searxng/searxng · error · SearxEngineXPathException

SearxEngineXPathException

Error message

SearxEngineXPathException

What it means

SearxEngineXPathException raised by eval_xpath when evaluating an already-compiled XPath against an element raises lxml XPathError — typically because the expression is evaluated on the wrong node type (e.g. an attribute node) or lxml runtime limits are hit, not a syntax error.

Source

Thrown at searx/utils.py:555

    ``TypeError``:
      Raised when ``xpath_spec`` is neither a :py:obj:`str` nor a
      :py:obj:`lxml.etree.XPath`.

    ``SearxXPathSyntaxException``:
      Raised when there is a syntax error in the *XPath* selector (``str``).

    ``SearxEngineXPathException:``
      Raised when the XPath can't be evaluated (masked
      :py:obj:`lxml.etree..XPathError`).
    """
    xpath: XPath = get_xpath(xpath_spec)
    try:
        # https://lxml.de/xpathxslt.html#xpath-return-values
        return xpath(element)
    except XPathError as e:
        arg = ' '.join([str(i) for i in e.args])
        raise SearxEngineXPathException(xpath_spec, arg) from e


def eval_xpath_list(element: ElementType, xpath_spec: XPathSpecType, min_len: int | None = None) -> list[t.Any]:
    """Same as :py:obj:`searx.utils.eval_xpath`, but additionally ensures the
    return value is a :py:obj:`list`.  The minimum length of the list is also
    checked (if ``min_len`` is set)."""

    result: list[t.Any] = eval_xpath(element, xpath_spec)
    if not isinstance(result, list):
        raise SearxEngineXPathException(xpath_spec, 'the result is not a list')
    if min_len is not None and min_len > len(result):
        raise SearxEngineXPathException(xpath_spec, 'len(xpath_str) < ' + str(min_len))
    return result


def eval_xpath_getindex(
    element: ElementType,
    xpath_spec: XPathSpecType,

View on GitHub (pinned to 9fea41204f)

Solutions

1. Evaluate on element nodes: select '//a/@href' text separately but path expressions on elements
2. Re-check the expression with SearxXPathSyntaxException causes ruled out — test directly: element.xpath(expr)
3. Pin/upgrade lxml to a version compatible with your SearXNG release

try:
    res = eval_xpath(dom, xp)
except SearxEngineXPathException as e:
    logger.warning('xpath eval failed for %s: %s', xp, e)
    res = None

When it happens

Trigger: Calling eval_xpath(element, xpath) where xpath(element) raises XPathError, e.g. expressions expecting element context applied to attribute/text nodes, or very unusual evaluation failures inside lxml.

Common situations: Engine code evaluating element-relative paths on nodes obtained from attribute selection, or lxml version changes altering evaluation behavior on certain node types.

Related errors

• invalid
• extract_text(None, allow_none=False)
• unsupported type
• Empty url resultset
• URL not found

AI-assisted analysis of searxng/searxng@9fea41204f (2026-08-27). Data as JSON: /api/errors/722d7e31c3244100. Report an issue: GitHub.