Get article from Telegra.ph and transform resulting html

telegraph_article.py

import re
import os
import argparse
from urllib import request
from urllib.parse import urljoin
from html.parser import HTMLParser


class ArticleParser(HTMLParser):

    IMAGES_DIR = 'images'

    def __init__(self, base_url):
        super().__init__()

        self.base_url = base_url
        self.resulting_html = ''
        self._appending = False
        self._data_buf = ''
        self._tags_stack = []

    @property
    def html(self):
        return self.resulting_html

    @staticmethod
    def _wrap_in_tag(tag, data):
        return '<{tag}>{data}</{tag}>'.format(tag=tag, data=data.lstrip())

    def handle_starttag(self, tag, attrs):
        if tag == 'br' and self._appending:
            self._data_buf += '\n'

        if tag == 'br':
            return

        # Download images
        if tag == 'img':
            for attr, value in attrs:
                if attr == 'src':
                    filename = self.download_file(value)
                    new_url_path = os.path.join('/', self.IMAGES_DIR, filename)
                    self.resulting_html += self._wrap_in_tag('figure', '<img src="{}">'.format(new_url_path))
                    return

        # Handle section title
        if tag == 'h3':
            self._appending = True

        # Handle main text
        if tag == 'p':
            self._appending = True

        # Handle quotes
        if tag == 'blockquote':
            self._appending = True

        self._tags_stack.append(tag)

    def download_file(self, path):
        filename = os.path.basename(path)
        filepath = os.path.join(self.IMAGES_DIR, filename)
        url = urljoin(self.base_url, path)
        request.urlretrieve(url, filename=filepath)
        return filename

    def handle_endtag(self, tag):
        if tag == 'br':
            return

        if not self._tags_stack:
            raise ValueError('Open/closing tags are not balanced')

        current_tag = self._tags_stack.pop()

        if tag in ('p', 'h3', 'blockquote'):
            if current_tag != tag:
                raise ValueError('Invalid closing tag: %s. Current on stack: %s.', tag, current_tag)
            if self._data_buf:
                self.resulting_html += self._wrap_in_tag(current_tag, self._data_buf)

            self._appending = False
            self._data_buf = ''

    def handle_data(self, data):
        if self._appending:
            self._data_buf += data


def find_tag(tag_name, html_data):
    exp = '<{tag_name}[^>]*>(.*?)</{tag_name}>'.format(tag_name=tag_name)
    m = re.search(exp, html_data)
    result = m.group(1)  # Match within a tag
    return result


def parse_args():
    parser = argparse.ArgumentParser(description='Grab article from telegra.ph')
    parser.add_argument('--url', dest='url', required=True,
                        help='URL to the article')

    args = parser.parse_args()
    return args