Spaces:
Sleeping
Sleeping
File size: 16,920 Bytes
f228014 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 |
# app_fixed.py
# Phishing detector - improved version with heuristics, SPF/DKIM checks and optional OpenAI integration.
# Requires: gradio, tldextract, dnspython, dkimpy, requests
import os
import re
import json
import traceback
import requests
import tldextract
import gradio as gr
import dns.resolver
import dkim
from email import policy
from email.parser import BytesParser
from typing import List, Dict, Any
# ----------------------------
# Config
# ----------------------------
SHORTENER_DOMAINS = {
"bit.ly", "t.co", "tinyurl.com", "goo.gl", "ow.ly", "is.gd", "buff.ly",
"shorturl.at", "rb.gy", "tiny.one", "clk.im"
}
# Model fallback list: try in order
OPENAI_MODEL_FALLBACK = [
"gpt-4o-mini",
"gpt-4o",
"gpt-5-mini",
]
OPENAI_API_URL = "https://api.openai.com/v1/responses"
# ----------------------------
# Utilities
# ----------------------------
def extract_links(text: str) -> List[str]:
url_re = re.compile(r"(?i)\bhttps?://[^\s<>'\)\]]+")
found = set()
for m in url_re.finditer(text or ""):
url = m.group(0).rstrip('.,:;\")')
found.add(url)
return sorted(found)
def url_hostname(url: str) -> str:
try:
parsed = tldextract.extract(url)
if parsed.domain:
if parsed.suffix:
return f"{parsed.domain}.{parsed.suffix}"
return parsed.domain
m = re.match(r"https?://([^/]+)", url)
return m.group(1).lower() if m else url
except Exception:
return url
def is_shortener(url: str) -> bool:
host = url_hostname(url)
return any(host.endswith(s) for s in SHORTENER_DOMAINS)
def contains_ip(url: str) -> bool:
return bool(re.search(r"https?://(\d{1,3}(?:\.\d{1,3}){3})", url))
def contains_urgent_language(text: str) -> bool:
urgent_re = re.compile(r"\b(urgente|inmediatamente|verifique|actualice|pago|riesgo|suspendido|caduca|vencimiento|bloqueado|atenci[oó]n|urgencia)\b", re.I)
return bool(urgent_re.search(text or ""))
# ----------------------------
# Email parsing & checks
# ----------------------------
def parse_email_raw(raw_text: str) -> Dict[str, Any]:
"""Try to parse headers and body from a raw email text. Returns dict."""
out = {"from": None, "reply_to": None, "subject": None, "body": raw_text, "raw_bytes": None}
try:
# Ensure bytes for the BytesParser
if isinstance(raw_text, str):
raw_bytes = raw_text.encode('utf-8', errors='ignore')
else:
raw_bytes = raw_text
out['raw_bytes'] = raw_bytes
parser = BytesParser(policy=policy.default)
try:
msg = parser.parsebytes(raw_bytes)
except Exception:
msg = None
if msg:
out['from'] = str(msg.get('From') or "").strip()
out['reply_to'] = str(msg.get('Reply-To') or "").strip()
out['subject'] = str(msg.get('Subject') or "").strip()
# get body (prefer plain)
if msg.is_multipart():
parts = []
for part in msg.walk():
ctype = part.get_content_type()
disp = str(part.get_content_disposition() or "")
if ctype == 'text/plain' and disp != 'attachment':
try:
parts.append(part.get_content())
except Exception:
parts.append(part.get_payload(decode=True).decode('utf-8', errors='ignore'))
out['body'] = "\n\n".join(p for p in parts if p)
if not out['body']:
# fallback to first text part
for part in msg.walk():
if part.get_content_type().startswith('text/'):
try:
out['body'] = part.get_content()
break
except:
pass
else:
try:
out['body'] = msg.get_content()
except:
out['body'] = msg.get_payload(decode=True).decode('utf-8', errors='ignore') if msg.get_payload(decode=True) else raw_text
except Exception as e:
print("PARSE RAW ERROR:", repr(e))
traceback.print_exc()
return out
def spf_check(ip: str, domain: str) -> Dict[str, Any]:
"""Simple SPF presence check: queries TXT records for the domain and returns if spf record found."""
try:
answers = dns.resolver.resolve(domain, 'TXT', lifetime=5)
txts = [b"".join(r.strings).decode('utf-8', errors='ignore') for r in answers]
spf = [t for t in txts if t.lower().startswith('v=spf1')]
return {"ok": bool(spf), "records": txts}
except Exception as e:
return {"ok": False, "error": str(e)}
def dkim_check(raw_bytes: bytes) -> Dict[str, Any]:
"""Attempt DKIM verification using dkimpy; returns result dict."""
