docs: read the sample and quickstart schema from migrations

[omer-cengel]

Summary

The external Maven sample and the quickstart now point sql[].schema at a Flyway-style sql/migrations directory instead of a single schema snapshot. This shows sqlcj reading the schema an application already maintains through migrations. The later migration adds a column and an index, and the generated repository and the sample's output stay the same.

Changes

• Replace the sample's sql/schema.sql with V1__create_authors_and_books.sql, which creates authors (id, name, bio) and books, and V2__add_author_created_at_and_book_index.sql, which adds authors.created_at and creates an index on books (author_id).
• Set the sample's sqlcj.yaml to schema: sql/migrations and update the App Javadoc to describe the migrations being applied in version order.
• Make the release verification script generate from the migration directory, check that the generated row carries LocalDateTime createdAt, and apply both migrations with psql in version order under ON_ERROR_STOP before running the sample.
• Update the quickstart's project layout, configuration, schema section and database setup to use the two migration files. The schema section explains Flyway version ordering, how ALTER TABLE ... ADD COLUMN applies to the table created earlier, and why CREATE INDEX is accepted and ignored.

Scope and non-goals

• No compiler or runtime behavior changes. sqlcj still reads migration files and never executes them.
• No migration tool is added. The sample and the quickstart apply migrations with psql.
• The generated API, selected-column order, sample queries and sample runtime output are unchanged.