import pandas as pd
import gradio as gr
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import cosine_similarity
from openpyxl import load_workbook
from IPython.display import Markdown
import os
# Initialize empty DataFrame and TF-IDF matrix
data = pd.DataFrame()
tfidf_vectorizer = None
tfidf_matrix = None
# Helper: Load Data from File (Excel or CSV)
def load_file(file):
global data, tfidf_vectorizer, tfidf_matrix
try:
# Read file based on its extension
if file.endswith(".xlsx"):
# Read Excel file
workbook = load_workbook(filename=file, data_only=True)
sheet = workbook.active
headers = [cell.value if cell.value is not None else f"Unnamed Column {i}" for i, cell in enumerate(sheet[1])]
rows = sheet.iter_rows(min_row=2, values_only=True)
data = pd.DataFrame(rows, columns=headers)
# Extract hyperlinks from "Information" column if present
if "Information" in data.columns:
for i, row in enumerate(sheet.iter_rows(min_row=2)):
cell = row[data.columns.get_loc("Information")]
if cell.hyperlink:
data.at[i, "Information"] = cell.hyperlink.target
elif file.endswith(".csv"):
# Read CSV file
data = pd.read_csv(file)
else:
return "Unsupported file format. Please upload a .xlsx or .csv file."
# Initialize TF-IDF for employee search
tfidf_vectorizer = TfidfVectorizer(analyzer="char_wb", ngram_range=(2, 4))
tfidf_matrix = tfidf_vectorizer.fit_transform(data["Employee Name"].astype(str))
return "File loaded successfully!"
except Exception as e:
return f"Error loading file: {e}"
# Helper: Generate Report Dynamically
import os
def generate_dynamic_report(query):
if data.empty or tfidf_vectorizer is None:
return None
query_vec = tfidf_vectorizer.transform([query])
similarities = cosine_similarity(query_vec, tfidf_matrix).flatten()
best_match_idx = similarities.argmax()
if similarities[best_match_idx] > 0:
employee = data.iloc[best_match_idx]
employee_name = employee["Employee Name"]
report_file = f"{employee_name.replace(' ', '_')}_report.html"
# Generate HTML content dynamically
with open(report_file, "w") as file:
file.write(f"{employee_name} Report")
file.write(f"Employee Report for {employee_name}
")
file.write("
")
# Employee Details
file.write("Employee Details:
")
for column, value in employee.items():
if column == "Information" and isinstance(value, str):
if value.startswith("http") or value.startswith("file://"):
# Detect file type from hyperlink
file_type = os.path.splitext(value)[-1].lower()
if file_type == ".pdf":
label = "View PDF Document"
elif file_type in [".ppt", ".pptx"]:
label = "View PowerPoint Presentation"
elif file_type in [".doc", ".docx"]:
label = "View Word Document"
else:
label = "Download/View File"
value = f'{label}'
else:
# Add a clickable link for local paths
value = f'Download/View File'
file.write(f"- {column}: {value}
")
file.write("
")
# Customer Insights
file.write("Customer Insights:
")
if "Experience" in data.columns:
experience = employee.get("Experience", "No experience details available.")
file.write(f"- Experience in the field: {experience}
")
if "Skills" in data.columns:
skills = employee.get("Skills", "No skills information available.")
file.write(f"- Key skills: {skills}
")
if "Projects" in data.columns:
projects = employee.get("Projects", "No projects listed.")
file.write(f"- Notable projects: {projects}
")
file.write("
")
# Summary
file.write("Summary:
")
file.write(f"{employee_name} has shown notable contributions in the domain. Refer to the linked documents for more details.
")
file.write("Thank you for using our Employee Dashboard!
")
file.write("")
return report_file
else:
return None
# Helper: Search Employee
def search_employee(query):
if data.empty or tfidf_vectorizer is None:
return pd.DataFrame([{"Error": "No data available. Please upload a file first."}])
query_vec = tfidf_vectorizer.transform([query])
similarities = cosine_similarity(query_vec, tfidf_matrix).flatten()
best_match_idx = similarities.argmax()
if similarities[best_match_idx] > 0:
# Ensure output is a valid DataFrame with one row
employee = data.iloc[best_match_idx].to_frame().T # Convert Series to DataFrame
return employee
else:
return pd.DataFrame([{"Error": "No matching employee found."}])
# Gradio Interface
with gr.Blocks() as interface:
gr.Markdown("""
Employee Dashboard
Upload your Excel or CSV file to get started. Search employees, view metrics, and generate dynamic reports.
""")
with gr.Row():
file_upload = gr.File(label="Upload Excel or CSV File", type="filepath")
upload_status = gr.Textbox(label="Upload Status", interactive=False)
upload_button = gr.Button("Upload")
upload_button.click(load_file, inputs=[file_upload], outputs=[upload_status])
with gr.Row():
search_query = gr.Textbox(label="Search Employee", placeholder="Type partial name (e.g., 'Aar')")
employee_details = gr.Dataframe(label="Employee Details", interactive=True)
report_output = gr.File(label="Download Report")
search_button = gr.Button("Search")
search_button.click(
search_employee,
inputs=[search_query],
outputs=[employee_details],
)
search_button.click(
generate_dynamic_report,
inputs=[search_query],
outputs=[report_output],
)
if __name__ == "__main__":
interface.launch(share=True)