Fix unicode selectors.

Resolves #184.

Author: rushter

CHANGES.md

@@ -1,5 +1,8 @@
 # selectolax Changelog
 
+Unreleased
+
+- Fix CSS selectors that contain Unicode characters.
 
 # Version 0.4.0
 

selectolax/lexbor/selection.pxi

@@ -53,7 +53,7 @@ cdef class LexborCSSSelector:
             raise TypeError("Query must be a string.")
 
         bytes_query = query.encode(_ENCODING)
-        selectors_list = lxb_css_selectors_parse(self.parser, <lxb_char_t *> bytes_query, <size_t>len(query))
+        selectors_list = lxb_css_selectors_parse(self.parser, <lxb_char_t *> bytes_query, <size_t>len(bytes_query))
 
         if selectors_list == NULL:
             raise SelectolaxError("Can't parse CSS selector.")

tests/test_lexbor.py

@@ -50,3 +50,10 @@ def test_double_unwrap_does_not_segfault():
     inner_div.unwrap()
     some_set.add(outer_div.parent)
     some_set.add(outer_div.parent)
+
+
+def test_unicode_selector_works():
+    html = '<span data-original-title="Pneu renforcé"></span>'
+    tree = LexborHTMLParser(html)
+    node = tree.css_first('span[data-original-title="Pneu renforcé"]')
+    assert node.tag == "span"