Florent Guillaume wrote:
Martijn Faassen wrote: [snip]
What I'm worried about doing this for old code, but some steps will probably become clear during the brainstorming session. Migration tools that turn strings in the ZODB into unicode ones magically (with the ability to spell out exceptions and encoding)? Tricky...
I'd start by adding some deprecation warnings at the places where PlacelessTranslationService (or Localizer for that matter) do their workarounds so that normally incompatible types are mixed together.
With the most recent Silva under development, we're not using either anymore, so someone else would have to do this. Where would you put the deprecation warning in PTS? It monkey patches the ZPT engine, and in particular its StringIO object, to make sure that UTF-8 strings and unicode strings, when finally concatenated at the end after page template generation, can actually be concatenated as it will first encode the unicode strings to UTF-8. The output of the page template will then be UTF-8. In Zope 2 without this in place, the output of the page template will be unicode, and if you put in UTF-8 strings (that aren't just ascii), the system will give rise to a UnicodeError when this concatenation takes place. I guess you could put some detection code in PTS's concatenation functionality to see whether anything unicode is mixed with non-ascii strings, and in that case output one deprecation warning. I have a suspicion that you'll see this warning whenever someone views a page template in Plone. :) Regards, Martijn