dataannotation.py

from bs4 import BeautifulSoup as bs
import requests
import sys

def main(url):
    if not url:
        url = input("url:")
    rs = requests.get(url)
    if rs.status_code != 200:
        return
    soup = bs(rs.text, 'html.parser')
    co = soup.get_text(separator='\n').strip()
    start = False
    filtered = ""
    for l in co.splitlines():
        if start:
            filtered += l + "\n"
        if l.strip() == "y-coordinate":
            start = True
    ls = filtered.strip().splitlines()
    groups = [ls[i:i+3] for i in range(0, len(ls), 3)]
    max_x, max_y = 0, 0
    for g in groups:
        if len(g) != 3:
            continue
        ix = int(g[0])
        iy = int(g[2])
        if ix > max_x:
            max_x = ix
        if iy > max_y:
            max_y = iy
    grid = [[" " for _ in range(max_x + 1)] for _ in range(max_y + 1)]
    for g in groups:
        if len(g) != 3:
            continue
        x_val = int(g[0])
        char = g[1]
        y_val = int(g[2])
        grid[y_val][x_val] = char
    for row in grid:
        print("".join(row))

if __name__ == '__main__':
    if len(sys.argv) > 1:
        url = sys.argv[1]
    else:
        url = 'https://docs.google.com/document/d/e/2PACX-1vSHesOf9hv2sPOntssYrEdubmMQm8lwjfwv6NPjjmIRYs_FOYXtqrYgjh85jBUebK9swPXh_a5TJ5Kl/pub'
    main(url)