Creating app.py

app.py

@@ -0,0 +1,153 @@
+import streamlit as st  # Import Streamlit for creating a web app interface
+import pandas as pd  # Import pandas for data manipulation
+from serpapi import GoogleSearch  # Import SerpAPI to perform Google searches
+import re  # Import regex module for pattern matching
+import tempfile  # Import tempfile for creating temporary files
+import shutil  # Import shutil for file operations
+import os  # Import os for handling file paths
+from difflib import SequenceMatcher  # Import SequenceMatcher to calculate string similarity
+
+# Function to construct a Google search query from applicant data
+def construct_query(row):
+    """Constructs the Google search query using applicant data."""
+    query = str(row['Applicant Name'])  # Start with the applicant's name
+    print(f"Constructing query for Applicant Name: {row['Applicant Name']}")
+
+    # Additional fields to include in the search query if available
+    optional_fields = ['Job Title', 'State', 'City', 'Skills']
+    for field in optional_fields:
+        if field in row and pd.notna(row[field]):  # Check if the field exists and is not NaN
+            value = row[field]
+            if isinstance(value, str) and value.strip():  # Ensure the value is a non-empty string
+                query += f" {value.strip()}"  # Add the value to the query
+            elif not isinstance(value, str):  # Handle non-string values
+                query += f" {str(value).strip()}"
+    query += " linkedin"  # Append "linkedin" to focus search on LinkedIn profiles
+    print(f"Constructed query: {query}")
+    return query
+
+# Function to extract the name from a LinkedIn profile URL
+def get_name_from_url(link):
+    """Extracts the name part from a LinkedIn profile URL."""
+    print(f"Extracting name from LinkedIn URL: {link}")
+    match = re.search(r'linkedin\.com/in/([a-zA-Z0-9-]+)', link)  # Regex to find profile name
+    if match:
+        name = match.group(1).replace('-', ' ')  # Replace dashes with spaces for readability
+        print(f"Extracted name: {name}")
+        return name
+    print("No name extracted from URL.")
+    return None
+
+# Function to calculate similarity between two names
+def calculate_similarity(name1, name2):
+    """Calculates similarity between two names."""
+    similarity = SequenceMatcher(None, name1.lower().strip(), name2.lower().strip()).ratio()
+    print(f"Calculated similarity between '{name1}' and '{name2}': {similarity}")
+    return similarity
+
+# Function to fetch LinkedIn links using SerpAPI
+def fetch_linkedin_links(query, api_key, applicant_name):
+    """Fetches LinkedIn profile links and validates them against the applicant's name."""
+    linkedin_regex = r'https://(www|[a-z]{2})\.linkedin\.com/.*'  # Regex for LinkedIn links
+    try:
+        print(f"Fetching LinkedIn links for query: {query}")
+        search = GoogleSearch({
+            "q": query,  # The search query
+            "num": 5,  # Number of search results
+            "api_key": api_key  # API key for SerpAPI
+        })
+
+        # Execute the search and get results
+        results = search.get_dict()
+        organic_results = results.get("organic_results", [])  # Extract organic search results
+        print(f"Raw search results: {organic_results}")
+
+        # Iterate through results to find LinkedIn links
+        for result in organic_results:
+            link = result.get("link")  # Get the URL of the search result
+            print(f"Checking link: {link}")
+            if re.match(linkedin_regex, link):  # Check if the link matches LinkedIn regex
+                profile_name = get_name_from_url(link)  # Extract the name from the URL
+                if profile_name:
+                    similarity = calculate_similarity(applicant_name, profile_name)  # Validate name similarity
+                    if similarity >= 0.5:  # Accept link if similarity is above the threshold
+                        print(f"Valid LinkedIn link found: {link} (Similarity: {similarity})")
+                        return link
+                    else:
+                        print(f"Rejected link: {link} (Similarity: {similarity})")
+            else:
+                print(f"Link does not match LinkedIn regex: {link}")
+        
+        print("No valid LinkedIn link found.")
+        return None
+    except Exception as e:
+        print(f"Error fetching link for query '{query}': {e}")
+        st.error(f"Error fetching link for query '{query}': {e}")
+        return None
+
+# Function to process the uploaded Excel file
+def process_file(file, api_key):
+    """Processes the uploaded Excel file to fetch LinkedIn profile links."""
+    try:
+        print("Reading uploaded Excel file...")
+        df = pd.read_excel(file)  # Read the Excel file into a pandas DataFrame
+        print(f"Initial DataFrame:\n{df.head()}")
+
+        # Filter out rows with empty or missing applicant names
+        df = df[df['Applicant Name'].notna()]
+        df = df[df['Applicant Name'].str.strip() != '']
+        print(f"Filtered DataFrame:\n{df.head()}")
+
+        # Generate search queries for each applicant
+        df['Search Query'] = df.apply(construct_query, axis=1)
+        print(f"DataFrame with Search Queries:\n{df[['Applicant Name', 'Search Query']].head()}")
+
+        # Fetch LinkedIn links for each applicant
+        df['LinkedIn Link'] = df.apply(
+            lambda row: fetch_linkedin_links(row['Search Query'], api_key, row['Applicant Name']),
+            axis=1
+        )
+        print(f"DataFrame with LinkedIn Links:\n{df.head()}")
+
+        # Save the updated DataFrame to a temporary file
+        temp_dir = tempfile.mkdtemp()  # Create a temporary directory
+        output_file = os.path.join(temp_dir, "updated_with_linkedin_links.csv")
+        df.to_csv(output_file, index=False)  # Save as CSV
+        print(f"CSV file created at: {output_file}")
+
+        return output_file
+    except Exception as e:
+        print(f"Error processing file: {e}")
+        st.error(f"Error processing file: {e}")
+        return None
+
+# Streamlit UI setup
+st.title("LinkedIn Profile Link Scraper")  # App title
+st.markdown("Upload an Excel file with applicant details, and get a CSV with LinkedIn profile links.")  # Description
+
+# Input for SerpAPI Key
+api_key = st.text_input("Enter your SerpAPI Key", type="password")  # Input for SerpAPI key
+
+# File uploader widget
+uploaded_file = st.file_uploader("Upload Excel File", type=["xlsx"])  # File uploader for Excel files
+
+# Process the file if both file and API key are provided
+if uploaded_file and api_key:
+    st.write("Processing file...")
+    output_file = process_file(uploaded_file, api_key)  # Process the uploaded file
+
+    if output_file:
+        with open(output_file, "rb") as f:  # Open the CSV for download
+            st.download_button(
+                label="Download Updated CSV",
+                data=f,
+                file_name="updated_with_linkedin_links.csv",
+                mime="text/csv"
+            )
+        print("File ready for download.")
+
+        # Clean up the temporary directory after download
+        shutil.rmtree(os.path.dirname(output_file))
+        print("Temporary files cleaned up.")
+elif not api_key:
+    st.warning("Please enter your SerpAPI key to proceed.")  # Warning for missing API key