Write tests for YouTube metadata

spotdl/metadata/providers/spotify.py

@@ -2,7 +2,7 @@ import spotipy
 import spotipy.oauth2 as oauth2
 
 from spotdl.metadata import ProviderBase
-
+from spotdl.metadata.exceptions import SpotifyMetadataNotFoundError
 
 class ProviderSpotify(ProviderBase):
     def __init__(self, spotify=None):
@@ -17,8 +17,14 @@ class ProviderSpotify(ProviderBase):
         return self.metadata_to_standard_form(metadata)
 
     def from_query(self, query):
-        metadata = self.spotify.search(query, limit=1)["tracks"]["items"][0]
-        return self.metadata_to_standard_form(metadata)
+        tracks = self.spotify.search(query, limit=1)["tracks"]["items"]
+        if tracks is None:
+            raise SpotifyMetadataNotFoundError(
+                'Could not find any tracks matching the given search query ("{}")'.format(
+                    query,
+                )
+            )
+        return self.metadata_to_standard_form(tracks[0])
 
     def _generate_token(self, client_id, client_secret):
         """ Generate the token. """

spotdl/metadata/providers/tests/test_spotify.py

@@ -0,0 +1,13 @@
+from spotdl.metadata import ProviderBase
+from spotdl.metadata.exceptions import SpotifyMetadataNotFoundError
+from spotdl.metadata.providers import ProviderSpotify
+
+class TestProviderSpotify:
+    def test_subclass(self):
+        assert issubclass(ProviderSpotify, ProviderBase)
+
+    # def test_metadata_not_found_error(self):
+    #     provider = ProviderSpotify(spotify=spotify)
+    #     with pytest.raises(SpotifyMetadataNotFoundError):
+    #         provider.from_query("This track doesn't exist on Spotify.")
+

