代码示例

pip install sopdf

import sopdf

with sopdf.open("document.pdf") as doc:
    img_bytes = doc[0].render(dpi=150)          # 返回 PNG 字节
    doc[0].render_to_file("page-1.png", dpi=300)

import sopdf

with sopdf.open("document.pdf") as doc:
    images = sopdf.render_pages(doc.pages, dpi=150, parallel=True)
    print(f"rendered: {len(images)} pages")

import sopdf

with sopdf.open("document.pdf") as doc:
    text = doc[0].get_text()
    blocks = doc[0].get_text_blocks()  # list[TextBlock]，包含边界框
    print(text[:200])
    print(len(blocks))

import sopdf

with sopdf.open("document.pdf") as doc:
    hits = doc[0].search("invoice", match_case=False)  # list[Rect]
    print(f"hit count: {len(hits)}")

import sopdf

with sopdf.open("source.pdf") as doc:
    chapter = doc.split(pages=[0, 1, 2], output="chapter-1.pdf")
    doc.split_each(output_dir="pages")

sopdf.merge(["intro.pdf", "body.pdf"], output="book.pdf")

import sopdf

with sopdf.open("source.pdf") as doc:
    doc.save("output.pdf", compress=True, garbage=True)
    raw_bytes = doc.to_bytes()  # 不落盘，直接得到 bytes
    print(len(raw_bytes))

import sopdf

with sopdf.open("document.pdf") as doc:
    doc[0].rotation = 90
    doc.save("rotated.pdf")

import sopdf

with open("document.pdf", "rb") as f:
    data = f.read()

with sopdf.open(stream=data) as doc:
    print(doc.page_count)

import sopdf

with sopdf.open("document.pdf") as doc:
    print(doc.metadata.title)
    doc.metadata.title = "Updated Title"
    print(doc.metadata.creation_datetime)

    for item in doc.outline.items:
        print(f"[p{item.page + 1}] {item.title}")

    flat = doc.outline.to_list()  # PyMuPDF 兼容的扁平目录结构
    print(len(flat))

import sopdf

with sopdf.open("protected.pdf", password="hunter2") as doc:
    doc.save("unlocked.pdf")

with sopdf.open("corrupted.pdf") as doc:
    doc.save("repaired.pdf")