Update app.py

app.py

@@ -1,180 +1,186 @@
-import requests
-from bs4 import BeautifulSoup
-import pandas as pd
-from sentence_transformers import SentenceTransformer
-from langchain.vectorstores import FAISS
-from langchain.embeddings import HuggingFaceEmbeddings
-import gradio as gr
-
-# Step 1: Scrape data from Analytics Vidhya's free courses page
-def scrape_courses():
-    page_url = "https://courses.analyticsvidhya.com/pages/all-free-courses"
-    response = requests.get(page_url)
-    response.raise_for_status()
-
-    html_parser = BeautifulSoup(response.content, 'html.parser')
-    course_data = []
-
-    # Extract course details
-    for section in html_parser.find_all('div', class_='course-cards__container'):
-        section_heading = section.find('h3', class_='section__heading')
-        category_name = section_heading.get_text(strip=True) if section_heading else "Miscellaneous"
-
-        for card in section.find_all('a', class_='course-card'):
-            course_title = card.find('h3')
-            course_image = card.find('img', class_='course-card__img')
-
-            if course_title and course_image:
-                title = course_title.get_text(strip=True)
-                image_url = course_image['src']
-                link = card['href']
-
-                # Ensure full URL
-                if not link.startswith('http'):
-                    link = 'https://courses.analyticsvidhya.com' + link
-
-                course_data.append({
-                    'course_name': title,
-                    'course_category': category_name,
-                    'thumbnail': image_url,
-                    'course_url': link
-                })
-    return pd.DataFrame(course_data)
-
-# Step 2: Generate embeddings and create FAISS index using HuggingFace embeddings
-def create_faiss_index(course_texts):
-    try:
-        # Load HuggingFace Embedding model
-        embeddings = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
-
-        # Add course metadata (index) for each course
-        metadata = [{"index": idx} for idx in range(len(course_texts))]
-
-        # Create FAISS vector store with metadata
-        vector_store = FAISS.from_texts(course_texts, embeddings, metadatas=metadata)
-        print("FAISS index created successfully!")
-        return vector_store
-    except Exception as e:
-        print(f"Error during FAISS index creation: {str(e)}")
-        return None
-
-# Step 3: Define search function
-def search_courses(user_query):
-    if vector_store is None:
-        return '<p class="error">Error: FAISS index not initialized. Please check the embedding setup.</p>'
-    try:
-        # Perform similarity search
-        results = vector_store.similarity_search(user_query, k=10)
-        if results:
-            output_html = '<div class="search-results">'
-            for result in results:
-                course_idx = result.metadata['index']  # Access the index from metadata
-                course = course_df.iloc[course_idx]
-                output_html += f'''
-                <div class="course-box">
-                    <img src="{course['thumbnail']}" alt="{course['course_name']}" class="course-thumbnail"/>
-                    <div class="course-details">
-                        <h3>{course['course_name']}</h3>
-                        <p><strong>Category:</strong> {course['course_category']}</p>
-                        <a href="{course['course_url']}" target="_blank" class="view-course">Explore Course</a>
-                    </div>
-                </div>'''
-            output_html += '</div>'
-            return output_html
-        else:
-            return '<p class="no-matches">No matching courses found. Please refine your query.</p>'
-    except Exception as e:
-        return f'<p class="error">Error during search: {str(e)}</p>'
-
-# Step 4: Initialize scraping, FAISS index, and Gradio interface
-course_df = scrape_courses()
-course_texts = course_df['course_name'].tolist()  # Texts for embeddings
-vector_store = create_faiss_index(course_texts)
-
