확장자, 포트 수정

This commit is contained in:
2025-02-14 12:13:05 +09:00
parent 0d34c20882
commit c6a71467e8
6 changed files with 43 additions and 14 deletions

View File

@@ -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)