Merge PDFs into one and rescale them all to A4 width (21cm or 595px)

merge_rescale_pdf.py

from PIL import Image
from PyPDF2 import PdfWriter, PdfReader
import tempfile

PREPEND_FILE = "打分框.pdf"


def scale_pdf_to_a4(input):
    reader = input
    if not isinstance(input, PdfReader):
        reader = PdfReader(input)
    for page in reader.pages:
        scale = 595.0 / float(page.mediabox.width)
        page.scale_by(scale)
    return reader


def merge_files_into_pdf(fs, output):
    merger = PdfWriter()

    def append(x):
        if isinstance(x, str):
            x = PdfReader(x)
        x = scale_pdf_to_a4(x)
        for page in x.pages:
            merger.add_page(page)

    append(PREPEND_FILE)
    for f in fs:
        if f.endswith(".pdf"):
            append(f)
        elif (
            f.endswith(".png")
            or f.endswith(".jpg")
            or f.endswith(".jpeg")
            or f.endswith(".jfif")
            or f.endswith(".pjpeg")
            or f.endswith(".pjp")
            or f.endswith(".svg")
            or f.endswith(".gif")
            or f.endswith(".webp")
            or f.endswith(".bmp")
            or f.endswith(".ico")
            or f.endswith(".tif")
            or f.endswith(".tiff")
        ):
            # create a temporary file
            temp = tempfile.NamedTemporaryFile(suffix=".pdf")
            # convert image into pdf and save it to temp
            img = Image.open(f)
            width, height = img.size
            desired_resolution = 418 / 3456 * width
            img.save(temp, "PDF", resolution=desired_resolution, save_all=True)
            # read temp and append it to merger
            temp.seek(0)
            append(temp)
    merger.write(output)