try:
# dkim.verify expects full message bytes
res = dkim.verify(raw_bytes)
return {"ok": bool(res)}
except Exception as e:
return {"ok": False, "error": str(e)}
# ----------------------------
# Heuristics
# ----------------------------
def analyze_heuristics(raw_text: str, from_header: str = "") -> Dict[str, Any]:
links = extract_links(raw_text)
reasons = []
score = 0
# domain mismatch
from_dom = ""
if from_header:
m = re.search(r"@([\w\.-]+)", from_header)
from_dom = m.group(1).lower() if m else ""
for u in links:
host = url_hostname(u)
if from_dom and host and from_dom not in host:
reasons.append("Dominio de enlaces distinto al dominio del remitente")
score += 20
break
if any(contains_ip(u) for u in links):
reasons.append("Enlaces con IP en vez de dominio")
score += 20
if any(is_shortener(u) for u in links):
reasons.append("Enlace acortado sospechoso")
score += 15
if contains_urgent_language(raw_text):
reasons.append("Lenguaje de urgencia / presión")
score += 15
if re.search(r'\.(exe|scr|bat|cmd|msi|zip)\b', raw_text, re.I):
reasons.append("Adjunto ejecutable o extensión peligrosa detectada")
score += 15
# reply-to different
m_reply = re.search(r"Reply-To:\s*(.+)", raw_text, re.I)
m_from = re.search(r"From:\s*(.+)", raw_text, re.I)
if m_reply and m_from:
reply = m_reply.group(1).strip()
frm = m_from.group(1).strip()
if reply and frm and (reply.lower() not in frm.lower()):
reasons.append("Reply-To diferente al From")
score += 10
# normalize
score = max(0, min(100, score))
return {"score": score, "reasons": reasons, "links": links, "from_domain": from_dom}
# ----------------------------
# OpenAI helper with fallbacks & robust error messages
# ----------------------------
def call_openai(prompt_text: str, api_key: str, models=None, timeout=20):
if models is None:
models = OPENAI_MODEL_FALLBACK
headers = {"Authorization": f"Bearer {api_key}", "Content-Type": "application/json"}
for model in models:
payload = {"model": model, "input": prompt_text}
try:
resp = requests.post(OPENAI_API_URL, headers=headers, json=payload, timeout=timeout)
except Exception as e:
print("AI CALL ERROR (connection):", repr(e))
traceback.print_exc()
return False, f"Error de conexión: {e}"
if resp.status_code == 200:
try:
j = resp.json()
# extract output text from Responses API
out = ""
if "output" in j:
if isinstance(j["output"], list):
parts = []
for item in j["output"]:
if isinstance(item, dict):
c = item.get("content") or item.get("text") or item.get("output_text")
if isinstance(c, str):
parts.append(c)
elif isinstance(c, list):
for el in c:
if isinstance(el, dict):
txt = el.get("text") or el.get("output_text") or el.get("content")
if txt:
parts.append(str(txt))
else:
parts.append(str(el))
out = "\n\n".join(parts).strip()
elif isinstance(j["output"], str):
out = j["output"].strip()
if not out and "choices" in j and isinstance(j.get("choices"), list) and j["choices"]:
ch = j["choices"][0]
out = ch.get("text") or ch.get("message", {}).get("content", {}).get("text") or ""
if not out:
out = json.dumps(j, ensure_ascii=False)[:4000]
return True, out
except Exception as e:
print("AI CALL ERROR (parse):", repr(e))
traceback.print_exc()
return False, f"Error al parsear respuesta de OpenAI: {e}"
else:
try:
err_json = resp.json()
except Exception:
err_json = {"status_code": resp.status_code, "text": resp.text}
print(f"AI CALL HTTP ERROR model={model}: status={resp.status_code} body={str(err_json)[:1000]}")
if resp.status_code == 401:
return False, "AuthenticationError (401): clave inválida o revocada. Revoca y crea una nueva en platform.openai.com"
if resp.status_code == 429:
return False, "RateLimitError (429): cuota superada o límite de velocidad en OpenAI."
# model not found? try next
msg = ""
if isinstance(err_json, dict):
msg = err_json.get("error", {}).get("message") or err_json.get("message") or str(err_json)
if msg and "model" in msg.lower():
# try next model
continue
return False, f"Error HTTP {resp.status_code} al llamar a OpenAI: {msg or resp.text}"
return False, "Ningún modelo disponible o permitido en la cuenta de OpenAI."
# ----------------------------
# Main analyze function
# ----------------------------
def analyze_email(raw_text: str, use_ai: bool = False, do_spf: bool = False, do_dkim: bool = False) -> Dict[str, Any]:
result = {"heuristic": None, "spf": None, "dkim": None, "ai": None}
try:
parsed = parse_email_raw(raw_text or "")
heur = analyze_heuristics(parsed.get('body', raw_text), parsed.get('from') or parsed.get('reply_to') or "")
result['heuristic'] = heur
# technical checks
# SPF: try to extract an IP from Received headers (simple heuristic)
if do_spf:
# find first Received header IP
m = re.search(r"Received: .*\[?(\d{1,3}(?:\.\d{1,3}){3})\]?", raw_text or "", re.I)
ip = m.group(1) if m else None
domain = heur.get('from_domain') or (parsed.get('from') and re.search(r"@([\w\.-]+)", parsed.get('from')) and re.search(r"@([\w\.-]+)", parsed.get('from')).group(1))
if domain and ip:
result['spf'] = spf_check(ip, domain)
else:
result['spf'] = {"ok": False, "error": "No se pudo extraer IP o dominio para SPF"}
if do_dkim:
raw_bytes = parsed.get('raw_bytes')
if raw_bytes:
result['dkim'] = dkim_check(raw_bytes)
else:
result['dkim'] = {"ok": False, "error": "No raw bytes disponibles para DKIM"}
# AI
if use_ai:
key = os.environ.get('OPENAI_API_KEY')
if not key:
result['ai'] = {"error": "OPENAI_API_KEY no configurada en Settings → Variables and secrets."}
else:
prompt = (
"Eres un detector de phishing. Recibiste este correo (incluye cabeceras y cuerpo):\n\n" +
(raw_text or "") +
"\n\nResponde con JSON válido con campos: verdict ('phishing'|'suspicious'|'legitimate'), score (float 0-1), reasons (lista de strings). SOLO devuelve JSON puro."
