PdfProcessing/add_watermark.py

import os
import sys

from PyPDF2 import PdfReader, PdfWriter
import requests
import PyPDF2


def add_watermark_to_pdf(pdf_file_in, pdf_file_mark, pdf_file_out):
    """把水印添加到pdf中"""

    input_stream = open(pdf_file_in, 'rb')
    pdf_input = PdfReader(input_stream, strict=False)

    # 获取PDF文件的页数
    # pageNum = pdf_input.getNumPages()
    pageNum = len(pdf_input.pages)

    pdf_output = PdfWriter()
    # 读入水印pdf文件
    pdf_watermark = PdfReader(open(pdf_file_mark, 'rb'), strict=False)
    # 给每一页打水印
    for i in range(pageNum):
        # page = pdf_input.getPage(i)
        page = pdf_input.pages[i]
        page.merge_page(pdf_watermark.pages[0])
        page.compress_content_streams()  # 压缩内容
        pdf_output.add_page(page)
    pdf_output.write(open(pdf_file_out, 'wb'))



if __name__ == '__main__':
   
    if len(sys.argv) < 3:
        print("Usage: add_watermark.py <url> <pdf_file_in>  <pdf_file_out>  <pdf_file_mark>")
        sys.exit(1)

    pdf_file_url = sys.argv[1]
    pdf_file_in = sys.argv[2] # 要加水印的文件名
    pdf_file_out = sys.argv[3] # 加好水印的结果文件
    pdf_file_mark = sys.argv[4]
    add_watermark(pdf_file_url, pdf_file_in, pdf_file_mark, pdf_file_out)
    print(pdf_file_out)