#!/usr/bin/env python3

import time
import json
import yaml
import requests
import re
from datetime import datetime, timedelta, timezone
from packaging.version import Version, InvalidVersion

BASE_URL = "https://euvdservices.enisa.europa.eu/api/search"
HEADERS = {
    "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) EUVD-Auditor/1.0"
}


def parse_single_version(v_str):
    """
    Limpia y extrae la versión semántica limpia de una cadena.
    Ejemplos: 'v1.8.0' -> '1.8.0', 'openjdk 1.8.0_20' -> '1.8.0'
    """
    if not v_str:
        return None
    cleaned = re.sub(r'^(openjdk|v|version)\s*', '', v_str.strip(), flags=re.IGNORECASE)
    match = re.search(r'\b\d+(\.\d+)+', cleaned)
    if match:
        try:
            return Version(match.group(0))
        except InvalidVersion:
            return None
    return None


def matches_product_version(target_version_str, product_version_str):
    """
    Evalúa matemáticamente si target_version_str se ve afectada por product_version_str.
    Soporta:
      - Expresiones con operadores: "<19.1.1", "<=19.1.1", ">19.0", ">=19.0"
      - Rangos combinados: "19.1 <19.1.1", "19.1 <19.2.3"
      - Rangos con palabra "to": "1.7.0 to 1.7.10"
      - Coincidencia exacta o comodines '*'
    """
    if not product_version_str or product_version_str.lower() in ["n/a", "unknown", "*"]:
        return True, "Versión indeterminada en la fuente (Revisar manualmente)"

    t_ver = parse_single_version(target_version_str)
    if not t_ver:
        if target_version_str.strip().lower() in product_version_str.lower():
            return True, f"Coincidencia directa por cadena '{product_version_str}'"
        return False, f"No coincide con '{product_version_str}'"

    p_str = product_version_str.strip()

    # 1. Manejo de rangos con palabra "to" (ej. "1.7.0 to 1.8.2")
    if " to " in p_str.lower():
        parts = re.split(r'\s+to\s+', p_str, flags=re.IGNORECASE)
        if len(parts) == 2:
            min_v = parse_single_version(parts[0])
            max_v = parse_single_version(parts[1])
            if min_v and max_v:
                if min_v <= t_ver <= max_v:
                    return True, f"Afecta al rango {min_v} <= {t_ver} <= {max_v}"
                else:
                    return False, f"Fuera del rango ({min_v} a {max_v})"

    # 2. Extraer condiciones/operadores de la cadena product_version
    tokens = re.findall(r'(<=|>=|<|>)?\s*(\d+(?:\.\d+)+[a-zA-Z0-9_\.-]*)', p_str)

    if not tokens:
        if t_ver.public in p_str:
            return True, f"Coincidencia textual directa con '{product_version_str}'"
        return False, f"No coincide con '{product_version_str}'"

    all_conditions_met = True
    reasons = []

    for op, ver_str in tokens:
        cond_ver = parse_single_version(ver_str)
        if not cond_ver:
            continue

        if op == "<":
            met = t_ver < cond_ver
            reasons.append(f"{t_ver} < {cond_ver} ({'Sí' if met else 'No'})")
        elif op == "<=":
            met = t_ver <= cond_ver
            reasons.append(f"{t_ver} <= {cond_ver} ({'Sí' if met else 'No'})")
        elif op == ">":
            met = t_ver > cond_ver
            reasons.append(f"{t_ver} > {cond_ver} ({'Sí' if met else 'No'})")
        elif op == ">=":
            met = t_ver >= cond_ver
            reasons.append(f"{t_ver} >= {cond_ver} ({'Sí' if met else 'No'})")
        else:
            met = (t_ver == cond_ver) or (t_ver.public.startswith(cond_ver.public))
            reasons.append(f"Rama/Coincidencia {cond_ver} ({'Sí' if met else 'No'})")

        if not met:
            all_conditions_met = False

    rule_msg = " | ".join(reasons)
    if all_conditions_met:
        return True, f"Cumple condiciones: {rule_msg}"
    else:
        return False, f"No cumple condiciones: {rule_msg}"


def fetch_fast_enisa_vulnerabilities(vendor, product, from_score=7.0, days_back=730):
    vulnerabilities = []
    from_date = (datetime.now(timezone.utc) - timedelta(days=days_back)).strftime("%Y-%m-%d")
    
