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# - custom post header
 15
 16import os
 17import sys
 18import json
 19from datetime import datetime
 20
 21
 22def print_post_header(post_title, post_date, post_tag):
 23    # TODO: handle post tag/tags
 24    # TODO: support for custom header
 25    post_header = '---\ntitle: {title}\ndate: {date}\n\
 26tag: {tag}\nlayout: post\n---\n'.format(\
 27            title=post_title, date=post_date, tag=post_tag)
 28
 29    return post_header
 30
 31
 32def parse_post_photo(post, media_dir):
 33    post_photo_src = post['photo'][7:]
 34    post_photo_src = media_dir + '/' + post_photo_src
 35    post_photo = '![image]({src})\n\n'.format(\
 36            src=post_photo_src)
 37
 38    return post_photo
 39
 40
 41# def md_str(string):
 42    # string = string.replace('\n','\n\n')
 43    # string = string.replace('. ', '.\n')
 44
 45    # return string
 46
 47
 48def text_format(string, fmt):
 49    if fmt in ('*', '**', '***', '`', '```'):
 50        output = '{fmt}{txt}{fmt}'
 51    elif fmt == '```':
 52        output = '{fmt}\n{txt}\n{fmt}'
 53    else:
 54        output = '<{fmt}>{txt}</{fmt}>'
 55
 56    output = output.format(fmt=fmt, txt=string.strip())
 57    output += '\n' * string.split('\n').count('') * string.endswith('\n')
 58    return output
 59
 60def text_link_format(text, link):
 61    link_fmt = '[{text}]({href})'
 62    link_fmt = link_fmt.format(text=text.strip(), href=link)
 63    link_fmt += '\n' * text.count('\n') * text.endswith('\n')
 64    return link_fmt
 65
 66
 67def parse_text_object(obj):
 68
 69    obj_type = obj['type']
 70    obj_text = obj['text']
 71
 72    if obj_type == 'hashtag':
 73        post_tag = obj_text
 74        return post_tag
 75
 76    elif obj_type == 'text_link':
 77        return text_link_format(obj_text, obj['href'])
 78
 79    elif obj_type == 'link' or obj_type == 'email':
 80        post_link = '<{href}>'.format(href=obj_text.strip())
 81        return post_link
 82
 83    elif obj_type == 'phone':
 84        return obj_text
 85
 86    elif obj_type == 'italic':
 87        return text_format(obj_text, '*')
 88
 89    elif obj_type == 'bold':
 90        return text_format(obj_text, '**')
 91
 92    elif obj_type == 'code':
 93        return text_format(obj_text, '`')
 94
 95    elif obj_type == 'pre':
 96        return text_format(obj_text, '```')
 97
 98    elif obj_type == 'underline':
 99        return text_format(obj_text, 'u')
100
101    elif obj_type == 'strikethrough':
102        return text_format(obj_text, 's')
103
104
105def parse_post_text(post):
106    # TODO: handle reply-to
107    post_raw_text = post['text']
108    post_parsed_text = ''
109
110    if type(post_raw_text) == str:
111        return str(post_raw_text)
112
113    else:
114        for obj in post_raw_text:
115            if type(obj) == str:
116                post_parsed_text += obj
117            else:
118                post_parsed_text += str(parse_text_object(obj))
119
120        return post_parsed_text
121
122
123def parse_post_media(post, media_dir):
124    # get filename without parent directory
125    post_media_src = post['file'][post['file'].rfind("/") + 1:]
126
127    # add parent directory
128    post_media_src = media_dir + '/' + post_media_src
129    post_media = '\n<audio controls>\n \
130        <source src="{src}" type="{mime_type}">\n \
131        </audio>'.format(src=post_media_src, mime_type=post['mime_type'])
132
133    return post_media
134    
135
136def parse_post(post):
137    post_output = ''
138    
139    # optional image
140    photo_dir = '/photos'
141    if 'photo' in post:
142        post_output += str(parse_post_photo(post, photo_dir))
143
144    # post text
145    post_output += str(parse_post_text(post))
146
147    # optional media
148    media_dir = '/files'
149    if 'media_type' in post:
150        post_output += str(parse_post_media(post, media_dir))
151
152    return post_output
153
154
155def main():
156    # try directory from first argument
157    try:
158        input_dir = sys.argv[1]
159    except IndexError as e:
160        # if it's not specified, use current directory
161        input_dir = '.'
162
163    # create output directory
164    out_dir = input_dir + '/' + 'formatted_posts'
165    try:
166        os.mkdir(out_dir)
167    except FileExistsError as e:
168        pass
169
170    # load json file
171    json_path = input_dir + '/' + 'result.json'
172    try:
173        with open(json_path, 'r') as f:
174            data = json.load(f)
175    except FileNotFoundError as e:
176        sys.exit('result.json not found.\nPlease, specify right directory')
177
178    # load only messages
179    raw_posts = data['messages']
180
181    for post in raw_posts:
182    # TODO: handle forwarded posts
183        if post['type'] == 'message' and 'forwarded_from' not in post:
184
185            post_date = datetime.fromisoformat(post['date'])
186            post_id = post['id']
187            post_filename = out_dir + '/' + str(post_date.date()) + '-' \
188                    + str(post_id) + '.md'
189
190            with open (post_filename, 'w') as f:
191                print(print_post_header(
192                    post_id, post_date, None), 
193                    file=f)
194                print(parse_post(post), file=f)
195
196
197if __name__ == '__main__':
198    main()
199