFromDefinition` (the build path both the
PSL interpreter and the TS DSL converge on).
2. `sql-contract-serializer-base.ts` — `hydrateSqlStorage` previously
re-injected the slot on every deserialize round-trip, defeating the
build-time fix during emit. An `abstract get defaultNamespaceId():
string` was added to the base class and implemented in the PG, SQLite,
and Mongo (`SqlContractSerializer` fallback) subclasses.
Regenerates the affected fixtures and baked extension-pack contracts
(pgvector, paradedb, postgis, supabase, audit, feature-flags) to drop
the spurious `__unbound__` slot. Updates the contract-builder.namespaces
tests that encoded the bug as expected behavior. Adds an IR-level test
in interpreter.namespaces that pins the un-namespaced PG → public
invariant end-to-end.
The diamond fixture chain under examples/prisma-next-demo/fixtures/* is
not regenerated by this PR: it has no per-step contract.prisma source
so the regen script cannot rebuild it, and a `grep` of the codebase
shows zero consumers. Left stale; follow-up to add a regen mechanism.
Extends `regen-extension-migrations.mjs` to cover the
multi-extension-monorepo sub-extension packages (audit, feature-flags)
that also carry on-disk migrations + head refs.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Signed-off-by: willbot <w.a.madden+machine@gmail.com>
Signed-off-by: Will Madden <madden@prisma.io>
…struction declarations - postgres-contract-serializer.test.ts: RegistryDispatchProbeSerializer needs to implement the new abstract defaultNamespaceId getter. - skills/upgrade + skills/extension-author 0.13-to-0.14/instructions.md: add TML-2916 incidental-substrate-diff markers so per-pr-declaration CI gate sees the upgrade-instruction files were touched. Signed-off-by: willbot <w.a.madden+machine@gmail.com> Signed-off-by: Will Madden <madden@prisma.io>
…-2918 Post-rebase regen: TML-2918 (#837) landed schema-qualified op ids (`column.${schema}.${table}.${column}`) for the PG AddColumn op. The prisma-next-demo migration ops.json files now carry `column.public.*` ids; migrationHash updates accordingly. Signed-off-by: willbot <w.a.madden+machine@gmail.com> Signed-off-by: Will Madden <madden@prisma.io>
wmadden-electric
force-pushed
the
tml-2916-un-namespaced-pg-defaults-to-public
branch
from
85695a1 to
99d8dc6
Compare
wmadden-electric
added a commit
that referenced
this pull request
Jun 18, 2026
Main shipped the 0.14.0 publish bump (#843) and several follow-ups (TML-2916 #838, TML-2503 #839, TML-2787 #683, TML-2893 #849, ci-fix #847) while slice 1 was in review. Merge to pick them up. Resolution: - `skills/extension-author/.../0.13-to-0.14/instructions.md` — union of the slice's TML-2919 entry and main's TML-2916 entry as two distinct HTML-comment blocks, matching the file's per-ticket convention. Per the skill spec's rebase scenario (lines 122-130): existing entries in the just-shipped 0.13-to-0.14 directory may stay in place; we do not relocate the TML-2919 entry to 0.14-to-0.15. - Lockfile + example fixtures + apps + all other non-slice files — take main's side cleanly (the branch hadn't touched them; auto-merge resolved to main's content). `pnpm install` confirms the lockfile is consistent. - No `src/` conflicts — only the one instructions.md conflict above. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Signed-off-by: willbot <w.a.madden+machine@gmail.com> Signed-off-by: Will Madden <madden@prisma.io>
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
2 participants
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.
