Closed heuer closed 9 years ago
I think in test_kanji_enforce_binary there is a misunderstanding for test 2. According to the standard, the smallest possible mode should always be chosen. In other words, if it is possible to store 16 bits as 13 bits, then the mode should be kanji. The encoding parameter should specify what encoding the contents are in, not what they will become. I adjusted this test to use the mode parameter to force binary. Basically, testing that overriding the guessed content type works.
Still looking at the other tests...
I believe I fixed everything. All the tests pass.
BTW, "QuickMark" can handle Kanji