In that case I'll just use a container construction as that is the cleanest solution for this problem as I see it now.
...but if the use of '[' and ']' will give the appropropriate escaping/changing of fieldnames that means the string is being parsed anyway isn't it? Why would you want to _not_ have your strings converted to the correct version? Or is this more about where it should be done than how/if it should be done?
Simon
Kees Jongenburger wrote:
<quote> I found bug #5228 that reports this and contains a maintainer notice with a workaround, but shouldn't this be fixed? The problem is particularly nasty if you develop and deploy on different databases so you won't notice it at first. </quote>
It's nasty but the real problem is that the constraints are in a String maybe it's better to use containers?
Anyway everybody seems to think the translation from a to m_a and vice-versa
is supposed to happen in a different place (query,database,bridge,taglib)
my opinion GIGA (Garbadge in, Garbadge out) als long as the constraints have to be parsed we will keep having such problems
