lyrics-transcriber 0.32.2__py3-none-any.whl → 0.33.0__py3-none-any.whl

lyrics_transcriber/cli/cli_main.py

@@ -60,19 +60,24 @@ def create_arg_parser() -> argparse.ArgumentParser:
     output_group.add_argument(
         "--cache_dir",
         type=Path,
-        help="Directory to cache downloaded/generated files. Default: /tmp/lyrics-transcriber-cache/",
+        help="Directory to cache downloaded/generated files. Default: ~/lyrics-transcriber-cache/",
     )
     output_group.add_argument(
         "--output_styles_json",
         type=Path,
         help="JSON file containing output style configurations for CDG and video generation",
     )
-    output_group.add_argument("--generate_cdg", action="store_true", help="Generate CDG karaoke files")
 
-    # Video options
-    video_group = parser.add_argument_group("Video Options")
-    video_group.add_argument("--render_video", action="store_true", help="Render a karaoke video with the generated lyrics")
-    video_group.add_argument(
+    # Feature control group
+    feature_group = parser.add_argument_group("Feature Control")
+    feature_group.add_argument("--skip_lyrics_fetch", action="store_true", help="Skip fetching lyrics from online sources")
+    feature_group.add_argument("--skip_transcription", action="store_true", help="Skip audio transcription process")
+    feature_group.add_argument("--skip_correction", action="store_true", help="Skip lyrics correction process")
+    feature_group.add_argument("--skip_plain_text", action="store_true", help="Skip generating plain text output files")
+    feature_group.add_argument("--skip_lrc", action="store_true", help="Skip generating LRC file")
+    feature_group.add_argument("--skip_cdg", action="store_true", help="Skip generating CDG karaoke files")
+    feature_group.add_argument("--skip_video", action="store_true", help="Skip rendering karaoke video")
+    feature_group.add_argument(
         "--video_resolution", choices=["4k", "1080p", "720p", "360p"], default="360p", help="Resolution of the karaoke video. Default: 360p"
     )
 
@@ -86,7 +91,7 @@ def parse_args(parser: argparse.ArgumentParser, args_list: list[str] | None = No
 
     # Set default cache_dir if not provided
     if not hasattr(args, "cache_dir") or args.cache_dir is None:
-        args.cache_dir = Path(os.getenv("LYRICS_TRANSCRIBER_CACHE_DIR", "/tmp/lyrics-transcriber-cache/"))
+        args.cache_dir = Path(os.getenv("LYRICS_TRANSCRIBER_CACHE_DIR", os.path.join(os.path.expanduser("~"), "lyrics-transcriber-cache")))
 
     return args
 
@@ -135,9 +140,14 @@ def create_configs(args: argparse.Namespace, env_config: Dict[str, str]) -> tupl
         output_styles_json=str(args.output_styles_json),
         output_dir=str(args.output_dir) if args.output_dir else os.getcwd(),
         cache_dir=str(args.cache_dir),
-        render_video=args.render_video,
-        generate_cdg=args.generate_cdg,
         video_resolution=args.video_resolution,
+        fetch_lyrics=not args.skip_lyrics_fetch,
+        run_transcription=not args.skip_transcription,
+        run_correction=not args.skip_correction,
+        generate_plain_text=not args.skip_plain_text,
+        generate_lrc=not args.skip_lrc,
+        generate_cdg=not args.skip_cdg,
+        render_video=not args.skip_video,
     )
 
     return transcriber_config, lyrics_config, output_config

lyrics_transcriber/core/config.py

@@ -28,8 +28,18 @@ class OutputConfig:
     max_line_length: int = 36
     styles: Dict[str, Any] = field(default_factory=dict)
     output_dir: Optional[str] = os.getcwd()
-    cache_dir: str = os.getenv("LYRICS_TRANSCRIBER_CACHE_DIR", "/tmp/lyrics-transcriber-cache/")
-    render_video: bool = False
-    generate_cdg: bool = False
-    video_resolution: str = "360p"
+    cache_dir: str = os.getenv(
+        "LYRICS_TRANSCRIBER_CACHE_DIR",
+        os.path.join(os.path.expanduser("~"), "lyrics-transcriber-cache")
+    )
+
+    fetch_lyrics: bool = True
+    run_transcription: bool = True
+    run_correction: bool = True
     enable_review: bool = True
+
+    generate_plain_text: bool = True
+    generate_lrc: bool = True
+    generate_cdg: bool = True
+    render_video: bool = True
+    video_resolution: str = "360p"