spotdl/metadata/providers/tests/test_youtube.py

@@ -0,0 +1,242 @@
+from spotdl.metadata.providers.youtube import YouTubeSearch
+from spotdl.metadata.providers.youtube import YouTubeStreams
+from spotdl.metadata.providers import youtube
+from spotdl.metadata.providers import ProviderYouTube
+from spotdl.metadata.exceptions import YouTubeMetadataNotFoundError
+
+import pytube
+import urllib.request
+import pickle
+import sys
+import os
+import pytest
+
+
+@pytest.fixture(scope="module")
+def track():
+    return "selena gomez wolves"
+
+
+@pytest.fixture(scope="module")
+def no_result_track():
+    return "n0 v1d305 3x157 f0r 7h15 53arc4 qu3ry"
+
+
+@pytest.fixture(scope="module")
+def expect_search_results():
+    return [
+        "https://www.youtube.com/watch?v=cH4E_t3m3xM",
+        "https://www.youtube.com/watch?v=xrbY9gDVms0",
+        "https://www.youtube.com/watch?v=jX0n2rSmDbE",
+        "https://www.youtube.com/watch?v=nVzA1uWTydQ",
+        "https://www.youtube.com/watch?v=rQ6jcpwzQZU",
+        "https://www.youtube.com/watch?v=-grLLLTza6k",
+        "https://www.youtube.com/watch?v=j0AxZ4V5WQw",
+        "https://www.youtube.com/watch?v=zbWsb36U0uo",
+        "https://www.youtube.com/watch?v=3B1aY9Ob8r0",
+        "https://www.youtube.com/watch?v=hd2SGk90r9k",
+    ]
+
+
+@pytest.fixture(scope="module")
+def expect_mock_search_results():
+    return [
+        "https://www.youtube.com/watch?v=cH4E_t3m3xM",
+        "https://www.youtube.com/watch?v=xrbY9gDVms0",
+        "https://www.youtube.com/watch?v=jX0n2rSmDbE",
+        "https://www.youtube.com/watch?v=rQ6jcpwzQZU",
+        "https://www.youtube.com/watch?v=nVzA1uWTydQ",
+        "https://www.youtube.com/watch?v=-grLLLTza6k",
+        "https://www.youtube.com/watch?v=zbWsb36U0uo",
+        "https://www.youtube.com/watch?v=rykH1BkGwTo",
+        "https://www.youtube.com/watch?v=j0AxZ4V5WQw",
+        "https://www.youtube.com/watch?v=RyxsaKfu-ZY",
+    ]
+
+
+class TestYouTubeSearch:
+    @pytest.fixture(scope="module")
+    def youtube_searcher(self):
+        return YouTubeSearch()
+
+    def test_generate_search_url(self, track, youtube_searcher):
+        url = youtube_searcher.generate_search_url(track)
+        expect_url = "https://www.youtube.com/results?sp=EgIQAQ%253D%253D&q=selena%20gomez%20wolves"
+        assert url == expect_url
+
+    @pytest.mark.network
+    def test_search(self, track, youtube_searcher, expect_search_results):
+        results = youtube_searcher.search(track)
+        assert results == expect_search_results
+
+    class MockHTTPResponse:
+        response_file = ""
+
+        def __init__(self, url):
+            pass
+
+        def read(self):
+            module_directory = os.path.dirname(__file__)
+            mock_html = os.path.join(module_directory, "data", self.response_file)
+            with open(mock_html, "r") as fin:
+                html = fin.read()
+            return html
+
+    # @pytest.mark.mock
+    def test_mock_search(self, track, youtube_searcher, expect_mock_search_results, monkeypatch):
+        self.MockHTTPResponse.response_file = "youtube_search_results.html"
+        monkeypatch.setattr(urllib.request, "urlopen", self.MockHTTPResponse)
+        self.test_search(track, youtube_searcher, expect_mock_search_results)
+
+    @pytest.mark.network
+    def test_no_videos_search(self, no_result_track, youtube_searcher):
+        results = youtube_searcher.search(no_result_track)
+        assert results == []
+
+    def test_mock_no_videos_search(self, no_result_track, youtube_searcher, monkeypatch):
+        self.MockHTTPResponse.response_file = "youtube_no_search_results.html"
+        monkeypatch.setattr(urllib.request, "urlopen", self.MockHTTPResponse)
+        self.test_no_videos_search(no_result_track, youtube_searcher)
+
+
+@pytest.fixture(scope="module")
+def content():
+    return pytube.YouTube("https://www.youtube.com/watch?v=cH4E_t3m3xM")
+
+
+class MockYouTube:
+    def __init__(self, url):
+        self.watch_html = '\\"category\\":\\"Music\\",\\"publishDate\\":\\"2017-11-18\\",\\"ownerChannelName\\":\\"SelenaGomezVEVO\\",'
+        self.title = "Selena Gomez, Marshmello - Wolves"
+        self.author = "SelenaGomezVEVO"
+        self.length = 213
+        self.watch_url = "https://youtube.com/watch?v=cH4E_t3m3xM"
+        self.thumbnail_url = "https://i.ytimg.com/vi/cH4E_t3m3xM/maxresdefault.jpg"
+
+    @property
+    def streams(self):
+        module_directory = os.path.dirname(__file__)
+        mock_streams = os.path.join(module_directory, "data", "streams.dump")
+        with open(mock_streams, "rb") as fin:
+            streams_dump = pickle.load(fin)
+        return streams_dump
+
+
+@pytest.fixture(scope="module")
+def mock_content():
+    return MockYouTube("https://www.youtube.com/watch?v=cH4E_t3m3xM")
+
+
+@pytest.fixture(scope="module")
+def expect_formatted_streams():
+        return [
+            {"bitrate": 160, "download_url": None, "encoding": "opus", "filesize": 3614184},
+            {"bitrate": 128, "download_url": None, "encoding": "mp4a.40.2", "filesize": 3444850},
+            {"bitrate": 70, "download_url": None, "encoding": "opus", "filesize": 1847626},
+            {"bitrate": 50, "download_url": None, "encoding": "opus", "filesize": 1407962}
+        ]
+
+
+class TestYouTubeStreams:
+    @pytest.mark.network
+    def test_streams(self, content, expect_formatted_streams):
+        formatted_streams = YouTubeStreams(content.streams)
+        for index in range(len(formatted_streams.all)):
+            assert isinstance(formatted_streams.all[index]["download_url"], str)
+            formatted_streams.all[index]["download_url"] = None
+
+        assert formatted_streams.all == expect_formatted_streams
+
+    # @pytest.mark.mock
+    def test_mock_streams(self, mock_content, expect_formatted_streams):
+        self.test_streams(mock_content, expect_formatted_streams)
+
+    @pytest.mark.network
+    def test_getbest(self, content):
+        formatted_streams = YouTubeStreams(content.streams)
+        best_stream = formatted_streams.getbest()
+        best_stream["download_url"] = None
+        assert best_stream == {
+            "bitrate": 160,
+            "download_url": None,
+            "encoding": "opus",
+            "filesize": 3614184
+        }
+
+    # @pytest.mark.mock
+    def test_mock_getbest(self, mock_content):
+        self.test_getbest(mock_content)
+
+    @pytest.mark.network
+    def test_getworst(self, content):
+        formatted_streams = YouTubeStreams(content.streams)
+        worst_stream = formatted_streams.getworst()
+        worst_stream["download_url"] = None
+        assert worst_stream == {
+            "bitrate": 50,
+            "download_url": None,
+            "encoding": 'opus',
+            "filesize": 1407962
+        }
+
+    # @pytest.mark.mock
+    def test_mock_getworst(self, mock_content):
+        self.test_getworst(mock_content)
+
+
+class TestProviderYouTube:
+    @pytest.fixture(scope="module")
+    def youtube_provider(self):
+        return ProviderYouTube()
+
+    class MockYouTubeSearch:
+        watch_urls = []
+        def search(self, query):
+            return self.watch_urls
+
+    @pytest.mark.network
+    def test_from_query(self, track, youtube_provider):
+        metadata = youtube_provider.from_query(track)
+        assert isinstance(metadata["streams"], YouTubeStreams)
+
+        metadata["streams"] = []
+        assert metadata == {
+          'album': {'artists': [{'name': None}],
+                    'images': [{'url': 'https://i.ytimg.com/vi/cH4E_t3m3xM/maxresdefault.jpg'}],
+                    'name': None},
+          'artists': [{'name': 'SelenaGomezVEVO'}],
+          'copyright': None,
+          'disc_number': 1,
+          'duration': 213,
+          'external_ids': {'isrc': None},
+          'external_urls': {'youtube': 'https://youtube.com/watch?v=cH4E_t3m3xM'},
+          'genre': None,
+          'lyrics': None,
+          'name': 'Selena Gomez, Marshmello - Wolves',
+          'provider': 'youtube',
+          'publisher': None,
+          'release_date': '2017-11-1',
+          'streams': [],
+          'total_tracks': 1,
+          'track_number': 1,
+          'type': 'track',
+          'year': '2017'
+       }
+
+    def test_mock_from_query(self, track, youtube_provider, expect_mock_search_results, monkeypatch):
+        self.MockYouTubeSearch.watch_urls = expect_mock_search_results
+        monkeypatch.setattr(youtube, "YouTubeSearch", self.MockYouTubeSearch)
+        monkeypatch.setattr(pytube, "YouTube", MockYouTube)
+        self.test_from_query(track, youtube_provider)
+
+    @pytest.mark.network
+    def test_error_exception_from_query(self, no_result_track, youtube_provider):
+        with pytest.raises(YouTubeMetadataNotFoundError):
+            youtube_provider.from_query(no_result_track)
+
+    def test_mock_error_exception_from_query(self, no_result_track, youtube_provider, monkeypatch):
+        self.MockYouTubeSearch.watch_urls = []
+        monkeypatch.setattr(youtube, "YouTubeSearch", self.MockYouTubeSearch)
+        monkeypatch.setattr(pytube, "YouTube", MockYouTube)
+        self.test_error_exception_from_query(no_result_track, youtube_provider)
+

