app.py

from flask import Flask, render_template, request
import matplotlib

matplotlib.use('Agg')
import matplotlib.pyplot as plt
import io
import base64
import pandas as pd
import google.generativeai as genai
import os
from docx import Document
import plotly.express as px
import plotly.io as pio

app = Flask(__name__)
app.config['UPLOAD_FOLDER'] = 'uploads'
app.config['MAX_CONTENT_LENGTH'] = 16 * 1024 * 1024  # 16MB max file size

# Configure Gemini API
GOOGLE_API_KEY = 'AIzaSyBLcWuSj6N1bkhQsTF4kt3_hFh4ibH11pQ'  # Replace with your actual API key
genai.configure(api_key=GOOGLE_API_KEY)
model = genai.GenerativeModel('gemini-2.0-flash')


def ensure_upload_folder():
    if not os.path.exists(app.config['UPLOAD_FOLDER']):
        os.makedirs(app.config['UPLOAD_FOLDER'])


def extract_text_from_docx(file_path):
    doc = Document(file_path)
    full_text = []
    for paragraph in doc.paragraphs:
        full_text.append(paragraph.text)
    return '\n'.join(full_text)


def extract_data_using_gemini(text):
    prompt = """

    Extract the event counts from the following table format in the text:



    2022-2023

    Cultural competitions/events: NUMBER

    Sports competitions/events: NUMBER

    Technical fest/Academic fest: NUMBER

    Social activities/events: NUMBER

    Any other events through Active clubs and forums: NUMBER



    2021-2022

    Cultural competitions/events: NUMBER

    Sports competitions/events: NUMBER

    Technical fest/Academic fest: NUMBER

    Social activities/events: NUMBER

    Any other events through Active clubs and forums: NUMBER



    2020-2021

    Cultural competitions/events: NUMBER

    Sports competitions/events: NUMBER

    Technical fest/Academic fest: NUMBER

    Social activities/events: NUMBER

    Any other events through Active clubs and forums: NUMBER



    2019-2020

    Cultural competitions/events: NUMBER

    Sports competitions/events: NUMBER

    Technical fest/Academic fest: NUMBER

    Social activities/events: NUMBER

    Any other events through Active clubs and forums: NUMBER



    2018-2019

    Cultural competitions/events: NUMBER

    Sports competitions/events: NUMBER

    Technical fest/Academic fest: NUMBER

    Social activities/events: NUMBER

    Any other events through Active clubs and forums: NUMBER



    Look for these exact numbers in the text. The data appears in a table with years and categories.

    For each year, find:

    - Number of Cultural competitions/events

    - Number of Sports competitions/events

    - Number of Technical fest/Academic fest events

    - Number of Social activities/events

    - Number of "Any other events through Active clubs and forums"



    Return the data in this exact Python dictionary format:

    {

        '2022-2023': {'Cultural': 11, 'Sports': 10, 'Technical': 29, 'Social': 15, 'Other': 20},

        '2021-2022': {'Cultural': 7, 'Sports': 8, 'Technical': 13, 'Social': 12, 'Other': 15},

        '2020-2021': {'Cultural': 7, 'Sports': 9, 'Technical': 15, 'Social': 10, 'Other': 17},

        '2019-2020': {'Cultural': 12, 'Sports': 17, 'Technical': 21, 'Social': 14, 'Other': 11},

        '2018-2019': {'Cultural': 8, 'Sports': 17, 'Technical': 15, 'Social': 11, 'Other': 9}

    }



    Important:

    - Use the EXACT numbers from the document

    - Include ALL years from 2018-2019 to 2022-2023

    - Make sure to find the correct table in the document that has these numbers

    - Return only the Python dictionary, no other text

    """

    try:
        # Print the first part of the text for debugging
        print("\nSearching in text:")
        print("=" * 50)
        # Look for specific patterns in text
        import re
        years = re.findall(r'(20\d{2}-20\d{2})', text)
        print(f"Found years: {years}")

        # Look for numbers near key terms
        cultural = re.findall(r'Cultural competitions/events\s*(\d+)', text)
        sports = re.findall(r'Sports competitions/events\s*(\d+)', text)
        technical = re.findall(r'Technical fest/Academic fest\s*(\d+)', text)
        other = re.findall(r'Any other events.*?(\d+)', text)
        social = re.findall(r'Social activities/events\s*(\d+)', text)

        print(f"Found cultural numbers: {cultural}")
        print(f"Found sports numbers: {sports}")
        print(f"Found technical numbers: {technical}")
        print(f"Found other numbers: {other}")
        print(f"Found social numbers: {social}")
        print("=" * 50)

        response = model.generate_content(text + "\n" + prompt)
        response_text = response.text.strip()

        # Debug print
        print("Raw response:", response_text)

        # Remove any markdown formatting
        if '' in response_text:
            response_text = response_text.split('')[1]
            if 'python' in response_text.split('\n')[0]:
                response_text = '\n'.join(response_text.split('\n')[1:])

        # Clean the response text
        response_text = response_text.strip()
        print("Cleaned response:", response_text)

