Files
amb_rag/backend/app/processors/parsers/docx_parser.py
T
2026-09-01 13:00:36 +08:00

55 lines
1.8 KiB
Python

"""DOCX 解析器(python-docx fallback)。"""
from app.core.logging import get_logger
logger = get_logger(__name__)
def parse_docx_with_python_docx(file_content: bytes) -> str | None:
"""用 python-docx 解析 DOCX,返回 Markdown 或 None(失败时)。"""
try:
import io
from docx import Document
doc = Document(io.BytesIO(file_content))
md_parts = []
for para in doc.paragraphs:
text = para.text.strip()
if not text:
continue
style = para.style.name.lower()
if style.startswith("heading 1"):
md_parts.append(f"# {text}")
elif style.startswith("heading 2"):
md_parts.append(f"## {text}")
elif style.startswith("heading 3"):
md_parts.append(f"### {text}")
elif style.startswith("heading"):
md_parts.append(f"#### {text}")
elif style.startswith("list"):
md_parts.append(f"- {text}")
else:
md_parts.append(text)
# 处理表格
for table in doc.tables:
rows = []
for row in table.rows:
cells = [cell.text.strip() for cell in row.cells]
rows.append("| " + " | ".join(cells) + " |")
if rows:
# 添加表头分隔行
if len(rows) > 1:
sep = "| " + " | ".join(["---"] * len(table.columns)) + " |"
rows.insert(1, sep)
md_parts.append("\n".join(rows))
markdown = "\n\n".join(md_parts)
return markdown if markdown.strip() else None
except Exception as e:
logger.warning("python-docx 解析失败: %s", e)
return None