diff options
| -rw-r--r-- | NEWS.txt | 3 | ||||
| -rw-r--r-- | bs4/tests/test_soup.py | 22 |
2 files changed, 22 insertions, 3 deletions
@@ -3,6 +3,9 @@ * Renamed Tag.nsprefix to Tag.prefix, for consistency with NamespacedAttribute. +* Fixed a test failure that occured on Python 3.x when chardet was + installed. + = 4.0.0b8 (20110224) = * All tree builders now preserve namespace information in the diff --git a/bs4/tests/test_soup.py b/bs4/tests/test_soup.py index 33ab0fa..d8584b7 100644 --- a/bs4/tests/test_soup.py +++ b/bs4/tests/test_soup.py @@ -8,9 +8,18 @@ from bs4.element import ( NamespacedAttribute, ) from bs4.dammit import EntitySubstitution, UnicodeDammit -from bs4.testing import SoupTest +from bs4.testing import ( + SoupTest, + skipIf, +) import warnings +try: + import chardet + CHARDET_PRESENT = True +except ImportError, e: + CHARDET_PRESENT = False + class TestDeprecatedConstructorArguments(SoupTest): def test_parseOnlyThese_renamed_to_parse_only(self): @@ -212,16 +221,23 @@ class TestUnicodeDammit(unittest.TestCase): self.assertEqual( "euc-jp", dammit.original_encoding) + @skipIf( + CHARDET_PRESENT, + "Not testing last-ditch entity replacement because chardet is present and will find an encoding.") def test_last_ditch_entity_replacement(self): # This is a UTF-8 document that contains bytestrings - # completely incompatible with UTF-8 (encoded with some other + # completely incompatible with UTF-8 (ie. encoded with some other # encoding). # # Since there is no consistent encoding for the document, # Unicode, Dammit will eventually encode the document as UTF-8 # and encode the incompatible characters as REPLACEMENT # CHARACTER. - + # + # If chardet is installed, it will detect that the document + # can be converted into ISO-8859-1 without errors. This happens + # to be the wrong encoding, but it is a consistent encoding, so the + # code we're testing here won't run. doc = b"""\357\273\277<?xml version="1.0" encoding="UTF-8"?> <html><b>\330\250\330\252\330\261</b> <i>\310\322\321\220\312\321\355\344</i></html>""" |