import io
import os
import logging
import warnings
from pathlib import Path
from typing import Union, Optional
try:
import pandas as pd
except ImportError:
pd = None
try:
from docx import Document
except ImportError:
Document = None
try:
import PyPDF2
except ImportError:
PyPDF2 = None
try:
import pdfplumber
except ImportError:
pdfplumber = None
# 抑制PDF处理警告
warnings.filterwarnings('ignore', category=UserWarning, module='pdfminer')
warnings.filterwarnings('ignore', message='.*CropBox missing.*')
# 设置pdfminer日志级别
logging.getLogger("pdfminer").setLevel(logging.ERROR)
logging.getLogger("pdfplumber").setLevel(logging.ERROR)
class FileConverter:
"""文档格式转换器"""
@staticmethod
def docx_to_markdown(file_path: Union[str, Path]) -> str:
"""
将DOCX文件转换为Markdown格式
Args:
file_path: DOCX文件路径
Returns:
str: Markdown格式的文本
Raises:
ImportError: 如果python-docx库未安装
Exception: 文件读取错误
"""
if Document is None:
raise ImportError("需要安装python-docx库: pip install python-docx")
try:
doc = Document(file_path)
markdown_content = []
for paragraph in doc.paragraphs:
text = paragraph.text.strip()
if text:
# 简单的格式转换
if paragraph.style.name.startswith("Heading"):
level = (
int(paragraph.style.name.split()[-1])
if paragraph.style.name.split()[-1].isdigit()
else 1
)
markdown_content.append(f"{'#' * level} {text}")
else:
markdown_content.append(text)
markdown_content.append("") # 添加空行
# 处理表格
for table in doc.tables:
markdown_content.append("") # 表格前空行
for i, row in enumerate(table.rows):
cells = [cell.text.strip() for cell in row.cells]
markdown_content.append("| " + " | ".join(cells) + " |")
if i == 0: # 添加表头分隔线
markdown_content.append(
"| " + " | ".join(["---"] * len(cells)) + " |"
)
markdown_content.append("") # 表格后空行
return "\n".join(markdown_content)
except Exception as e:
raise Exception(f"DOCX文件转换失败: {e}")
@staticmethod
def xlsx_to_markdown(file_path: Union[str, Path]) -> str:
"""
将XLSX文件转换为Markdown格式
Args:
file_path: XLSX文件路径
Returns:
str: Markdown格式的文本
Raises:
ImportError: 如果pandas库未安装
Exception: 文件读取错误
"""
if pd is None:
raise ImportError("需要安装pandas库: pip install pandas openpyxl")
try:
# 读取所有工作表
excel_file = pd.ExcelFile(file_path)
markdown_content = []
for sheet_name in excel_file.sheet_names:
markdown_content.append(f"# {sheet_name}")
markdown_content.append("")
df = pd.read_excel(file_path, sheet_name=sheet_name)
# 将DataFrame转换为Markdown表格
if not df.empty:
# 创建表头
headers = df.columns.tolist()
markdown_content.append(
"| " + " | ".join(str(h) for h in headers) + " |"
)
markdown_content.append(
"| " + " | ".join(["---"] * len(headers)) + " |"
)
# 添加数据行
for _, row in df.iterrows():
cells = [str(cell) if pd.notna(cell) else "" for cell in row]
markdown_content.append("| " + " | ".join(cells) + " |")
markdown_content.append("") # 工作表之间的空行
return "\n".join(markdown_content)
except Exception as e:
raise Exception(f"XLSX文件转换失败: {e}")
@staticmethod
def pdf_to_markdown(
file_path: Union[str, Path], use_pdfplumber: bool = True
) -> str:
"""
将PDF文件转换为Markdown格式
Args:
file_path: PDF文件路径
use_pdfplumber: 是否使用pdfplumber库（更好的文本提取）
Returns:
str: Markdown格式的文本
Raises:
