[Community] Add Genius Lyrics Document Loader

libs/community/langchain_community/document_loaders/__init__.py

@@ -205,6 +205,9 @@
     from langchain_community.document_loaders.gcs_file import (
         GCSFileLoader,
     )
+    from langchain_community.document_loaders.genius import (
+        GeniusLoader,
+    )
     from langchain_community.document_loaders.geodataframe import (
         GeoDataFrameLoader,
     )
@@ -814,6 +817,7 @@ def __getattr__(name: str) -> Any:
     "GlueCatalogLoader",
     "GCSFileLoader",
     "GeoDataFrameLoader",
+    "GeniusLoader",
     "GitHubIssuesLoader",
     "GitLoader",
     "GitbookLoader",

libs/community/langchain_community/document_loaders/genius.py

@@ -0,0 +1,55 @@
+from typing import Iterator
+
+from langchain_core.document_loaders import BaseLoader
+from langchain_core.documents import Document
+
+
+class GeniusLoader(BaseLoader):
+    """Load lyrics using the Genius API.
+
+    This loader utilizes the `lyricsgenius` Python package to fetch song lyrics
+    and metadata. You need a Genius API token, which can be generated at
+    https://genius.com/api-clients.
+    """
+
+    def __init__(self, search_query: str, api_token: str = None):
+        """Initialize with search query and API token.
+
+        Args:
+            search_query: The search query (e.g., "Imagine Dragons - Radioactive").
+            api_token: Genius API Token. If not provided, looks for GENIUS_ACCESS_TOKEN env var.
+        """
+        self.search_query = search_query
+        self.api_token = api_token
+
+    def lazy_load(self) -> Iterator[Document]:
+        """Load lyrics and metadata."""
+        try:
+            import lyricsgenius
+        except ImportError:
+            raise ImportError(
+                "lyricsgenius package not found, please install it with "
+                "`pip install lyricsgenius`"
+            )
+
+        # Initialize Genius client
+        genius = lyricsgenius.Genius(self.api_token)
+
+        # Search for the song (we use the first best match)
+        song = genius.search_song(self.search_query)
+
+        # If no song is found, yield nothing (empty iterator)
+        if not song:
+            return
+
+        # Create the LangChain Document
+        metadata = {
+            "source": "Genius",
+            "title": song.title,
+            "artist": song.artist,
+            "album": song.album,
+            "url": song.url,
+            "id": song.id,
+        }
+
+        yield Document(page_content=song.lyrics, metadata=metadata)

test_run.py

@@ -0,0 +1,36 @@
+# We try to import it from the top-level package
+# If this works, it means your __init__.py edit was successful
+try:
+    from langchain_community.document_loaders import GeniusLoader
+
+    print("✅ STEP 1 PASSED: GeniusLoader was imported successfully!")
+except ImportError as e:
+    print("❌ STEP 1 FAILED: Could not import GeniusLoader.")
+    print(e)
+    exit()
+
+# Now we try to use it
+# Note: If you don't have a real token, we expect an error, but NOT a crash.
+try:
+    print("attempting to initialize loader...")
+    # Replace 'fake_token' with a real one if you want actual lyrics
+    loader = GeniusLoader("Taylor Swift", api_token="fake_token")
+    print("✅ STEP 2 PASSED: Loader initialized!")
+
+    print("Attempting to load data...")
+    docs = list(loader.lazy_load())
+
+    if docs:
+        print(f"✅ STEP 3 PASSED: Found song: {docs[0].metadata['title']}")
+    else:
+        print("⚠️ STEP 3: No docs found (Expected if token is fake).")
+
+except Exception as e:
+    # If it fails because of the API token, that is actually GOOD.
+    # It means your code ran and tried to hit the API.
+    if "401" in str(e) or "403" in str(e) or "Token" in str(e):
+        print(
+            "✅ STEP 3 PASSED: Code ran! (API rejected the fake token, which is expected)."
+        )
+    else:
+        print(f"❌ STEP 3 FAILED with unexpected error: {e}")