Update app.py

app.py

@@ -146,6 +146,73 @@ def zip_to_many(zip_upload):
     os.rmdir(temp_dir)
     return "Done.", output_path
 
+def bulk_split_zip_files(zip_upload, mindur2, maxdur2, name_for_split_files2, strict):
+    if zip_upload is None:
+        raise gr.Error("Zip file cannot be empty!")
+    if mindur2 == maxdur2:
+        raise gr.Error(f"Cannot split mindur={mindur2} and maxdur={maxdur2}, min and max are the same number.")
+    elif mindur2 > maxdur2:
+        raise gr.Error(f"Cannot split mindur={mindur2} and maxdur={maxdur2}, mindur is higher than maxdur.")
+    elif name_for_split_files2 is None or name_for_split_files2 == "":
+        raise gr.Error("Split files name cannot be empty!")
+
+    temp_dir = "bulk_zip_temp"
+    os.makedirs(temp_dir, exist_ok=True)
+
+    with zipfile.ZipFile(zip_upload, "r") as zip_ref:
+        zip_ref.extractall(temp_dir)
+
+    split_files = []
+
+    for fname in os.listdir(temp_dir):
+        if not fname.lower().endswith((".wav", ".mp3")):
+            continue
+
+        audio_path = os.path.join(temp_dir, fname)
+        audio_regions = auditok.split(
+            audio_path,
+            min_dur=mindur2,
+            max_dur=maxdur2,
+            max_silence=0.3,
+            energy_threshold=45,
+            strict_min_dur=True if strict else False
+        )
+
+        base = os.path.splitext(os.path.basename(audio_path))[0]
+        for i, r in enumerate(audio_regions):
+            out_name = f"{name_for_split_files2}-{base}-{i+1}.wav"
+            filename = r.save(out_name)
+            split_files.append(filename)
+
+    if not split_files:
+        for root, dirs, files in os.walk(temp_dir, topdown=False):
+            for f in files:
+                os.remove(os.path.join(root, f))
+            for d in dirs:
+                os.rmdir(os.path.join(root, d))
+        os.rmdir(temp_dir)
+        raise gr.Error("No audio segments were created from the provided ZIP.")
+
+    zip_file_name = "bulk_split_audio_files.zip"
+    with zipfile.ZipFile(zip_file_name, "w") as zip_file:
+        for fpath in split_files:
+            zip_file.write(fpath, os.path.basename(fpath))
+
+    for fpath in split_files:
+        if os.path.exists(fpath):
+            os.remove(fpath)
+
+    for root, dirs, files in os.walk(temp_dir, topdown=False):
+        for f in files:
+            os.remove(os.path.join(root, f))
+        for d in dirs:
+            os.rmdir(os.path.join(root, d))
+    os.rmdir(temp_dir)
+
+    os.remove(zip_upload)
+
+    return f"File split successfully from ZIP! Amount created: {len(split_files)}", zip_file_name
+
 with gr.Blocks(theme='ParityError/Interstellar', title="Global Dataset Maker") as app:
     gr.HTML(
         "<h1> Welcome to the Cafeteria (formally GDMGS)!</h1>"
@@ -183,6 +250,22 @@ with gr.Blocks(theme='ParityError/Interstellar', title="Global Dataset Maker") a
                             [audiofileuploader, mindur2, maxdur2, name_for_split_files2, strict],
                             [gr.Text(label="Output"), gr.File(label="Zipped files")]
                         )
+            with gr.Tab("Bulk file splitter from ZIP"):
+                gr.Markdown("Upload a ZIP containing multiple audio files, and each of them will be split.")
+                with gr.Row():
+                    with gr.Column():
+                        with gr.Row():
+                            zip_file_bulk = gr.File(file_count='single', file_types=[".zip"], label="Zip file")
+                            mindur_bulk = gr.Number(label="Min duration", minimum=1, maximum=10, value=1)
+                            maxdur_bulk = gr.Number(label="Max duration", minimum=1, maximum=10, value=5)
+                            name_for_split_audio_files_bulk = gr.Textbox(label="Name prefix for split files")
+                            strict_bulk = gr.Checkbox(False, label="Enable strict?", info="Same behavior for single file uploading")
+                            bulk_split_btn = gr.Button("Bulk split", variant='primary')
+                        bulk_split_btn.click(
+                            bulk_split_zip_files,
+                            [zip_file_bulk, mindur_bulk, maxdur_bulk, name_for_split_audio_files_bulk, strict_bulk],
+                            [gr.Text(label="Output"), gr.File(label="Zipped split files")]
+                        )
 
             with gr.Tab("Split audio file by duration"):
                 gr.Markdown("If you have an audio file thats too long for MVSEP (or any other use cases), use this section to split the audio by duration.")
@@ -338,6 +421,7 @@ with gr.Blocks(theme='ParityError/Interstellar', title="Global Dataset Maker") a
                         )
 
         with gr.TabItem("Changelog"):
+            gr.Markdown("v1.5 - Added new tool - Bulk split audio files from a zip.")
             gr.Markdown("v1.4 - Added more MVSep and changed default model.")
             gr.Markdown("v1.3 - Added more MVSep models.")
             gr.Markdown("v1.2 - Added an option (in mp4 to mp3/wav converter) to have the output file be a random filename.")