Merge pull request spotDL#1361 from spotDL/dev

* bugfix: fixed m3u issues (spotDL#1357)


* bugfix: remove duplicate songs from songs_list (spotDL#1356)

* bugfix: fixed ytdl error reporting (spotDL#1360)

* Remembered to bump version number to 3.7.2
Are you proud of me? I actually did it before merging :)

Co-authored-by: Jakub Kot <[email protected]>

Author: Silverarmor

setup.cfg

@@ -1,5 +1,5 @@
 [metadata]
-version = 3.7.1
+version = 3.7.2
 
 name = spotdl
 url = https://github.com/spotDL/spotify-downloader

spotdl/download/downloader.py

@@ -11,6 +11,7 @@
 from spotdl.search import SongObject
 from spotdl.download.progress_ui_handler import YTDLLogger
 from spotdl.download import ffmpeg, set_id3_data, DisplayManager, DownloadTracker
+from spotdl.providers.provider_utils import _get_converted_file_path
 
 
 class DownloadManager:
@@ -176,7 +177,6 @@ async def download_song(self, song_object: SongObject) -> None:
                     "outtmpl": f"{str(temp_folder)}/%(id)s.%(ext)s",
                     "quiet": True,
                     "no_warnings": True,
-                    "ignoreerrors": True,
                     "logger": YTDLLogger(),
                     "progress_hooks": [display_progress_tracker.ytdl_progress_hook]
                     if display_progress_tracker
@@ -276,81 +276,8 @@ def _perform_audio_download(
         except Exception as e:  # noqa:E722
             # ! This is equivalent to a failed download, we do nothing, the song remains on
             # ! download_trackers download queue and all is well...
-
             temp_files = Path(temp_folder).glob(f"{converted_file_name}.*")
             for temp_file in temp_files:
                 temp_file.unlink()
 
             raise e
-
-
-# ========================
-# === Helper function ===
-# ========================
-
-
-def _sanitize_filename(input_str: str) -> str:
-    output = input_str
-
-    # ! this is windows specific (disallowed chars)
-    output = "".join(char for char in output if char not in "/?\\*|<>")
-
-    # ! double quotes (") and semi-colons (:) are also disallowed characters but we would
-    # ! like to retain their equivalents, so they aren't removed in the prior loop
-    output = output.replace('"', "'").replace(":", "-")
-
-    return output
-
-
-def _get_smaller_file_path(input_song: SongObject, output_format: str) -> Path:
-    # Only use the first artist if the song path turns out to be too long
-    smaller_name = f"{input_song.contributing_artists[0]} - {input_song.song_name}"
-
-    smaller_name = _sanitize_filename(smaller_name)
-
-    try:
-        return Path(f"{smaller_name}.{output_format}").resolve()
-    except (OSError, WindowsError):
-        # Expected to happen in the rare case when the saved path is too long,
-        # even with the short filename
-        raise OSError("Cannot save song due to path issues.")
-
-
-def _get_converted_file_path(song_obj: SongObject, output_format: str = None) -> Path:
-
-    # ! we eliminate contributing artist names that are also in the song name, else we
-    # ! would end up with things like 'Jetta, Mastubs - I'd love to change the world
-    # ! (Mastubs REMIX).mp3' which is kinda an odd file name.
-
-    # also make sure that main artist is included in artistStr even if they
-    # are in the song name, for example
-    # Lil Baby - Never Recover (Lil Baby & Gunna, Drake).mp3
-
-    artists_filtered = []
-
-    if output_format is None:
-        output_format = "mp3"
-
-    for artist in song_obj.contributing_artists:
-        if artist.lower() not in song_obj.song_name:
-            artists_filtered.append(artist)
-        elif artist.lower() is song_obj.contributing_artists[0].lower():
-            artists_filtered.append(artist)
-
-    artist_str = ", ".join(artists_filtered)
-
-    converted_file_name = _sanitize_filename(
-        f"{artist_str} - {song_obj.song_name}.{output_format}"
-    )
-
-    converted_file_path = Path(converted_file_name)
-
-    # ! Checks if a file name is too long (256 max on both linux and windows)
-    try:
-        if len(str(converted_file_path.resolve().name)) > 256:
-            print("Path was too long. Using Small Path.")
-            return _get_smaller_file_path(song_obj, output_format)
-    except (OSError, WindowsError):
-        return _get_smaller_file_path(song_obj, output_format)
-
-    return converted_file_path

spotdl/download/ffmpeg.py

@@ -102,7 +102,7 @@ async def convert(
 
     proc_out = await process.communicate()
 
-    if proc_out[0] and proc_out[1]:
+    if proc_out[0] or proc_out[1]:
         out = str(b"".join(proc_out))
     else:
         out = ""

spotdl/download/progress_ui_handler.py

@@ -42,7 +42,7 @@ def warning(self, msg):
         pass
 
     def error(self, msg):
-        pass
+        raise Exception(msg)
 
 
 class SizedTextColumn(ProgressColumn):

spotdl/parsers/query_parser.py

@@ -26,7 +26,15 @@ def parse_query(
         # linefeed to visually separate output for each query
         print()
 
-    return songs_list
+    # remove duplicates
+    seen_songs = set()
+    songs = []
+    for song in songs_list:
+        if song.file_name not in seen_songs:
+            songs.append(song)
+            seen_songs.add(song.file_name)
+
+    return songs
 
 
 def parse_request(
@@ -64,7 +72,7 @@ def parse_request(
     elif "open.spotify.com" in request and "album" in request:
         print("Fetching Album...")
         song_list = song_gatherer.from_album(
-            request, output_format, use_youtube, threads
+            request, output_format, use_youtube, generate_m3u, threads
         )
     elif "open.spotify.com" in request and "playlist" in request:
         print("Fetching Playlist...")