lyrics_transcriber/core/controller.py

@@ -1,12 +1,10 @@
+import difflib
+import json
 import os
 import logging
 from dataclasses import dataclass, field
 from typing import Dict, Optional, List
-from lyrics_transcriber.types import (
-    LyricsData,
-    TranscriptionResult,
-    CorrectionResult,
-)
+from lyrics_transcriber.types import LyricsData, PhraseType, TranscriptionResult, CorrectionResult, AnchorSequence, GapSequence, PhraseScore
 from lyrics_transcriber.transcribers.base_transcriber import BaseTranscriber
 from lyrics_transcriber.transcribers.audioshake import AudioShakeTranscriber, AudioShakeConfig
 from lyrics_transcriber.transcribers.whisper import WhisperTranscriber, WhisperConfig
@@ -83,6 +81,16 @@ class LyricsTranscriber:
         self.lyrics_config = lyrics_config or LyricsConfig()
         self.output_config = output_config or OutputConfig()
 
+        # Check if styles JSON is available for CDG and video features
+        if not self.output_config.output_styles_json or not os.path.exists(self.output_config.output_styles_json):
+            if self.output_config.generate_cdg or self.output_config.render_video:
+                self.logger.warning(
+                    f"Output styles JSON file not found: {self.output_config.output_styles_json}. "
+                    "CDG and video generation will be disabled."
+                )
+                self.output_config.generate_cdg = False
+                self.output_config.render_video = False
+
         # Basic settings
         self.audio_filepath = audio_filepath
         self.artist = artist
@@ -106,6 +114,18 @@ class LyricsTranscriber:
         self.corrector = corrector or LyricsCorrector(cache_dir=self.output_config.cache_dir, logger=self.logger)
         self.output_generator = output_generator or self._initialize_output_generator()
 
+        # Log enabled features
+        self.logger.info("Enabled features:")
+        self.logger.info(f"  Lyrics fetching: {'enabled' if self.output_config.fetch_lyrics else 'disabled'}")
+        self.logger.info(f"  Transcription: {'enabled' if self.output_config.run_transcription else 'disabled'}")
+        self.logger.info(f"  Lyrics correction: {'enabled' if self.output_config.run_correction else 'disabled'}")
+        self.logger.info(f"  Plain text output: {'enabled' if self.output_config.generate_plain_text else 'disabled'}")
+        self.logger.info(f"  LRC file generation: {'enabled' if self.output_config.generate_lrc else 'disabled'}")
+        self.logger.info(f"  CDG file generation: {'enabled' if self.output_config.generate_cdg else 'disabled'}")
+        self.logger.info(f"  Video rendering: {'enabled' if self.output_config.render_video else 'disabled'}")
+        if self.output_config.render_video:
+            self.logger.info(f"    Video resolution: {self.output_config.video_resolution}")
+
     def _initialize_transcribers(self) -> Dict[str, BaseTranscriber]:
         """Initialize available transcription services."""
         transcribers = {}
@@ -175,26 +195,21 @@ class LyricsTranscriber:
         return OutputGenerator(config=self.output_config, logger=self.logger)
 
     def process(self) -> LyricsControllerResult:
-        """
-        Main processing method that orchestrates the entire workflow.
-
-        Returns:
-            LyricsControllerResult containing all outputs and generated files.
+        """Main processing method that orchestrates the entire workflow."""
 
-        Raises:
-            Exception: If a critical error occurs during processing.
-        """
-        # Step 1: Fetch lyrics if artist and title are provided
-        if self.artist and self.title:
+        # Step 1: Fetch lyrics if enabled and artist/title are provided
+        if self.output_config.fetch_lyrics and self.artist and self.title:
             self.fetch_lyrics()
 
-        # Step 2: Run transcription
-        self.transcribe()
+        # Step 2: Run transcription if enabled
+        if self.output_config.run_transcription:
+            self.transcribe()
 
-        # Step 3: Process and correct lyrics
-        self.correct_lyrics()
+        # Step 3: Process and correct lyrics if enabled
+        if self.output_config.run_correction:
+            self.correct_lyrics()
 
-        # Step 4: Generate outputs
+        # Step 4: Generate outputs based on what's enabled and available
         self.generate_outputs()
 
         self.logger.info("Processing completed successfully")
@@ -239,7 +254,32 @@ class LyricsTranscriber:
         """Run lyrics correction using transcription and internet lyrics."""
         self.logger.info("Starting lyrics correction process")
 