)
ok, out = call_openai(prompt, key)
if not ok:
result['ai'] = {"error": out}
else:
# try to parse json
parsed_ai = None
try:
parsed_ai = json.loads(out)
except Exception:
# try to find JSON substring
s = out.find('{')
e = out.rfind('}')
if s != -1 and e != -1 and e > s:
try:
parsed_ai = json.loads(out[s:e+1])
except Exception:
parsed_ai = {"raw": out}
else:
parsed_ai = {"raw": out}
result['ai'] = parsed_ai
return result
except Exception as e:
print("ANALYZE ERROR:", repr(e))
traceback.print_exc()
return {"error": True, "message": str(e)}
# ----------------------------
# UI
# ----------------------------
def format_result_html(res: Dict[str, Any]) -> str:
if res.get('error'):
return f"<b>Error:</b> {res.get('message')}"
parts = []
heur = res.get('heuristic') or {}
parts.append(f"<h3>Resultado del análisis</h3>")
parts.append(f"<b>Riesgo heurístico:</b> {heur.get('score',0)}%")
parts.append("<h4>Heurísticas</h4>")
if heur.get('reasons'):
parts.append("<ul>")
for r in heur.get('reasons'):
parts.append(f"<li>{r}</li>")
parts.append("</ul>")
else:
parts.append("<p>No se detectaron heurísticas sospechosas.</p>")
parts.append("<h4>Enlaces detectados</h4>")
links = heur.get('links') or []
if links:
parts.append("<ul>")
for u in links:
parts.append(f"<li><a href=\"{u}\" target=\"_blank\">{u}</a></li>")
parts.append("</ul>")
else:
parts.append("<p>-</p>")
parts.append("<h4>Comprobaciones técnicas</h4>")
if res.get('spf') is not None:
spf = res['spf']
if spf.get('ok'):
parts.append(f"<p>SPF: <b>Encontrado</b> (registros: {len(spf.get('records',[]))})</p>")
else:
parts.append(f"<p>SPF: <b>No verificado</b> - {spf.get('error') or ''}</p>")
if res.get('dkim') is not None:
d = res['dkim']
if d.get('ok'):
parts.append("<p>DKIM: <b>Firma válida</b></p>")
else:
parts.append(f"<p>DKIM: <b>No válido</b> - {d.get('error') or ''}</p>")
parts.append("<h4>Veredicto IA</h4>")
if res.get('ai') is None:
parts.append("<p>IA no activada.</p>")
elif isinstance(res.get('ai'), dict) and res.get('ai').get('error'):
parts.append(f"<p style='color:crimson;'><b>Error IA:</b> {res['ai'].get('error')}</p>")
else:
parts.append("<pre style='white-space:pre-wrap;background:#111;padding:10px;border-radius:6px;color:#d6d6d6;'>")
parts.append(json.dumps(res.get('ai'), ensure_ascii=False, indent=2))
parts.append("</pre>")
return '\\n'.join(parts)
with gr.Blocks(css=".gradio-container .output_html { color: #ddd; }", analytics_enabled=False) as demo:
gr.Markdown("## 🔎 Detector de Phishing — Mejorado (heurísticas + SPF/DKIM + OpenAI opcional)")
with gr.Row():
with gr.Column(scale=7):
inp = gr.Textbox(label="Correo (RAW o contenido)", lines=20, placeholder="Pega aquí el correo (ideal: RAW con cabeceras)")
use_ai = gr.Checkbox(label="Usar IA (OpenAI)", value=False)
do_spf = gr.Checkbox(label="Comprobar SPF (intentará extraer IP desde Received)", value=False)
do_dkim = gr.Checkbox(label="Comprobar DKIM (si pegas el RAW completo)", value=False)
btn = gr.Button("Analizar")
with gr.Column(scale=5):
out_html = gr.HTML("<i>Resultado aparecerá aquí</i>")
def run(raw, use_ai_flag, spf_flag, dkim_flag):
res = analyze_email(raw or "", use_ai=bool(use_ai_flag), do_spf=bool(spf_flag), do_dkim=bool(dkim_flag))
return format_result_html(res)
btn.click(run, inputs=[inp, use_ai, do_spf, do_dkim], outputs=[out_html])
if __name__ == '__main__':
demo.launch(server_name='0.0.0.0', server_port=7860)
|