Prepare to release.

[youtubedl] / youtube_dl / extractor / udemy.py

diff --git a/youtube_dl/extractor/udemy.py b/youtube_dl/extractor/udemy.py
index f5b5e7fd6775b4194ca18265f4c78f639a5f092a..cce29c6e07134c58425f5486b15ce6d9e017d0d7 100644 (file)
--- a/youtube_dl/extractor/udemy.py
+++ b/youtube_dl/extractor/udemy.py
@@ -1,23 +1,38 @@
 from __future__ import unicode_literals
 
 from __future__ import unicode_literals
 

+import re
+

 from .common import InfoExtractor
 from ..compat import (
     compat_HTTPError,
 from .common import InfoExtractor
 from ..compat import (
     compat_HTTPError,

-    compat_urllib_parse,
+    compat_str,

     compat_urllib_request,
     compat_urllib_request,

+    compat_urlparse,

 )
 from ..utils import (
 )
 from ..utils import (

+    determine_ext,
+    extract_attributes,

     ExtractorError,
     float_or_none,
     int_or_none,
     sanitized_Request,
     unescapeHTML,
     ExtractorError,
     float_or_none,
     int_or_none,
     sanitized_Request,
     unescapeHTML,

+    urlencode_postdata,

 )
 
 
 class UdemyIE(InfoExtractor):
     IE_NAME = 'udemy'
 )
 
 
 class UdemyIE(InfoExtractor):
     IE_NAME = 'udemy'

-    _VALID_URL = r'https?://www\.udemy\.com/(?:[^#]+#/lecture/|lecture/view/?\?lectureId=)(?P<id>\d+)'
+    _VALID_URL = r'''(?x)
+                    https?://
+                        www\.udemy\.com/
+                        (?:
+                            [^#]+\#/lecture/|
+                            lecture/view/?\?lectureId=|
+                            [^/]+/learn/v4/t/lecture/
+                        )
+                        (?P<id>\d+)
+                    '''

     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
     _ORIGIN_URL = 'https://www.udemy.com'
     _NETRC_MACHINE = 'udemy'
     _LOGIN_URL = 'https://www.udemy.com/join/login-popup/?displayType=ajax&showSkipButton=1'
     _ORIGIN_URL = 'https://www.udemy.com'
     _NETRC_MACHINE = 'udemy'


@@ -33,36 +48,55 @@ class UdemyIE(InfoExtractor):
             'duration': 579.29,
         },
         'skip': 'Requires udemy account credentials',
             'duration': 579.29,
         },
         'skip': 'Requires udemy account credentials',

+    }, {
+        # new URL schema
+        'url': 'https://www.udemy.com/electric-bass-right-from-the-start/learn/v4/t/lecture/4580906',
+        'only_matching': True,

     }]
 
     }]
 

-    def _enroll_course(self, webpage, course_id):
+    def _extract_course_info(self, webpage, video_id):
+        course = self._parse_json(
+            unescapeHTML(self._search_regex(
+                r'ng-init=["\'].*\bcourse=({.+?});', webpage, 'course', default='{}')),
+            video_id, fatal=False) or {}
+        course_id = course.get('id') or self._search_regex(
+            (r'&quot;id&quot;\s*:\s*(\d+)', r'data-course-id=["\'](\d+)'),
+            webpage, 'course id')
+        return course_id, course.get('title')
+
+    def _enroll_course(self, base_url, webpage, course_id):
+        def combine_url(base_url, url):
+            return compat_urlparse.urljoin(base_url, url) if not url.startswith('http') else url
+

         checkout_url = unescapeHTML(self._search_regex(
         checkout_url = unescapeHTML(self._search_regex(

-            r'href=(["\'])(?P<url>https?://(?:www\.)?udemy\.com/payment/checkout/.+?)\1',
+            r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/payment/checkout/.+?)\1',

             webpage, 'checkout url', group='url', default=None))
         if checkout_url:
             raise ExtractorError(
                 'Course %s is not free. You have to pay for it before you can download. '
             webpage, 'checkout url', group='url', default=None))
         if checkout_url:
             raise ExtractorError(
                 'Course %s is not free. You have to pay for it before you can download. '

