import os
import json
from typing import List
from urllib.parse import quote
import logging
import gradio as gr
import pandas as pd
from server import cv_processor, job_processor, applicant_evaluator

logger = logging.getLogger(__name__)
logging.basicConfig(level=logging.INFO)


logo_base64 = cv_processor.encode_base64(
    "static/AIRecruiterAgent.png"
)

def evaluate_applicants(
    cv_files: List[str],
    job_description: str,
    #progress=gr.Progress(visible=True, label="Evaluating Applicants...")
) -> pd.DataFrame:
    """
    Evaluate applicants' CVs against the job description.

    Parameters
    ----------
    cv_files: List[str]
        List of CV file paths to evaluate.
    job_description: str
        The job description text to evaluate against.

    Returns
    -------
        pd.DataFrame: DataFrame containing evaluation results with match scores and reasoning.
    """
    # TODO: Add progress bar support with batch processing
    # TODO: Add error handling for file processing and evaluation
    # if not cv_files:
    #     gr.Error("Please upload applicants CV files in PDF format.")
    # if not job_description:
    #     gr.Error("Please provide the job description text for evaluation.")
    
    # Get job annotation
    logger.info("Getting job annotation from job description.")
    job_annotation = job_processor.get_job_content(job_description)
           
    evaluation_res = []
    for cv_file in cv_files:
        # Get CV annotation
        logger.info("Getting cv annotation from CV file: %s", cv_file.name)
        cv_annotation = cv_processor.get_cv_content(cv_file.name)
        
        # Evaluate the applicant against the job description
        logger.info("Evaluating applicant CV against job description.")
        res = applicant_evaluator.evaluate_applicant(
            cv_annotation["cv"]["annotation"], 
            job_annotation["job"]["annotation"]
        )
        evaluation = json.loads(res["evaluation"])
        cv_base64 = cv_processor.encode_base64(cv_file.name)

        score = float(evaluation["match_score"])
        if score >= 0.8:
            match_labels = "Strong Matched"
        elif score >= 0.5:
            match_labels = "Partially Matched"
        else:
            match_labels = "Not Matched"
        evaluation_res.append(
            {
                "Applicant": os.path.basename(cv_file.name),
                "Match Score": evaluation["match_score"],
                "Match Labels": match_labels,
                "Match Reasoning": evaluation["match_reasoning"],
                "CV Base64": cv_base64,
                "CV Url": f"gradio_api/file={cv_file.name}"
            }
        )
    #logger.info(f"Evaluation results: {response}")
    evaluation_res = sorted(
        evaluation_res, 
        key=lambda d: d['Match Score'], 
        reverse=True
    )
    return pd.DataFrame.from_records(evaluation_res)


def df_select_callback(df: pd.DataFrame, evt: gr.SelectData):
    selected_row = evt.row_value
    if not selected_row:
        return "No row selected.", ""

    match_score = selected_row[1]  # .get('Match Score', 'N/A')
    match_labels = selected_row[2]  # .get('Match Labels', 'N/A')
    match_reasoning = selected_row[3]  # .get('Match Reasoning', 'N/A')
    
    cv_base64 = selected_row[4]  # .get('CV Base64', '')
    pdf_url = selected_row[5]  # .get('CV Url', '')
    pdf_encoded_url = f"https://agents-mcp-hackathon-airecruiteragent.hf.space/{pdf_url}"
    
    if cv_base64:
        pdf_encoded = gr.HTML(
            """
            <!-- The Modal -->
            <div id="myModal" class="modal" style="display: none;">

            <!-- Modal content -->
            <div class="modal-content">
                <div class="modal-header">
                <span class="close" onclick="hide_pdfviewer()">&times;</span>
                <h2>Candidate CV</h2>
                </div>
                <div class="modal-body">
                
                    <div id="my-pdf" class="pdfobject-container" style="height: 100%">
                    <object data="data:application/pdf;base64,{cv_base64}" type="application/pdf" style="height: 100%; width: 100%;">
                        <iframe src="https://docs.google.com/viewer?url={pdf_encoded_url}&embedded=true" style="height: 100%; width: 100%;" frameborder="0" allow="accelerometer; ambient-light-sensor; autoplay; battery; camera; document-domain; encrypted-media; fullscreen; geolocation; gyroscope; layout-animations; legacy-image-formats; magnetometer; microphone; midi; oversized-images; payment; picture-in-picture; publickey-credentials-get; sync-xhr; usb; vr ; wake-lock; xr-spatial-tracking" sandbox="allow-forms allow-modals allow-popups allow-popups-to-escape-sandbox allow-same-origin allow-scripts allow-downloads"></iframe>
                    </object>
                    </div>
                </div>
                <div class="modal-footer">
                
                </div>
            </div>

            </div>
            
            """.format(
                title=selected_row[0],
                cv_base64=cv_base64,
                pdf_url=pdf_url, 
                pdf_encoded_url=pdf_encoded_url
            ),
            label="CV PDF Viewer",
            elem_id="pdf_viewer",
            min_height="0px",
            max_height="100%",
            visible=True,
        )
    else:
        pdf_encoded = gr.HTML(
            "", 
            label="CV PDF Viewer",
            elem_id="pdf_viewer", 
            min_height="0px",
            max_height="100%",
            visible=False,
        )

    return match_reasoning, pdf_encoded


head = """
<script src="https://unpkg.com/pdfobject@2.3.1/pdfobject.min.js"></script>
<script type="text/javascript">
    function hide_pdfviewer(){
    document.getElementById('myModal').style.display="none";
    }

    function show_pdfviewer(){
    document.getElementById('myModal').style.display="block";
    }
</script>
"""

css = """
.logo-container {
  width:100%;
  height:auto;
  padding:1%;
}

.logo-img {
  margin-left:2%;
  float:left; 
  height:40px;
  width:40px;    
} 

/* The Modal (background) */
.modal {
  display: none;
  position: fixed; 
  z-index: 1000; /* Sit on top */
  padding-top: 100px; 
  left: 0;
  top: 0;
  width: 100%; 
  height: 100%; 
  overflow: auto; /* Enable scroll if needed */
  background-color: rgb(0,0,0); /* Fallback color */
  background-color: rgba(0,0,0,0.4); /* Black w/ opacity */
}

