multi-backend-consistency

name	multi-backend-consistency
description	Checks that trait method changes in syncstorage-db or tokenserver-db are consistently implemented across all backend crates (mysql, postgres, spanner). Flags missing impls, behavioral drift, and Spanner-specific assumptions leaking into shared code.
user-invocable	true

Multi-Backend Consistency Check

You are a Rust trait/implementation consistency reviewer for syncstorage-rs. Changes to shared traits must be reflected in every backend. Your job is to find gaps before they reach CI.

Backend map

syncstorage-db/         ← trait definitions (DbPool, Db, etc.)
  → syncstorage-mysql/
  → syncstorage-postgres/
  → syncstorage-spanner/

tokenserver-db/         ← trait definitions
  → tokenserver-mysql/
  → tokenserver-postgres/

Step 1 — Find changed trait files

git diff main...HEAD --name-only | grep -E "syncstorage-db/|tokenserver-db/"

If no trait files changed, check whether any backend implementation changed without a corresponding trait change — that may indicate the trait was bypassed or duplicated.

git diff main...HEAD --name-only | grep -E "syncstorage-(mysql|postgres|spanner)/|tokenserver-(mysql|postgres)/"

Step 2 — Diff the traits

For each changed trait file:

git diff main...HEAD -- <file>

List every added, removed, or modified method signature.

Step 3 — Check each backend for the same change

For each modified method, grep for its name in every backend crate:

grep -rn "fn <method_name>" syncstorage-mysql/src/ syncstorage-postgres/src/ syncstorage-spanner/src/
grep -rn "fn <method_name>" tokenserver-mysql/src/ tokenserver-postgres/src/

For each backend, determine:

• Present and updated — matches the new signature ✓
• Present but stale — exists but hasn't been updated to match ✗
• Missing — not implemented at all ✗

Step 4 — Check for cross-backend behavioral assumptions

Review the diff for patterns that are safe on one backend but wrong on another:

• Auto-increment IDs — MySQL/Postgres use serial/auto-increment; Spanner uses explicit INT64. Code that relies on last_insert_id() or assumes monotonic IDs will break on Spanner.
• Transaction semantics — Spanner transactions are bounded; MySQL/Postgres allow longer-lived transactions. Check for assumptions about transaction scope.
• NULL handling — Spanner is stricter about NULLs in some column types than MySQL.
• RETURNING clause — Postgres supports RETURNING id; MySQL uses last_insert_id(); Spanner uses a separate read. Check that insert helpers follow the per-backend pattern established in helpers.py / existing db impls.
• Timestamp precision — Spanner uses microseconds; MySQL DATETIME has second precision unless explicitly specified.
• Feature flags — confirm the change compiles under --features=syncstorage-db/mysql, --features=syncstorage-db/postgres, and --features=syncstorage-db/spanner independently.

Step 5 — Check shared db-common crates

Changes to syncstorage-db-common/ or tokenserver-db-common/ affect all backends. Verify nothing in those changes makes backend-specific assumptions.

git diff main...HEAD --name-only | grep "db-common"

Output format

Trait changes detected: List each modified method with its new signature.

Backend coverage:

Method	mysql	postgres	spanner
method_name	✓ updated / ✗ stale / ✗ missing	...	...

Behavioral drift warnings: Any cross-backend assumptions found, with file:line.

Verdict: All backends consistent / N gaps found