pdf_create.py

from pathlib import Path
import sys

from PIL import Image
from PyPDF4 import PdfFileMerger, PdfFileReader, PdfFileWriter

from general_functions import try_remove

def png_2_pdf(path: Path):
    """
        Converts all png files in path to PDFs
    """
    for file in path.iterdir():
        file = path / file
        if file.suffix == ".png":
            pdf_name = file.with_suffix(".pdf")
            img_rgba = Image.open(file)
            img_rgb = Image.new("RGB", img_rgba.size, (255, 255, 255))
            img_rgb.paste(img_rgba, mask=img_rgba.split()[3])
            img_rgb.save(pdf_name, "PDF", resolution=100)


def merge_pdfs(path: Path):
    """
        Merge all PDF files in path into one results.pdf.
        Rotates all files that start with 'x'. These are output of dotter and
        need to be rotated for better readability.
        Deletes all used PDF files afterwards.
    """
    out_file = path / "results.pdf"
    if out_file.is_file():
        try_remove(out_file)
    merger = PdfFileMerger()
    delete_afterwards = []
    for file_name in sorted(path.iterdir()):
        file = path / file_name
        if file.suffix == ".pdf" and file.stat().st_size > 0:
            delete_afterwards.append(file)
            if (fn:=file.name).startswith("x"):
                #rotate (as dotter renders weird pdfs)
                pdf_in = PdfFileReader(open(file, 'rb'))
                pdf_writer = PdfFileWriter()
                pdf_writer.addPage(pdf_in.getPage(0).rotateClockwise(90))
                new_file_name = file.parent / fn.replace("x", "dotplot")
                pdf_writer.write(open(new_file_name, "wb"))
                # delete_afterwards.append(new_file_name)
                file = new_file_name
            merger.append(str(file))

    merger.write(str(out_file))
    merger.close()

    for file in delete_afterwards:
        try_remove(file)


if __name__ == "__main__":
    main_dir = Path(sys.argv[1])
    for conflict_dir in main_dir.iterdir():
        conflict_dir = main_dir / conflict_dir
        png_2_pdf(conflict_dir)
        merge_pdfs(conflict_dir)