Add pscale import d1 for Cloudflare D1 offline migration

[no-itsbackpack]

Summary

Adds pscale import d1 — an offline migration path from Cloudflare D1 (SQLite export) into PlanetScale Postgres.

Workflow: export with wrangler → lint/plan → import → verify → complete

CLI commands (pscale import d1 …)

Command	Purpose
doctor	Check prerequisites (pgloader, sqlite3, etc.)
lint	Analyze a D1 SQL export for migration issues
convert-schema	Convert SQLite schema to PostgreSQL DDL
start	Lint + plan + load into Postgres (--dry-run for preview only)
verify	Row counts, sequences, and content checks
status	Show local migration state
complete	Mark migration finished (teardown alias)

Export is done with wrangler, not the CLI:

wrangler d1 export <name> --remote --output ./d1-export.sql

Example import flow:

pscale import d1 start mydb dev --input ./d1-export.sql --method pgloader
pscale import d1 verify mydb dev --migration-id <id> --input ./d1-export.sql
pscale import d1 complete mydb dev --migration-id <id>

Also includes

• Core library: internal/import/d1/ — schema conversion, lint/plan, pgloader import, verify, local migration state
• Slim Postgres helpers: internal/postgres/ — connection URI parsing, pgx open, psql discovery/version check (no logical replication / pg_dump pipeline)
• Slack notifications: lifecycle events are reported to the api
• Tests: unit tests under internal/import/d1/ and internal/cmd/importcmd/

---

How to test

Prerequisites: pscale auth, a Postgres database branch, pgloader, sqlite3, and a D1 SQL export (from wrangler or the repo fixture).

1. Smoke test (no PlanetScale load)

pscale import d1 doctor
pscale import d1 lint --input internal/import/d1/testdata/sample_d1_export.sql --format json
pscale import d1 start mydb --input internal/import/d1/testdata/sample_d1_export.sql --dry-run --format json

Dry-run returns a migration_id — save it for verify/complete.

2. Full import (against a dev branch)

pscale import d1 start mydb dev --input ./d1-export.sql --method pgloader
pscale import d1 verify mydb dev --migration-id <id> --input ./d1-export.sql
pscale import d1 complete mydb dev --migration-id <id>

3. Automated tests

go test ./internal/import/d1/... ./internal/cmd/importcmd/...

4. Slack notifications (optional)

Run against local api-bb with Sidekiq running. Import lifecycle events should appear in #d1-migrations. Pass --no-notify on start/verify/complete to skip.

---

Test plan

• doctor passes with pgloader + sqlite3 installed
• lint and start --dry-run succeed on the sample fixture
• Full start → verify → complete against a dev Postgres branch
• go test ./internal/import/d1/... ./internal/cmd/importcmd/... passes in CI
• (Optional) API import notifications

[cursor]

Cursor Bugbot has reviewed your changes using default effort and found 1 potential issue.

^{❌ Bugbot Autofix is OFF. To automatically fix reported issues with cloud agents, have a team admin enable autofix in the Cursor dashboard.}

^{Reviewed by Cursor Bugbot for commit 541a200. Configure here.}

[cursor]

Partial schema blocks import retry

Medium Severity

Schema is applied via psql -f without a wrapping transaction, so a mid-file error can leave some import tables on the branch while schema_applied stays false. A later start retry resets progress and hits errExistingImportTables against those partial tables, blocking resume unless tables are dropped manually.

Additional Locations (1)

• internal/import/d1/import.go#L128-L134

 

^{Reviewed by Cursor Bugbot for commit 541a200. Configure here.}