lyrics-transcriber 0.41.0__tar.gz → 0.43.0__tar.gz

{lyrics_transcriber-0.41.0 → lyrics_transcriber-0.43.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: lyrics-transcriber
-Version: 0.41.0
+Version: 0.43.0
 Summary: Automatically create synchronised lyrics files in ASS and MidiCo LRC formats with word-level timestamps, using Whisper and lyrics from Genius and Spotify
 License: MIT
 Author: Andrew Beveridge
@@ -19,10 +19,12 @@ Requires-Dist: karaoke-lyrics-processor (>=0.4)
 Requires-Dist: lyricsgenius (>=3)
 Requires-Dist: metaphone (>=0.6)
 Requires-Dist: nltk (>=3.9)
+Requires-Dist: ollama (>=0.4.7,<0.5.0)
 Requires-Dist: pydub (>=0.25)
 Requires-Dist: python-dotenv (>=1)
 Requires-Dist: python-levenshtein (>=0.26)
 Requires-Dist: python-slugify (>=8)
+Requires-Dist: shortuuid (>=1.0.13,<2.0.0)
 Requires-Dist: spacy (>=3.8)
 Requires-Dist: spacy-syllables (>=3)
 Requires-Dist: syllables (>=1)

{lyrics_transcriber-0.41.0 → lyrics_transcriber-0.43.0}/lyrics_transcriber/core/controller.py

@@ -1,4 +1,3 @@
-import difflib
 import os
 import logging
 from dataclasses import dataclass, field
@@ -21,7 +20,7 @@ class LyricsControllerResult:
     """Holds the results of the transcription and correction process."""
 
     # Results from different sources
-    lyrics_results: List[LyricsData] = field(default_factory=list)
+    lyrics_results: dict[str, LyricsData] = field(default_factory=dict)
     transcription_results: List[TranscriptionResult] = field(default_factory=list)
 
     # Corrected results
@@ -260,7 +259,7 @@ class LyricsTranscriber:
             try:
                 result = provider.fetch_lyrics(self.artist, self.title)
                 if result:
-                    self.results.lyrics_results.append(result)
+                    self.results.lyrics_results[name] = result
                     self.logger.info(f"Successfully fetched lyrics from {name}")
 
             except Exception as e:
@@ -303,12 +302,10 @@ class LyricsTranscriber:
                 self.results.transcription_corrected = CorrectionResult(
                     original_segments=best_transcription.result.segments,
                     corrected_segments=best_transcription.result.segments,
-                    corrected_text="",  # Will be generated from segments
                     corrections=[],  # No corrections made
                     corrections_made=0,  # No corrections made
                     confidence=1.0,  # Full confidence since we're using original
-                    transcribed_text="",  # Will be generated from segments
-                    reference_texts={},
+                    reference_lyrics={},
                     anchor_sequences=[],
                     gap_sequences=[],
                     resized_segments=[],  # Will be populated later
@@ -320,15 +317,24 @@ class LyricsTranscriber:
                 )
             return
 
-        # Run correction if we have reference lyrics
-        corrected_data = self.corrector.run(
-            transcription_results=self.results.transcription_results, lyrics_results=self.results.lyrics_results
-        )
+        # Create metadata dict with song info
+        metadata = {
+            "artist": self.artist,
+            "title": self.title,
+            "full_reference_texts": {source: lyrics.get_full_text() for source, lyrics in self.results.lyrics_results.items()},
+        }
+
+        # Get enabled handlers from metadata if available
+        enabled_handlers = metadata.get("enabled_handlers", None)
+
+        # Create corrector with enabled handlers
+        corrector = LyricsCorrector(cache_dir=self.output_config.cache_dir, enabled_handlers=enabled_handlers, logger=self.logger)
 
-        # Add audio filepath to metadata
-        if not corrected_data.metadata:
-            corrected_data.metadata = {}
-        corrected_data.metadata["audio_filepath"] = self.audio_filepath
+        corrected_data = corrector.run(
+            transcription_results=self.results.transcription_results,
+            lyrics_results=self.results.lyrics_results,
+            metadata=metadata,
+        )
 
         # Store corrected results
         self.results.transcription_corrected = corrected_data
@@ -336,49 +342,21 @@ class LyricsTranscriber:
 
         # Add human review step
         if self.output_config.enable_review:
-            from ..review import start_review_server
-            import json
-            from copy import deepcopy
+            from lyrics_transcriber.review.server import ReviewServer
 
             self.logger.info("Starting human review process")
 
-            def normalize_data(data_dict):
-                """Normalize numeric values in the data structure before JSON conversion."""
-                if isinstance(data_dict, dict):
-                    return {k: normalize_data(v) for k, v in data_dict.items()}
-                elif isinstance(data_dict, list):
-                    return [normalize_data(item) for item in data_dict]
-                elif isinstance(data_dict, float):
-                    # Convert whole number floats to integers
-                    if data_dict.is_integer():
-                        return int(data_dict)
-                    return data_dict
-                return data_dict
-
-            # Normalize and convert auto-corrected data
-            auto_data = normalize_data(deepcopy(self.results.transcription_corrected.to_dict()))
-            auto_corrected_json = json.dumps(auto_data, indent=4).splitlines()
-
-            # Pass through review server
-            reviewed_data = start_review_server(self.results.transcription_corrected)
-
-            # Normalize and convert reviewed data
-            human_data = normalize_data(deepcopy(reviewed_data.to_dict()))
-            human_corrected_json = json.dumps(human_data, indent=4).splitlines()
-
-            self.logger.info("Human review completed")
-
-            # Compare the normalized JSON strings
-            diff = list(
-                difflib.unified_diff(auto_corrected_json, human_corrected_json, fromfile="auto-corrected", tofile="human-corrected")
+            # Create and start review server
+            review_server = ReviewServer(
+                correction_result=self.results.transcription_corrected,
+                output_config=self.output_config,
+                audio_filepath=self.audio_filepath,
+                logger=self.logger,
             )
+            reviewed_data = review_server.start()
 
-            if diff:
-                self.logger.warning("Changes made by human review:")
-                for line in diff:
-                    self.logger.warning(line.rstrip())
-
-            # exit(1)
+            self.logger.info("Human review completed, updated transcription_corrected with reviewed_data")
+            self.results.transcription_corrected = reviewed_data
 
     def generate_outputs(self) -> None:
         """Generate output files based on enabled features and available data."""