Various search improvements

.github/workflows/tests.yml

@@ -7,7 +7,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.7, 3.8, 3.9]
+        python-version: [3.8, 3.9, 3.10, 3.11, 3.12]
 
     steps:
     - uses: actions/checkout@v2

requirements.txt

@@ -3,4 +3,6 @@ yt-dlp>=2023.3.4
 spotipy~=2.21
 mutagen~=1.45
 rich~=12.0
-urllib3~=1.26
+urllib3~=1.26
+ytmusicapi~=1.6.0
+Levenshtein~=0.25.1

setup.py

@@ -12,7 +12,7 @@
 setup(
     name="spotify_dl",
     version=VERSION,
-    python_requires=">=3.7",
+    python_requires=">=3.8",
     install_requires=requirements,
     author="Sathya Bhat",
     author_email="[email protected]",
@@ -36,10 +36,11 @@
         "Operating System :: OS Independent",
         "Programming Language :: Python",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.7",
         "Programming Language :: Python :: 3.8",
         "Programming Language :: Python :: 3.9",
         "Programming Language :: Python :: 3.10",
+        "Programming Language :: Python :: 3.11",
+        "Programming Language :: Python :: 3.12",
         "Topic :: Internet",
         "Topic :: Software Development :: Libraries",
         "Topic :: Software Development :: Libraries :: Python Modules",

spotify_dl/utils.py

@@ -1,3 +1,20 @@
+import Levenshtein
+
+
+def get_closest_match(results, expected) -> str:
+    """
+    Returns closest matching result based on Levenshtein edit distance.
+    """
+    best_r = ""
+    min_distance = float('inf')
+    for r in results:
+        curr_distance = Levenshtein.distance(r, expected)
+        if (curr_distance < min_distance):
+            min_distance = curr_distance
+            best_r = r
+    return best_r
+
+
 def sanitize(name, replace_with=""):
     """
     Removes some of the reserved characters from the name so it can be saved

spotify_dl/youtube.py

@@ -7,11 +7,12 @@
 import mutagen
 import csv
 import yt_dlp
+import ytmusicapi
 from mutagen.easyid3 import EasyID3
 from mutagen.id3 import APIC, ID3
 from mutagen.mp3 import MP3
 from spotify_dl.scaffold import log
-from spotify_dl.utils import sanitize
+from spotify_dl.utils import sanitize, get_closest_match
 from spotify_dl.constants import DOWNLOAD_LIST
 
 
@@ -33,16 +34,26 @@ def dump_json(songs):
     :param songs: the songs for which the JSON should be output
     """
     for song in songs:
-        query = f"{song.get('artist')} - {song.get('name')} Lyrics".replace(
+        query = f"{song.get('artist')} - {song.get('name')}".replace(
             ":", ""
         ).replace('"', "")
 
         ydl_opts = {"quiet": True}
 
         with yt_dlp.YoutubeDL(ydl_opts) as ydl:
             try:
-                ytJson = ydl.extract_info("ytsearch:" + query, False)
-                print(json.dumps(ytJson.get("entries")))
+                ytJson = {}
+                with ytmusicapi.YTMusic() as ym:
+                    # Reduce results to array of titles and video IDs
+                    result_titles, result_ids = zip(*map(
+                        lambda d: (f"{d['artists'][0]['name']} - {d['title']}".replace(":", "").replace('"', ""), d["videoId"]),
+                        ym.search(query, filter="songs")
+                    ))
+                    # Get ID of closest matching result by finding index in titles list
+                    videoId = result_ids[result_titles.index(get_closest_match(result_titles, query))]
+
+                    ytJson = ydl.extract_info(f"https://music.youtube.com/watch?v={videoId}", False)
+                print(json.dumps([ytJson]))  # insert into array so that the format stays the same
             except Exception as e:  # skipcq: PYL-W0703
                 log.debug(e)
                 print(
@@ -144,7 +155,7 @@ def set_tags(temp, filename, kwargs):
 def find_and_download_songs(kwargs):
     """
     function handles actual download of the songs
-    the youtube_search lib is used to search for songs and get best url
+    the ytmusicapi lib is used to search for songs and get best url via YT Music
     :param kwargs: dictionary of key value arguments to be used in download
     """
     sponsorblock_postprocessor = []
@@ -160,7 +171,7 @@ def find_and_download_songs(kwargs):
                 int(temp[-1].replace("\n", "")),
             )
 
-            query = f"{artist} - {name} Lyrics".replace(":", "").replace('"', "")
+            query = f"{artist} - {name}".replace(":", "").replace('"', "")
             print(f"Initiating download for {query}.")
 
             file_name = kwargs["file_name_f"](
@@ -200,6 +211,15 @@ def find_and_download_songs(kwargs):
                 print(f"File {mp3file_path} already exists, we do not overwrite it ")
                 continue
 
+            with ytmusicapi.YTMusic() as ym:
+                # Reduce search results to array of titles and video IDs
+                result_titles, result_ids = zip(*map(
+                    lambda d: (f"{d['artists'][0]['name']} - {d['title']}".replace(":", "").replace('"', ""), d["videoId"]),
+                    ym.search(query, filter="songs")
+                ))
+                # Get ID of closest matching result by finding index in titles list
+                video_id = result_ids[result_titles.index(get_closest_match(result_titles, query))]
+
             outtmpl = f"{file_path}.%(ext)s"
             ydl_opts = {
                 "proxy": kwargs.get("proxy"),
@@ -227,7 +247,7 @@ def find_and_download_songs(kwargs):
                 ydl_opts["postprocessors"].append(mp3_postprocess_opts.copy())
             with yt_dlp.YoutubeDL(ydl_opts) as ydl:
                 try:
-                    ydl.download([query])
+                    ydl.download([f"https://music.youtube.com/watch?v={video_id}"])
                 except Exception as e:  # skipcq: PYL-W0703
                     log.debug(e)
                     print(f"Failed to download {name}, make sure yt_dlp is up to date")