"""
DABstep Benchmark Utilities
Adapted from: https://huggingface.co/spaces/adyen/DABstep/blob/main/dabstep_benchmark/utils.py
"""
from __future__ import annotations

import re

import pandas as pd

from dabstep_benchmark.evaluation.scorer import question_scorer


def format_error(msg: str) -> str:
    """Format an error message in red."""
    return f"<p style='color: red; font-size: 20px; text-align: center;'>{msg}</p>"


def format_warning(msg: str) -> str:
    """Format a warning message in orange."""
    return f"<p style='color: orange; font-size: 20px; text-align: center;'>{msg}</p>"


def format_log(msg: str) -> str:
    """Format a log message in green."""
    return f"<p style='color: green; font-size: 20px; text-align: center;'>{msg}</p>"


def model_hyperlink(link: str, model_name: str) -> str:
    """Create a hyperlink for a model."""
    return f'<a target="_blank" href="{link}" style="color: var(--link-text-color); text-decoration: underline;text-decoration-style: dotted;">{model_name}</a>'


def is_valid_https_url(url: str) -> bool:
    """Validate that a URL is a valid HTTPS URL."""
    pattern = re.compile(
        r'^https://'  # URL must start with 'https://'
        r'(?!10(?:\.\d{1,3}){3})'  # Exclude private IP 10.x.x.x
        r'(?!127(?:\.\d{1,3}){3})'  # Exclude loopback IP 127.x.x.x
        r'(?!169\.254(?:\.\d{1,3}){2})'  # Exclude link-local IP 169.254.x.x
        r'(?!192\.168(?:\.\d{1,3}){2})'  # Exclude private IP 192.168.x.x
        r'(?!172\.(?:1[6-9]|2[0-9]|3[0-1])(?:\.\d{1,3}){2})'  # Exclude private IP 172.16.x.x - 172.31.x.x
        r'(?:(?:[a-zA-Z0-9-]+\.)+[a-zA-Z]{2,})'  # Match domain name
        r'(?::\d{2,5})?'  # Optional port
        r'(?:/[^\s]*)?$',  # Optional path
        re.IGNORECASE
    )
    return re.match(pattern, url) is not None


def evaluate(
    agent_answers: pd.DataFrame,
    tasks_with_gt: pd.DataFrame,
    submission_id: str = ""
) -> list[dict]:
    """
    Evaluate agent answers against ground truth.
    
    Args:
        agent_answers: DataFrame with columns 'task_id' and 'agent_answer'
        tasks_with_gt: DataFrame with columns 'task_id', 'answer', and 'level'
        submission_id: Identifier for the submission
    
    Returns:
        List of score dictionaries for each task
    """
    task_scores = []
    
    for _, row in tasks_with_gt.iterrows():
        correct_answer = row["answer"]
        level = str(row["level"])
        task_id = str(row["task_id"])
        
        if task_id not in agent_answers["task_id"].values:
            raise KeyError(f"Task ID: {task_id} not found. Are you sure you submitted the correct file?")
        
        agent_answer = agent_answers.loc[agent_answers.task_id == task_id, "agent_answer"].values[0]
        score = question_scorer(agent_answer, correct_answer)
        
        task_scores.append({
            "submission_id": submission_id,
            "task_id": task_id,
            "score": score,
            "level": level,
            "agent_answer": agent_answer,
        })
    
    return task_scores