Update app.py

app.py

@@ -1,4 +1,5 @@
 import os
+import subprocess
 import gradio as gr
 import spaces
 from infer_rvc_python import BaseLoader
@@ -26,6 +27,57 @@ import sys
 import torch
 import fairseq
 
+
+def get_supported_audio_video_extensions():
+    try:
+        subtitle_codecs = set()
+        decoders_res = subprocess.run(
+            ["ffmpeg", "-decoders"], capture_output=True, text=True, check=True
+        )
+        for line in decoders_res.stdout.splitlines():
+            line_str = line.strip()
+            if len(line_str) > 7 and line_str[0] in ("V", "A", "S"):
+                parts = line_str.split()
+                if len(parts) >= 2:
+                    media_type = line_str[0]
+                    codec_name = parts[1].lower()
+                    if media_type == "S":
+                        subtitle_codecs.add(codec_name)
+
+        demuxers_res = subprocess.run(
+            ["ffmpeg", "-demuxers"], capture_output=True, text=True, check=True
+        )
+        extensions = set()
+        for line in demuxers_res.stdout.splitlines():
+            line_str = line.strip()
+            if line_str.startswith("D"):
+                parts = line_str.split(maxsplit=2)
+                if len(parts) >= 2:
+                    demux_names = parts[1].split(",")
+                    description = parts[2].lower() if len(parts) >= 3 else ""
+
+                    if any(
+                        sub_kw in description
+                        for sub_kw in ["subtitle", "caption", "teletext", "lyrics"]
+                    ):
+                        continue
+
+                    for ext in demux_names:
+                        clean_ext = ext.strip().lower()
+                        if clean_ext in subtitle_codecs:
+                            continue
+                        if clean_ext and clean_ext.isalnum():
+                            extensions.add(f".{clean_ext}")
+
+        return sorted(list(extensions))
+    except Exception as e:
+        print(f"Error querying ffmpeg: {e}")
+        return [".mp3", ".wav", ".flac", ".m4a", ".aac", ".ogg", ".opus", ".wma", ".aiff", ".aif", ".alac", ".caf", ".amr"]
+
+
+supported_extensions = get_supported_audio_video_extensions()
+print("Supported extensions found:", supported_extensions)
+
 parser = argparse.ArgumentParser(description="Run the app with optional sharing")
 parser.add_argument(
     '--share',
@@ -154,10 +206,10 @@ def find_my_model(a_, b_):
 
     for ff in end_files:
         if ff.endswith(".pth"):
-            model = os.path.join(directory, ff)
+            model = ff
             gr.Info(f"Model found: {ff}")
         if ff.endswith(".index"):
-            index = os.path.join(directory, ff)
+            index = ff
             gr.Info(f"Index found: {ff}")
 
     if not model:
@@ -368,6 +420,7 @@ def run(
     audio_effects,
     type_output,
     steps,
+    play_audio=False,
 ):
     if not audio_files:
         raise ValueError("The audio pls")
@@ -381,7 +434,7 @@ def run(
     except Exception as e:
         print(e)
 
-    if file_m is not None and file_m.endswith(".txt"):
+    if file_m is not None and (file_m.endswith(".txt") or file_m.endswith(".zip")):
         file_m, file_index = find_my_model(file_m, file_index)
         print(file_m, file_index)
 
@@ -409,7 +462,9 @@ def run(
     if audio_effects:
         result = add_audio_effects(result, type_output)
 
-    return result
+    player_audio = result[0] if (play_audio and result and len(result) > 0) else None
+
+    return result, player_audio
 
 
 def audio_conf():
@@ -418,6 +473,7 @@ def audio_conf():
         file_count="multiple",
         type="filepath",
         container=True,
+        file_types=supported_extensions,
     )
 
 
@@ -426,6 +482,7 @@ def model_conf():
         label="Model file",
         type="filepath",
         height=130,
+        file_types=[".txt", ".pth", ".zip"],
     )
 
 
@@ -456,6 +513,7 @@ def index_conf():
         label="Index file",
         type="filepath",
         height=130,
+        file_types=[".index", ".txt"],
     )
 
 
@@ -550,13 +608,12 @@ def tts_button_conf():
     )
 
 
-def tts_play_conf():
+def player_conf():
     return gr.Checkbox(
         False,
-        label="Play",
-        # info="",
+        label="Audio Player",
         container=False,
-        visible=False,
+        visible=True,
     )
 
