app.py

import spaces
import gradio as gr
import time
import threading
import logging
from gradio.themes.utils import sizes
from main import run_repository_ranking  # Import the repository ranking function

# ---------------------------
# Global Logging Buffer Setup
# ---------------------------
LOG_BUFFER = []
LOG_BUFFER_LOCK = threading.Lock()

class BufferLogHandler(logging.Handler):
    def emit(self, record):
        log_entry = self.format(record)
        with LOG_BUFFER_LOCK:
            LOG_BUFFER.append(log_entry)

root_logger = logging.getLogger()
if not any(isinstance(h, BufferLogHandler) for h in root_logger.handlers):
    handler = BufferLogHandler()
    formatter = logging.Formatter("%(asctime)s - %(levelname)s - %(message)s")
    handler.setFormatter(formatter)
    root_logger.addHandler(handler)

def filter_logs(logs):
    filtered = []
    last_was_fetching = False
    for log in logs:
        if "HTTP Request:" in log:
            if not last_was_fetching:
                filtered.append("Fetching repositories...")
                last_was_fetching = True
        else:
            filtered.append(log)
            last_was_fetching = False
    return filtered

def parse_result_to_html(raw_result: str) -> str:
    """
    Parses the raw string output from run_repository_ranking to an HTML table.
    Only the top 10 results are displayed.
    """
    entries = raw_result.strip().split("Final Rank:")
    entries = entries[1:11]  # Use only the first 10 entries
    if not entries:
        return "<p>No repositories found for your query.</p>"
    html = """
    <table border="1" style="width:80%; margin: auto; border-collapse: collapse;">
        <thead>
            <tr>
                <th>Rank</th>
                <th>Title</th>
                <th>Link</th>
                <th>Combined Score</th>
            </tr>
        </thead>
        <tbody>
    """
    for entry in entries:
        lines = entry.strip().split("\n")
        data = {}
        data["Final Rank"] = lines[0].strip() if lines else ""
        for line in lines[1:]:
            if ": " in line:
                key, val = line.split(": ", 1)
                data[key.strip()] = val.strip()
        html += f"""
            <tr>
                <td>{data.get('Final Rank', '')}</td>
                <td>{data.get('Title', '')}</td>
                <td><a href="{data.get('Link', '#')}" target="_blank">GitHub</a></td>
                <td>{data.get('Combined Score', '')}</td>
            </tr>
        """
    html += "</tbody></table>"
    return html

# ---------------------------
# GPU-enabled Wrapper for Repository Ranking
# ---------------------------
@spaces.GPU(duration=180)
def gpu_run_repo(topic: str):
    return run_repository_ranking(topic)

def run_lite_workflow(topic, result_container):
    result = gpu_run_repo(topic)
    result_container["raw_result"] = result

def stream_lite_workflow(topic):
    with LOG_BUFFER_LOCK:
        LOG_BUFFER.clear()
    result_container = {}
    workflow_thread = threading.Thread(target=run_lite_workflow, args=(topic, result_container))
    workflow_thread.start()
    
    last_index = 0
    while workflow_thread.is_alive() or (last_index < len(LOG_BUFFER)):
        with LOG_BUFFER_LOCK:
            new_logs = LOG_BUFFER[last_index:]
            last_index = len(LOG_BUFFER)
        if new_logs:
            filtered_logs = filter_logs(new_logs)
            status_msg = filtered_logs[-1]
            detail_msg = "<br/>".join(filtered_logs)
            yield status_msg, detail_msg
        time.sleep(0.5)
    
    workflow_thread.join()
    with LOG_BUFFER_LOCK:
        final_logs = LOG_BUFFER[:]
    raw_result = result_container.get("raw_result", "No results returned.")
    html_result = parse_result_to_html(raw_result)
    yield "", html_result

def lite_runner(topic):
    # Provide immediate feedback upon button press.
    yield "Workflow started", "<p>Processing your request. Please wait...</p>"
    for status, details in stream_lite_workflow(topic):
        yield status, details

# ---------------------------
# App UI Setup Using Gradio Soft Theme with Centered Layout
# ---------------------------
with gr.Blocks(
    theme=gr.themes.Soft(text_size=sizes.text_md), 
    title="DeepGit Lite",
    css="""
        /* Center header and footer */
        #header { text-align: center; margin-bottom: 20px; }
        #main-container { max-width: 800px; margin: auto; }
        #footer { text-align: center; margin-top: 20px; }
    """
) as demo:
    gr.Markdown(
        """
        <div style="padding-top: 60px;">
            <div style="display: flex; align-items: center; justify-content: center;">
                <img src="https://img.icons8.com/?size=100&id=118557&format=png&color=000000" 
                     style="width: 60px; height: 60px; margin-right: 12px;">
                <h1 style="margin: 0; font-size: 2.5em; font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;">
                    DeepGit Lite
                </h1>
            </div>
            <div style="text-align: center; margin-top: 20px; font-size: 1.1em; font-family: 'Segoe UI', Tahoma, Geneva, Verdana, sans-serif;">
                <p>
                    ✨ DeepGit Lite is the lightweight pro version of <strong>DeepGit</strong>.<br>
                    It harnesses advanced deep semantic search to explore GitHub repositories and deliver curated results.<br>
                    Under the hood, it leverages a hybrid ranking approach combining dense retrieval, BM25 scoring, and cross-encoder re-ranking for optimal discovery.<br>
                    If the agent returns no repositories found, it means no chain was invoked due to GPU unavailability. Please duplicate the space and re-run. 
                </p>
                <p>
                    🚀 Check out the full DeepGit version on 
                    <a href="https://github.com/zamalali/DeepGit" target="_blank">GitHub</a> and ⭐ 
                    <strong>Star DeepGit</strong> on GitHub!
                </p>
            </div>
        </div>
        """,
        elem_id="header"
    )

    with gr.Column(elem_id="main-container"):
        research_input = gr.Textbox(
            label="Research Query",
            placeholder="Enter your research topic here, e.g., 'data augmentation pipelines for LLM fine-tuning'",
            lines=3
        )
        run_button = gr.Button("Run DeepGit Lite", variant="primary")
        status_display = gr.Markdown(label="Status")
        detail_display = gr.HTML(label="Results")
    
    run_button.click(
        fn=lite_runner,
        inputs=[research_input],
        outputs=[status_display, detail_display],
        api_name="deepgit_lite",
        show_progress=True
    )
    
    research_input.submit(
        fn=lite_runner,
        inputs=[research_input],
        outputs=[status_display, detail_display],
        api_name="deepgit_lite_submit",
        show_progress=True
    )
    
    gr.HTML(
        """
        <div id="footer">
            Made with ❤️ by <b>Zamal</b>
        </div>
        """
    )

demo.queue(max_size=10).launch()