-                'Use this URL to confirm purchase: %s' % (course_id, checkout_url), expected=True)
+                'Use this URL to confirm purchase: %s'
+                % (course_id, combine_url(base_url, checkout_url)),
+                expected=True)

 
         enroll_url = unescapeHTML(self._search_regex(
 
         enroll_url = unescapeHTML(self._search_regex(

-            r'href=(["\'])(?P<url>https?://(?:www\.)?udemy\.com/course/subscribe/.+?)\1',
+            r'href=(["\'])(?P<url>(?:https?://(?:www\.)?udemy\.com)?/course/subscribe/.+?)\1',

             webpage, 'enroll url', group='url', default=None))
         if enroll_url:
             webpage, 'enroll url', group='url', default=None))
         if enroll_url:

-            webpage = self._download_webpage(enroll_url, course_id, 'Enrolling in the course')
+            webpage = self._download_webpage(
+                combine_url(base_url, enroll_url),
+                course_id, 'Enrolling in the course',
+                headers={'Referer': base_url})

             if '>You have enrolled in' in webpage:
                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 
     def _download_lecture(self, course_id, lecture_id):
         return self._download_json(
             if '>You have enrolled in' in webpage:
                 self.to_screen('%s: Successfully enrolled in the course' % course_id)
 
     def _download_lecture(self, course_id, lecture_id):
         return self._download_json(

-            'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?%s' % (
-                course_id, lecture_id, compat_urllib_parse.urlencode({
-                    'video_only': '',
-                    'auto_play': '',
-                    'fields[lecture]': 'title,description,asset',
-                    'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
-                    'instructorPreviewMode': 'False',
-                })),
-            lecture_id, 'Downloading lecture JSON')
+            'https://www.udemy.com/api-2.0/users/me/subscribed-courses/%s/lectures/%s?'
+            % (course_id, lecture_id),
+            lecture_id, 'Downloading lecture JSON', query={
+                'fields[lecture]': 'title,description,view_html,asset',
+                'fields[asset]': 'asset_type,stream_url,thumbnail_url,download_urls,data',
+            })

 
     def _handle_error(self, response):
         if not isinstance(response, dict):
 
     def _handle_error(self, response):
         if not isinstance(response, dict):


@@ -75,7 +109,7 @@ class UdemyIE(InfoExtractor):
                 error_str += ' - %s' % error_data.get('formErrors')
             raise ExtractorError(error_str, expected=True)
 
                 error_str += ' - %s' % error_data.get('formErrors')
             raise ExtractorError(error_str, expected=True)
 

-    def _download_json(self, url_or_request, video_id, note='Downloading JSON metadata'):
+    def _download_json(self, url_or_request, *args, **kwargs):

         headers = {
             'X-Udemy-Snail-Case': 'true',
             'X-Requested-With': 'XMLHttpRequest',
         headers = {
             'X-Udemy-Snail-Case': 'true',
             'X-Requested-With': 'XMLHttpRequest',


@@ -93,7 +127,7 @@ class UdemyIE(InfoExtractor):
         else:
             url_or_request = sanitized_Request(url_or_request, headers=headers)
 
         else:
             url_or_request = sanitized_Request(url_or_request, headers=headers)
 

-        response = super(UdemyIE, self)._download_json(url_or_request, video_id, note)
+        response = super(UdemyIE, self)._download_json(url_or_request, *args, **kwargs)

         self._handle_error(response)
         return response
 
         self._handle_error(response)
         return response
 


@@ -109,7 +143,9 @@ class UdemyIE(InfoExtractor):
             self._LOGIN_URL, None, 'Downloading login popup')
 
         def is_logged(webpage):
             self._LOGIN_URL, None, 'Downloading login popup')
 
         def is_logged(webpage):

-            return any(p in webpage for p in ['href="https://www.udemy.com/user/logout/', '>Logout<'])
+            return any(re.search(p, webpage) for p in (
+                r'href=["\'](?:https://www\.udemy\.com)?/user/logout/',
+                r'>Logout<'))

