Merge branch 'master' into feature/80-add-connectors-and-metadata

Author: ehddnr301

.gitignore

@@ -10,4 +10,4 @@ test_lhm/
 .cursorignore
 .vscode
 table_info_db
-ko_reranker_local
+ko_reranker_local

Dockerfile

@@ -0,0 +1,32 @@
+# Python 3.10 slim 이미지 기반
+FROM python:3.12-slim
+
+# 시스템 라이브러리 설치
+RUN apt-get update && apt-get install -y \
+    build-essential \
+    curl \
+    software-properties-common \
+    git \
+    libpq-dev \
+    && rm -rf /var/lib/apt/lists/*
+
+# 작업 디렉토리 설정
+WORKDIR /app
+
+# 의존성 파일 복사 및 설치
+COPY requirements.txt .
+RUN pip install --no-cache-dir -r requirements.txt
+
+# 전체 서비스 코드 복사
+COPY . .
+
+# Python 환경 설정
+ENV PYTHONPATH=/app
+ENV PYTHONUNBUFFERED=1
+
+# Streamlit 포트 노출
+EXPOSE 8501
+
+# Streamlit 실행 명령
+CMD ["python", "-c", "from llm_utils.tools import set_gms_server; import os; set_gms_server(os.getenv('DATAHUB_SERVER', 'http://localhost:8080'))"]
+CMD ["streamlit", "run", "./interface/streamlit_app.py", "--server.port=8501"] 

README.md

@@ -1,16 +1,29 @@
 """
-Datahub GMS 서버 URL을 설정하고, 필요 시 Streamlit 인터페이스를 실행하는 CLI 프로그램입니다.
+Lang2SQL CLI 프로그램입니다.
+이 프로그램은 Datahub GMS 서버 URL을 설정하고, 필요 시 Streamlit 인터페이스를 실행합니다.
+
+명령어 예시: lang2sql --datahub_server http://localhost:8080 --run-streamlit
 """
 
+import logging
 import subprocess
 
 import click
 
+from llm_utils.check_server import CheckServer
 from llm_utils.tools import set_gms_server
+from version import __version__
+
+logging.basicConfig(
+    level=logging.INFO,
+    format="%(asctime)s [%(levelname)s] %(message)s",
+    datefmt="%Y-%m-%d %H:%M:%S",
+)
+logger = logging.getLogger(__name__)
 
 
 @click.group()
-@click.version_option(version="0.1.4")
+@click.version_option(version=__version__)
 @click.pass_context
 @click.option(
     "--datahub_server",
@@ -64,11 +77,20 @@ def cli(
         'set_gms_server' 함수에서 ValueError가 발생할 경우, 프로그램은 비정상 종료(exit code 1)합니다.
     """
 
-    try:
+    logger.info(
+        "Initialization started: GMS server = %s, run_streamlit = %s, port = %d",
+        datahub_server,
+        run_streamlit,
+        port,
+    )
+
+    if CheckServer.is_gms_server_healthy(url=datahub_server):
         set_gms_server(datahub_server)
-    except ValueError as e:
-        click.secho(f"GMS 서버 URL 설정 실패: {str(e)}", fg="red")
+        logger.info("GMS server URL successfully set: %s", datahub_server)
+    else:
+        logger.error("GMS server health check failed. URL: %s", datahub_server)
         ctx.exit(1)
+
     if run_streamlit:
         run_streamlit_command(port)
 
@@ -89,6 +111,8 @@ def run_streamlit_command(port: int) -> None:
         - subprocess 호출 실패 시 예외가 발생할 수 있습니다.
     """
 
+    logger.info("Starting Streamlit application on port %d...", port)
+
     try:
         subprocess.run(
             [
@@ -100,8 +124,9 @@ def run_streamlit_command(port: int) -> None:
             ],
             check=True,
         )
+        logger.info("Streamlit application started successfully.")
     except subprocess.CalledProcessError as e:
-        click.echo(f"Streamlit 실행 실패: {e}")
+        logger.error("Failed to start Streamlit application: %s", e)
         raise
 
 
@@ -132,4 +157,5 @@ def run_streamlit_cli_command(port: int) -> None:
         - Streamlit 실행에 실패할 경우 subprocess 호출에서 예외가 발생할 수 있습니다.
     """
 
+    logger.info("Executing 'run-streamlit' command on port %d...", port)
     run_streamlit_command(port)

cli/__init__.py

@@ -1 +0,0 @@
-# data_utils 패키지 초기화 파일

data_utils/__init__.py

@@ -24,6 +24,7 @@
 else:
     print(f"⚠️  환경변수 파일(.env)이 {os.getcwd()}에 없습니다!")
 
