Plot PDF text/rect objects' using rows + Pillow

pdf_plot.py

# pip install pillow cached-property pdfminer.six https://github.com/turicas/rows/archive/develop.zip
import argparse

from rows.plugins.plugin_pdf import (
    RectObject,
    TextObject,
    PDFMinerBackend,
    group_objects,
    YGroupsAlgorithm,
    plot_objects,
    split_object_lines,
)
from PIL.ImageShow import register, DisplayViewer

register(DisplayViewer(), 0)  # Use `display` command on Image.show()

parser = argparse.ArgumentParser()
parser.add_argument("--width")
parser.add_argument("--height")
parser.add_argument("--page-number", type=int, default=1)
parser.add_argument("pdf_filename")
args = parser.parse_args()

doc = PDFMinerBackend(args.pdf_filename)
selected_page = None
for counter, page in enumerate(doc.objects(), start=1):
    if counter == args.page_number:
        selected_page = page
        break
img1 = plot_objects(selected_page, width=args.width, height=args.height)
img1.show()

page_split = []
for obj in selected_page:
    if isinstance(obj, TextObject):
        page_split.extend(split_object_lines(obj))
    else:
        page_split.append(obj)
img2 = plot_objects(page_split, width=args.width, height=args.height)
img2.show()