Debian Codegraph @

recipes/telerama.recipe

raw
1 #!/usr/bin/env python
2 # vim:fileencoding=utf-8
3 from __future__ import absolute_import, division, print_function, unicode_literals
4
5 __license__ = 'GPL v3'
6 __copyright__ = '2026, Kabonix'
7
8 import json
9 import re
10 from urllib.parse import urlparse
11
12 from calibre.web.feeds.news import BasicNewsRecipe
13
14
15 class TeleramaPremium(BasicNewsRecipe):
16 title = 'Télérama'
17 __author__ = 'Kabonix'
18 description = 'Édition complète (API Bypass) - Cover HD & Lecture Pure'
19 publisher = 'Télérama'
20 language = 'fr'
21 encoding = 'utf-8'
22
23 oldest_article = 7
24 max_articles_per_feed = 50
25 no_stylesheets = True
26 ignore_duplicate_articles = {'title', 'url'}
27
28 # On laisse les images des articles tranquilles
29 scale_news_images = None
30
31 # --- API ---
32 headers = {'User-Agent': 'Telerama/4.3.5 (Android; 14)', 'X-Lmd-Token': 'TWPLMOLMO', 'Accept': 'application/json'}
33
34 def get_browser(self, *args, **kwargs):
35 br = BasicNewsRecipe.get_browser(self, *args, **kwargs)
36 for name, val in self.headers.items():
37 br.addheaders.append((name, val))
38 return br
39
40 # --- COUVERTURE DYNAMIQUE (Ta demande) ---
41 def get_cover_url(self):
42 cover_url = None
43 try:
44 self.log('🔍 Recherche de la dernière couverture...')
45 # On va sur la page kiosque
46 soup = self.index_to_soup('https://www.telerama.fr/kiosque/telerama')
47
48 # On cherche le premier élément "popin-link" avec une data-cover-url
49 # Le premier de la liste est toujours le dernier numéro paru
50 link = soup.find('a', attrs={'class': 'popin-link', 'data-cover-url': True})
51
52 if link:
53 url = link['data-cover-url']
54 # L'URL contient /180/ (basse déf). On passe en HD /1200/
55 # Ex: .../0/0/180/0/... -> .../0/0/1200/0/...
56 cover_url = url.replace('/180/', '/1200/')
57 self.log(f'✅ Couverture trouvée : {cover_url}')
58 else:
59 self.log('⚠️ Aucune couverture trouvée dans le kiosque.')
60 except Exception as e:
61 self.log(f'❌ Erreur récupération couverture : {e}')
62
63 return cover_url
64
65 # --- BYPASS API ---
66 def get_article_url(self, article):
67 url = article.get('link', article.get('url', ''))
68 path = urlparse(url).path
69 return 'https://apps.telerama.fr/tlr/v1/premium-android-phone/element?id={}'.format(path)
70
71 # --- JSON TO HTML ---
72 def preprocess_raw_html(self, raw_html, url):
73 if '/tlr/v1/' in url:
74 try:
75 data = json.loads(raw_html)
76 content = ''
77 title = 'Télérama'
78
79 if 'templates' in data and 'raw_content' in data['templates']:
80 content = data['templates']['raw_content']['content']
81 elif 'body' in data:
82 content = data['body']
83
84 if 'template_vars' in data:
85 title = data['template_vars'].get('share_title', 'Article Télérama')
86
87 if not content:
88 return '<html><body><h2>Contenu vide</h2></body></html>'
89
90 # Nettoyage préventif
91 content = content.replace('{{{ scripts_bottom }}}', '')
92 content = re.sub(r'>\s*[•·]\s*<', '><', content)
93
94 # Fix images
95 content = content.replace('{{width}}', '1200').replace('{{height}}', '')
96 content = content.replace('%7B%7Bwidth%7D%7D', '1200')
97
98 html = f'<html><head><title>{title}</title></head><body><h1 id="main-title">{title}</h1>{content}</body></html>'
99 return html
