Hello. At Fri, 09 Dec 2005 10:11:42 -0500, Jim Fulton wrote:
- Common approach to Unicode
In particular, In Zope 3, all text is stored and managed as Unicode. The publisher decodes request data and encodes response data. The vast majority of application and library code can ignore encoding issues. (The exceptions are applications and frameworks that need to exhange text with non-Unicode-aware external systems.) This has provided great simplifications and allowed us to avoid common pitfals from mixing Unicode and encoded text.
We need to migrate Zope 2 to use a similar strategy. We need volunteers to brainstorm how this can be done and make one or more proposals. This is likely a prerequisite for finishing the publisher and ZPT work.
I want to tackle this problem, because I use japanese and I have experienced the encode/decode error pitfalls in a last few years. I'm wrong, I didn't touch zope development and didn't read the list carefully. In zope2.8, there is a zpt problem and this is a same pitfall. (unicode and string are mixed in TAL) -- TAHARA Yusei tahara@timedia.co.jp