diff options
| author | Leonard Richardson <leonard.richardson@canonical.com> | 2012-03-01 13:37:42 -0500 |
|---|---|---|
| committer | Leonard Richardson <leonard.richardson@canonical.com> | 2012-03-01 13:37:42 -0500 |
| commit | e533b599457f713944a17d29739308f09bfd5aef (patch) | |
| tree | 70fa05e8de9e6457c0df6bfb594ac3cf04abd38d /bs4/testing.py | |
| parent | 483286bfbb40bfabe4c48c9f31c59ef7449d64bb (diff) | |
In HTML5-style <meta charset="foo"> tags, the value of the "charset" attribute is now replaced with the appropriate encoding on output. [bug=942714]
Diffstat (limited to 'bs4/testing.py')
| -rw-r--r-- | bs4/testing.py | 18 |
1 files changed, 18 insertions, 0 deletions
diff --git a/bs4/testing.py b/bs4/testing.py index a3e0b38..1b73160 100644 --- a/bs4/testing.py +++ b/bs4/testing.py @@ -368,6 +368,24 @@ class HTMLTreeBuilderSmokeTest(object): # For the rest of the story, see TestSubstitutions in # test_tree.py. + def test_html5_style_meta_tag_reflects_current_encoding(self): + # Here's the <meta> tag saying that a document is + # encoded in Shift-JIS. + meta_tag = ('<meta id="encoding" charset="x-sjis" />') + + # Here's a document incorporating that meta tag. + shift_jis_html = ( + '<html><head>\n%s\n' + '<meta http-equiv="Content-language" content="ja"/>' + '</head><body>Shift-JIS markup goes here.') % meta_tag + soup = self.soup(shift_jis_html) + + # Parse the document, and the charset is replaced with a + # generic value. + parsed_meta = soup.find('meta', id="encoding") + self.assertEqual('%SOUP-ENCODING%', parsed_meta['charset']) + self.assertEqual(True, parsed_meta.contains_substitutions) + class XMLTreeBuilderSmokeTest(object): def test_docstring_generated(self): |