davelists2@peoplemerge.com wrote at 2005-1-21 16:58 -0800:
We have a ton of AT objects ... ... UnicodeDecodeError: 'ascii' codec can't decode byte 0xc3 in position 5: ordinal not in range(128).
The traceback shows the field which is causing the error, and I don't see any unicode chars there!
The current Archetypes has a strange handling of text data: Internally, it stores it as unicode (good!) but it tries hard to deliver the content encoded. I am not even sure, there is a way to tell it to deliver unicode; but at least you can pass in the encoding your want ("encoding" keyword argument). AT 1.3.1 uses "instance.getCharset()" as "encoding", if not specified explicitely. This is better than previous versions that used the "original encoding" (which could mix different encodings in a single page). By the way: the Archetypes mailing lists are probably better suited to discuss Archetype related questions. -- Dieter