Merge pull request #143 from nixxo/gedi

[Gedi] Add new extractor

3 files changed, 276 insertions, 0 deletions

diff --git a/youtube_dlc/extractor/extractors.py b/youtube_dlc/extractor/extractors.py
index b7ebc6951..f5894504e 100644
--- a/youtube_dlc/extractor/extractors.py
+++ b/youtube_dlc/extractor/extractors.py
@@ -420,6 +420,10 @@ from .gamestar import GameStarIE
 from .gaskrank import GaskrankIE
 from .gazeta import GazetaIE
 from .gdcvault import GDCVaultIE
+from .gedi import (
+    GediIE,
+    GediEmbedsIE,
+)
 from .generic import GenericIE
 from .gfycat import GfycatIE
 from .giantbomb import GiantBombIE
diff --git a/youtube_dlc/extractor/gedi.py b/youtube_dlc/extractor/gedi.py
new file mode 100644
index 000000000..9d9d4acc2
--- /dev/null
+++ b/youtube_dlc/extractor/gedi.py
@@ -0,0 +1,266 @@
+# coding: utf-8
+from __future__ import unicode_literals
+
+import re
+
+from .common import InfoExtractor
+from ..compat import compat_str
+from ..utils import (
+    base_url,
+    url_basename,
+    urljoin,
+)
+
+
+class GediBaseIE(InfoExtractor):
+    @staticmethod
+    def _clean_audio_fmts(formats):
+        unique_formats = []
+        for f in formats:
+            if 'acodec' in f:
+                unique_formats.append(f)
+        formats[:] = unique_formats
+
+    def _real_extract(self, url):
+        video_id = self._match_id(url)
+
+        webpage = self._download_webpage(url, video_id)
+        player_data = re.findall(
+            r'PlayerFactory\.setParam\(\'(?P<type>.+?)\',\s*\'(?P<name>.+?)\',\s*\'(?P<val>.+?)\'\);',
+            webpage)
+
+        formats = []
+        audio_fmts = []
+        hls_fmts = []
+        http_fmts = []
+        title = ''
+        thumb = ''
+
+        fmt_reg = r'(?P<t>video|audio)-(?P<p>rrtv|hls)-(?P<h>[\w\d]+)(?:-(?P<br>[\w\d]+))?$'
+        br_reg = r'video-rrtv-(?P<br>\d+)-'
+
+        for t, n, v in player_data:
+            if t == 'format':
+                m = re.match(fmt_reg, n)
+                if m:
+                    # audio formats
+                    if m.group('t') == 'audio':
+                        if m.group('p') == 'hls':
+                            audio_fmts.extend(self._extract_m3u8_formats(
+                                v, video_id, 'm4a', m3u8_id='hls', fatal=False))
+                        elif m.group('p') == 'rrtv':
+                            audio_fmts.append({
+                                'format_id': 'mp3',
+                                'url': v,
+                                'tbr': 128,
+                                'ext': 'mp3',
+                                'vcodec': 'none',
+                                'acodec': 'mp3',
+                            })
+
+                    # video formats
+                    elif m.group('t') == 'video':
+                        # hls manifest video
+                        if m.group('p') == 'hls':
+                            hls_fmts.extend(self._extract_m3u8_formats(
+                                v, video_id, 'mp4', m3u8_id='hls', fatal=False))
+                        # direct mp4 video
+                        elif m.group('p') == 'rrtv':
+                            if not m.group('br'):
+                                mm = re.search(br_reg, v)
+                            http_fmts.append({
+                                'format_id': 'https-' + m.group('h'),
+                                'protocol': 'https',
+                                'url': v,
+                                'tbr': int(m.group('br')) if m.group('br') else
+                                (int(mm.group('br')) if mm.group('br') else 0),
+                                'height': int(m.group('h'))
+                            })
+
+            elif t == 'param':
+                if n == 'videotitle':
+                    title = v
+                if n == 'image_full_play':
+                    thumb = v
+
+        title = self._og_search_title(webpage) if title == '' else title
+
+        # clean weird char
+        title = compat_str(title).encode('utf8', 'replace').replace(b'\xc3\x82', b'').decode('utf8', 'replace')
