about summary refs log tree commit diff
path: root/.venv/lib/python3.12/site-packages/lxml/cssselect.py
diff options
context:
space:
mode:
authorS. Solomon Darnell2025-03-28 21:52:21 -0500
committerS. Solomon Darnell2025-03-28 21:52:21 -0500
commit4a52a71956a8d46fcb7294ac71734504bb09bcc2 (patch)
treeee3dc5af3b6313e921cd920906356f5d4febc4ed /.venv/lib/python3.12/site-packages/lxml/cssselect.py
parentcc961e04ba734dd72309fb548a2f97d67d578813 (diff)
downloadgn-ai-4a52a71956a8d46fcb7294ac71734504bb09bcc2.tar.gz
two version of R2R are here HEAD master
Diffstat (limited to '.venv/lib/python3.12/site-packages/lxml/cssselect.py')
-rw-r--r--.venv/lib/python3.12/site-packages/lxml/cssselect.py101
1 files changed, 101 insertions, 0 deletions
diff --git a/.venv/lib/python3.12/site-packages/lxml/cssselect.py b/.venv/lib/python3.12/site-packages/lxml/cssselect.py
new file mode 100644
index 00000000..54cd75ac
--- /dev/null
+++ b/.venv/lib/python3.12/site-packages/lxml/cssselect.py
@@ -0,0 +1,101 @@
+"""CSS Selectors based on XPath.
+
+This module supports selecting XML/HTML tags based on CSS selectors.
+See the `CSSSelector` class for details.
+
+This is a thin wrapper around cssselect 0.7 or later.
+"""
+
+
+from . import etree
+try:
+    import cssselect as external_cssselect
+except ImportError:
+    raise ImportError(
+        'cssselect does not seem to be installed. '
+        'See https://pypi.org/project/cssselect/')
+
+
+SelectorSyntaxError = external_cssselect.SelectorSyntaxError
+ExpressionError = external_cssselect.ExpressionError
+SelectorError = external_cssselect.SelectorError
+
+
+__all__ = ['SelectorSyntaxError', 'ExpressionError', 'SelectorError',
+           'CSSSelector']
+
+
+class LxmlTranslator(external_cssselect.GenericTranslator):
+    """
+    A custom CSS selector to XPath translator with lxml-specific extensions.
+    """
+    def xpath_contains_function(self, xpath, function):
+        # Defined there, removed in later drafts:
+        # http://www.w3.org/TR/2001/CR-css3-selectors-20011113/#content-selectors
+        if function.argument_types() not in (['STRING'], ['IDENT']):
+            raise ExpressionError(
+                "Expected a single string or ident for :contains(), got %r"
+                % function.arguments)
+        value = function.arguments[0].value
+        return xpath.add_condition(
+            'contains(__lxml_internal_css:lower-case(string(.)), %s)'
+            % self.xpath_literal(value.lower()))
+
+
+class LxmlHTMLTranslator(LxmlTranslator, external_cssselect.HTMLTranslator):
+    """
+    lxml extensions + HTML support.
+    """
+
+
+def _make_lower_case(context, s):
+    return s.lower()
+
+ns = etree.FunctionNamespace('http://codespeak.net/lxml/css/')
+ns.prefix = '__lxml_internal_css'
+ns['lower-case'] = _make_lower_case
+
+
+class CSSSelector(etree.XPath):
+    """A CSS selector.
+
+    Usage::
+
+        >>> from lxml import etree, cssselect
+        >>> select = cssselect.CSSSelector("a tag > child")
+
+        >>> root = etree.XML("<a><b><c/><tag><child>TEXT</child></tag></b></a>")
+        >>> [ el.tag for el in select(root) ]
+        ['child']
+
+    To use CSS namespaces, you need to pass a prefix-to-namespace
+    mapping as ``namespaces`` keyword argument::
+
+        >>> rdfns = 'http://www.w3.org/1999/02/22-rdf-syntax-ns#'
+        >>> select_ns = cssselect.CSSSelector('root > rdf|Description',
+        ...                                   namespaces={'rdf': rdfns})
+
+        >>> rdf = etree.XML((
+        ...     '<root xmlns:rdf="%s">'
+        ...       '<rdf:Description>blah</rdf:Description>'
+        ...     '</root>') % rdfns)
+        >>> [(el.tag, el.text) for el in select_ns(rdf)]
+        [('{http://www.w3.org/1999/02/22-rdf-syntax-ns#}Description', 'blah')]
+
+    """
+    def __init__(self, css, namespaces=None, translator='xml'):
+        if translator == 'xml':
+            translator = LxmlTranslator()
+        elif translator == 'html':
+            translator = LxmlHTMLTranslator()
+        elif translator == 'xhtml':
+            translator = LxmlHTMLTranslator(xhtml=True)
+        path = translator.css_to_xpath(css)
+        super().__init__(path, namespaces=namespaces)
+        self.css = css
+
+    def __repr__(self):
+        return '<%s %x for %r>' % (
+            self.__class__.__name__,
+            abs(id(self)),
+            self.css)