[O1sumitkumar]

Summary

The "Tables" section of indexer/README.md listed snake_case names (indexer_cursor, invocations, contracts, storage_changes, …). The schema has no @@map, so Prisma creates the tables under the model names, and querying the documented names fails with relation "..." does not exist.

This PR:

• renames each heading to the real table name: "IndexerCursor", "Invocation", "Contract", "BackfillCursor", "StorageChange", "ContractEventRecord", "IndexerGap";
• adds a short note and example query showing that the names (and the camelCase columns) must be double-quoted in SQL, since Postgres folds unquoted names to lower case;
• adds the missing column table for "BackfillCursor", taken from prisma/schema.prisma and src/db.ts (singleton row with id 1).

Closes #317

Checks

Docs-only change, so the build/test checks don't apply. To check the names, I applied the three migrations in indexer/prisma/migrations/ in order to a fresh Postgres (PGlite, Postgres compiled to WASM, since I had no local server) and ran:

• SELECT count(*) FROM "<name>" for all seven documented names: each one works.
• SELECT count(*) FROM invocations, FROM Invocation and FROM indexer_cursor: each fails with relation "..." does not exist, which is the problem the issue describes and why the note about quoting is there.
• information_schema.columns for BackfillCursor, which matches the new column table.

Notes

I didn't touch the descriptions of the other tables' columns, only their names.

[vercel]

@O1sumitkumar is attempting to deploy a commit to the thewealthyplace100-8633's projects Team on Vercel.

A member of the Team first needs to authorize it.