-        # Run correction
+        # Check if we have reference lyrics to work with
+        if not self.results.lyrics_results:
+            self.logger.warning("No reference lyrics available for correction - using raw transcription")
+            # Use the highest priority transcription result as the "corrected" version
+            if self.results.transcription_results:
+                sorted_results = sorted(self.results.transcription_results, key=lambda x: x.priority)
+                best_transcription = sorted_results[0]
+
+                # Create a CorrectionResult with no corrections
+                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={},
+                    anchor_sequences=[],
+                    gap_sequences=[],
+                    resized_segments=[],  # Will be populated later
+                    metadata={"correction_type": "none", "reason": "no_reference_lyrics"},
+                )
+            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
         )
@@ -249,19 +289,60 @@ class LyricsTranscriber:
         self.logger.info("Lyrics correction completed")
 
         # Add human review step
-        if self.output_config.enable_review:  # We'll need to add this config option
+        if self.output_config.enable_review:
             from ..review import start_review_server
+            import json
+            from copy import deepcopy
 
             self.logger.info("Starting human review process")
-            self.results.transcription_corrected = start_review_server(corrected_data)
+
+            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")
+            )
+
+            if diff:
+                self.logger.warning("Changes made by human review:")
+                for line in diff:
+                    self.logger.warning(line.rstrip())
+
+            # exit(1)
+
     def generate_outputs(self) -> None:
-        """Generate output files."""
+        """Generate output files based on enabled features and available data."""
         self.logger.info("Generating output files")
 
+        # Only proceed with outputs that make sense based on what we have
+        has_correction = bool(self.results.transcription_corrected)
+
         output_files = self.output_generator.generate_outputs(
-            transcription_corrected=self.results.transcription_corrected,
+            transcription_corrected=self.results.transcription_corrected if has_correction else None,
             lyrics_results=self.results.lyrics_results,
             output_prefix=self.output_prefix,
             audio_filepath=self.audio_filepath,
@@ -269,7 +350,7 @@ class LyricsTranscriber:
             title=self.title,
         )
 
-        # Store all output paths in results
+        # Store results
         self.results.lrc_filepath = output_files.lrc
         self.results.ass_filepath = output_files.ass
         self.results.video_filepath = output_files.video

lyrics_transcriber/correction/phrase_analyzer.py

@@ -21,10 +21,19 @@ class PhraseAnalyzer:
         try:
             self.nlp = spacy.load(language_code)
         except OSError:
-            self.logger.error(f"Failed to load language model: {language_code}")
-            raise OSError(
-                f"Language model '{language_code}' not found. " f"Please install it with: python -m spacy download {language_code}"
-            )
+            self.logger.info(f"Language model {language_code} not found. Attempting to download...")
+            import subprocess
+
+            try:
+                subprocess.check_call(["python", "-m", "spacy", "download", language_code])
+                self.nlp = spacy.load(language_code)
+                self.logger.info(f"Successfully downloaded and loaded {language_code}")
+            except subprocess.CalledProcessError as e:
+                self.logger.error(f"Failed to download language model: {language_code}")
+                raise OSError(
+                    f"Language model '{language_code}' could not be downloaded. "
+                    f"Please install it manually with: python -m spacy download {language_code}"
+                ) from e
 
     def score_phrase(self, words: List[str], context: str) -> PhraseScore:
         """Score a phrase based on grammatical completeness and natural breaks.

lyrics_transcriber/output/segment_resizer.py

@@ -123,8 +123,9 @@ class SegmentResizer:
                 LyricsSegment(text="Here's another one.", ...)
             ]
         """
-        self.logger.info(f"Processing oversized segment {segment_idx}: '{segment.text}'")
         segment_text = self._clean_text(segment.text)
+
+        self.logger.info(f"Processing oversized segment {segment_idx}: '{segment_text}'")
         split_lines = self._process_segment_text(segment_text)
         self.logger.debug(f"Split into {len(split_lines)} lines: {split_lines}")
 
@@ -163,7 +164,7 @@ class SegmentResizer:
                     if word_pos != -1:
                         line_words.append(words_to_process.pop(0))
                         # Remove the word and any following spaces from remaining line
-                        remaining_line = remaining_line[word_pos + len(word_clean):].strip()
+                        remaining_line = remaining_line[word_pos + len(word_clean) :].strip()
                         continue
 
                 # If we can't find the word in the remaining line, we're done with this line

lyrics_transcriber/review/server.py

@@ -2,7 +2,7 @@ import logging
 from fastapi import FastAPI, Body
 from fastapi.middleware.cors import CORSMiddleware
 from typing import Optional, Dict, Any
-from ..types import CorrectionResult
+from ..types import CorrectionResult, WordCorrection, LyricsSegment
 import time
 import subprocess
 import os
@@ -64,16 +64,22 @@ async def complete_review(updated_data: Dict[str, Any] = Body(...)):
     Mark the review as complete and update the correction data.
 
