WIP: indexing Chinese better

apps/server/src/core/search/search.service.ts

@@ -36,7 +36,7 @@ export class SearchService {
         'createdAt',
         'updatedAt',
         sql<number>`ts_rank(tsv, to_tsquery(${searchQuery}))`.as('rank'),
-        sql<string>`ts_headline('english', text_content, to_tsquery(${searchQuery}), 'MinWords=9, MaxWords=10, MaxFragments=10')`.as(
+        sql<string>`ts_headline('chinese_zh', text_content, to_tsquery(${searchQuery}), 'MinWords=9, MaxWords=10, MaxFragments=10')`.as(
           'highlight',
         ),
       ])

apps/server/src/database/migrations/20240324T086800-pages-tsvector-trigger.ts

@@ -4,8 +4,8 @@ export async function up(db: Kysely<any>): Promise<void> {
   await sql`CREATE OR REPLACE FUNCTION pages_tsvector_trigger() RETURNS trigger AS $$
         begin
             new.tsv :=
-                      setweight(to_tsvector('english', coalesce(new.title, '')), 'A') ||
-                      setweight(to_tsvector('english', coalesce(new.text_content, '')), 'B');
+                      setweight(to_tsvector('chinese_zh', coalesce(new.title, '')), 'A') ||
+                      setweight(to_tsvector('chinese_zh', coalesce(new.text_content, '')), 'B');
             return new;
         end;
         $$ LANGUAGE plpgsql;`.execute(db);

docker-compose.yml

@@ -18,7 +18,8 @@ services:
       - docmost:/app/data/storage
 
   db:
-    image: postgres:16-alpine
+    #image: postgres:16-alpine
+    image: abcfy2/zhparser:16-alpine
     environment:
       POSTGRES_DB: docmost
       POSTGRES_USER: docmost