spotdl/metadata/providers/youtube.py

@@ -5,6 +5,7 @@ import urllib.request
 
 from spotdl.metadata import StreamsBase
 from spotdl.metadata import ProviderBase
+from spotdl.metadata.exceptions import YouTubeMetadataNotFoundError
 
 BASE_URL = "https://www.youtube.com/results?sp=EgIQAQ%253D%253D&q={}"
 
@@ -77,6 +78,9 @@ class YouTubeStreams(StreamsBase):
     def __init__(self, streams):
         audiostreams = streams.filter(only_audio=True).order_by("abr").desc()
         self.all = [{
+            # Store only the integer part. For example the given
+            # bitrate would be "192kbps", we store only the integer
+            # part here and drop the rest.
             "bitrate": int(stream.abr[:-4]),
             "download_url": stream.url,
             "encoding": stream.audio_codec,
@@ -93,6 +97,11 @@ class YouTubeStreams(StreamsBase):
 class ProviderYouTube(ProviderBase):
     def from_query(self, query):
         watch_urls = YouTubeSearch().search(query)
+        if not watch_urls:
+            raise YouTubeMetadataNotFoundError(
+                'YouTube returned nothing for the given search '
+                'query ("{}")'.format(query)
+            )
         return self.from_url(watch_urls[0])
 
     def from_url(self, url):
@@ -111,7 +120,6 @@ class ProviderYouTube(ProviderBase):
 
     def metadata_to_standard_form(self, content):
         """ Fetch a song's metadata from YouTube. """
-        streams = []
         publish_date = self._fetch_publish_date(content)
         metadata = {
             "name": content.title,