Update app.py

app.py

@@ -1,54 +1,42 @@
 import streamlit as st
 import speech_recognition as sr
 from pydub import AudioSegment
-import tempfile
-import os
 
 # Title and description
-st.title("Arabic Audio Trimming App")
-st.write("Upload an Arabic audio file and enter the desired text. The app will trim the audio to match the specified text.")
+st.title("Audio Trimming App")
+st.write("Upload an audio file and enter the desired text. The app will trim the audio to match the specified text.")
 
 # Upload an audio file
-uploaded_audio = st.file_uploader("Upload an Arabic audio file", type=["wav", "mp3"])
+uploaded_audio = st.file_uploader("Upload an audio file", type=["wav", "mp3"])
 
 # Input text box
-input_text = st.text_area("Enter the desired Arabic text")
+input_text = st.text_area("Enter the desired text")
 
 # Function to trim audio
 def trim_audio(input_audio, input_text):
     if not input_audio:
         return
 
-    # Create a temporary directory to store audio files
-    temp_dir = tempfile.TemporaryDirectory()
+    # Load the audio file and recognize text
+    recognizer = sr.Recognizer()
+    with sr.AudioFile(input_audio) as source:
+        audio = recognizer.record(source)
+        audio_text = recognizer.recognize_google(audio)
 
-    try:
-        # Save uploaded audio to a temporary file
-        temp_audio_path = os.path.join(temp_dir.name, "temp_audio.wav")
-        with open(temp_audio_path, "wb") as audio_file:
-            audio_file.write(input_audio.read())
+    # Process the audio and create a trimmed audio
+    trimmed_audio = AudioSegment.silent(duration=0)
+    text_to_match = input_text.lower()
+    audio_text = audio_text.lower()
 
-        # Initialize the recognizer with the Arabic language
-        recognizer = sr.Recognizer(language="ar-AE")
+    for word in audio_text.split():
+        if word in text_to_match:
+            trimmed_audio += AudioSegment.from_file(input_audio)
 
-        # Load the audio file and recognize text
-        with sr.AudioFile(temp_audio_path) as source:
-            audio = recognizer.record(source)
-            audio_text = recognizer.recognize_google(audio, language="ar-AE")
+    # Export the trimmed audio to a file
+    trimmed_audio.export("trimmed_audio.mp3", format="mp3")
 
-        # Process the audio and create a trimmed audio
-        trimmed_audio = AudioSegment.silent(duration=0)
-        text_to_match = input_text.lower()
-        audio_text = audio_text.lower()
-
-        for word in audio_text.split():
-            if word in text_to_match:
-                trimmed_audio += AudioSegment.from_file(temp_audio_path, format="wav")
-
-        st.audio(trimmed_audio.export(format="wav"), format="audio/wav")
-
-    finally:
-        temp_dir.cleanup()
+    # Display the trimmed audio
+    st.audio("trimmed_audio.mp3", format="audio/mp3")
 
 # Process button
 if st.button("Process"):