-# Custom CSS for Gradio interface
-custom_style = """
-body {
-    font-family: 'Roboto', sans-serif;
-    background-color: #f7f9fc;
-    margin: 0;
-    padding: 0;
-}
-.search-results {
-    display: flex;
-    flex-wrap: wrap;
-    gap: 15px;
-    justify-content: center;
-}
-.course-box {
-    background-color: #fff;
-    border: 1px solid #e3e3e3;
-    border-radius: 8px;
-    overflow: hidden;
-    width: 300px;
-    box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
-    transition: transform 0.3s;
-}
-.course-box:hover {
-    transform: translateY(-5px);
-}
-.course-thumbnail {
-    width: 100%;
-    height: 150px;
-    object-fit: cover;
-}
-.course-details {
-    padding: 15px;
-}
-.course-details h3 {
-    margin: 0 0 10px;
-    font-size: 18px;
-    color: #333;
-}
-.course-details p {
-    margin: 0 0 15px;
-    font-size: 14px;
-    color: #555;
-}
-.view-course {
-    display: inline-block;
-    padding: 10px 20px;
-    background-color: #007bff;
-    color: #fff;
-    text-decoration: none;
-    border-radius: 5px;
-    font-size: 14px;
-}
-.view-course:hover {
-    background-color: #0056b3;
-}
-.no-matches {
-    text-align: center;
-    font-size: 16px;
-    color: #666;
-    margin: 20px 0;
-}
-.error {
-    color: red;
-    font-size: 16px;
-    text-align: center;
-    margin: 20px;
-}
-"""
-
-# Gradio Interface
-tool_interface = gr.Interface(
-    fn=search_courses,
-    inputs=gr.Textbox(label="Search for Free Courses", placeholder="Type keywords like 'Data Science' or 'Python'"),
-    outputs=gr.HTML(label="Search Results"),
-    title="Find Free Courses",
-    description="Quickly find free courses avalable on Analytics Vidhya using this tool.",
-    css=custom_style,
-    examples=[ 
-        ["Generative AI"],
-        ["Business Analytics"],
-        ["Python Programming"]
-    ]
-)
-
-if __name__ == "__main__":
-    tool_interface.launch()
+import requests
+from bs4 import BeautifulSoup
+import pandas as pd
+from langchain.vectorstores import FAISS
+from langchain.embeddings import HuggingFaceEmbeddings
+import gradio as gr
+
+# Step 1: Scrape data from Analytics Vidhya's free courses page
+def scrape_courses():
+    page_url = "https://courses.analyticsvidhya.com/pages/all-free-courses"
+    response = requests.get(page_url)
+    response.raise_for_status()
+
+    html_parser = BeautifulSoup(response.content, 'html.parser')
+    course_data = []
+
+    # Extract course details
+    for section in html_parser.find_all('div', class_='course-cards__container'):
+        section_heading = section.find('h3', class_='section__heading')
+        category_name = section_heading.get_text(strip=True) if section_heading else "Miscellaneous"
+
+        for card in section.find_all('a', class_='course-card'):
+            course_title = card.find('h3')
+            course_image = card.find('img', class_='course-card__img')
+
+            if course_title and course_image:
+                title = course_title.get_text(strip=True)
+                image_url = course_image['src']
+                link = card['href']
+
+                # Ensure full URL
+                if not link.startswith('http'):
+                    link = 'https://courses.analyticsvidhya.com' + link
+
+                course_data.append({
+                    'course_name': title,
+                    'course_category': category_name,
+                    'thumbnail': image_url,
+                    'course_url': link
+                })
+    return pd.DataFrame(course_data)
+
+# Step 2: Generate embeddings and create FAISS index using HuggingFace embeddings
+def create_faiss_index(course_texts):
+    try:
+        # Load HuggingFace Embedding model
+        embeddings = HuggingFaceEmbeddings(model_name="all-MiniLM-L6-v2")
+
+        # Add course metadata (index) for each course
+        metadata = [{"index": idx} for idx in range(len(course_texts))]
+
+        # Create FAISS vector store with metadata
+        vector_store = FAISS.from_texts(course_texts, embeddings, metadatas=metadata)
+        print("FAISS index created successfully!")