+
 def get_db_connector(db_type: Optional[str] = None, config: Optional[DBConfig] = None):
     """
     Return the appropriate DB connector instance.
@@ -43,23 +44,25 @@ def get_db_connector(db_type: Optional[str] = None, config: Optional[DBConfig] =
     if db_type is None:
         db_type = os.getenv("DB_TYPE")
         if not db_type:
-            raise ValueError("DB type must be provided or set in environment as DB_TYPE.")
+            raise ValueError(
+                "DB type must be provided or set in environment as DB_TYPE."
+            )
 
     db_type = db_type.lower()
 
     if config is None:
         config = load_config_from_env(db_type.upper())
 
         connector_map = {
-        "clickhouse": ClickHouseConnector,
-        "postgresql": PostgresConnector,
-        "mysql": MySQLConnector,
-        "mariadb": MariaDBConnector,
-        "oracle": OracleConnector,
-        "duckdb": DuckDBConnector,
-        "databricks": DatabricksConnector,
-        "snowflake": SnowflakeConnector,
-    }
+            "clickhouse": ClickHouseConnector,
+            "postgresql": PostgresConnector,
+            "mysql": MySQLConnector,
+            "mariadb": MariaDBConnector,
+            "oracle": OracleConnector,
+            "duckdb": DuckDBConnector,
+            "databricks": DatabricksConnector,
+            "snowflake": SnowflakeConnector,
+        }
 
     if db_type not in connector_map:
         logger.error(f"Unsupported DB type: {db_type}")
@@ -87,7 +90,6 @@ def get_db_connector(db_type: Optional[str] = None, config: Optional[DBConfig] =
     return connector_map[db_type](config)
 
 
-
 def load_config_from_env(prefix: str) -> DBConfig:
     """
     Load DBConfig from environment variables with a given prefix.
@@ -107,7 +109,9 @@ def load_config_from_env(prefix: str) -> DBConfig:
     # Extract standard values
     config = {
         "host": os.getenv(f"{prefix}_HOST"),
-        "port": int(os.getenv(f"{prefix}_PORT")) if os.getenv(f"{prefix}_PORT") else None,
+        "port": (
+            int(os.getenv(f"{prefix}_PORT")) if os.getenv(f"{prefix}_PORT") else None
+        ),
         "user": os.getenv(f"{prefix}_USER"),
         "password": os.getenv(f"{prefix}_PASSWORD"),
         "database": os.getenv(f"{prefix}_DATABASE"),
@@ -117,11 +121,11 @@ def load_config_from_env(prefix: str) -> DBConfig:
     extra = {}
     for key, value in os.environ.items():
         if key.startswith(f"{prefix}_"):
-            suffix = key[len(f"{prefix}_"):]
+            suffix = key[len(f"{prefix}_") :]
             if suffix.upper() not in base_keys:
                 extra[suffix.lower()] = value
 
     if extra:
         config["extra"] = extra
 
-    return DBConfig(**config)
+    return DBConfig(**config)

db_utils/__init__.py

@@ -3,10 +3,12 @@
 import pandas as pd
 from db_utils import DBConfig, logger
 
+
 class ClickHouseConnector(BaseConnector):
     """
     Connect to ClickHouse and execute SQL queries.
     """
+
     client = None
 
     def __init__(self, config: DBConfig):
@@ -54,8 +56,8 @@ def run_sql(self, sql: str) -> pd.DataFrame:
             self.connect()
 
         try:
-            result = self.client.query(sql).result()
-            return pd.DataFrame(result)
+            result = self.client.query_dataframe(sql)
+            return result
         except Exception as e:
             logger.error(f"Failed to execute SQL query: {e}")
             raise

db_utils/clickhouse_connector.py

@@ -0,0 +1,28 @@
+services:
+  streamlit:
+    build: .
+    ports:
+      - "8501:8501"
+    volumes:
+      - .:/app
+    env_file:
+      - .env
+    environment:
+      - DATABASE_URL=postgresql://postgres:password@db:5432/streamlit_db
+    depends_on:
+      - db
+
+  db:
+    image: pgvector/pgvector:pg17
+    container_name: pgvector-db
+    environment:
+      POSTGRES_USER: postgres
+      POSTGRES_PASSWORD: password
+      POSTGRES_DB: streamlit_db
+    ports:
+      - "5432:5432"
+    volumes:
+      - pgdata:/var/lib/postgresql/data
+      - ./postgres/schema.sql:/docker-entrypoint-initdb.d/schema.sql
+volumes:
+  pgdata: