GHSA-V7CF-C9RM-WM3J

Vulnerability from github – Published: 2026-03-17 14:07 – Updated: 2026-03-17 14:07
VLAI
Summary
Uncontrolled recursion DoS in JustHTML() via deeply nested HTML
Details

Summary

justhtml through 1.9.1 allows denial of service via deeply nested HTML. During parsing, JustHTML.__init__() always reaches TreeBuilder.finish(), which unconditionally calls _populate_selectedcontent(). That function recursively traverses the DOM via _find_elements() / _find_element() without a depth bound, allowing attacker-controlled deeply nested input to trigger an unhandled RecursionError on CPython. Depending on the host application's exception handling, this can abort parsing, fail requests, or terminate a worker/process.

Details

TreeBuilder.finish() (treebuilder.py#L476) unconditionally calls _populate_selectedcontent(self.document) at line 494. _populate_selectedcontent() (treebuilder.py#L1243) calls _find_elements() (treebuilder.py#L1280) to recursively search the DOM tree for <select> elements:

def _find_elements(self, node: Any, name: str, result: list[Any]) -> None:
    """Recursively find all elements with given name."""
    if node.name == name:
        result.append(node)
    if node.has_child_nodes():
        for child in node.children:
            self._find_elements(child, name, result)  # recursive call

When the DOM tree depth exceeds CPython's default recursion limit (1000), this raises an unhandled RecursionError. The full call path is:

JustHTML(html)tokenizer.run()tree_builder.finish()_populate_selectedcontent(document)_find_elements(root, "select", selects) (recursive)

Deeply nested DOM trees can be produced by nesting <div> tags ~1000 levels deep. On CPython with the default recursion limit, approximately 11 KB of <div> nesting is sufficient to trigger the error. The exact depth threshold is environment-dependent (CPython version, recursion limit setting, call stack depth at invocation).

Additional recursive functions are affected on already-parsed deep trees: - Node.clone_node(deep=True) (node.py#L523) — called during sanitization - _node_to_html() (serialize.py#L580) — used by to_html(pretty=True) - _to_markdown_walk() (node.py#L817) — used by to_markdown()

Note: the library already uses iterative traversal in several comparable functions (e.g., _node_to_html_compact at serialize.py#L197, _to_text_collect at node.py#L161, _is_blocky_element at serialize.py#L405, apply_to_children at transforms.py#L1642), demonstrating the correct pattern.

PoC

from justhtml import JustHTML

html = "<div>" * 1000 + "x" + "</div>" * 1000
doc = JustHTML(html)  # raises RecursionError

Test environment: CPython 3.14.3, macOS ARM64 (Apple Silicon), justhtml 1.9.1, default recursion limit (1000)

Input Size Result
<div> × 500 5,501 bytes OK
<div> × 800 8,801 bytes OK
<div> × 1000 11,001 bytes RecursionError

The error occurs with both sanitize=True (default) and sanitize=False.

Impact

An attacker who can supply HTML for parsing can trigger an unhandled RecursionError during JustHTML() construction. The error is triggered during construction and is not avoided by justhtml configuration alone; mitigating it requires host-application exception handling or input constraints. Depending on the host application's exception handling, this can abort parsing, fail requests, or terminate a worker/process.

Suggested Fix

Convert the recursive tree traversal functions to iterative implementations using an explicit stack. Example for _find_elements:

def _find_elements(self, node: Any, name: str, result: list[Any]) -> None:
    stack = [node]
    while stack:
        current = stack.pop()
        if current.name == name:
            result.append(current)
        if current.has_child_nodes():
            stack.extend(reversed(current.children))

The same conversion should be applied to _find_element, clone_node(deep=True), _node_to_html(), and _to_markdown_walk().

Show details on source website

{
  "affected": [
    {
      "database_specific": {
        "last_known_affected_version_range": "\u003c= 1.9.1"
      },
      "package": {
        "ecosystem": "PyPI",
        "name": "justhtml"
      },
      "ranges": [
        {
          "events": [
            {
              "introduced": "0"
            },
            {
              "fixed": "1.10.0"
            }
          ],
          "type": "ECOSYSTEM"
        }
      ]
    }
  ],
  "aliases": [],
  "database_specific": {
    "cwe_ids": [
      "CWE-674"
    ],
    "github_reviewed": true,
    "github_reviewed_at": "2026-03-17T14:07:38Z",
    "nvd_published_at": null,
    "severity": "HIGH"
  },
  "details": "### Summary\n\njusthtml through 1.9.1 allows denial of service via deeply nested HTML. During parsing, `JustHTML.__init__()` always reaches `TreeBuilder.finish()`, which unconditionally calls `_populate_selectedcontent()`. That function recursively traverses the DOM via `_find_elements()` / `_find_element()` without a depth bound, allowing attacker-controlled deeply nested input to trigger an unhandled `RecursionError` on CPython. Depending on the host application\u0027s exception handling, this can abort parsing, fail requests, or terminate a worker/process.\n\n### Details\n\n`TreeBuilder.finish()` ([`treebuilder.py#L476`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/treebuilder.py#L476)) unconditionally calls `_populate_selectedcontent(self.document)` at [line 494](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/treebuilder.py#L494). `_populate_selectedcontent()` ([`treebuilder.py#L1243`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/treebuilder.py#L1243)) calls `_find_elements()` ([`treebuilder.py#L1280`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/treebuilder.py#L1280)) to recursively search the DOM tree for `\u003cselect\u003e` elements:\n\n```python\ndef _find_elements(self, node: Any, name: str, result: list[Any]) -\u003e None:\n    \"\"\"Recursively find all elements with given name.\"\"\"\n    if node.name == name:\n        result.append(node)\n    if node.has_child_nodes():\n        for child in node.children:\n            self._find_elements(child, name, result)  # recursive call\n```\n\nWhen the DOM tree depth exceeds CPython\u0027s default recursion limit (1000), this raises an unhandled `RecursionError`. The full call path is:\n\n`JustHTML(html)` \u2192 `tokenizer.run()` \u2192 `tree_builder.finish()` \u2192 `_populate_selectedcontent(document)` \u2192 `_find_elements(root, \"select\", selects)` (recursive)\n\nDeeply nested DOM trees can be produced by nesting `\u003cdiv\u003e` tags ~1000 levels deep. On CPython with the default recursion limit, approximately 11 KB of `\u003cdiv\u003e` nesting is sufficient to trigger the error. The exact depth threshold is environment-dependent (CPython version, recursion limit setting, call stack depth at invocation).\n\nAdditional recursive functions are affected on already-parsed deep trees:\n- `Node.clone_node(deep=True)` ([`node.py#L523`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/node.py#L523)) \u2014 called during sanitization\n- `_node_to_html()` ([`serialize.py#L580`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/serialize.py#L580)) \u2014 used by `to_html(pretty=True)`\n- `_to_markdown_walk()` ([`node.py#L817`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/node.py#L817)) \u2014 used by `to_markdown()`\n\nNote: the library already uses iterative traversal in several comparable functions (e.g., `_node_to_html_compact` at [`serialize.py#L197`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/serialize.py#L197), `_to_text_collect` at [`node.py#L161`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/node.py#L161), `_is_blocky_element` at [`serialize.py#L405`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/serialize.py#L405), `apply_to_children` at [`transforms.py#L1642`](https://github.com/EmilStenstrom/justhtml/blob/a866b6077770d9ec4cb6b6f9bfe7c918f98455e4/src/justhtml/transforms.py#L1642)), demonstrating the correct pattern.\n\n### PoC\n\n```python\nfrom justhtml import JustHTML\n\nhtml = \"\u003cdiv\u003e\" * 1000 + \"x\" + \"\u003c/div\u003e\" * 1000\ndoc = JustHTML(html)  # raises RecursionError\n```\n\nTest environment: CPython 3.14.3, macOS ARM64 (Apple Silicon), justhtml 1.9.1, default recursion limit (1000)\n\n| Input | Size | Result |\n|-------|------|--------|\n| `\u003cdiv\u003e` \u00d7 500 | 5,501 bytes | OK |\n| `\u003cdiv\u003e` \u00d7 800 | 8,801 bytes | OK |\n| `\u003cdiv\u003e` \u00d7 1000 | 11,001 bytes | RecursionError |\n\nThe error occurs with both `sanitize=True` (default) and `sanitize=False`.\n\n### Impact\n\nAn attacker who can supply HTML for parsing can trigger an unhandled `RecursionError` during `JustHTML()` construction. The error is triggered during construction and is not avoided by `justhtml` configuration alone; mitigating it requires host-application exception handling or input constraints. Depending on the host application\u0027s exception handling, this can abort parsing, fail requests, or terminate a worker/process.\n\n### Suggested Fix\n\nConvert the recursive tree traversal functions to iterative implementations using an explicit stack. Example for `_find_elements`:\n\n```python\ndef _find_elements(self, node: Any, name: str, result: list[Any]) -\u003e None:\n    stack = [node]\n    while stack:\n        current = stack.pop()\n        if current.name == name:\n            result.append(current)\n        if current.has_child_nodes():\n            stack.extend(reversed(current.children))\n```\n\nThe same conversion should be applied to `_find_element`, `clone_node(deep=True)`, `_node_to_html()`, and `_to_markdown_walk()`.",
  "id": "GHSA-v7cf-c9rm-wm3j",
  "modified": "2026-03-17T14:07:38Z",
  "published": "2026-03-17T14:07:38Z",
  "references": [
    {
      "type": "WEB",
      "url": "https://github.com/EmilStenstrom/justhtml/security/advisories/GHSA-v7cf-c9rm-wm3j"
    },
    {
      "type": "PACKAGE",
      "url": "https://github.com/EmilStenstrom/justhtml"
    },
    {
      "type": "WEB",
      "url": "https://github.com/EmilStenstrom/justhtml/releases/tag/v1.10.0"
    }
  ],
  "schema_version": "1.4.0",
  "severity": [
    {
      "score": "CVSS:4.0/AV:N/AC:L/AT:N/PR:N/UI:P/VC:N/VI:N/VA:H/SC:N/SI:N/SA:N",
      "type": "CVSS_V4"
    }
  ],
  "summary": "Uncontrolled recursion DoS in JustHTML() via deeply nested HTML"
}


Log in or create an account to share your comment.




Tags
Taxonomy of the tags.


Loading…

Loading…

Loading…

Forecast uses a logistic model when the trend is rising, or an exponential decay model when the trend is falling. Fitted via linearized least squares.

Sightings

Author Source Type Date Other

Nomenclature

  • Seen: The vulnerability was mentioned, discussed, or observed by the user.
  • Confirmed: The vulnerability has been validated from an analyst's perspective.
  • Published Proof of Concept: A public proof of concept is available for this vulnerability.
  • Exploited: The vulnerability was observed as exploited by the user who reported the sighting.
  • Patched: The vulnerability was observed as successfully patched by the user who reported the sighting.
  • Not exploited: The vulnerability was not observed as exploited by the user who reported the sighting.
  • Not confirmed: The user expressed doubt about the validity of the vulnerability.
  • Not patched: The vulnerability was not observed as successfully patched by the user who reported the sighting.

Loading…

Detection rules are retrieved from Rulezet.

Loading…

Loading…