확장자, 포트 수정

workspace/config.py

@@ -5,5 +5,6 @@ UPLOAD_DIR = os.getenv("UPLOAD_DIR", "data")
 OUTPUT_DIR = os.getenv("OUTPUT_DIR", "converted")
 
 # Redis 연결 정보
+# local 연결시 redis -> localhost로 변경
 REDIS_HOST = os.getenv("REDIS_HOST", "redis")
 REDIS_PORT = int(os.getenv("REDIS_PORT", "6379"))

workspace/convert_obj_to_md.py

@@ -13,21 +13,19 @@ def convert_hwp_to_md(input_path: str, output_path: str):
     for doc in docs:
         try:
             text = doc.page_content if hasattr(doc, "page_content") else str(doc)
-            text = text.encode("utf-8", "ignore").decode(
-                "utf-8"
-            )  # UTF-8로 변환하면서 깨진 문자 제거
+            text = text.encode("utf-8", "ignore").decode("utf-8")
             docs_as_text.append(text)
         except Exception as e:
             print(f"인코딩 변환 중 오류 발생: {e}")
 
     with open(output_path, "w", encoding="utf-8") as f:
-        f.write("\n".join(docs_as_text))  # ✅ 변환된 리스트를 파일에 저장
+        f.write("\n".join(docs_as_text))
 
     return None
 
 
 def convert_to_md(input_path: str, output_path: str):
-    md = MarkItDown(docintel_endpoint="<document_intelligence_endpoint>")
+    md = MarkItDown()
     result = md.convert(input_path)
     with open(output_path, "w", encoding="utf-8") as f:
         f.write(result.text_content)

workspace/main.py

@@ -49,7 +49,15 @@ async def convert_files():
 
     for file in files:
         file_ext = file.split(".")[-1].lower()
-        if file_ext in ["txt", "html", "docx", "pdf", "hwp"]:
+        if file_ext in [
+            "txt",
+            "html",
+            "pdf",
+            "hwp",
+            "pptx",
+            "xlsx",
+            "docx",
+        ]:
             task = {
                 "filename": file,
                 "extension": file_ext,

workspace/worker.py

@@ -34,5 +34,5 @@ def process_task(task):
 if __name__ == "__main__":
     listen = ["task_queue1"]
 
-    worker = Worker(listen, connection=redis_client)  # 최신 방식
+    worker = Worker(listen, connection=redis_client)
     worker.work()