Content document & index — ContentDocument record + Meilisearch IndexSettings bean

[herbie-bot]

Summary

Define the canonical ContentDocument record (the shape of one indexed page) and register the Meilisearch IndexSettings bean for the content index. Searchable attributes weighted title > excerpt > body; filterable source/locale/author/categories/publishedDate; publishedDate sortable. No crawling yet — this defines the data contract and index schema.

Spec

• Spec folder: docs/specs/003-content-document-index/
• Design: docs/specs/003-content-document-index/design.md
• Behaviors: docs/specs/003-content-document-index/behaviors.md

Acceptance criteria (from behaviors.md)

• ContentDocument.id(source, url) is deterministic; different URLs yield different ids
• id contains only Meilisearch-legal chars [A-Za-z0-9_-] even for URLs with / : .
• toMap() contains all indexed fields; null optional fields map cleanly
• IndexSettings bean registered with correct searchable/filterable/sortable attributes and primaryKey=id
• Settings applied at startup by MeilisearchIndexSettingsInitializer
• Index UID honors the configured index-prefix

Notes

• Roadmap Phase 1 step 3; design doc §4, §8. Builds on specs 001 (Project skeleton — standalone Spring Boot app on spring-services #1), 002 (Content source config — ContentSource abstraction + @ConfigurationProperties + Ant-glob URL matching #3).