+
+        if audio_fmts:
+            self._clean_audio_fmts(audio_fmts)
+            self._sort_formats(audio_fmts)
+        if hls_fmts:
+            self._sort_formats(hls_fmts)
+        if http_fmts:
+            self._sort_formats(http_fmts)
+
+        formats.extend(audio_fmts)
+        formats.extend(hls_fmts)
+        formats.extend(http_fmts)
+
+        return {
+            'id': video_id,
+            'title': title,
+            'description': self._html_search_meta('twitter:description', webpage),
+            'thumbnail': thumb,
+            'formats': formats,
+        }
+
+
+class GediIE(GediBaseIE):
+    _VALID_URL = r'''(?x)https?://video\.
+                    (?:
+                        (?:espresso\.)?repubblica
+                        |lastampa
+                        |huffingtonpost
+                        |ilsecoloxix
+                        |iltirreno
+                        |messaggeroveneto
+                        |ilpiccolo
+                        |gazzettadimantova
+                        |mattinopadova
+                        |laprovinciapavese
+                        |tribunatreviso
+                        |nuovavenezia
+                        |gazzettadimodena
+                        |lanuovaferrara
+                        |corrierealpi
+                        |lasentinella
+                    )
+                    (?:\.gelocal)?\.it/(?!embed/).+?/(?P<id>[\d/]+)(?:\?|\&|$)'''
+    _TESTS = [{
+        'url': 'https://video.lastampa.it/politica/il-paradosso-delle-regionali-la-lega-vince-ma-sembra-aver-perso/121559/121683',
+        'md5': '84658d7fb9e55a6e57ecc77b73137494',
+        'info_dict': {
+            'id': '121559/121683',
+            'ext': 'mp4',
+            'title': 'Il paradosso delle Regionali: ecco perché la Lega vince ma sembra aver perso',
+            'description': 'md5:de7f4d6eaaaf36c153b599b10f8ce7ca',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }, {
+        'url': 'https://video.repubblica.it/motori/record-della-pista-a-spa-francorchamps-la-pagani-huayra-roadster-bc-stupisce/367415/367963',
+        'md5': 'e763b94b7920799a0e0e23ffefa2d157',
+        'info_dict': {
+            'id': '367415/367963',
+            'ext': 'mp4',
+            'title': 'Record della pista a Spa Francorchamps, la Pagani Huayra Roadster BC stupisce',
+            'description': 'md5:5deb503cefe734a3eb3f07ed74303920',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }, {
+        'url': 'https://video.ilsecoloxix.it/sport/cassani-e-i-brividi-azzurri-ai-mondiali-di-imola-qui-mi-sono-innamorato-del-ciclismo-da-ragazzino-incredibile-tornarci-da-ct/66184/66267',
+        'md5': 'e48108e97b1af137d22a8469f2019057',
+        'info_dict': {
+            'id': '66184/66267',
+            'ext': 'mp4',
+            'title': 'Cassani e i brividi azzurri ai Mondiali di Imola: \\"Qui mi sono innamorato del ciclismo da ragazzino, incredibile tornarci da ct\\"',
+            'description': 'md5:fc9c50894f70a2469bb9b54d3d0a3d3b',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }, {
+        'url': 'https://video.iltirreno.gelocal.it/sport/dentro-la-notizia-ferrari-cosa-succede-a-maranello/141059/142723',
+        'md5': 'a6e39f3bdc1842bbd92abbbbef230817',
+        'info_dict': {
+            'id': '141059/142723',
+            'ext': 'mp4',
+            'title': 'Dentro la notizia - Ferrari, cosa succede a Maranello',
+            'description': 'md5:9907d65b53765681fa3a0b3122617c1f',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }]
+
+
+class GediEmbedsIE(GediBaseIE):
+    _VALID_URL = r'''(?x)https?://video\.
+                    (?:
+                        (?:espresso\.)?repubblica
+                        |lastampa
+                        |huffingtonpost
+                        |ilsecoloxix
+                        |iltirreno
+                        |messaggeroveneto
+                        |ilpiccolo
+                        |gazzettadimantova
+                        |mattinopadova
+                        |laprovinciapavese