 
         # already logged in
         if is_logged(login_popup):
 
         # already logged in
         if is_logged(login_popup):


@@ -118,17 +154,17 @@ class UdemyIE(InfoExtractor):
         login_form = self._form_hidden_inputs('login-form', login_popup)
 
         login_form.update({
         login_form = self._form_hidden_inputs('login-form', login_popup)
 
         login_form.update({

-            'email': username.encode('utf-8'),
-            'password': password.encode('utf-8'),
+            'email': username,
+            'password': password,

         })
 
         })
 

-        request = sanitized_Request(
-            self._LOGIN_URL, compat_urllib_parse.urlencode(login_form).encode('utf-8'))
-        request.add_header('Referer', self._ORIGIN_URL)
-        request.add_header('Origin', self._ORIGIN_URL)
-

         response = self._download_webpage(
         response = self._download_webpage(

-            request, None, 'Logging in as %s' % username)
+            self._LOGIN_URL, None, 'Logging in as %s' % username,
+            data=urlencode_postdata(login_form),
+            headers={
+                'Referer': self._ORIGIN_URL,
+                'Origin': self._ORIGIN_URL,
+            })

 
         if not is_logged(response):
             error = self._html_search_regex(
 
         if not is_logged(response):
             error = self._html_search_regex(


@@ -143,15 +179,14 @@ class UdemyIE(InfoExtractor):
 
         webpage = self._download_webpage(url, lecture_id)
 
 
         webpage = self._download_webpage(url, lecture_id)
 

-        course_id = self._search_regex(
-            r'data-course-id=["\'](\d+)', webpage, 'course id')
+        course_id, _ = self._extract_course_info(webpage, lecture_id)

 
         try:
             lecture = self._download_lecture(course_id, lecture_id)
         except ExtractorError as e:
             # Error could possibly mean we are not enrolled in the course
             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:
 
         try:
             lecture = self._download_lecture(course_id, lecture_id)
         except ExtractorError as e:
             # Error could possibly mean we are not enrolled in the course
             if isinstance(e.cause, compat_HTTPError) and e.cause.code == 403:

-                self._enroll_course(webpage, course_id)
+                self._enroll_course(url, webpage, course_id)

                 lecture = self._download_lecture(course_id, lecture_id)
             else:
                 raise
                 lecture = self._download_lecture(course_id, lecture_id)
             else:
                 raise


@@ -161,56 +196,105 @@ class UdemyIE(InfoExtractor):
 
         asset = lecture['asset']
 
 
         asset = lecture['asset']
 

-        asset_type = asset.get('assetType') or asset.get('asset_type')
+        asset_type = asset.get('asset_type') or asset.get('assetType')

         if asset_type != 'Video':
             raise ExtractorError(
                 'Lecture %s is not a video' % lecture_id, expected=True)
 
         if asset_type != 'Video':
             raise ExtractorError(
                 'Lecture %s is not a video' % lecture_id, expected=True)
 

-        stream_url = asset.get('streamUrl') or asset.get('stream_url')
+        stream_url = asset.get('stream_url') or asset.get('streamUrl')

         if stream_url:
             youtube_url = self._search_regex(
                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
             if youtube_url:
                 return self.url_result(youtube_url, 'Youtube')
 
         if stream_url:
             youtube_url = self._search_regex(
                 r'(https?://www\.youtube\.com/watch\?v=.*)', stream_url, 'youtube URL', default=None)
             if youtube_url:
                 return self.url_result(youtube_url, 'Youtube')
 

-        video_id = asset['id']
-        thumbnail = asset.get('thumbnailUrl') or asset.get('thumbnail_url')
+        video_id = compat_str(asset['id'])
+        thumbnail = asset.get('thumbnail_url') or asset.get('thumbnailUrl')

         duration = float_or_none(asset.get('data', {}).get('duration'))
         duration = float_or_none(asset.get('data', {}).get('duration'))

-        outputs = asset.get('data', {}).get('outputs', {})

 
         formats = []
 
         formats = []

