app.py

import gradio as gr
from docx import Document
from datetime import datetime
import re
import tempfile
import pytesseract
from PIL import Image
from groq import Groq
import io
import os

# Initialize APIs
client = Groq(api_key=os.environ.get("GROQ_API_KEY"))
pytesseract.pytesseract.tesseract_cmd = r'/usr/bin/tesseract'  # Update path as needed

# Helper Functions
def extract_text_from_image(image_path):
    try:
        img = Image.open(image_path)
        text = pytesseract.image_to_string(img)
        return text.strip() if text else "No text found in image"
    except Exception as e:
        return f"OCR Error: {str(e)}"

def groq_processing(prompt):
    try:
        completion = client.chat.completions.create(
            model="mixtral-8x7b-32768",
            messages=[
                {
                    "role": "user",
                    "content": prompt
                }
            ]
        )
        return completion.choices[0].message.content
    except Exception as e:
        return f"AI Processing Error: {str(e)}"

def analyze_sentiment(text):
    prompt = f"""Analyze the sentiment of this text with detailed emotions:
    {text}
    
    Format response as: 
    Primary Emotion: [emotion]
    Secondary Emotions: [comma-separated list]
    Confidence Level: [percentage]"""
    return groq_processing(prompt)

def categorize_content(text):
    prompt = f"""Categorize this text into specific themes from the following options:
    Technology, Business, Lifestyle, Education, Politics, Health, Entertainment, Sports, Art, Science
    
    Text: {text}
    
    Respond with top 3 relevant categories in order of relevance."""
    return groq_processing(prompt)

# Document Processing Functions
def create_word_document(posts):
    doc = Document()
    doc.add_heading('Social Media Data Extraction Report', 0)
    
    for idx, post in enumerate(posts):
        doc.add_heading(f'Post {idx+1}', level=1)
        
        data = [
            ("Date of Post", post.get('date', 'N/A')),
            ("Media Type", post.get('media_type', 'N/A')),
            ("Number of Pictures", post.get('num_pictures', 'N/A')),
            ("Likes", post.get('likes', 'N/A')),
            ("Comments", post.get('comments', 'N/A')),
            ("Caption", post.get('caption', 'N/A')),
            ("OCR Text", post.get('ocr_text', 'N/A')),
            ("Language", post.get('language', 'N/A')),
            ("Sentiment Analysis", post.get('sentiment', 'N/A')),
            ("Content Categories", post.get('categories', 'N/A')),
            ("Hashtags", ', '.join(post.get('hashtags', [])) if post.get('hashtags') else 'N/A'),
            ("Concept Keywords", ', '.join(post.get('concepts', [])) if post.get('concepts') else 'N/A')
        ]
        
        for label, value in data:
            doc.add_paragraph(f"{label}: {value}")
        
        doc.add_page_break()
    
    temp_file = tempfile.NamedTemporaryFile(delete=False, suffix=".docx")
    doc.save(temp_file.name)
    return temp_file.name

# Processing Functions 
def process_social_media(profile_link, hashtags, concepts, platform, date_range, num_posts):
    # Mock data - Replace with actual social media API calls
    mock_image = Image.new('RGB', (800, 600), color='white')
    mock_image_path = tempfile.mktemp(suffix='.jpg')
    mock_image.save(mock_image_path)
    
    posts = [{
        'date': datetime.now().strftime("%Y-%m-%d"),
        'media_type': 'Image',
        'num_pictures': 1,
        'likes': 150,
        'comments': 20,
        'caption': 'Example caption with #technology',
        'ocr_text': extract_text_from_image(mock_image_path),
        'hashtags': ['#technology'],
        'concepts': [c.strip() for c in concepts.split(',')],
        'sentiment': analyze_sentiment('Example caption with #technology'),
        'categories': categorize_content('Example caption with #technology')
    }]
    
    os.remove(mock_image_path)
    return create_word_document(posts[:num_posts])

def process_word_file(file, concepts):
    doc = Document(file.name)
    posts = []
    
    for para in doc.paragraphs:
        if para.text.startswith('Post'):
            posts.append({'caption': ''})
        elif posts:
            posts[-1]['caption'] += para.text + '\n'
    
    processed_posts = []
    for post in posts:
        caption = post.get('caption', '')
        processed_post = {
            'date': datetime.now().strftime("%Y-%m-%d"),
            'media_type': 'Text',
            'caption': caption,
            'sentiment': analyze_sentiment(caption),
            'categories': categorize_content(caption),
            'concepts': [c.strip() for c in concepts.split(',')],
            'hashtags': re.findall(r'#\w+', caption)
        }
        processed_posts.append(processed_post)
    
    return create_word_document(processed_posts)

# Gradio Interface
with gr.Blocks(title="Social Media Analyzer") as app:
    gr.Markdown("# Social Media Data Extraction Tool")
    
    with gr.Tabs():
        with gr.TabItem("Social Media Extraction"):
            gr.Markdown("## Analyze Social Media Profiles")
            with gr.Row():
                with gr.Column():
                    profile_link = gr.Textbox(label="Profile URL")
                    hashtags = gr.Textbox(label="Hashtags (comma separated)")
                    concepts = gr.Textbox(label="Concept Keywords (comma separated)")
                    platform = gr.Radio(["Instagram", "Twitter", "Facebook"], label="Platform")
                    date_range = gr.Textbox(label="Date Range (YYYY-MM-DD to YYYY-MM-DD)")
                    num_posts = gr.Slider(1, 100, value=10, label="Number of Posts")
                    sm_submit = gr.Button("Analyze Profile", variant="primary")
                
                with gr.Column():
                    sm_output = gr.File(label="Download Analysis Report")

        with gr.TabItem("Document Processing"):
            gr.Markdown("## Analyze Word Documents")
            with gr.Row():
                with gr.Column():
                    word_file = gr.File(label="Upload Word Document")
                    wp_concepts = gr.Textbox(label="Concept Keywords (comma separated)")
                    wp_submit = gr.Button("Analyze Document", variant="primary")
                
                with gr.Column():
                    wp_output = gr.File(label="Download Analysis Report")

    sm_submit.click(
        fn=process_social_media,
        inputs=[profile_link, hashtags, concepts, platform, date_range, num_posts],
        outputs=sm_output
    )
    
    wp_submit.click(
        fn=process_word_file,
        inputs=[word_file, wp_concepts],
        outputs=wp_output
    )

if __name__ == "__main__":
    app.launch()