+        return vector_store
+    except Exception as e:
+        print(f"Error during FAISS index creation: {str(e)}")
+        return None
+
+# Step 3: Define search function
+def search_courses(user_query):
+    if vector_store is None:
+        return '<p class="error">Error: FAISS index not initialized. Please check the embedding setup.</p>'
+    try:
+        # Perform similarity search
+        results = vector_store.similarity_search(user_query, k=10)
+        if results:
+            output_html = '<div class="search-results">'
+            for result in results:
+                course_idx = result.metadata['index']  # Access the index from metadata
+                course = course_df.iloc[course_idx]
+                output_html += f'''
+                <div class="course-box">
+                    <img src="{course['thumbnail']}" alt="{course['course_name']}" class="course-thumbnail"/>
+                    <div class="course-details">
+                        <h3>{course['course_name']}</h3>
+                        <p><strong>Category:</strong> {course['course_category']}</p>
+                        <a href="{course['course_url']}" target="_blank" class="view-course">Explore Course</a>
+                    </div>
+                </div>'''
+            output_html += '</div>'
+            return output_html
+        else:
+            return '<p class="no-matches">No matching courses found. Please refine your query.</p>'
+    except Exception as e:
+        return f'<p class="error">Error during search: {str(e)}</p>'
+
+# Step 4: Create a simpler FAISS search for Gradio interface
+def search(query):
+    # Perform a similarity search in the FAISS index
+    results = vector_store.similarity_search(query, k=3)  # You can adjust the 'k' value as needed
+    # Return the results in a readable format
+    return "\n".join([f"Result {i+1}: {result['text']}" for i, result in enumerate(results)])
+
+# Step 5: Initialize scraping, FAISS index, and Gradio interface
+course_df = scrape_courses()
+course_texts = course_df['course_name'].tolist()  # Texts for embeddings
+vector_store = create_faiss_index(course_texts)
+
+# Custom CSS for Gradio interface
+custom_style = """
+body {
+    font-family: 'Roboto', sans-serif;
+    background-color: #f7f9fc;
+    margin: 0;
+    padding: 0;
+}
+.search-results {
+    display: flex;
+    flex-wrap: wrap;
+    gap: 15px;
+    justify-content: center;
+}
+.course-box {
+    background-color: #fff;
+    border: 1px solid #e3e3e3;
+    border-radius: 8px;
+    overflow: hidden;
+    width: 300px;
+    box-shadow: 0 4px 6px rgba(0, 0, 0, 0.1);
+    transition: transform 0.3s;
+}
+.course-box:hover {
+    transform: translateY(-5px);
+}
+.course-thumbnail {
+    width: 100%;
+    height: 150px;
+    object-fit: cover;
+}
+.course-details {
+    padding: 15px;
+}
+.course-details h3 {
+    margin: 0 0 10px;
+    font-size: 18px;
+    color: #333;
+}
+.course-details p {
+    margin: 0 0 15px;
+    font-size: 14px;
+    color: #555;
+}
+.view-course {
+    display: inline-block;
+    padding: 10px 20px;
+    background-color: #007bff;
+    color: #fff;
+    text-decoration: none;
+    border-radius: 5px;
+    font-size: 14px;
+}
+.view-course:hover {
+    background-color: #0056b3;
+}
+.no-matches {
+    text-align: center;
+    font-size: 16px;
+    color: #666;
+    margin: 20px 0;
+}
+.error {
+    color: red;
+    font-size: 16px;
+    text-align: center;
+    margin: 20px;
+}
+"""
+
+# Gradio Interface
+tool_interface = gr.Interface(
+    fn=search_courses,
+    inputs=gr.Textbox(label="Search for Free Courses", placeholder="Type keywords like 'Data Science' or 'Python'"),
+    outputs=gr.HTML(label="Search Results"),
+    title="Find Free Courses",
+    description="Quickly find free courses available on Analytics Vidhya using this tool.",
+    css=custom_style,
+    examples=[ 
+        ["Generative AI"],
+        ["Business Analytics"],
+        ["Python Programming"]
+    ]
+)
+
+# Launch Gradio interface
+tool_interface.launch()