From: Jesse P. <je...@st...> - 2004-09-05 21:18:16
|
>Pros >1) We would need to evaluate how much further it gets the translation >compared to not using. > > I see it as allowing more accurate string/message translations however it will require a little more code for the developers and a little extra instruction to a translator (how to move around variables). >Cons >1) The possible slowdowns of using the class instead of built in php >processor gettext code > > This class allows use of php's built-in gettext libraries if they are available. This is important as I am pretty sure we could inter-mix it's use with the standard gettext function, therefor only having to use it when we needed a string with vars in it. >2) What's with the setVar? Does that mean we need to add setVars for every >var used in a string? That will be some more work. > > > Yes, and yes it will be a little more work. >3) Do the tolls for making the po's such as xgettext all work smoothly with >this enhancement? > > huh? It appears to use the same POs gen'd by xgettext if that's what you are asking about. >4) If setVar is not done on a variable then I'm thinking what will happen is >the variable name with $ will appear instead of the variable's value. This >could get easily missed by various people at various times. > >5) How stable and used is this class? > > They use it and claim it is stable enough for their use. We will have to test it/read docs to determine it's how it functions when mis-used. The main point here, though, is that I haven't seen a more simple way to get things fully translated (as opposed to in pieces) than a wrapper like this or through complicated (s)printfs. If others think the pieced together translations will suffice, then so be it. I just would hate to have gone through this trouble and find later that people really can't stand the translations that have been done and to realize that with a little extra work at the beginning they could have been better. jesse |