Code Editor for app.py

import streamlit as st
import pandas as pd
from profanity_check import predict, predict_prob
import nltk
from nltk.tokenize import sent_tokenize

# Download required NLTK data
try:
    nltk.data.find('tokenizers/punkt')
except LookupError:
    nltk.download('punkt')

st.title("Content Safety Detector")
st.write("This app helps detect potentially harmful or inappropriate content in text.")

# Text input area
text_input = st.text_area(
    "Enter the text you want to analyze:",
    height=200,
    placeholder="Type or paste your text here..."
)

if text_input:
    st.subheader("Analysis Results")
    
    # Split text into sentences
    sentences = sent_tokenize(text_input)
    
    # Analyze each sentence
    results = []
    for sentence in sentences:
        if sentence.strip():  # Skip empty sentences
            probability = predict_prob([sentence])[0]
            is_harmful = predict([sentence])[0]
            results.append({
                'Sentence': sentence,
                'Harmful Content Probability': f"{probability:.2%}",
                'Is Harmful': '⚠️ Yes' if is_harmful else '✅ No'
            })
    
    # Create a DataFrame
    df = pd.DataFrame(results)
    
    # Display overall statistics
    total_sentences = len(results)
    harmful_sentences = sum(1 for r in results if '⚠️ Yes' in r['Is Harmful'])
    
    col1, col2 = st.columns(2)
    with col1:
        st.metric("Total Sentences", total_sentences)
    with col2:
        st.metric("Potentially Harmful Sentences", harmful_sentences)
    
    if harmful_sentences > 0:
        st.warning(
            f"⚠️ This text contains {harmful_sentences} potentially harmful sentences. "
            "Please review the content carefully."
        )
    else:
        st.success("✅ No harmful content detected in the text.")
    
    # Display detailed results
    st.subheader("Detailed Analysis")
    st.dataframe(
        df,
        column_config={
            "Sentence": st.column_config.TextColumn("Sentence", width="large"),
            "Harmful Content Probability": st.column_config.TextColumn("Probability", width="medium"),
            "Is Harmful": st.column_config.TextColumn("Status", width="small"),
        },
        hide_index=True,
    )
    
    st.caption("""
    Note: This tool uses machine learning to detect potentially harmful content. 
    It may not be 100% accurate and should be used as a guidance rather than a definitive assessment.
    Please use your judgment when reviewing the results.
    """)

# Sidebar with information
with st.sidebar:
    st.title("About")
    st.info("""
    This content detector helps identify potentially harmful or inappropriate content in text. 
    It analyzes the text sentence by sentence and provides:
    
    - Overall statistics
    - Probability scores
    - Sentence-level analysis
    
    Use this tool responsibly and remember that machine learning models 
    may not catch all harmful content or may flag safe content incorrectly.
    """)