1#!/usr/bin/env python
  2
  3# This script parses output from Telegram channel and converts each post to 
  4# jekyll-applicable post in markdown.
  5#
  6# Telegram creates result.json file, as well as different directories containing
  7# multimedia, photos, etc. This script creates new directory and populates it 
  8# with formatted posts ready to publish.
  9#
 10# TODO summary:
 11# - replies
 12# - single/muliple tags
 13# - forwarded posts
 14
 15import os
 16# import sys
 17import json
 18from datetime import datetime
 19
 20# post:
 21# header
 22# [photo?]
 23# text
 24# [media?]
 25
 26# text:
 27# [str|list(str|obj, ...)]
 28
 29def print_post_header(post_title, post_date, post_tag):
 30    # TODO: handle post tag/tags
 31    post_header = '---\ntitle: {title}\ndate: {date}\n\
 32tag: {tag}\nlayout: post\n---\n'.format(\
 33            title=post_title, date=post_date, tag=post_tag)
 34
 35    return post_header
 36
 37def parse_post_photo(post):
 38    post_photo = '![image]({src})\n\n'.format(src=post['photo'])
 39
 40    return post_photo
 41
 42def md_str(string):
 43    string = string.replace('\n','\n\n')
 44    string = string.replace('. ', '.\n')
 45
 46    return string
 47
 48
 49def parse_text_object(obj):
 50    '''
 51    Parse object from post text.
 52
 53    Objects are text links, plain links, underlined text, strikethrough text,
 54    italic text, bold text, code blocks and hashtags.
 55
 56    This is a mess, but what is better?
 57    '''
 58
 59    obj_type = obj['type']
 60    obj_text = obj['text']
 61
 62    if obj_type == 'hashtag':
 63        post_tag = obj_text
 64        return post_tag
 65
 66    elif obj_type == 'text_link':
 67        post_text_link = '[{text}]({href})'.format(text=obj_text, \
 68                href=obj['href'])
 69        return post_text_link
 70
 71    elif obj_type == 'link':
 72        post_link = '[link]({href})'.format(href=obj_text)
 73        return post_link
 74
 75    # I dunno how this appeared, but it seems like hyphenated numbers
 76    # are treated as phone numbers, so return them as plain text.
 77    elif obj_type == 'phone':
 78        return obj_text
 79
 80    elif obj_type == 'bold':
 81        post_inline_bold = '**{text}**'.format(text=obj_text)
 82        return post_inline_bold
 83
 84    elif obj_type == 'italic':
 85        post_inline_italic = '*{text}*'.format(text=obj_text)
 86        return post_inline_italic
 87
 88    elif obj_type == 'underline':
 89        post_inline_underline = '<u>{text}</u>'.format(text=obj_text)
 90        return post_inline_underline
 91
 92    elif obj_type == 'strikethrough':
 93        post_inline_strike = '<s>{text}</s>'.format(text=obj_text)
 94        return post_inline_strike
 95
 96    elif obj_type == 'code':
 97        post_inline_code = '```\n{text}\n```'.format(text=obj_text)
 98        return post_inline_code
 99
100
101def parse_post_text(post):
102    # TODO: handle reply-to
103    post_raw_text = post['text']
104    post_parsed_text = ''
105
106    if type(post_raw_text) == str:
107        return str(post_raw_text)
108
109    else:
110        for obj in post_raw_text:
111            if type(obj) == str:
112                post_parsed_text += obj
113            else:
114                post_parsed_text += str(parse_text_object(obj))
115
116        return post_parsed_text
117
118
119def parse_post_media(post):
120    post_media = '<audio controls>\n \
121    <source src="{src}" type="{mime_type}">\n \
122    </audio>'.format(src=post['file'], mime_type=post['mime_type'])
123
124    return post_media
125    
126
127def parse_post(post):
128    post_output = ''
129    
130    # optional image
131    if 'photo' in post:
132        post_output += str(parse_post_photo(post))
133
134    # post text
135    post_output += md_str(parse_post_text(post))
136
137    # optional media
138    if 'media_type' in post:
139        post_output += str(parse_post_media(post))
140
141    return post_output
142
143
144def main():
145    # create output directory
146    out_dir = './formatted_posts'
147    try:
148        os.mkdir(out_dir)
149    except FileExistsError as e:
150        pass
151
152    # load json file
153    with open('result.json', 'r') as f:
154        data = json.load(f)
155
156    # load only messages
157    raw_posts = data['messages']
158
159    for post in raw_posts:
160    # TODO: handle forwarded posts
161        if post['type'] == 'message' and 'forwarded_from' not in post:
162
163            post_date = datetime.fromisoformat(post['date'])
164            post_id = post['id']
165            post_filename = out_dir + '/' + str(post_date.date()) + '-' \
166                    + str(post_id) + '.md'
167
168            with open (post_filename, 'w') as f:
169                print(print_post_header(
170                    post_id, post_date.date(), None), 
171                    file=f)
172                print(parse_post(post), file=f)
173
174
175if __name__ == '__main__':
176    main()
177