Python使用PyPDF进行PDF操作的代码示例

pip install pypdf

from pypdf import PdfMerger

# 初始化合并器
merger = PdfMerger()

# 添加需要合并的 PDF 文件
merger.append("file1.pdf")
merger.append("file2.pdf")

# 保存合并后的文件
merger.write("merged.pdf")
merger.close()
print("PDF 合并完成！")

from pypdf import PdfReader, PdfWriter

# 读取 PDF 文件
reader = PdfReader("input.pdf")

# 拆分每一页
for i, page in enumerate(reader.pages):
    writer = PdfWriter()
    writer.add_page(page)
    with open(f"page_{i+1}.pdf", "wb") as output_file:
        writer.write(output_file)
print("PDF 拆分完成！")

from pypdf import PdfReader

# 读取 PDF 文件
reader = PdfReader("input.pdf")

# 提取每页的文本
for page in reader.pages:
    print(page.extract_text())

from pypdf import PdfReader, PdfWriter

reader = PdfReader("input.pdf")
writer = PdfWriter()

# 复制所有页面到新 PDF
writer.add_pages(reader.pages)

# 修改元信息
writer.metadata = {
    "/Title": "新的标题",
    "/Author": "作者名",
    "/Subject": "主题描述"
}

with open("output.pdf", "wb") as output_file:
    writer.write(output_file)
print("元信息修改完成！")

from pypdf import PdfWriter

writer = PdfWriter()
writer.append("input.pdf")

# 设置密码
writer.encrypt(user_password="user123", owner_password="owner123")

with open("encrypted.pdf", "wb") as output_file:
    writer.write(output_file)
print("PDF 加密完成！")

from pypdf import PdfReader

reader = PdfReader("encrypted.pdf")

# 提供密码解密
reader.decrypt("user123")

for page in reader.pages:
    print(page.extract_text())

from pypdf import PdfReader, PdfWriter

reader = PdfReader("input.pdf")
writer = PdfWriter()

# 旋转每一页
for page in reader.pages:
    page.rotate(90)  # 顺时针旋转 90 度
    writer.add_page(page)

with open("rotated.pdf", "wb") as output_file:
    writer.write(output_file)
print("页面旋转完成！")

from pypdf import PdfReader, PdfWriter

reader = PdfReader("input.pdf")
writer = PdfWriter()

for page in reader.pages:
    # 设置裁剪框 (左, 下, 右, 上)
    page.mediabox.lower_left = (50, 50)
    page.mediabox.upper_right = (500, 700)
    writer.add_page(page)

with open("cropped.pdf", "wb") as output_file:
    writer.write(output_file)
print("页面裁剪完成！")