     Args:
-        updated_data: The complete correction result data with any modifications
+        updated_data: Dictionary containing corrections and corrected_segments
     """
     global review_completed, current_review
 
     logger.info("Received updated correction data")
 
     try:
-        # Update the current review with modified data
-        # We use from_dict to ensure the data is properly structured
-        current_review = CorrectionResult.from_dict(updated_data)
+        # Only update the specific fields that were modified
+        if current_review is None:
+            raise ValueError("No review in progress")
+
+        # Update only the corrections and corrected_segments
+        current_review.corrections = [WordCorrection.from_dict(c) for c in updated_data["corrections"]]
+        current_review.corrected_segments = [LyricsSegment.from_dict(s) for s in updated_data["corrected_segments"]]
+        current_review.corrections_made = len(current_review.corrections)
+
         logger.info(f"Successfully updated correction data with {len(current_review.corrections)} corrections")
 
         review_completed = True

{lyrics_transcriber-0.32.2.dist-info → lyrics_transcriber-0.33.0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: lyrics-transcriber
-Version: 0.32.2
+Version: 0.33.0
 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

{lyrics_transcriber-0.32.2.dist-info → lyrics_transcriber-0.33.0.dist-info}/RECORD

@@ -1,9 +1,9 @@
 lyrics_transcriber/__init__.py,sha256=JpdjDK1MH_Be2XiSQWnb4i5Bbil1uPMA_KcuDZ3cyUI,240
 lyrics_transcriber/cli/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-lyrics_transcriber/cli/cli_main.py,sha256=pquvPfhw6brNgUyMuAzXfCUXNN0NM5tP_MyxlLWqNPc,8968
+lyrics_transcriber/cli/cli_main.py,sha256=TFB7CwzgLuwPfoV7ggPPe5dh4WKNcWRoZkCu_WWUcLQ,9818
 lyrics_transcriber/core/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-lyrics_transcriber/core/config.py,sha256=eddocVn_VoBGRqgBzfSvO7EZp3NuoGfSmd6mcT8wa74,941
-lyrics_transcriber/core/controller.py,sha256=eQ0M67SWIA-hr23fkw6F8hmqKkklOHsxOsJnYQLXFBE,12184
+lyrics_transcriber/core/config.py,sha256=y6MsAL0gFz7zRErtRRF81Z0vFOrySIrCw2aKDHExBz8,1160
+lyrics_transcriber/core/controller.py,sha256=t1jUCSJWHZgYurdGKEbyAqDNG5PTNkgiVueDqIzG06k,17149
 lyrics_transcriber/correction/anchor_sequence.py,sha256=YpKyY24Va5i4JgzP9ssqlOIkaYu060KaldiehbfgTdk,22200
 lyrics_transcriber/correction/corrector.py,sha256=VOM6YhbANu00rYs6JpKHGZXnZtD5fxArnYtRrsp1YM4,12998
 lyrics_transcriber/correction/handlers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -17,7 +17,7 @@ lyrics_transcriber/correction/handlers/sound_alike.py,sha256=mAmnpRpO29rHaP96V-U
 lyrics_transcriber/correction/handlers/syllables_match.py,sha256=5M7-0A6G-eu4nyzxT0wiuUpN5zbqXq9d-zeTP7AdLfg,8377
 lyrics_transcriber/correction/handlers/word_count_match.py,sha256=zbyZ01VE_6azaFpi8rS0Ato7c_VBxM2KV83VnDH5t3c,2522
 lyrics_transcriber/correction/handlers/word_operations.py,sha256=2COTaJsEwpSWyXHXmGgjfcf2x7tbAnsQ0dIW0qyHYK4,5141