 
@@ -564,11 +621,10 @@ def sound_gui():
     return gr.Audio(
         value=None,
         type="filepath",
-        # format="mp3",
         autoplay=True,
-        visible=True,
+        visible=False,
         interactive=False,
-        elem_id="audio_tts",
+        label="Audio Player",
     )
 
 
@@ -610,7 +666,7 @@ def effects_conf():
     )
 
 
-def infer_tts_audio(tts_voice, tts_text, play_tts):
+def infer_tts_audio(tts_voice, tts_text, play_audio):
     out_dir = "output"
     folder_tts = "USER_"+str(random.randint(10000, 99999))
 
@@ -619,9 +675,8 @@ def infer_tts_audio(tts_voice, tts_text, play_tts):
     out_path = os.path.join(out_dir, folder_tts, "tts.mp3")
 
     asyncio.run(edge_tts.Communicate(tts_text, "-".join(tts_voice.split('-')[:-1])).save(out_path))
-    if play_tts:
-        return [out_path], out_path
-    return [out_path], None
+    player_audio = out_path if play_audio else None
+    return [out_path], player_audio
 
 
 def show_components_tts(value_active):
@@ -631,8 +686,6 @@ def show_components_tts(value_active):
         visible=value_active
     ), gr.update(
         visible=value_active
-    ), gr.update(
-        visible=value_active
     )
 
 
@@ -672,15 +725,7 @@ def show_components_down(value_active):
         visible=value_active
     )
 
-CSS = """
-#audio_tts {
-  visibility: hidden;   /* invisible but still takes space */
-  height: 0px;
-  width: 0px;
-  max-width: 0px;
-  max-height: 0px;
-}
-"""
+CSS = ""
 
 def get_gui(theme):
     with gr.Blocks(theme=theme, css=CSS, fill_width=True, fill_height=False, delete_cache=delete_cache_time) as app:
@@ -696,26 +741,18 @@ def get_gui(theme):
                     with gr.Column():
                         with gr.Row():
                             tts_voice = tts_voice_conf()
-                            tts_active_play = tts_play_conf()
 
                 tts_button = tts_button_conf()
-                tts_play = sound_gui()
 
         active_tts.change(
             fn=show_components_tts,
             inputs=[active_tts],
-            outputs=[tts_voice, tts_text, tts_button, tts_active_play],
+            outputs=[tts_voice, tts_text, tts_button],
         )
 
         aud = audio_conf()
         # gr.HTML("<hr>")
 
-        tts_button.click(
-            fn=infer_tts_audio,
-            inputs=[tts_voice, tts_text, tts_active_play],
-            outputs=[aud, tts_play],
-        )
-
         down_active_gui = down_active_conf()
         down_info = gr.Markdown(
             f"Provide a link to a zip file, like this one: `https://huggingface.co/MrDawg/ToothBrushing/resolve/main/ToothBrushing.zip?download=true`, or separate links with a comma for the .pth and .index files, like this: `{test_model}`",
@@ -758,6 +795,22 @@ def get_gui(theme):
                     with gr.Row():
                         denoise_gui = denoise_conf()
                         effects_gui = effects_conf()
+                        player_gui = player_conf()
+
+        player_audio = sound_gui()
+
+        player_gui.change(
+            fn=lambda val: gr.update(visible=val),
+            inputs=[player_gui],
+            outputs=[player_audio],
+        )
+
+        tts_button.click(
+            fn=infer_tts_audio,
+            inputs=[tts_voice, tts_text, player_gui],
+            outputs=[aud, player_audio],
+        )
+
         button_base = button_conf()
         output_base = output_conf()
 
@@ -777,8 +830,9 @@ def get_gui(theme):
                 effects_gui,
                 format_out,
                 steps_gui,
+                player_gui,
             ],
-            outputs=[output_base],
+            outputs=[output_base, player_audio],
         )
 
         gr.Examples(
@@ -830,7 +884,7 @@ def get_gui(theme):
                 envel_r,
                 const,
             ],
-            outputs=[output_base],
+            outputs=[output_base, player_audio],
             cache_examples=False,
         )
         gr.Markdown(RESOURCES)
@@ -856,4 +910,4 @@ if __name__ == "__main__":
         quiet=False,
         debug=IS_COLAB,
         ssr_mode=False,
-    )
+    )