At a glance
Postgres contracts were emitting an empty
__unbound__storage namespace slot alongsidepublicon every contract — in violation of ADR 223 — Target-owned default namespace, which makestarget.defaultNamespaceId(="public"for Postgres) the static, target-owned default. Surfaced during the #825 (TML-2889) review.The decision
Two sites in the authoring/emit pipeline injected the spurious slot. Both are now gated on
defaultNamespaceId === UNBOUND_NAMESPACE_ID— descriptor-driven, notargetId === 'postgres'branch anywhere (per ADR 223). Only SQLite and Mongo (whose default IS unbound) keep the slot.packages/2-sql/2-authoring/contract-ts/src/build-contract.ts:748— end ofbuildSqlContractFromDefinition, the path both the PSL interpreter and the TS DSL converge onensureUnboundNamespaceSlotcall ondefaultNamespaceId === UNBOUND_NAMESPACE_ID.packages/2-sql/9-family/src/core/ir/sql-contract-serializer-base.ts—hydrateSqlStoragere-injected the slot on every deserialize round-trip, defeating site 1 during emitprotected abstract get defaultNamespaceId(): stringtoSqlContractSerializerBase; implemented in PG, SQLite, and theSqlContractSerializerfamily-level fallback (which preserves legacy behaviour for callers without a target descriptor). Mongo uses its own base class and is unaffected.Why two sites
D1 (diagnosis dispatch) identified site 1 only; the in-memory contract built via the TS DSL was already clean (only
public). D2's emit traced throughexecuteContractEmit→familyInstance.deserializeContract(enrichedIR)→hydrateSqlStorage, where the round-trip re-introduced the slot. Sites 1 and 2 are independent; both need the gate.What's regenerated
src/contract.json+.d.tslose the__unbound__slot; their migrationhead.jsonrefs andmigration.tshashes are rewritten bymigrations:regen.scripts/regen-extension-migrations.mjswas extended to walk these (they live underexamples/multi-extension-monorepo/packages/*, outside the script's originalpackages/3-extensions/walk).apps/telemetry-backend,examples/{bundle-size, paradedb-demo, prisma-next-cloudflare-worker, multi-extension-monorepo/app, prisma-next-demo, prisma-next-postgis-demo}— emitted contracts + migration chains lose the slot; migration hashes update.packages/2-sql/4-lanes/sql-builder/test/fixtures/generated/,packages/3-extensions/sql-orm-client/test/fixtures/generated/,test/e2e/framework/test/fixtures/generated/,test/integration/test/{fixtures,namespaced-accessors/fixtures/generated}/— same.What was NOT regenerated
The diamond + 6 sibling fixture chains under
examples/prisma-next-demo/fixtures/*/migrations/app/. The regen script (regen-example-migrations.mjs) requires a per-stepcontract.prismafor each migration directory to recompute the end-state contract; these chains store onlymigration.tsper step plus a single rootcontract.prismafor the final state. Agrep -rln "fixtures/diamond"(excludingnode_modules,dist,.json,.d.ts) returns zero hits outside the slice's own docs — no test or script reads these fixtures. Left stale on this branch; follow-up to add a regen mechanism or rewrite them.Tests
packages/2-sql/2-authoring/contract-psl/test/interpreter.namespaces.test.ts— a baremodel user { id String @id }interpreted against the PG target producesdomain.namespaces.public.models.user+storage.namespaces.public, with no__unbound__namespace. Goes red onmainwithout the fix at site 1.packages/2-sql/2-authoring/contract-ts/test/contract-builder.namespaces.test.tsthat asserted__unbound__was present in PG storage — flipped to assert post-fix shape with a// TML-2916marker.packages/3-extensions/pgvector/test/descriptor.test.ts:100asserted the install-vector pre/postcheck SQL contained literalextname = 'vector'. The regen produces parameterised SQL ("extname" = \$1+params: ["vector"]), a downstream effect of TML-2889's typed-AST routing of verification queries. Test updated to assert both the parameter placeholder and the params array.ADR-223 grep gate
Returns the same 3 hits as before — all input validation guards rejecting reserved namespace names. No new namespace-defaulting branches.
Related
MigrationPG methods require explicitschema). This PR is the planner/interpreter half.Closes TML-2916.
Summary by CodeRabbit
Release Notes
Refactor
publicnamespace where applicable.Chores
Tests
Documentation