        # Parse the response
        try:
            import ast
            data = ast.literal_eval(response_text)
        except:
            # Fallback to JSON parsing if ast fails
            response_text = response_text.replace("'", '"')
            import json
            data = json.loads(response_text)

        # Validate data structure
        if not isinstance(data, dict):
            raise ValueError("Response is not a dictionary")

        # Ensure all years are present
        expected_years = ['2022-2023', '2021-2022', '2020-2021', '2019-2020', '2018-2019']
        if not all(year in data for year in expected_years):
            raise ValueError("Missing some years in the data")

        # Ensure all categories are present for each year
        required_categories = {'Cultural', 'Sports', 'Technical', 'Social', 'Other'}
        for year in data:
            if not all(cat in data[year] for cat in required_categories):
                raise ValueError(f"Missing categories in year {year}")

        return data

    except Exception as e:
        print(f"Error processing with Gemini: {str(e)}")
        print(f"Response text was: {response_text if 'response_text' in locals() else 'No response text'}")
        return None


def get_graph_insights(data, plot_type):
    """Generate detailed insights including SWOT analysis for different types of plots."""
    df = pd.DataFrame(data).T

    if plot_type == 'bar':
        total_by_category = df.sum()
        max_category = total_by_category.idxmax()
        min_category = total_by_category.idxmin()
        avg_events = total_by_category.mean()

        insights = {
            'main_insight': f"The most frequent event category overall is {max_category} with {int(total_by_category[max_category])} events, while {min_category} has the least with {int(total_by_category[min_category])} events.",
            'swot': {
                'strengths': [
                    f"Strong performance in {max_category} events",
                    f"Diverse range of events across categories",
                    f"Average of {avg_events:.1f} events per category"
                ],
                'weaknesses': [
                    f"Low participation in {min_category} events",
                    f"Uneven distribution across categories",
                    "Potential resource allocation issues"
                ],
                'opportunities': [
                    f"Room for growth in {min_category} category",
                    "Potential for cross-category events",
                    "Scope for balanced development"
                ],
                'threats': [
                    "Risk of over-dependence on dominant category",
                    "Resource strain in peak periods",
                    "Sustainability challenges"
                ]
            },
            'recommendations': [
                f"Consider boosting {min_category} events",
                "Implement balanced resource allocation",
                "Develop cross-category initiatives"
            ]
        }
        return insights

    elif plot_type == 'pie':
        latest_year = '2022-2023'
        year_data = data[latest_year]
        total = sum(year_data.values())
        max_cat = max(year_data.items(), key=lambda x: x[1])
        min_cat = min(year_data.items(), key=lambda x: x[1])
        percentage = (max_cat[1] / total) * 100

        insights = {
            'main_insight': f"In {latest_year}, {max_cat[0]} events dominated with {max_cat[1]} events ({percentage:.1f}% of total events).",
            'swot': {
                'strengths': [
                    f"Strong presence in {max_cat[0]} category",
                    "Clear category leadership",
                    "Established event structure"
                ],
                'weaknesses': [
                    f"Under-representation in {min_cat[0]} category",
                    "Imbalanced distribution",
                    "Resource concentration risks"
                ],
                'opportunities': [
                    "Potential for category diversification",
                    "Growth in underserved categories",
                    "New event type development"
                ],
                'threats': [
                    "Category saturation risk",
                    "Resource allocation challenges",
                    "Sustainability concerns"
                ]
            },
            'recommendations': [
                "Diversify event portfolio",
                f"Strengthen {min_cat[0]} category",
                "Implement balanced growth strategy"
            ]
        }
        return insights

    elif plot_type == 'line':
        trend = "increasing" if df.iloc[-1].mean() > df.iloc[0].mean() else "decreasing"
        growth_rate = ((df.iloc[-1].mean() - df.iloc[0].mean()) / df.iloc[0].mean() * 100)

        insights = {
            'main_insight': f"The overall trend shows a {trend} pattern with a {growth_rate:.1f}% change in event frequency over the years.",
            'swot': {
                'strengths': [
                    f"Consistent {trend} trend",
                    "Clear growth trajectory",
                    "Established pattern"
                ],
                'weaknesses': [
                    "Fluctuations in growth rate",
                    "Periodic inconsistencies",
                    "Resource scaling challenges"
                ],
                'opportunities': [
                    "Growth optimization potential",
                    "Pattern regularization",
                    "Strategic planning possibilities"
                ],
                'threats': [
                    "Sustainability of growth rate",
                    "Resource management challenges",
                    "Market saturation risks"
                ]
            },
            'recommendations': [
                "Develop sustainable growth plan",
                "Implement resource scaling strategy",
                "Monitor growth patterns"
            ]
        }
        return insights

    elif plot_type == 'growth':
        growth_rates = df.pct_change() * 100
        avg_growth = growth_rates.mean().mean()
        max_growth = growth_rates.max().max()
        min_growth = growth_rates.min().min()

