added downloader and bam as default

app.py

@@ -1,61 +1,38 @@
 import gradio as gr
 from asr import transcribe_multiple_files, ASR_LANGUAGES, model
 from lid import identify, LID_EXAMPLES
+from csv_processor import CSV_FILE_PATH
 import logging
 import soundfile as sf
+import os
 
 # Set up logging
 logging.basicConfig(level=logging.DEBUG)
 logger = logging.getLogger(__name__)
 
-# def safe_process_file(file_obj):
-#     try:
-#         logger.debug(f"Processing file: {file_obj.name}")
-        
-#         # Check if the file_obj has a 'name' attribute (NamedString)
-#         if hasattr(file_obj, 'name'):
-#             file_path = file_obj.name
-#         else:
-#             logger.error("Unexpected file object type")
-#             raise ValueError("Unexpected file object type")
-
-#         logger.debug(f"Loading audio from file path: {file_path}")
-        
-#         # Use soundfile to read the audio file
-#         audio, sr = sf.read(file_path)
-        
-#         safe_name = f"audio_{hash(file_path)}.wav"
-#         logger.debug(f"File processed successfully: {safe_name}")
-#         return audio, sr, safe_name
-#     except Exception as e:
-#         logger.error(f"Error processing file {getattr(file_obj, 'name', 'unknown')}: {str(e)}")
-#         raise
-
-# def transcribe_multiple_files(audio_files, lang, transcription):
-#     transcriptions = []
-#     for audio_file in audio_files:
-#         try:
-#             audio, sr, safe_name = safe_process_file(audio_file)
-#             logger.debug(f"Transcribing file: {safe_name}")
-#             logger.debug(f"Language selected: {lang}")
-#             logger.debug(f"User-provided transcription: {transcription}")
-            
-#             result = transcribe_multiple_files(model, audio, lang, transcription)
-#             logger.debug(f"Transcription result: {result}")
-            
-#             transcriptions.append(f"File: {safe_name}\nTranscription: {result}\n")
-#         except Exception as e:
-#             logger.error(f"Error in transcription process: {str(e)}")
-#             transcriptions.append(f"Error processing file: {str(e)}\n")
-#     return "\n".join(transcriptions)
-
+def download_csv():
+    file_path = CSV_FILE_PATH  # Путь к вашему файлу
+    if os.path.exists(file_path):  # Проверяем, существует ли файл
+        return file_path  # Если файл существует, возвращаем путь для скачивания
+    else:
+        logger.error(f"file {file_path} not found!")
+        full_path = "/home/user/app/"+ file_path
+        exist_full_path = os.path.exists(full_path)
+        logger.error(f"file {exist_full_path} {"found" if exist_full_path else "not found"}!")
+        return None  # Возвращаем None, если файла нет
+    
 # Prepare language options for Dropdown
 language_options = [f"{k} ({v})" for k, v in ASR_LANGUAGES.items()]
-# bam_index = 0
-# try:
-#     bam_index = language_options.index("bam")
-# except ValueError:
-#     bam_index = 0
+bam_val = "bam (Bamanankan)"
+bam_index = 0 if bam_val not in language_options else language_options.index(bam_val)
+
+download_interface = gr.Interface(
+    fn=download_csv,
+    inputs=[],
+    outputs=gr.File(label="Скачать CSV"),
+    title="Скачать CSV файл",
+    description="Скачайте файл audio_plus_hash_uniq_07102024.csv"
+)
 
 
 mms_transcribe = gr.Interface(
@@ -65,7 +42,7 @@ mms_transcribe = gr.Interface(
         gr.Dropdown(
             choices=language_options,
             label="Language",
-            value=language_options[0] if language_options else None,
+            value=language_options[bam_index] if language_options else None,
         ),
         gr.Textbox(label="Optional: Provide your own transcription"),
     ],
@@ -86,8 +63,8 @@ mms_identify = gr.Interface(
 )
 
 tabbed_interface = gr.TabbedInterface(
-    [mms_transcribe, mms_identify],
-    ["Speech-to-text", "Language Identification"],
+    [mms_transcribe, mms_identify, download_interface],
+    ["Speech-to-text", "Language Identification", "Load to laptop CSV"],
 )
 
 with gr.Blocks() as demo:

asr.py

@@ -49,7 +49,7 @@ def transcribe_multiple_files(audio_files, lang, transcription):
     # for audio_file in audio_files:
     try:
         audio_samples, sr, safe_name = safe_process_file(audio_files)
-        logger.debug(f"Transcribing file: {safe_name}")
+        logger.debug(f"Transcribing file {audio_files}: {safe_name}")
         logger.debug(f"Language selected: {lang}")
         logger.debug(f"User-provided transcription: {transcription}")
         
@@ -61,7 +61,7 @@ def transcribe_multiple_files(audio_files, lang, transcription):
     except Exception as e:
         logger.error(f"Error in transcription process: {str(e)}")
         transcriptions.append(f"Error processing file: {str(e)}\n")
-    # return "\n".join(transcriptions)
+    return "\n".join(transcriptions)
 
 @spaces.GPU
 def transcribe_file(model, audio_samples, lang, user_transcription):

csv_processor.py

@@ -1,5 +1,6 @@
 import hashlib
 import csv
+import os
 
 
 def hash_md5(filepath):
@@ -11,16 +12,16 @@ def hash_md5(filepath):
         return md5_hash.hexdigest()
     except Exception as e:
         return str(e)
-
+CSV_FILE_PATH = "audio_plus_hash_uniq_07102024.csv"
 
 def update_csv_with_files(csv_file_path, audio_old_path, audio_16000_path, new_transcription):
     hash_old = hash_md5(audio_old_path)
     hash_new = hash_md5(audio_16000_path)
-    update_csv(csv_file_path, hash_old, hash_new)
+    update_csv(csv_file_path, hash_old, hash_new, audio_old_path, new_transcription)
 
-def update_csv(file_path, search_hash, hash_16000, new_path, new_transcription):
+def update_csv(csv_file_path, search_hash, hash_16000, new_path, new_transcription):
     # Use read/write mode to modify the relevant line or append if not found
-    with open(file_path, mode='r+', newline='', encoding='utf-8') as file:
+    with open(csv_file_path, mode='r+', newline='', encoding='utf-8') as file:
         reader = csv.DictReader(file)
         fieldnames = reader.fieldnames
         rows = list(reader)
@@ -42,7 +43,7 @@ def update_csv(file_path, search_hash, hash_16000, new_path, new_transcription):
             writer.writerows(rows)  # Write all rows back, with the updated one
         return 
     # Append a new row if the hash is not found
-    with open(file_path, mode='a', newline='', encoding='utf-8') as append_file:
+    with open(csv_file_path, mode='a', newline='', encoding='utf-8') as append_file:
         writer = csv.DictWriter(append_file, fieldnames=fieldnames)
         writer.writerow({
             'hash': search_hash,