feat: postgres 컨테이너에 pgvector, mecab 설치 및 mecab 적용

boostcampwm-2024 · Feb 18, 2025 · cb168ea · cb168ea
1 parent d8c2cb9
commit cb168ea
Show file tree

Hide file tree

Showing 5 changed files with 69 additions and 8 deletions.
diff --git a/compose.local.yml b/compose.local.yml
@@ -2,7 +2,10 @@ version: "3.8"
 
 services:
   postgres:
-    image: pgvector/pgvector:0.8.0-pg16
+    build:
+      context: .
+      dockerfile: ./services/postgres/Dockerfile
+    image: postgres-pgvector-mecab
     env_file:
       - .env.local
     volumes:
@@ -55,6 +58,7 @@ services:
       - .env.local
     volumes:
       - ./apps/backend:/app/apps/backend
+      - ./libs:/app/libs
     depends_on:
       postgres:
         condition: service_healthy
@@ -74,6 +78,8 @@ services:
       - .env.local
     volumes:
       - ./apps/scheduler:/app/apps/scheduler
+      - ./libs:/app/libs
+
     depends_on:
       postgres:
         condition: service_healthy
@@ -93,6 +99,8 @@ services:
       - .env.local
     volumes:
       - ./apps/websocket:/app/apps/websocket
+      - ./libs:/app/libs
+
     depends_on:
       postgres:
         condition: service_healthy

diff --git a/libs/page/src/page.entity.ts b/libs/page/src/page.entity.ts
@@ -41,7 +41,7 @@ export class Page {
   @Column({
     generatedType: "STORED",
     type: "tsvector",
-    asExpression: `COALESCE(title, '') || ' ' || COALESCE(document, '')`,
+    asExpression: `to_tsvector('korean', COALESCE(title, '') || ' ' || COALESCE(document, ''))`,
     nullable: true,
   })
   fts: string;

diff --git a/libs/page/src/page.repository.ts b/libs/page/src/page.repository.ts
@@ -1,7 +1,7 @@
-import { Injectable, OnModuleInit } from '@nestjs/common';
-import { DataSource, Repository } from 'typeorm';
-import { Page } from './page.entity';
-import { InjectDataSource } from '@nestjs/typeorm';
+import { Injectable, OnModuleInit } from "@nestjs/common";
+import { DataSource, Repository } from "typeorm";
+import { Page } from "./page.entity";
+import { InjectDataSource } from "@nestjs/typeorm";
 
 @Injectable()
 export class PageRepository extends Repository<Page> implements OnModuleInit {
@@ -60,7 +60,7 @@ export class PageRepository extends Repository<Page> implements OnModuleInit {
 
     create or replace function hybrid_search(
       query_text text,
-      query_embedding vector(512),
+      query_embedding vector(384),
       match_count int,
       full_text_weight float = 1,
       semantic_weight float = 1,
@@ -119,7 +119,7 @@ export class PageRepository extends Repository<Page> implements OnModuleInit {
                    AND attname = 'fts') 
     THEN 
       ALTER TABLE page 
-      ADD COLUMN fts tsvector GENERATED ALWAYS AS (to_tsvector('english', document)) STORED;
+      ADD COLUMN fts tsvector GENERATED ALWAYS AS (to_tsvector('korean', COALESCE(title, '') || ' ' || COALESCE(document, ''))) STORED;
     END IF; 
   END $$;
 `);

diff --git a/services/postgres/Dockerfile b/services/postgres/Dockerfile
@@ -0,0 +1,52 @@
+# 베이스 이미지를 postgres:16으로 설정
+FROM postgres:16
+
+# 시스템 패키지 업데이트 및 설치
+RUN apt update -y && \
+    apt install -y \
+    wget \
+    build-essential \
+    postgresql-server-dev-16 \
+    automake \
+    unzip \
+    libmecab-dev
+
+# pgvector 설치
+RUN cd /tmp && \
+    wget https://github.com/pgvector/pgvector/archive/refs/tags/v0.8.0.tar.gz && \
+    tar -xvzf v0.8.0.tar.gz && \
+    cd pgvector-0.8.0 && \
+    make && \
+    make install
+
+# mecab-ko 설치
+RUN cd /tmp && \
+    wget https://bitbucket.org/eunjeon/mecab-ko/downloads/mecab-0.996-ko-0.9.2.tar.gz && \
+    tar xvfz mecab-0.996-ko-0.9.2.tar.gz && \
+    cd mecab-0.996-ko-0.9.2 && \
+    ./configure CC=gcc CXX=g++ CFLAGS="-m64" CXXFLAGS="-m64" && \
+    make && \
+    make install
+
+# mecab-ko-dic 설치
+RUN cd /tmp && \
+    wget https://bitbucket.org/eunjeon/mecab-ko-dic/downloads/mecab-ko-dic-2.1.1-20180720.tar.gz && \
+    tar xvfz mecab-ko-dic-2.1.1-20180720.tar.gz && \
+    cd mecab-ko-dic-2.1.1-20180720 && \
+    ./autogen.sh && \
+    ./configure && \
+    make && \
+    make install
+
+RUN cd /tmp && \
+    apt install git -y && \
+    git clone https://github.com/i0seph/textsearch_ko.git && \
+    cd textsearch_ko && \
+    make USE_PGXS=1 && \
+    make USE_PGXS=1 install && \
+    cp /tmp/textsearch_ko/ts_mecab_ko.sql /docker-entrypoint-initdb.d/
+
+COPY services/postgres/init.sql /docker-entrypoint-initdb.d/z_init.sql
+
+# PostgreSQL 컨테이너 기본 명령어 설정
+CMD ["postgres"]
diff --git a/services/postgres/init.sql b/services/postgres/init.sql
@@ -0,0 +1 @@
+CREATE EXTENSION IF NOT EXISTS vector;