/* Modal Content */
.modal-content {
  position: relative;
  background-color: #fefefe;
  margin: auto;
  padding: 0;
  border: 1px solid #888;
  width: 80%;
  height: 90%;
  overflow: auto;
  box-shadow: 0 4px 8px 0 rgba(0,0,0,0.2),0 6px 20px 0 rgba(0,0,0,0.19);
  -webkit-animation-name: animatetop;
  -webkit-animation-duration: 0.4s;
  animation-name: animatetop;
  animation-duration: 0.4s
}

/* Add Animation */
@-webkit-keyframes animatetop {
  from {top:-300px; opacity:0} 
  to {top:0; opacity:1}
}

@keyframes animatetop {
  from {top:-300px; opacity:0}
  to {top:0; opacity:1}
}

/* The Close Button */
.close {
  color: white;
  float: right;
  font-size: 28px;
  font-weight: bold;
}

.close:hover,
.close:focus {
  color: #000;
  text-decoration: none;
  cursor: pointer;
}

.modal-header {
  padding: 2px 16px;
  background-color: #5cb85c;
  color: white;
}

.modal-body {
  padding: 2px 16px;
  height: 90%;
}

.modal-footer {
  padding: 2px 16px;
  background-color: #5cb85c;
  color: white;
}
"""



# Create the Gradio interface
with gr.Blocks(head=head, css=css) as demo_app:
    # Title section
    with gr.Row():
        gr.HTML(
            """
            <div style="text-align:center; margin-bottom: 10px;">
            <div class="logo-container" style="display: inline-block; text-align: center;">
                <img src="data:image/png;base64,{logo_base64}" alt="AI Recruiter Agent" style="height: 100px; width: auto; " class="logo-img"> 
                <h1 style="color: #4A90E2; font-size: 2.5em;">
                    AI Recruiter Agent
                </h1>
                <h3>Revolutionizing recruitment with AI-driven insights.</h3>
            </div>
            
            </div>
        """.format(logo_base64=logo_base64)  # Assuming logo_base64 is defined elsewhere
        )

    with gr.Row():
        # Input section for resumes and job description
        with gr.Column(scale=1):
            gr.Markdown("### Upload Resumes and Job Description")
            gr.Markdown(
                "Upload multiple resumes in PDF format and provide the job description text for evaluation."
            )

            cv_files = gr.Files(
                file_count="multiple",
                file_types=[".pdf"],  # , ".docx", ".txt"
                label="Upload Candidate Resume files",
                height="150px",
            )
            job_description = gr.TextArea(
                placeholder="Enter Job Description text here...",
                label="Job Description",
                lines=12,
                max_lines=12,
            )
            with gr.Row():
                # Buttons for starting over and submitting
                start_over_button = gr.Button("Start Over", variant="secondary")
                submit_button = gr.Button("Match Applicants", variant="primary")

        # Output section for evaluation results
        with gr.Column(scale=1):
            gr.Markdown("### Evaluation Results")
            gr.Markdown(
                "Click on the results to view detailed evaluation of each applicant against the job description."
            )
            # Output area for evaluation results
            result_df = gr.DataFrame(
                headers=[
                    "Applicant",
                    "Match Score",
                    "Match Labels",
                ],
                #label="Evaluation Results",
                elem_id="myResult",
                type="pandas",
                interactive=False,
                show_row_numbers=True,
                column_widths=["50%", "25%", "25%", "0%", "0%", "0%"],
                wrap=False,
            )
            # output = gr.JSON(label="Evaluation Results")
            result_detail_textbox = gr.Textbox(
                label="Detailed Evaluation",
                placeholder="Click on a row to see detailed evaluation.",
                lines=8,
                max_lines=8,
            )
            show_pdf_button = gr.Button(
                "Show Selected Candidate CV", 
                variant="secondary", 
                elem_id="show_pdf_button",
            )

            pdf_viewer = gr.HTML(
                "", 
                label="CV PDF Viewer",
                elem_id="pdf_viewer", 
                min_height="0px",
                max_height="100%",
                visible=False,
            )

    # Button click handlers
    def reset_fields():
        return (
            [], "", [], "",  
            gr.HTML(
                "", 
                label="CV PDF Viewer",
                elem_id="pdf_viewer", 
                min_height="0px",
                max_height="100%",
                visible=False,
            )
            
        )

    # Reset function to clear inputs and outputs
    start_over_button.click(
        fn=reset_fields, 
        inputs=None, 
        outputs=[
            cv_files, 
            job_description, 
            result_df, 
            result_detail_textbox,
            pdf_viewer
        ]
    )
    submit_button.click(
        fn=evaluate_applicants, inputs=[cv_files, job_description], outputs=[result_df]
    )
    result_df.select(
        fn=df_select_callback,
        inputs=[result_df],
        outputs=[result_detail_textbox, pdf_viewer],
    )

    show_pdf_button.click(
        fn=lambda: gr.update(visible=True), 
        inputs=None, 
        outputs=pdf_viewer,
        js="show_pdfviewer()",
    )

# Launch the interface and MCP server
if __name__ == "__main__":
    demo_app.launch(
        mcp_server=True,
        allowed_paths=["static"]
    )