ImportError: 如果所需库未安装
Exception: 文件读取错误
"""
if use_pdfplumber and pdfplumber is not None:
return FileConverter._pdf_to_markdown_pdfplumber(file_path)
elif PyPDF2 is not None:
return FileConverter._pdf_to_markdown_pypdf2(file_path)
else:
raise ImportError(
"需要安装PDF处理库: pip install pdfplumber 或 pip install PyPDF2"
)
@staticmethod
def _pdf_to_markdown_pdfplumber(file_path: Union[str, Path]) -> str:
"""使用pdfplumber提取PDF文本，专注文本内容忽略媒体信息"""
try:
markdown_content = []
# 抑制特定的警告
with warnings.catch_warnings():
warnings.filterwarnings('ignore', message='.*CropBox missing.*')
warnings.filterwarnings('ignore', category=UserWarning, module='pdfminer')
with pdfplumber.open(file_path) as pdf:
for i, page in enumerate(pdf.pages):
if i > 0: # 添加页面分隔
markdown_content.append(f"\n---\n# 第 {i + 1} 页\n")
# 配置文本提取参数，专注文本内容
text = page.extract_text(
x_tolerance=3,
y_tolerance=3,
layout=False, # 不保持布局，简化输出
x_density=7.25,
y_density=13
)
if text:
# 清理和格式化文本
text = text.strip()
# 移除多余的空行
text = '\n'.join(line.strip() for line in text.split('\n') if line.strip())
# 简单的段落处理
paragraphs = text.split('\n\n')
for paragraph in paragraphs:
paragraph = paragraph.strip()
if paragraph and len(paragraph) > 3: # 过滤太短的内容
markdown_content.append(paragraph)
markdown_content.append("")
# 提取表格（如果需要）
try:
tables = page.extract_tables()
for table in tables:
if table and len(table) > 0:
markdown_content.append("")
for j, row in enumerate(table):
if row and any(cell for cell in row if cell): # 确保行不为空
cells = [str(cell).strip() if cell else "" for cell in row]
if any(cells): # 确保不是空行
markdown_content.append(
"| " + " | ".join(cells) + " |"
)
if j == 0: # 表头分隔线
markdown_content.append(
"| "
+ " | ".join(["---"] * len(cells))
+ " |"
)
markdown_content.append("")
except Exception:
# 如果表格提取失败，忽略表格但继续处理
pass
# 清理最终内容
final_content = '\n'.join(markdown_content).strip()
# 移除多余的连续空行
while '\n\n\n' in final_content:
final_content = final_content.replace('\n\n\n', '\n\n')
return final_content
except Exception as e:
raise Exception(f"PDF文件转换失败 (pdfplumber): {e}")
@staticmethod
def _pdf_to_markdown_pypdf2(file_path: Union[str, Path]) -> str:
"""使用PyPDF2提取PDF文本"""
try:
markdown_content = []
with open(file_path, "rb") as file:
pdf_reader = PyPDF2.PdfReader(file)
for i, page in enumerate(pdf_reader.pages):
if i > 0: # 添加页面分隔
markdown_content.append(f"\n---\n# 第 {i + 1} 页\n")
text = page.extract_text()
if text:
# 简单的段落处理
paragraphs = text.split("\n\n")
for paragraph in paragraphs:
paragraph = paragraph.strip()
if paragraph:
markdown_content.append(paragraph)
markdown_content.append("")
return "\n".join(markdown_content)
except Exception as e:
raise Exception(f"PDF文件转换失败 (PyPDF2): {e}")
@staticmethod
def get_converter_for_extension(file_extension: str):
"""
根据文件扩展名获取对应的转换器
Args:
file_extension: 文件扩展名（带点，如'.docx'）
Returns:
callable: 对应的转换函数，如果不需要转换则返回None
"""
converters = {
".docx": FileConverter.docx_to_markdown,
".xlsx": FileConverter.xlsx_to_markdown,
".xls": FileConverter.xlsx_to_markdown,
".pdf": FileConverter.pdf_to_markdown,
}
return converters.get(file_extension.lower())
