A lot of people would like to see these things cleaned up, but unfortunately unless it becomes a pressing need for someone it doesn't get done (for lack of time). If you have some experience with that, and have ideas on how to fix things (and check that they work ok), please do write some patches. Many of us will be grateful to you. Florent In article <20020728030851.29483.qmail@web20706.mail.yahoo.com> you write:
In PropertySheets.py of zope/lib/python/OFS
def xml_escape(v): ..... return unicode(v, "latin-1").encode("utf-8")
Are there anyway to substitute the "latin-1" with an attribute of the zope site's attribute say: "zope-charset" which a site maintainer can modify through the web?
I raise this question because the current-cvs version of zope is turned into unicode, but all the source everywhere is assumed that the data input is "latin-1" but I am very sure that it will not always the case. Some sites simply don't work in that assumption, turning many sites using zope stop upgrade or stop using zope. This will seriously hurt the zope community. The initial act to make zope understand unicode is to help zope internationalize, but the actual effect was not.
I urge the zope developer, the zope coporation as well as product developers (some products like ieDocument and worldPilot have "8859-1" specified in their source) alert on this issue, and take action before it is too late.
The best solution I could come up with is to allow individual site manager to choose its default charset and allow other site-branch manager to overwrite it with acquisition, thus the attribute zope-charset will be used to determine both the "Content-type: text/plain; charset="<dtml-var zope-charset>" and the code convertion of Zope like PropertySheets.py
-- Florent Guillaume, Nuxeo (Paris, France) +33 1 40 33 79 87 http://nuxeo.com mailto:fg@nuxeo.com