    params = {
        "vendor": vendor,
        "product": product,
        "fromScore": from_score,
        "fromDate": from_date,
        "page": 0,
        "size": 100
    }

    max_retries = 3
    for attempt in range(max_retries):
        try:
            response = requests.get(BASE_URL, headers=HEADERS, params=params, timeout=15)
            
            if response.status_code == 429:
                wait_time = (attempt + 1) * 5
                time.sleep(wait_time)
                continue

            response.raise_for_status()
            data = response.json()

            items = data.get("items", data.get("content", [])) if isinstance(data, dict) else data
            
            if not items and product:
                params_fallback = {
                    "text": product,
                    "fromScore": from_score,
                    "fromDate": from_date,
                    "page": 0,
                    "size": 100
                }
                res_fb = requests.get(BASE_URL, headers=HEADERS, params=params_fallback, timeout=15)
                if res_fb.status_code == 200:
                    items = res_fb.json().get("items", [])

            vulnerabilities.extend(items)
            break

        except Exception as e:
            if attempt == max_retries - 1:
                print(f"   ❌ Error final para {product}: {e}")
            time.sleep(2)

    return vulnerabilities


def generate_markdown_report(total_consolidados, total_analizadas, total_aplicables, 
                             total_no_aplicables, total_revision, resumen_objetivos, 
                             aplicables_severidad, output_file="informe_vulnerabilidades.md"):
    """
    Genera el informe ejecutivo en Markdown compatible con Pandoc/XeLaTeX.
    """
    now_iso = datetime.now(timezone.utc).strftime("%Y-%m-%dT%H:%M:%S.%f")[:-3] + "Z"

    md_content = f"""# Informe Ejecutivo de Vulnerabilidades

**Generado en:** {now_iso} | **Fuentes:** ENISA

* **Informes consolidados:** {total_consolidados}
* **Vulnerabilidades analizadas:** {total_analizadas}
* **Aplicables:** {total_aplicables}
* **No aplicables:** {total_no_aplicables}
* **Revisión manual:** {total_revision}

---

## Resumen por objetivo

| Fuente | Producto | Versión | Analizadas | Aplicables | No aplicables | Revisión manual |
|---|---|---|---|---|---|---|
"""

    for obj in resumen_objetivos:
        md_content += f"| {obj['Fuente']} | {obj['Producto']} | {obj['Versión']} | {obj['Analizadas']} | {obj['Aplicables']} | {obj['No aplicables']} | {obj['Revisión manual']} |\n"

    md_content += """
---

## Detalle de Vulnerabilidades Aplicables
"""

    if not aplicables_severidad:
        md_content += "\n*No se encontraron vulnerabilidades aplicables con los criterios seleccionados en el período indicado.*\n"
    else:
        aplicables_ordenadas = sorted(
            aplicables_severidad,
            key=lambda x: float(x["BaseScore"].split()[0]) if x.get("BaseScore") and x["BaseScore"].split()[0].replace('.', '', 1).isdigit() else 0.0,
            reverse=True
        )

        for app in aplicables_ordenadas:
            score_val = app['BaseScore'].split()[0] if app.get('BaseScore') else "0.0"
            try:
                score_num = float(score_val)
            except ValueError:
                score_num = 0.0

            if score_num >= 9.0:
                badge = "[**CRÍTICA**]"
            elif score_num >= 7.0:
                badge = "[**ALTA**]"
            elif score_num >= 4.0:
                badge = "[**MEDIA**]"
            else:
                badge = "[**BAJA**]"

            just = app['Justificación'].strip()
            score_vector = app.get('BaseScoreVector', 'N/A')

            md_content += f"""
### {badge} - {app['ID']} - {app['CVE']}

* **Producto / Versión:** `{app['Producto']}` (v`{app['Versión']}`) | **Fuente:** {app['Fuente']}
* **BaseScore:** {app['BaseScore']}
* **Vector CVSS:** `{score_vector}`
* **Regla versión:** {app['Regla versión']}
* **Justificación / Descripción:**
  > {just}