-        for format_ in asset.get('download_urls', {}).get('Video', []):
-            video_url = format_.get('file')
-            if not video_url:
-                continue
-            format_id = format_.get('label')
-            f = {
-                'url': format_['file'],
-                'height': int_or_none(format_id),
+
+        def extract_output_format(src):
+            return {
+                'url': src['url'],
+                'format_id': '%sp' % (src.get('height') or format_id),
+                'width': int_or_none(src.get('width')),
+                'height': int_or_none(src.get('height')),
+                'vbr': int_or_none(src.get('video_bitrate_in_kbps')),
+                'vcodec': src.get('video_codec'),
+                'fps': int_or_none(src.get('frame_rate')),
+                'abr': int_or_none(src.get('audio_bitrate_in_kbps')),
+                'acodec': src.get('audio_codec'),
+                'asr': int_or_none(src.get('audio_sample_rate')),
+                'tbr': int_or_none(src.get('total_bitrate_in_kbps')),
+                'filesize': int_or_none(src.get('file_size_in_bytes')),

             }
             }

-            if format_id:
-                # Some videos contain additional metadata (e.g.
-                # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
-                output = outputs.get(format_id)
-                if isinstance(output, dict):
-                    f.update({
-                        'format_id': '%sp' % (output.get('label') or format_id),
-                        'width': int_or_none(output.get('width')),
-                        'height': int_or_none(output.get('height')),
-                        'vbr': int_or_none(output.get('video_bitrate_in_kbps')),
-                        'vcodec': output.get('video_codec'),
-                        'fps': int_or_none(output.get('frame_rate')),
-                        'abr': int_or_none(output.get('audio_bitrate_in_kbps')),
-                        'acodec': output.get('audio_codec'),
-                        'asr': int_or_none(output.get('audio_sample_rate')),
-                        'tbr': int_or_none(output.get('total_bitrate_in_kbps')),
-                        'filesize': int_or_none(output.get('file_size_in_bytes')),
-                    })
+
+        outputs = asset.get('data', {}).get('outputs')
+        if not isinstance(outputs, dict):
+            outputs = {}
+
+        def add_output_format_meta(f, key):
+            output = outputs.get(key)
+            if isinstance(output, dict):
+                output_format = extract_output_format(output)
+                output_format.update(f)
+                return output_format
+            return f
+
+        download_urls = asset.get('download_urls')
+        if isinstance(download_urls, dict):
+            video = download_urls.get('Video')
+            if isinstance(video, list):
+                for format_ in video:
+                    video_url = format_.get('file')
+                    if not video_url:
+                        continue
+                    format_id = format_.get('label')
+                    f = {
+                        'url': format_['file'],
+                        'format_id': '%sp' % format_id,
+                        'height': int_or_none(format_id),
+                    }
+                    if format_id:
+                        # Some videos contain additional metadata (e.g.
+                        # https://www.udemy.com/ios9-swift/learn/#/lecture/3383208)
+                        f = add_output_format_meta(f, format_id)
+                    formats.append(f)
+
+        view_html = lecture.get('view_html')
+        if view_html:
+            view_html_urls = set()
+            for source in re.findall(r'<source[^>]+>', view_html):
+                attributes = extract_attributes(source)
+                src = attributes.get('src')
+                if not src:
+                    continue
+                res = attributes.get('data-res')
+                height = int_or_none(res)
+                if src in view_html_urls:
+                    continue
+                view_html_urls.add(src)
+                if attributes.get('type') == 'application/x-mpegURL' or determine_ext(src) == 'm3u8':
+                    m3u8_formats = self._extract_m3u8_formats(
+                        src, video_id, 'mp4', entry_protocol='m3u8_native',
+                        m3u8_id='hls', fatal=False)
+                    for f in m3u8_formats:
+                        m = re.search(r'/hls_(?P<height>\d{3,4})_(?P<tbr>\d{2,})/', f['url'])
+                        if m:
+                            if not f.get('height'):
+                                f['height'] = int(m.group('height'))
+                            if not f.get('tbr'):
+                                f['tbr'] = int(m.group('tbr'))
+                    formats.extend(m3u8_formats)

                 else:
                 else:

