Spaces:
Runtime error
Runtime error
Create app.py
Browse files
app.py
ADDED
|
@@ -0,0 +1,71 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
import PyPDF2
|
| 2 |
+
from docx import Document
|
| 3 |
+
import os
|
| 4 |
+
from unittest.mock import patch
|
| 5 |
+
from TTS.api import TTS
|
| 6 |
+
import torch
|
| 7 |
+
import gradio as gr
|
| 8 |
+
|
| 9 |
+
|
| 10 |
+
os.environ["COQUI_TOS_AGREED"] = "1"
|
| 11 |
+
|
| 12 |
+
# Function to always return 'y'
|
| 13 |
+
def always_yes(*args, **kwargs):
|
| 14 |
+
return 'y'
|
| 15 |
+
|
| 16 |
+
#Get device
|
| 17 |
+
device="cuda:0" if torch.cuda.is_available() else "cpu"
|
| 18 |
+
tts = TTS(model_name="tts_models/multilingual/multi-dataset/xtts_v2", progress_bar=False).to(device)
|
| 19 |
+
|
| 20 |
+
#################################
|
| 21 |
+
# Path to the audio file
|
| 22 |
+
SAMPLE_AUDIOS = ["sampleaudio.wav"]
|
| 23 |
+
|
| 24 |
+
################################
|
| 25 |
+
|
| 26 |
+
|
| 27 |
+
def text_to_speech(text):
|
| 28 |
+
save_path = "outputz.wav"
|
| 29 |
+
try:
|
| 30 |
+
tts.tts_to_file(
|
| 31 |
+
text,
|
| 32 |
+
speaker_wav=SAMPLE_AUDIOS,
|
| 33 |
+
language="en",
|
| 34 |
+
file_path=save_path,
|
| 35 |
+
split_sentences=True,
|
| 36 |
+
)
|
| 37 |
+
return save_path
|
| 38 |
+
except Exception as e:
|
| 39 |
+
return str(e)
|
| 40 |
+
|
| 41 |
+
def extract_text_from_pdf(file_path):
|
| 42 |
+
with open(file_path, 'rb') as file:
|
| 43 |
+
reader = PyPDF2.PdfReader(file)
|
| 44 |
+
text = "".join([page.extract_text() for page in reader.pages])
|
| 45 |
+
return text
|
| 46 |
+
|
| 47 |
+
def extract_text_from_doc(file_path):
|
| 48 |
+
doc = Document(file_path)
|
| 49 |
+
return "\n".join([para.text for para in doc.paragraphs])
|
| 50 |
+
|
| 51 |
+
def file_to_audio(file_path):
|
| 52 |
+
file_type = file_path.split('.')[-1]
|
| 53 |
+
if file_type == 'pdf':
|
| 54 |
+
text = extract_text_from_pdf(file_path)
|
| 55 |
+
elif file_type in ['doc', 'docx']:
|
| 56 |
+
text = extract_text_from_doc(file_path)
|
| 57 |
+
else:
|
| 58 |
+
return "Unsupported file format"
|
| 59 |
+
|
| 60 |
+
return text_to_speech(text)
|
| 61 |
+
def handle_inputs(text, file):
|
| 62 |
+
if text and file:
|
| 63 |
+
return None, "Please provide only one input at a time: text or file."
|
| 64 |
+
elif text:
|
| 65 |
+
audio_path = text_to_speech(text)
|
| 66 |
+
return audio_path, "Thanks for your text input"
|
| 67 |
+
elif file:
|
| 68 |
+
audio_path = file_to_audio(file)
|
| 69 |
+
return audio_path, 'Thanks for your File'
|
| 70 |
+
else:
|
| 71 |
+
return None, "No input provided."
|