spotdl/providers/provider_utils.py

@@ -3,6 +3,7 @@
 from typing import List
 from rapidfuzz import fuzz
 from bs4 import BeautifulSoup
+from pathlib import Path
 
 
 def _match_percentage(str1: str, str2: str, score_cutoff: float = 0) -> float:
@@ -60,7 +61,7 @@ def _parse_duration(duration: str) -> float:
 
 def _create_song_title(song_name: str, song_artists: List[str]) -> str:
     joined_artists = ", ".join(song_artists)
-    return f"{joined_artists} - {song_name}".lower()
+    return f"{joined_artists} - {song_name}"
 
 
 def _get_song_lyrics(song_name: str, song_artists: List[str]) -> str:
@@ -102,3 +103,70 @@ def _get_song_lyrics(song_name: str, song_artists: List[str]) -> str:
         return ""
     except:  # noqa: E722
         return ""
+
+
+def _sanitize_filename(input_str: str) -> str:
+    output = input_str
+
+    # ! this is windows specific (disallowed chars)
+    output = "".join(char for char in output if char not in "/?\\*|<>")
+
+    # ! double quotes (") and semi-colons (:) are also disallowed characters but we would
+    # ! like to retain their equivalents, so they aren't removed in the prior loop
+    output = output.replace('"', "'").replace(":", "-")
+
+    return output
+
+
+def _get_smaller_file_path(input_song, output_format: str) -> Path:
+    # Only use the first artist if the song path turns out to be too long
+    smaller_name = f"{input_song.contributing_artists[0]} - {input_song.song_name}"
+
+    smaller_name = _sanitize_filename(smaller_name)
+
+    try:
+        return Path(f"{smaller_name}.{output_format}").resolve()
+    except (OSError, WindowsError):
+        # Expected to happen in the rare case when the saved path is too long,
+        # even with the short filename
+        raise OSError("Cannot save song due to path issues.")
+
+
+def _get_converted_file_path(song_obj, output_format: str = None) -> Path:
+
+    # ! we eliminate contributing artist names that are also in the song name, else we
+    # ! would end up with things like 'Jetta, Mastubs - I'd love to change the world
+    # ! (Mastubs REMIX).mp3' which is kinda an odd file name.
+
+    # also make sure that main artist is included in artistStr even if they
+    # are in the song name, for example
+    # Lil Baby - Never Recover (Lil Baby & Gunna, Drake).mp3
+
+    artists_filtered = []
+
+    if output_format is None:
+        output_format = "mp3"
+
+    for artist in song_obj.contributing_artists:
+        if artist.lower() not in song_obj.song_name:
+            artists_filtered.append(artist)
+        elif artist.lower() is song_obj.contributing_artists[0].lower():
+            artists_filtered.append(artist)
+
+    artist_str = ", ".join(artists_filtered)
+
+    converted_file_name = _sanitize_filename(
+        f"{artist_str} - {song_obj.song_name}.{output_format}"
+    )
+
+    converted_file_path = Path(converted_file_name)
+
+    # ! Checks if a file name is too long (256 max on both linux and windows)
+    try:
+        if len(str(converted_file_path.resolve().name)) > 256:
+            print("Path was too long. Using Small Path.")
+            return _get_smaller_file_path(song_obj, output_format)
+    except (OSError, WindowsError):
+        return _get_smaller_file_path(song_obj, output_format)
+
+    return converted_file_path

spotdl/providers/yt_provider.py

@@ -45,7 +45,7 @@ def search_and_get_best_match(
             if isrc_result is not None and isrc_result.watch_url is not None:
                 return isrc_result.watch_url
 
-    song_title = _create_song_title(song_name, song_artists)
+    song_title = _create_song_title(song_name, song_artists).lower()
 
     # Query YTM by songs only first, this way if we get correct result on the first try
     # we don't have to make another request to ytmusic api that could result in us
@@ -126,7 +126,7 @@ def _order_yt_results(
             continue
 
         artist_match = (artist_match_number / len(song_artists)) * 100
-        song_title = _create_song_title(song_name, song_artists)
+        song_title = _create_song_title(song_name, song_artists).lower()
         name_match = round(
             _match_percentage(
                 unidecode(result.title.lower()), unidecode(song_title), 60

spotdl/providers/ytm_provider.py

@@ -61,7 +61,7 @@ def search_and_get_best_match(
             ):
                 return isrc_result["link"]
 
-    song_title = _create_song_title(song_name, song_artists)
+    song_title = _create_song_title(song_name, song_artists).lower()
 
     # Query YTM by songs only first, this way if we get correct result on the first try
     # we don't have to make another request to ytmusic api that could result in us
@@ -85,7 +85,7 @@ def search_and_get_best_match(
     # We didn't find the correct song on the first try so now we get video type results
     # add them to song_results, and get the result with highest score
     video_results = _query_and_simplify(
-        _create_song_title(song_name, song_artists), filter="videos"
+        _create_song_title(song_name, song_artists).lower(), filter="videos"
     )
 
     # Order video results
@@ -194,7 +194,7 @@ def _order_ytm_results(
 
         artist_match = (artist_match_number / len(song_artists)) * 100
 
-        song_title = _create_song_title(song_name, song_artists)
+        song_title = _create_song_title(song_name, song_artists).lower()
 
         # Find name match and drop results below 60%
         # this needs more testing