[svn r3835] r4520@delle: sbehnel | 2008-06-20 09:48:39 +0200

CSS selector parser accidentally merged non-adjacent classes --HG-- branch : trunk
scrapy · Jun 20, 2008 · 87bf4d3 · 87bf4d3
1 parent 3af9e22
commit 87bf4d3
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 0 deletions.
diff --git a/src/lxml/cssselect.py b/src/lxml/cssselect.py
@@ -738,6 +738,8 @@ def parse_simple_selector(stream):
                 result = Pseudo(result, type, ident)
             continue
         else:
+            if peek == ' ':
+                stream.next()
             break
         # FIXME: not sure what "negation" is
     return result
@@ -823,7 +825,10 @@ def tokenize(s):
     while 1:
         match = _whitespace_re.match(s, pos=pos)
         if match:
+            preceding_whitespace_pos = pos
             pos = match.end()
+        else:
+            preceding_whitespace_pos = 0
         if pos >= len(s):
             return
         match = _count_re.match(s, pos=pos)
@@ -839,6 +844,8 @@ def tokenize(s):
             pos += 2
             continue
         if c in '>+~,.*=[]()|:#':
+            if c in '.#' and preceding_whitespace_pos > 0:
+                yield Token(' ', preceding_whitespace_pos)
             yield Token(c, pos)
             pos += 1
             continue

diff --git a/src/lxml/tests/test_css.py b/src/lxml/tests/test_css.py
@@ -46,6 +46,14 @@ class CSSTestCase(HelperTestCase):
         ('div.dialog', 51),
         ('div .dialog', 51),
         ('div.character, div.dialog', 99),
+        ('div.direction.dialog', 0),
+        ('div.dialog.direction', 0),
+        ('div.dialog.scene', 1),
+        ('div.scene.scene', 1),
+        ('div.scene .scene', 0),
+        ('div.direction .dialog ', 0),
+        ('div .dialog .direction', 4),
+        ('div.dialog .dialog .direction', 4),
         ('#speech5', 1),
         ('div#speech5', 1),
         ('div #speech5', 1),