// the find
pgroonga/pgroonga
PGroonga is a PostgreSQL extension to use Groonga as index. PGroonga makes PostgreSQL fast full text search platform for all languages!
PGroonga wraps the Groonga full-text search engine as a PostgreSQL extension, giving you fast full-text search across any language without the tokenization headaches of pg_tsvector. It's aimed at teams building multilingual search (especially CJK) who want to stay in Postgres rather than spinning up Elasticsearch.
First-class CJK tokenization out of the box — no need to configure MeCab or custom dictionaries separately, Groonga handles it. The upgrade path is unusually well-maintained: 100+ sequential migration SQL files going back to v1.0.0 mean you can upgrade from any version without data dumps. Bidirectional migration scripts (e.g., 3.2.1→3.2.0) exist, which is rare and useful when a release breaks something. Active maintenance with a recent push in June 2026 and cross-platform CI for Linux, macOS, and Windows.
The README is essentially a pointer to an external website — zero quickstart, no code examples, no install instructions in the repo itself. If the website goes down or moves, new users have nothing. Groonga is a separate daemon dependency that PostgreSQL calls into; this means your failure surface is larger than a pure-Postgres solution and Groonga tuning is a separate skill set most Postgres DBAs don't have. Only 730 stars and 29 forks after years of development signals a narrow audience — community support, StackOverflow answers, and third-party tooling are sparse. No native vector/semantic search integration, so if you're already using pgvector in the same database, you're maintaining two separate indexing systems for search.