-lyrics_transcriber/correction/phrase_analyzer.py,sha256=D94Ytlo7U2eTHYJ_nCVLhafmtTYExKx6v2_aczOJaQ8,16455
+lyrics_transcriber/correction/phrase_analyzer.py,sha256=dtO_2LjxnPdHJM7De40mYIdHCkozwhizVVQp5XGO7x0,16962
 lyrics_transcriber/correction/text_utils.py,sha256=VkOqgZHa9wEqLJdVNi4-KLFojQ6d4lWOGl_Y_vknenU,808
 lyrics_transcriber/lyrics/base_lyrics_provider.py,sha256=i4wxzu8nk2a3NDtnB_4r6rOGBZ7WvJFVlcEBjAkUYgI,5511
 lyrics_transcriber/lyrics/genius.py,sha256=M4rs3yk5RKW-RYfMm9w-UxwKQ8itgYeM-kVS6LCn8D0,3295
@@ -68,19 +68,19 @@ lyrics_transcriber/output/fonts/verdana.ttf,sha256=lu0UlJyktzks_yNbnEHVXBJTgqu-D
 lyrics_transcriber/output/generator.py,sha256=HQa3Ft8SKJie9-cYO0NKDbAU2-h_YnnH5wACxj0qFKw,7482
 lyrics_transcriber/output/lyrics_file.py,sha256=_KQyQjCOMIwQdQ0115uEAUIjQWTRmShkSfQuINPKxaw,3741
 lyrics_transcriber/output/plain_text.py,sha256=3mYKq0BLYz1rGBD6ROjG2dn6BPuzbn5dxIQbWZVi4ao,3689
-lyrics_transcriber/output/segment_resizer.py,sha256=xkKCNt4CTdTErUTYsYtjmllKY8YHny1srqQMrJQYbK8,17141
+lyrics_transcriber/output/segment_resizer.py,sha256=b553FCdcjYAl9T1IA5K6ya0pcn1-irD5spmxSc26wnI,17143
 lyrics_transcriber/output/subtitles.py,sha256=BQy7N_2zdBBWEiHL0NWFz3ZgAerWqQvTLALgxxK3Etk,16920
 lyrics_transcriber/output/video.py,sha256=kYGeEMYtoJvrGnMuyNpuSmu2DTskGDXBNlrv6ddvC8I,8485
 lyrics_transcriber/review/__init__.py,sha256=_3Eqw-uXZhOZwo6_sHZLhP9vxAVkLF9EBXduUvPdLjQ,57
-lyrics_transcriber/review/server.py,sha256=iAG0WUkGrqnAF7dI4ZQQayp2qaamqGGYT6rWJF9OysI,4397
+lyrics_transcriber/review/server.py,sha256=xUW55PhAeCKldXFm6F2X7waYid5vI_BsiPSoF4KnO0g,4744
 lyrics_transcriber/storage/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 lyrics_transcriber/storage/dropbox.py,sha256=Dyam1ULTkoxD1X5trkZ5dGp5XhBGCn998moC8IS9-68,9804
 lyrics_transcriber/transcribers/audioshake.py,sha256=QzKGimVa6BovlvYFj35CbGpaGePI_DApAJGEBR_JQLc,8709
 lyrics_transcriber/transcribers/base_transcriber.py,sha256=yPzUWPTCGmzE97H5Rz6g61e-qEGL77ZzUoiBOmswhts,5973
 lyrics_transcriber/transcribers/whisper.py,sha256=P0kas2_oX16MO1-Qy7U5gl5KQN-RuUIJZz7LsEFLUiE,12906
 lyrics_transcriber/types.py,sha256=xGf3hkTRcGZTTAjMVIev2i2DOU6co0QGpW8NxvaBQAA,16759
-lyrics_transcriber-0.32.2.dist-info/LICENSE,sha256=BiPihPDxhxIPEx6yAxVfAljD5Bhm_XG2teCbPEj_m0Y,1069
-lyrics_transcriber-0.32.2.dist-info/METADATA,sha256=_eW3ZTF3vCxCPtfJRzKgedrQ8VmaL1b244mAlxO71tY,5856
-lyrics_transcriber-0.32.2.dist-info/WHEEL,sha256=IYZQI976HJqqOpQU6PHkJ8fb3tMNBFjg-Cn-pwAbaFM,88
-lyrics_transcriber-0.32.2.dist-info/entry_points.txt,sha256=ChnmR13YoalGnC3sHW0TppX5FbhEXntYIha24tVQJ1M,104
-lyrics_transcriber-0.32.2.dist-info/RECORD,,
+lyrics_transcriber-0.33.0.dist-info/LICENSE,sha256=BiPihPDxhxIPEx6yAxVfAljD5Bhm_XG2teCbPEj_m0Y,1069
+lyrics_transcriber-0.33.0.dist-info/METADATA,sha256=2iT3ynJ2po1YYE1db_WX1XZBRrpQ-Z2-JJTkCP7nUAY,5856
+lyrics_transcriber-0.33.0.dist-info/WHEEL,sha256=IYZQI976HJqqOpQU6PHkJ8fb3tMNBFjg-Cn-pwAbaFM,88
+lyrics_transcriber-0.33.0.dist-info/entry_points.txt,sha256=ChnmR13YoalGnC3sHW0TppX5FbhEXntYIha24tVQJ1M,104
+lyrics_transcriber-0.33.0.dist-info/RECORD,,