---
"""

    with open(output_file, "w", encoding="utf-8") as f:
        f.write(md_content)

    print(f"📝 Informe Markdown generado exitosamente en '{output_file}'")


def process_targets(config_path="targets.yaml"):
    try:
        with open(config_path, "r", encoding="utf-8") as f:
            config = yaml.safe_load(f)
    except FileNotFoundError:
        print(f"❌ Archivo '{config_path}' no encontrado.")
        return

    resumen_objetivos = []
    aplicables_severidad = []

    total_consolidados = 0
    total_analizadas = 0
    total_aplicables = 0
    total_no_aplicables = 0
    total_revision = 0

    print("🚀 Iniciando análisis de precisión en ENISA EUVD...\n")

    for target in config.get("targets", []):
        vendor = target.get("vendor", "")
        product = target.get("product", "")
        version = target.get("version", "")

        total_consolidados += 1
        print(f"👉 Analizando: {vendor}/{product} (v{version})")

        items = fetch_fast_enisa_vulnerabilities(vendor=vendor, product=product, from_score=0.0, days_back=730)

        num_analizadas = len(items)
        num_aplicables = 0
        num_no_aplicables = 0
        num_revision = 0

        for item in items:
            score = item.get("baseScore", 0.0) or 0.0
            score_ver = item.get("baseScoreVersion", "3.1")
            score_vector = item.get("baseScoreVector", "N/A")
            euvd_id = item.get("id", "N/A")
            
            aliases = item.get("aliases", "")
            aliases_list = aliases.strip().split("\n") if isinstance(aliases, str) else []
            cve_id = aliases_list[0] if aliases_list and aliases_list[0] else euvd_id

            desc = item.get("description", "Sin descripción disponible.")
            desc_short = desc[:180].strip() + "..." if len(desc) > 180 else desc

            enisa_products = item.get("enisaIdProduct", [])
            is_affected = False
            match_rule = "Sin información de versión en la API"

            if enisa_products:
                for prod_entry in enisa_products:
                    p_version = prod_entry.get("product_version", "")
                    affected, rule_msg = matches_product_version(version, p_version)
                    if affected:
                        is_affected = True
                        match_rule = f"`product_version`: \"{p_version}\" -> ({rule_msg})"
                        break
            else:
                is_affected = True
                match_rule = "Evaluación por coincidencia general de producto"

            if is_affected:
                num_aplicables += 1
                aplicables_severidad.append({
                    "Fuente": "ENISA",
                    "ID": euvd_id,
                    "CVE": cve_id,
                    "Producto": product,
                    "Versión": version,
                    "BaseScore": f"{score} {score_ver}".strip(),
                    "BaseScoreVector": score_vector,
                    "Regla versión": match_rule,
                    "Justificación": desc_short
                })
            else:
                num_no_aplicables += 1

        total_analizadas += num_analizadas
        total_aplicables += num_aplicables
        total_no_aplicables += num_no_aplicables
        total_revision += num_revision

        resumen_objetivos.append({
            "Fuente": "ENISA",
            "Producto": product,
            "Versión": version,
            "Analizadas": num_analizadas,
            "Aplicables": num_aplicables,
            "No aplicables": num_no_aplicables,
            "Revisión manual": num_revision
        })

        print(f"   📊 Analizadas: {num_analizadas} | Aplicables: {num_aplicables} | Descartadas por versión: {num_no_aplicables}\n")
        time.sleep(0.5)

    reporte_json = {
        "generado_en": datetime.now(timezone.utc).isoformat(),
        "fuentes": ["ENISA"],
        "resumen_por_objetivo": resumen_objetivos,
        "aplicables_por_severidad": aplicables_severidad
    }
    with open("reporte_vulnerabilidades.json", "w", encoding="utf-8") as f:
        json.dump(reporte_json, f, indent=2, ensure_ascii=False)
        
    generate_markdown_report(
        total_consolidados=total_consolidados,
        total_analizadas=total_analizadas,
        total_aplicables=total_aplicables,
        total_no_aplicables=total_no_aplicables,
        total_revision=total_revision,
        resumen_objetivos=resumen_objetivos,
        aplicables_severidad=aplicables_severidad,
        output_file="informe_vulnerabilidades.md"
    )        

    print("✅ Análisis finalizado y archivo 'informe_vulnerabilidades.md' actualizado con éxito.")

if __name__ == "__main__":
    process_targets()