        insights = {
            'main_insight': f"The average year-over-year growth rate is {avg_growth:.1f}%, with peaks of {max_growth:.1f}% and lows of {min_growth:.1f}%.",
            'swot': {
                'strengths': [
                    "Positive average growth rate",
                    "Strong peak performance periods",
                    "Growth momentum"
                ],
                'weaknesses': [
                    "Growth rate volatility",
                    "Negative growth periods",
                    "Inconsistent patterns"
                ],
                'opportunities': [
                    "Growth stabilization potential",
                    "Performance optimization",
                    "Strategic growth planning"
                ],
                'threats': [
                    "Growth sustainability",
                    "Resource scaling challenges",
                    "Market fluctuations"
                ]
            },
            'recommendations': [
                "Stabilize growth patterns",
                "Develop contingency plans",
                "Implement growth monitoring"
            ]
        }
        return insights

    elif plot_type == 'area':
        total_growth = ((df.iloc[-1].sum() - df.iloc[0].sum()) / df.iloc[0].sum() * 100)
        avg_yearly_growth = total_growth / (len(df) - 1)

        insights = {
            'main_insight': f"The cumulative events show a {total_growth:.1f}% total change, averaging {avg_yearly_growth:.1f}% yearly growth.",
            'swot': {
                'strengths': [
                    "Consistent cumulative growth",
                    "Strong overall trajectory",
                    "Clear progress pattern"
                ],
                'weaknesses': [
                    "Growth rate variations",
                    "Resource scaling challenges",
                    "Potential sustainability issues"
                ],
                'opportunities': [
                    "Long-term growth potential",
                    "Pattern optimization",
                    "Strategic expansion"
                ],
                'threats': [
                    "Scaling challenges",
                    "Resource constraints",
                    "Market saturation"
                ]
            },
            'recommendations': [
                "Develop long-term growth strategy",
                "Implement resource planning",
                "Monitor cumulative trends"
            ]
        }
        return insights

    return {
        'main_insight': "No specific insights available for this visualization.",
        'swot': {
            'strengths': [],
            'weaknesses': [],
            'opportunities': [],
            'threats': []
        },
        'recommendations': []
    }


def create_plots(data):
    plots = {}
    df = pd.DataFrame(data).T

    # Bar Chart
    fig1 = px.bar(df, barmode='group', title='Events Distribution Across Years')
    plots['bar'] = {
        'plot': pio.to_html(fig1, full_html=False),
        'insight': get_graph_insights(data, 'bar')
    }

    # Pie Chart
    latest_year = '2022-2023'
    fig2 = px.pie(names=data[latest_year].keys(), values=data[latest_year].values(),
                  title=f'Event Distribution for {latest_year}')
    plots['pie'] = {
        'plot': pio.to_html(fig2, full_html=False),
        'insight': get_graph_insights(data, 'pie')
    }

    # Line Chart
    fig3 = px.line(df, markers=True, title='Event Trends Over Years')
    plots['line'] = {
        'plot': pio.to_html(fig3, full_html=False),
        'insight': get_graph_insights(data, 'line')
    }

    # Growth Rate Chart
    growth_rates = df.pct_change() * 100
    fig4 = px.bar(growth_rates, title='Year-over-Year Growth Rate by Category')
    plots['growth'] = {
        'plot': pio.to_html(fig4, full_html=False),
        'insight': get_graph_insights(data, 'growth')
    }

    # Area Chart
    fig5 = px.area(df, title='Cumulative Events Distribution')
    plots['area'] = {
        'plot': pio.to_html(fig5, full_html=False),
        'insight': get_graph_insights(data, 'area')
    }

    # Statistical Analysis
    stats = {
        'total_events': df.sum().sum(),
        'avg_events_per_year': df.sum(axis=1).mean().round(2),
        'most_active_year': df.sum(axis=1).idxmax(),
        'most_common_category': df.sum().idxmax(),
        'growth_analysis': {
            'total_growth': ((df.iloc[-1].sum() - df.iloc[0].sum()) / df.iloc[0].sum() * 100).round(2),
            'category_growth': ((df.iloc[-1] - df.iloc[0]) / df.iloc[0] * 100).round(2).to_dict()
        }
    }
    plots['stats'] = stats

    return plots


@app.route('/', methods=['GET', 'POST'])
def index():
    plots = None
    error_message = None

    if request.method == 'POST':
        if 'document' not in request.files:
            error_message = 'No file uploaded'
            return render_template('index.html', error=error_message)

        file = request.files['document']
        if file.filename == '':
            error_message = 'No file selected'
            return render_template('index.html', error=error_message)

        if file and file.filename.endswith('.docx'):
            ensure_upload_folder()
            file_path = os.path.join(app.config['UPLOAD_FOLDER'], file.filename)
            file.save(file_path)

            try:
                text = extract_text_from_docx(file_path)
                data = extract_data_using_gemini(text)
                print("Extracted data:", data)

                if data:
                    plots = create_plots(data)
                else:
                    error_message = 'Could not extract data from document. Please check the document format.'

                os.remove(file_path)

            except Exception as e:
                error_message = f'Error processing document: {str(e)}'
                print(f"Full error: {str(e)}")
        else:
            error_message = 'Please upload a .docx file'

    return render_template('index.html', plots=plots, error=error_message)


if __name__ == '__main__':
    app.run(debug=True, port=5001)