-                    f['format_id'] = '%sp' % format_id
-            formats.append(f)
+                    formats.append(add_output_format_meta({
+                        'url': src,
+                        'format_id': '%dp' % height if height else None,
+                        'height': height,
+                    }, res))

 
 

-        self._sort_formats(formats)
+        self._sort_formats(formats, field_preference=('height', 'width', 'tbr', 'format_id'))

 
         return {
             'id': video_id,
 
         return {
             'id': video_id,


@@ -224,7 +308,7 @@ class UdemyIE(InfoExtractor):
 
 class UdemyCourseIE(UdemyIE):
     IE_NAME = 'udemy:course'
 
 class UdemyCourseIE(UdemyIE):
     IE_NAME = 'udemy:course'

-    _VALID_URL = r'https?://www\.udemy\.com/(?P<id>[\da-z-]+)'
+    _VALID_URL = r'https?://(?:www\.)?udemy\.com/(?P<id>[^/?#&]+)'

     _TESTS = []
 
     @classmethod
     _TESTS = []
 
     @classmethod


@@ -236,29 +320,34 @@ class UdemyCourseIE(UdemyIE):
 
         webpage = self._download_webpage(url, course_path)
 
 
         webpage = self._download_webpage(url, course_path)
 

-        response = self._download_json(
-            'https://www.udemy.com/api-1.1/courses/%s' % course_path,
-            course_path, 'Downloading course JSON')
-
-        course_id = response['id']
-        course_title = response.get('title')
+        course_id, title = self._extract_course_info(webpage, course_path)

 
 

-        self._enroll_course(webpage, course_id)
+        self._enroll_course(url, webpage, course_id)

 
         response = self._download_json(
 
         response = self._download_json(

-            'https://www.udemy.com/api-1.1/courses/%s/curriculum' % course_id,
-            course_id, 'Downloading course curriculum')
+            'https://www.udemy.com/api-2.0/courses/%s/cached-subscriber-curriculum-items' % course_id,
+            course_id, 'Downloading course curriculum', query={
+                'fields[chapter]': 'title,object_index',
+                'fields[lecture]': 'title,asset',
+                'page_size': '1000',
+            })

 
         entries = []
 
         entries = []

-        chapter, chapter_number = None, None
-        for asset in response:
-            asset_type = asset.get('assetType') or asset.get('asset_type')
-            if asset_type == 'Video':
-                asset_id = asset.get('id')
-                if asset_id:
+        chapter, chapter_number = [None] * 2
+        for entry in response['results']:
+            clazz = entry.get('_class')
+            if clazz == 'lecture':
+                asset = entry.get('asset')
+                if isinstance(asset, dict):
+                    asset_type = asset.get('asset_type') or asset.get('assetType')
+                    if asset_type != 'Video':
+                        continue
+                lecture_id = entry.get('id')
+                if lecture_id:

                     entry = {
                         '_type': 'url_transparent',
                     entry = {
                         '_type': 'url_transparent',

-                        'url': 'https://www.udemy.com/%s/#/lecture/%s' % (course_path, asset['id']),
+                        'url': 'https://www.udemy.com/%s/learn/v4/t/lecture/%s' % (course_path, entry['id']),
+                        'title': entry.get('title'),

                         'ie_key': UdemyIE.ie_key(),
                     }
                     if chapter_number:
                         'ie_key': UdemyIE.ie_key(),
                     }
                     if chapter_number:


@@ -266,8 +355,8 @@ class UdemyCourseIE(UdemyIE):
                     if chapter:
                         entry['chapter'] = chapter
                     entries.append(entry)
                     if chapter:
                         entry['chapter'] = chapter
                     entries.append(entry)

-            elif asset.get('type') == 'chapter':
-                chapter_number = asset.get('index') or asset.get('object_index')
-                chapter = asset.get('title')
+            elif clazz == 'chapter':
+                chapter_number = entry.get('object_index')
+                chapter = entry.get('title')

 
 

-        return self.playlist_result(entries, course_id, course_title)
+        return self.playlist_result(entries, course_id, title)