Spaces:
Sleeping
Sleeping
File size: 6,525 Bytes
34362fa |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 |
import openai
import os
import streamlit as st
from PIL import Image
from gtts import gTTS
import tempfile
import shutil
import re
def translate_to_japanese(api_key, text):
"""
Translates English text to Japanese using OpenAI's API and provides pronunciation.
"""
# Validate input
if not api_key:
return "Error: API key is missing.", None
if not text:
return "Error: Input text is empty.", None
# Set the OpenAI API key
openai.api_key = api_key
# Define the messages for the chat model
messages_translation = [
{"role": "system", "content": "You are a helpful translator."},
{"role": "user", "content": f"Translate the following English text to Japanese:\n\n{text}"}
]
try:
# Call the OpenAI API to get the Japanese translation
response_translation = openai.ChatCompletion.create(
model="gpt-4o", # Use the correct endpoint for chat models
messages=messages_translation,
max_tokens=300,
temperature=0.5
)
# Extract the Japanese translation from the response
japanese_translation = response_translation.choices[0].message['content'].strip()
# Define the messages for the pronunciation (Romaji) request
messages_pronunciation = [
{"role": "system", "content": "You are a helpful assistant who provides the Romaji (Japanese pronunciation in Latin script) of Japanese text."},
{"role": "user", "content": f"Provide the Romaji pronunciation for the following Japanese text:\n\n{japanese_translation}"}
]
# Call the OpenAI API to get the pronunciation
response_pronunciation = openai.ChatCompletion.create(
model="gpt-4o",
messages=messages_pronunciation,
max_tokens=300,
temperature=0.5
)
# Extract the pronunciation (Romaji) from the response
pronunciation = response_pronunciation.choices[0].message['content'].strip()
return japanese_translation, pronunciation
except openai.error.OpenAIError as e:
return f"OpenAI API error: {str(e)}", None
except Exception as e:
return f"An unexpected error occurred: {str(e)}", None
# Function to clean pronunciation text
def clean_pronunciation(pronunciation_text):
# Remove introductory phrases like "Sure! The Romaji pronunciation..."
pronunciation_cleaned = re.sub(r"^Sure! The Romaji pronunciation for the Japanese text.*?is[:]*", "", pronunciation_text).strip()
return pronunciation_cleaned
# Function to generate audio file from text using gTTS
def generate_audio_from_text(text):
tts = gTTS(text, lang='ja') # 'ja' for Japanese language
# Save audio to a temporary file
temp_audio_file = tempfile.NamedTemporaryFile(delete=False, suffix=".mp3")
tts.save(temp_audio_file.name)
return temp_audio_file.name
# Streamlit UI
st.title("English to Japanese Translator with Pronunciation")
st.markdown("Translate English text into Japanese and get its pronunciation (Romaji).")
translateimg = Image.open("Untitled.png") # Ensure the file is in the correct directory
st.image(translateimg, use_container_width=True) # Adjust the size as per preference
# Access the API key from Hugging Face Secrets
api_key = os.getenv("OPENAI_API_KEY")
# Input field for the text
english_text = st.text_area("Enter the English text to translate")
# Initialize the progress bar and progress text above the translate button
progress_bar = st.progress(0)
progress_text = st.empty() # To show the progress text
# Button to trigger the translation
if st.button("Translate"):
if api_key and english_text:
try:
# Step 1: Request translation
progress_text.text("Translating text...")
progress_bar.progress(33) # Update progress bar to 33%
japanese_text, pronunciation = translate_to_japanese(api_key, english_text)
# Step 2: Check if translation was successful
if pronunciation:
progress_text.text("Generating Romaji pronunciation...")
progress_bar.progress(66) # Update progress bar to 66%
# Clean pronunciation (remove unnecessary parts)
cleaned_pronunciation = clean_pronunciation(pronunciation)
st.markdown("### Translation Result:")
st.write(f"**English Text:** {english_text}")
st.write(f"**Japanese Output:** {japanese_text}")
st.write(f"**Pronunciation:** {cleaned_pronunciation}")
# Save the result in a text file
result_text = f"English Text: {english_text}\n\nJapanese Translation: {japanese_text}\nPronunciation: {cleaned_pronunciation}"
# Write to a text file
with open("translation_result.txt", "w") as file:
file.write(result_text)
# Create a download button for the user to download the file
with open("translation_result.txt", "rb") as file:
st.download_button(
label="Download Translation Result",
data=file,
file_name="translation_result.txt",
mime="text/plain"
)
# Step 3: Generate audio for pronunciation
progress_text.text("Generating pronunciation audio...")
progress_bar.progress(100) # Update progress bar to 100%
audio_file_path = generate_audio_from_text(cleaned_pronunciation)
# Provide a button to play the pronunciation audio
st.audio(audio_file_path, format="audio/mp3")
translateimg2 = Image.open("v3.png") # Ensure the file is in the correct directory
st.image(translateimg2, width=150) # Adjust the size as per preference
else:
st.error(japanese_text) # Display error message if API call fails
except Exception as e:
st.error(f"An error occurred: {str(e)}")
else:
if not api_key:
st.error("API key is missing. Please add it as a secret in Hugging Face Settings.")
else:
st.error("Please provide text to translate.") |