+                        |tribunatreviso
+                        |nuovavenezia
+                        |gazzettadimodena
+                        |lanuovaferrara
+                        |corrierealpi
+                        |lasentinella
+                    )
+                    (?:\.gelocal)?\.it/embed/.+?/(?P<id>[\d/]+)(?:\?|\&|$)'''
+    _TESTS = [{
+        'url': 'https://video.huffingtonpost.it/embed/politica/cotticelli-non-so-cosa-mi-sia-successo-sto-cercando-di-capire-se-ho-avuto-un-malore/29312/29276?responsive=true&el=video971040871621586700',
+        'md5': 'f4ac23cadfea7fef89bea536583fa7ed',
+        'info_dict': {
+            'id': '29312/29276',
+            'ext': 'mp4',
+            'title': 'Cotticelli: \\"Non so cosa mi sia successo. Sto cercando di capire se ho avuto un malore\\"',
+            'description': 'md5:d41d8cd98f00b204e9800998ecf8427e',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }, {
+        'url': 'https://video.espresso.repubblica.it/embed/tutti-i-video/01-ted-villa/14772/14870&width=640&height=360',
+        'md5': '0391c2c83c6506581003aaf0255889c0',
+        'info_dict': {
+            'id': '14772/14870',
+            'ext': 'mp4',
+            'title': 'Festival EMERGENCY, Villa: «La buona informazione aiuta la salute» (14772-14870)',
+            'description': 'md5:2bce954d278248f3c950be355b7c2226',
+            'thumbnail': r're:^https://www\.repstatic\.it/video/photo/.+?-thumb-social-play\.jpg$',
+        },
+    }]
+
+    @staticmethod
+    def _sanitize_urls(urls):
+        # add protocol if missing
+        for i, e in enumerate(urls):
+            if e.startswith('//'):
+                urls[i] = 'https:%s' % e
+        # clean iframes urls
+        for i, e in enumerate(urls):
+            urls[i] = urljoin(base_url(e), url_basename(e))
+        return urls
+
+    @staticmethod
+    def _extract_urls(webpage):
+        entries = [
+            mobj.group('url')
+            for mobj in re.finditer(r'''(?x)
+            (?:
+                data-frame-src=|
+                <iframe[^\n]+src=
+            )
+            (["'])
+                (?P<url>https?://video\.
+                    (?:
+                        (?:espresso\.)?repubblica
+                        |lastampa
+                        |huffingtonpost
+                        |ilsecoloxix
+                        |iltirreno
+                        |messaggeroveneto
+                        |ilpiccolo
+                        |gazzettadimantova
+                        |mattinopadova
+                        |laprovinciapavese
+                        |tribunatreviso
+                        |nuovavenezia
+                        |gazzettadimodena
+                        |lanuovaferrara
+                        |corrierealpi
+                        |lasentinella
+                    )
+                    (?:\.gelocal)?\.it/embed/.+?)
+            \1''', webpage)]
+        return GediEmbedsIE._sanitize_urls(entries)
+
+    @staticmethod
+    def _extract_url(webpage):
+        urls = GediEmbedsIE._extract_urls(webpage)
+        return urls[0] if urls else None
diff --git a/youtube_dlc/extractor/generic.py b/youtube_dlc/extractor/generic.py
index 74760a180..e5d29f316 100644
--- a/youtube_dlc/extractor/generic.py
+++ b/youtube_dlc/extractor/generic.py
@@ -120,6 +120,7 @@ from .expressen import ExpressenIE
 from .zype import ZypeIE
 from .odnoklassniki import OdnoklassnikiIE
 from .kinja import KinjaEmbedIE
+from .gedi import GediEmbedsIE
 from .rcs import RCSEmbedsIE
 from .bitchute import BitChuteIE
 
@@ -3215,6 +3216,11 @@ class GenericIE(InfoExtractor):
                 zype_urls, video_id, video_title, ie=ZypeIE.ie_key())
 
         # Look for RCS media group embeds
+        gedi_urls = GediEmbedsIE._extract_urls(webpage)
+        if gedi_urls:
+            return self.playlist_from_matches(
+                gedi_urls, video_id, video_title, ie=GediEmbedsIE.ie_key())
+
         rcs_urls = RCSEmbedsIE._extract_urls(webpage)
         if rcs_urls:
             return self.playlist_from_matches(