import math
import pandas as pd
import numpy as np
import numpy.linalg as la
from config import Color, logger, hr

def check_and_fix_bounds(tickers, asset_min, asset_max, sector_limit, sector_map, macro=None, gross_leverage_cap=1.0, silent=False):
    """
    Validates and automatically resolves conflicting portfolio constraints.
    Integrates with HMM regime detection to dynamically disable leverage in high-volatility crashes.
    """
    n = len(tickers)
    safe_min = asset_min
    adj_gross_cap = gross_leverage_cap
    warns = []
    
    # ─────────────────────────────────────────────
    # REGIME DETECTION: Dynamic Risk Management
    # ─────────────────────────────────────────────
    is_high_vol = False
    if macro and "hmm_regime" in macro:
        is_high_vol = macro["hmm_regime"].get("is_high_vol", False)

    if is_high_vol and (safe_min < 0 or adj_gross_cap > 1.0):
        safe_min = max(0.0, safe_min) # Disable shorts
        adj_gross_cap = 1.0           # Disable leverage
        warns.append(f" {Color.YELLOW}⚠ HMM High Volatility Regime: Short selling and leverage dynamically disabled to protect capital.{Color.RESET}")
    
    # ─────────────────────────────────────────────
    # STANDARD BOUNDS LOGIC
    # ─────────────────────────────────────────────
    if safe_min < 0:
        if n * asset_max < 1.0:
            asset_max = math.ceil(100.0 / n) / 100.0
            warns.append(f" {Color.YELLOW}⚠ Maximum weight per asset raised to {asset_max:.0%} so the weights can add up to 100%.{Color.RESET}")
        max_gl = n * max(asset_max, abs(safe_min))
        if max_gl > 3.0:
            warns.append(f" {Color.YELLOW}⚠ Your settings allow up to {max_gl:.1f}× gross leverage — make sure this is intentional.{Color.RESET}")
    else:
        if safe_min * n > 0.999:
            safe_min = 0.999 / n
            warns.append(f" {Color.YELLOW}⚠ Minimum weight reduced to {safe_min:.2%} so all weights can still sum to 100%.{Color.RESET}")
        
        sector_groups = {}
        for t in tickers:
            sector = sector_map.get(t, "Other")
            if sector not in sector_groups:
                sector_groups[sector] = []
            sector_groups[sector].append(t)
        
        for sector, members in sector_groups.items():
            if sector == "Other":
                continue
            if safe_min * len(members) > sector_limit:
                reduced = (sector_limit / len(members)) * 0.95 
                safe_min = min(safe_min, reduced)
                warns.append(f" {Color.YELLOW}⚠ '{sector}' sector: minimum weight × {len(members)} tickers would exceed the {sector_limit:.0%} sector cap. Minimum reduced to {reduced:.2%}.{Color.RESET}")

    if safe_min >= asset_max:
        safe_min = 0.0
        warns.append(f" {Color.YELLOW}⚠ Minimum weight was ≥ maximum weight — reset minimum to 0 to avoid an impossible constraint.{Color.RESET}")

    present_sectors = set(sector_map.get(t, "Other") for t in tickers)
    num_sectors = max(1, len(present_sectors))
    if num_sectors * sector_limit < 0.99:
        old_lim = sector_limit
        sector_limit = min(1.0, (1.0 / num_sectors) + 0.02)  
        warns.append(f" {Color.YELLOW}⚠ Sector limit mathematically forced from {old_lim:.0%} to {sector_limit:.0%} to ensure 100% portfolio coverage.{Color.RESET}")

    if warns and not silent:
        hr("─", 65, Color.YELLOW)
        print(f"{Color.BOLD}{Color.YELLOW} Constraint Adjustments{Color.RESET}")
        for w in warns:
            print(w)
        hr("─", 65, Color.YELLOW)

    return safe_min, asset_max, adj_gross_cap, sector_limit


def make_nearest_psd(matrix):
    """Finds the nearest positive-definite matrix to input."""
    sym_matrix = (matrix + matrix.T) / 2
    eigval, eigvec = np.linalg.eigh(sym_matrix)
    eigval[eigval < 0] = 1e-8
    return eigvec @ np.diag(eigval) @ eigvec.T