| [news.eclipse.tools.pdt] Re: Code Completion: include require |
Rik wrote:
Yes, I said "low priority".
Whatever the case, is an error to have have the overall project in the code completion because it says: "Hey, you can use this and this" while is not true, is out of scope. In other words, "code completion" should be named "code I-Feel-Lucky"! :-)
When one try to use PDT to make big projects, the code completion showing the overall project, everywhere, make it useless to have.
Maybe include/require are not sufficient. Maybe can be decided by an option. Maybe there are other solution. In any case this is a big problem for big projects.
We are in year 2008, almost 2009. A good code completion is the basis of an editor of choice and today good editors make the difference from how many develop in one language and how many in another. Is not only a choice based on the mere language.
So, it is importat to take care, very much, of the code completion feature, sooner or later.
"ChanibaL" <eclipse@xxxxxxxxxxxx> ha scritto nel messaggio news:gb0af6$lmf$1@xxxxxxxxxxxxxxxxxxxxI vote against as PHP kind of doesn't work this way - you've got autoloaders, automaticly included files, lots of magic etc. Limiting it to the explicitly (inclu|require)ded files is not a good idea.
The only thing I can think of that could be usefull is to add some kind of directory metadata that files in this dir only have content assist from these directories (and a list or regex to select them). I'd prefer through that PDT developers focus on more important things first.