100
101 except Exception as e:
102 self.log(f'Erreur JSON : {e}')
103 return raw_html
104 return raw_html
105
106 # --- NETTOYAGE ---
107 def preprocess_html(self, soup):
108 # 1. Suppression doublons structurels
109 for header in soup.find_all(attrs={'class': re.compile(r'article__page-header|header__article', re.I)}):
110 header.decompose()
111 for ns in soup.find_all('noscript'):
112 ns.decompose()
113
114 # 2. Suppression "À lire aussi"
115 for p in soup.find_all(['p', 'h3', 'h4', 'div', 'aside']):
116 text = p.get_text().strip()
117 if re.search(r'^(À|A) lire aussi', text, re.IGNORECASE):
118 p.decompose()
119
120 # 3. Nettoyage Méta TV et Puces
121 for tag in soup.find_all(['p', 'div', 'span', 'li', 'ul']):
122 text = tag.get_text().strip()
123 normalized_text = re.sub(r'\s+', ' ', text)
124
125 # Puces seules
126 if re.match(r'^[\s\n\r•·|\-.]+$', text):
127 tag.decompose()
128 continue
129 # Mots clés TV seuls
130 if re.match(r'^(Direct|Inédit|Replay|\s)+$', normalized_text, re.IGNORECASE):
131 tag.decompose()
132 continue
133 # Mots clés TV avec séparateurs
134 if re.search(r'(Direct|Inédit|Replay)\s*[•·-]', text, re.IGNORECASE):
135 tag.decompose()
136
137 # 4. SUPPRESSION DES LIENS (Lecture Pure)
138 for a in soup.find_all('a'):
139 a.unwrap()
140
141 return soup
142
143 keep_only_tags = [
144 dict(name='h1', attrs={'id': 'main-title'}),
145 dict(attrs={'class': ['article__page-content', 'article-body']}),
146 ]
147
148 remove_tags = [
149 dict(attrs={'class': re.compile(r'paywall|premium-banner|banner|pubstack|marketing', re.I)}),
150 dict(attrs={'class': re.compile(r'sharing|social|bookmark|button|btn|openapp|listBtns', re.I)}),
151 dict(attrs={'class': re.compile(r'OUTBRAIN|forecast|overlay', re.I)}),
152 dict(name=['script', 'style', 'nav', 'footer', 'button', 'iframe']),
153 ]
154
155 extra_css = '''
156 h1 {
157 font-family: "Georgia", serif;
158 font-size: 1.5em;
159 font-weight: bold;
160 text-align: center;
161 margin-bottom: 0.5em;
162 color: #111;
163 }
164 .article__label-subscriber {
165 display: block; background-color: #ffe600; color: #000; font-weight: bold;
166 font-size: 0.8em; text-transform: uppercase; padding: 4px 8px;
167 margin: 0 auto 1em auto; width: fit-content; border-radius: 4px;
168 }
169 .article__chapeau { font-weight: bold; font-style: italic; margin: 1.5em 0; font-size: 1.1em; color: #444; }
170 p { text-align: justify; line-height: 1.5; margin-bottom: 1em; }
171 figure { margin: 1.5em 0; }
172 img { display: block; margin: 0 auto; max-width: 100%; height: auto; }
173 figcaption, .media__caption, .media__legend { font-size: 0.75em; color: #666; text-align: center; font-style: italic; margin-top: 0.5em; }
174 .author { font-weight: bold; margin-top: 2em; border-top: 1px solid #eee; padding-top: 1em; color: #333; }
175 a { color: inherit; text-decoration: none; pointer-events: none; }
176 '''
177
178 feeds = [
179 ('À la une', 'https://www.telerama.fr/rss/une.xml'),
180 ('Cinéma', 'https://www.telerama.fr/rss/cinema.xml'),
181 ('Séries', 'https://www.telerama.fr/rss/series.xml'),
182 ('Télévision', 'https://www.telerama.fr/rss/television.xml'),
183 ('Musique', 'https://www.telerama.fr/rss/musique.xml'),
184 ('Livres', 'https://www.telerama.fr/rss/livres.xml'),
185 ]