1.6K star！这个开源文本提取神器，5分钟搞定PDF/图片/Office文档！

from kreuzberg import batch_extract_file_sync

# 批量处理合同文档
results = batch_extract_file_sync([
    "采购合同.pdf",
    "报价单.xlsx",
    "签字扫描件.jpg"
])

for path, (content, _, _) in results.items():
    print(f"【{path}】解析结果：\n{content[:200]}...\n")

async def extract_research_paper():
    result = await extract_file(
        "论文.pdf",
        language="chi_sim+eng",  # 中英混合识别
        psm=PSMMode.SINGLE_COLUMN  # 适合学术论文版式
    )
    print(f"摘要章节：{result.content.split('摘要')[1][:500]}...")

def process_legal_docs(docs):
    for doc, mime_type in docs:
        result = await extract_bytes(
            doc,
            mime_type=mime_type,
            force_ocr=True  # 确保扫描件100%识别
        )
        save_to_database(
            content=result.content,
            metadata=result.metadata
        )

async def build_knowledge_base(files):
    results = await batch_extract_file(files)
    for content, _, meta in results:
        await vector_db.insert({
            "text": content,
            "source": meta.get("title"),
            "author": meta.get("creator")
        })

def search_across_formats(keyword):
    results = []
    for file in os.listdir("docs"):
        content, _, _ = extract_file_sync(file)
        if keyword in content:
            results.append(file)
    return results

# 配置德语+英语识别，使用单列版面分析
await extract_file(
    "german_doc.jpg",
    language="deu+eng",
    psm=PSMMode.SINGLE_COLUMN,
    ocr_config={"tessedit_pageseg_mode": 6}
)

from kruetzberg import extract_file, ParsingError

try:
    result = await extract_file(
        "encrypted.pdf",
        pdf_options={"password": "123456"}
    )
except ParsingError as e:
    print("密码错误或文件损坏！")

# 提取带Markdown格式的内容
result = await extract_file(
    "report.docx",
    pandoc_options={"output_format": "markdown"}
)
print(result.content)  # 输出带Markdown标记的文本

# 1. 安装库
pip install kreuzberg

# 2. 安装依赖（Ubuntu示例）
sudo apt-get install pandoc tesseract-ocr tesseract-ocr-chi-sim

# 3. 中文识别测试
echo "测试文本" > test.txt
python -c "from kruetzberg import extract_file_sync; print(extract_file_sync('test.txt').content)"

FROM python:3.10-slim

RUN apt-get update && \
    apt-get install -y pandoc tesseract-ocr tesseract-ocr-chi-sim && \
    rm -rf /var/lib/apt/lists/*

COPY requirements.txt .
RUN pip install -r requirements.txt

CMD ["python", "app.py"]