import gradio as gr
from toxicity_model import (
    check_toxicity,
    is_custom_toxic,
    check_sentiment,
    paraphrase_text,
    check_cyberbullying_with_hatebert,
    is_person_or_pronoun
)
import mysql.connector

# MySQL connection settings
db_config = {
    'host': 'cyberbullying.cbo2y20cex64.ap-southeast-2.rds.amazonaws.com',
    'user': 'root',
    'password': 'Beauisagoodboy',
    'database': 'cyberbullying'
}

def log_to_rds(text, is_bullying, tox_score, sentiment_score, suggested, person_or_pronoun, cyberbullying_flag, zone_db, likelihood, comment):
    """
    Logs analyzed message data into MySQL database.
    """
    try:
        connection = mysql.connector.connect(**db_config)
        cursor = connection.cursor()

        tox_score = float(tox_score)
        sentiment_score = float(sentiment_score)

        insert_query = """
        INSERT INTO MESSAGE 
        (text, is_bullying, toxicity_score, sentiment_score, suggested_text, person_or_pronoun, cyberbullying_flag, zone, likelihood, comment)
        VALUES (%s, %s, %s, %s, %s, %s, %s, %s, %s, %s)
        """
        cursor.execute(insert_query, (
            text,
            int(is_bullying),
            tox_score,
            sentiment_score,
            suggested or "",
            int(person_or_pronoun),
            int(cyberbullying_flag),
            zone_db,
            likelihood,
            comment
        ))

        connection.commit()
        cursor.close()
        connection.close()

    except mysql.connector.Error as err:
        print(f"Database error: {err}")


def analyze(text):
    """
    Performs cyberbullying analysis on text and returns the same outputs as FastAPI.
    """
    tox_score = round(check_toxicity(text), 2)
    sent_score = round(check_sentiment(text), 2)
    custom_flag = is_custom_toxic(text)
    person_or_pronoun = is_person_or_pronoun(text)
    cyberbullying_flag = check_cyberbullying_with_hatebert(text)

    is_bullying = custom_flag or tox_score >= 0.4 or cyberbullying_flag >= 0.5
    suggested_text = paraphrase_text(text) if is_bullying else None
    suggested_text = str(suggested_text) if suggested_text else ""

    # Determine severity zone
    if custom_flag or tox_score >= 0.7 or cyberbullying_flag >= 0.7:
        zone = '🔴 Red Zone'
        zone_db = 'Red Zone'
        likelihood = 'Very high likelihood of bullying'
        comment = 'Warning: this message looks very harmful. It may seriously hurt someone. This may cross the line into cyberbullying.'
    elif tox_score >= 0.4 or cyberbullying_flag >= 0.4:
        zone = '🟠 Orange Zone'
        zone_db = 'Orange Zone'
        likelihood = 'High likelihood of bullying'
        comment = 'This could hurt someone’s feelings — try to say it in a more positive way.'
    elif tox_score >= 0.2 or cyberbullying_flag >= 0.2:
        zone = '🟡 Yellow Zone'
        zone_db = 'Yellow Zone'
        likelihood = 'Medium likelihood of bullying'
        comment = 'Looks safe, but context matters — make sure it won’t hurt anyone.'
    else:
        zone = '🟢 Green Zone'
        zone_db = 'Green Zone'
        likelihood = 'Low likelihood of bullying'
        comment = 'Looks good! No red flags here. Nice one!'

    # Log to database
    log_to_rds(text, is_bullying, tox_score, sent_score, suggested_text, person_or_pronoun, cyberbullying_flag, zone_db, likelihood, comment)

    return zone, likelihood, comment, suggested_text


# ----------------------
# Gradio interface
# ----------------------
iface = gr.Interface(
    fn=analyze,
    inputs=gr.Textbox(label="Enter text", lines=3),
    outputs=[
        gr.Textbox(label="Zone"),
        gr.Textbox(label="Likelihood"),
        gr.Textbox(label="Comment"),
        gr.Textbox(label="Suggested Text")
    ],
    title="Cyberbullying Detection Tool"
)

if __name__ == "__main__":
    iface.launch(share=True)