[motherless] Fix review issues and improve extraction (closes #26495, closes #27450)

Less than a week old videos use a '20h ago' or '1d ago' format.

I kept the support for 'Ago' with uppercase start at is was already in the code.

.github/workflows/ci.yml

@@ -10,28 +10,52 @@ jobs:
         os: [ubuntu-latest]
         # TODO: python 2.6
         python-version: [2.7, 3.3, 3.4, 3.5, 3.6, 3.7, 3.8, 3.9, pypy-2.7, pypy-3.6, pypy-3.7]
+        python-impl: [cpython]
         ytdl-test-set: [core, download]
         run-tests-ext: [sh]
         include:
         # python 3.2 is only available on windows via setup-python
         - os: windows-latest
           python-version: 3.2
+          python-impl: cpython
           ytdl-test-set: core
           run-tests-ext: bat
         - os: windows-latest
           python-version: 3.2
+          python-impl: cpython
           ytdl-test-set: download
           run-tests-ext: bat
+        # jython
+        - os: ubuntu-latest
+          python-impl: jython
+          ytdl-test-set: core
+          run-tests-ext: sh
+        - os: ubuntu-latest
+          python-impl: jython
+          ytdl-test-set: download
+          run-tests-ext: sh
     steps:
     - uses: actions/checkout@v2
     - name: Set up Python ${{ matrix.python-version }}
       uses: actions/setup-python@v2
+      if: ${{ matrix.python-impl == 'cpython' }}
       with:
         python-version: ${{ matrix.python-version }}
+    - name: Set up Java 8
+      if: ${{ matrix.python-impl == 'jython' }}
+      uses: actions/setup-java@v1
+      with:
+        java-version: 8
+    - name: Install Jython
+      if: ${{ matrix.python-impl == 'jython' }}
+      run: |
+        wget http://search.maven.org/remotecontent?filepath=org/python/jython-installer/2.7.1/jython-installer-2.7.1.jar -O jython-installer.jar
+        java -jar jython-installer.jar -s -d "$HOME/jython"
+        echo "$HOME/jython/bin" >> $GITHUB_PATH
     - name: Install nose
       run: pip install nose
     - name: Run tests
-      continue-on-error: ${{ matrix.ytdl-test-set == 'download' }}
+      continue-on-error: ${{ matrix.ytdl-test-set == 'download' || matrix.python-impl == 'jython' }}
       env:
         YTDL_TEST_SET: ${{ matrix.ytdl-test-set }}
       run: ./devscripts/run_tests.${{ matrix.run-tests-ext }}

.travis.yml

@@ -1,50 +0,0 @@
-language: python
-python:
-  - "2.6"
-  - "2.7"
-  - "3.2"
-  - "3.3"
-  - "3.4"
-  - "3.5"
-  - "3.6"
-  - "pypy"
-  - "pypy3"
-dist: trusty
-env:
-  - YTDL_TEST_SET=core
-#  - YTDL_TEST_SET=download
-jobs:
-  include:
-    - python: 3.7
-      dist: xenial
-      env: YTDL_TEST_SET=core
-#    - python: 3.7
-#      dist: xenial
-#      env: YTDL_TEST_SET=download
-    - python: 3.8
-      dist: xenial
-      env: YTDL_TEST_SET=core
-#    - python: 3.8
-#      dist: xenial
-#      env: YTDL_TEST_SET=download
-    - python: 3.8-dev
-      dist: xenial
-      env: YTDL_TEST_SET=core
-#    - python: 3.8-dev
-#      dist: xenial
-#      env: YTDL_TEST_SET=download
-    - env: JYTHON=true; YTDL_TEST_SET=core
-#    - env: JYTHON=true; YTDL_TEST_SET=download
-    - name: flake8
-      python: 3.8
-      dist: xenial
-      install: pip install flake8
-      script: flake8 .
-  fast_finish: true
-  allow_failures:
-#    - env: YTDL_TEST_SET=download
-    - env: JYTHON=true; YTDL_TEST_SET=core
-#    - env: JYTHON=true; YTDL_TEST_SET=download
-before_install:
-  - if [ "$JYTHON" == "true" ]; then ./devscripts/install_jython.sh; export PATH="$HOME/jython/bin:$PATH"; fi
-script: ./devscripts/run_tests.sh

devscripts/install_jython.sh

@@ -1,5 +0,0 @@
-#!/bin/bash
-
-wget http://central.maven.org/maven2/org/python/jython-installer/2.7.1/jython-installer-2.7.1.jar
-java -jar jython-installer-2.7.1.jar -s -d "$HOME/jython"
-$HOME/jython/bin/jython -m pip install nose

youtube_dl/downloader/hls.py

@@ -172,8 +172,12 @@ class HlsFD(FragmentFD):
                         iv = decrypt_info.get('IV') or compat_struct_pack('>8xq', media_sequence)
                         decrypt_info['KEY'] = decrypt_info.get('KEY') or self.ydl.urlopen(
                             self._prepare_url(info_dict, info_dict.get('_decryption_key_url') or decrypt_info['URI'])).read()
-                        frag_content = AES.new(
-                            decrypt_info['KEY'], AES.MODE_CBC, iv).decrypt(frag_content)
+                        # Don't decrypt the content in tests since the data is explicitly truncated and it's not to a valid block
+                        # size (see https://github.com/ytdl-org/youtube-dl/pull/27660). Tests only care that the correct data downloaded,
+                        # not what it decrypts to.
+                        if not test:
+                            frag_content = AES.new(
+                                decrypt_info['KEY'], AES.MODE_CBC, iv).decrypt(frag_content)
                     self._append_fragment(ctx, frag_content)
                     # We only download the first fragment during the test
                     if test:

youtube_dl/extractor/motherless.py

@@ -61,6 +61,23 @@ class MotherlessIE(InfoExtractor):
         # no keywords
         'url': 'http://motherless.com/8B4BBC1',
         'only_matching': True,
+    }, {
+        # see https://motherless.com/videos/recent for recent videos with
+        # uploaded date in "ago" format
+        'url': 'https://motherless.com/3C3E2CF',
+        'info_dict': {
+            'id': '3C3E2CF',
+            'ext': 'mp4',
+            'title': 'a/ Hot Teens',
+            'categories': list,
+            'upload_date': '20210104',
+            'uploader_id': 'yonbiw',
+            'thumbnail': r're:https?://.*\.jpg',
+            'age_limit': 18,
+        },
+        'params': {
+            'skip_download': True,
+        },
     }]
 
     def _real_extract(self, url):
@@ -85,20 +102,28 @@ class MotherlessIE(InfoExtractor):
             or 'http://cdn4.videos.motherlessmedia.com/videos/%s.mp4?fs=opencloud' % video_id)
         age_limit = self._rta_search(webpage)
         view_count = str_to_int(self._html_search_regex(
-            (r'>(\d+)\s+Views<', r'<strong>Views</strong>\s+([^<]+)<'),
+            (r'>([\d,.]+)\s+Views<', r'<strong>Views</strong>\s+([^<]+)<'),
             webpage, 'view count', fatal=False))
         like_count = str_to_int(self._html_search_regex(
-            (r'>(\d+)\s+Favorites<', r'<strong>Favorited</strong>\s+([^<]+)<'),
+            (r'>([\d,.]+)\s+Favorites<',
+             r'<strong>Favorited</strong>\s+([^<]+)<'),
             webpage, 'like count', fatal=False))
 
-        upload_date = self._html_search_regex(
-            (r'class=["\']count[^>]+>(\d+\s+[a-zA-Z]{3}\s+\d{4})<',
-             r'<strong>Uploaded</strong>\s+([^<]+)<'), webpage, 'upload date')
-        if 'Ago' in upload_date:
-            days = int(re.search(r'([0-9]+)', upload_date).group(1))
-            upload_date = (datetime.datetime.now() - datetime.timedelta(days=days)).strftime('%Y%m%d')
-        else:
-            upload_date = unified_strdate(upload_date)
+        upload_date = unified_strdate(self._search_regex(
+            r'class=["\']count[^>]+>(\d+\s+[a-zA-Z]{3}\s+\d{4})<', webpage,
+            'upload date', default=None))
+        if not upload_date:
+            uploaded_ago = self._search_regex(
+                r'>\s*(\d+[hd])\s+[aA]go\b', webpage, 'uploaded ago',
+                default=None)
+            if uploaded_ago:
+                delta = int(uploaded_ago[:-1])
+                _AGO_UNITS = {
+                    'h': 'hours',
+                    'd': 'days',
+                }
+                kwargs = {_AGO_UNITS.get(uploaded_ago[-1]): delta}
+                upload_date = (datetime.datetime.utcnow() - datetime.timedelta(**kwargs)).strftime('%Y%m%d')
 
         comment_count = webpage.count('class="media-comment-contents"')
         uploader_id = self._html_search_regex(