Author |
Topic |
|
willdean
Tomato Guru
134 Posts |
Posted - Mar 19 2004 : 10:24:34 AM
|
I know I shall eventually get banned for repeating this ad naseum, but could we *please* have a reparse function which actually reparses properly?
For example, just now I added a definition of a local variable - an instance of a class. Because the appropriate header file wasn't included, the class name was coloured black. All fine so far.
So I went to the top of the file, added the #include, and came back to where I was working. Variable/type still not coloured. Never mind, I think, it's just the lazy reparse not keeping up with me (which it never seems to) a reparse will fix this. ALT-X (that's bound to reparse here). Nothing changes. ALT-X a few more times. (cf elevator buttons, 'C' on calculators, etc.) Still nothing.
Sit still, hands off the keyboard and curse WTS. And *now* the wretched thing gets coloured properly. Maybe the problem is that I don't spend long enough cursing WTS - but at the moment that feels unlikely.
PLEASE PLEASE PLEASE PLEASE can we have a manual reparse which reparses the current file and colours/intellisense-enables everything in it.
Could you PLEASE PLEASE PLEASE do this before worrying about whether the background reparse is being triggered properly.
Personally, I don't care if the background reparse works right, as long as I can manual reparse (via a keyboard binding) and have things sorted.
Probably related to the above problem, but slightly different is what I consider to be an absolutely typical process, which is to go source->header->add new class member variable->go back to source->try and use new variable.
In all recent builds, the new variable is just not finding its way into VA. I pretty much always have to switch back to the include file and back again to the source file.
I know that life *at WTS* might consist solely of making minute tweaks to a vast body of existing code, but for those of us trying to speedily write pages of new stuff, VAX just isn't good at the moment.
Oh, and can I add my vote to the call to have ALT-SPACE work in the middle of variables again?
Thanks,
Will
|
|
Stephen
Tomato Guru
United Kingdom
781 Posts |
Posted - Mar 19 2004 : 11:38:53 AM
|
One time I find reparse doesn't work is when I add a new string resource, then go back and use it in a .cpp file. Even after saving resource.h and then pressing the reparse button on the toolbar (yes, I put it back there), the symbol is still squiggly-red-underlined. |
Stephen Turner ClickTracks http://www.clicktracks.com/ Winner: ClickZ's Best Web Analytics Tool 2003 & 2004
|
|
|
willdean
Tomato Guru
134 Posts |
Posted - Mar 19 2004 : 12:13:17 PM
|
Although I whine all the time, I'm actually prepared to tolerate that not being very good (partly because it used to be much worse), but also because the symbol has been created in a completely different place to the code editor. You do have the 'check external files' option (or whatever the wording is) set, don't you?
It's where VA has seen and coloured a symbol in a header file but won't reflect that into the .CPP file that I start to get irate. (Actually I'm irate most of the time, just not with VAX, mostly.)
Will
|
|
|
Stephen
Tomato Guru
United Kingdom
781 Posts |
Posted - Mar 19 2004 : 12:24:41 PM
|
Are you referring to "Watch for externally modified header files"? Yes, it's set.
Anyway, resource.h is local to my project and is #include'd in the cpp file, so it should be able to find it on a forced reparse. That's what I think, anyway. |
Stephen Turner ClickTracks http://www.clicktracks.com/ Winner: ClickZ's Best Web Analytics Tool 2003 & 2004
|
|
|
Uniwares
Tomato Guru
Portugal
2322 Posts |
Posted - Mar 19 2004 : 12:26:12 PM
|
What I can see at the moment is that reparse works just fine, but there seems to be something interfering with it AFTER reparsing. Not sure if that is just the on-screen-coloring or something else. I can see symbols appearing properly colored right after reparsing and then vanishing shortly after. Happens rarely for me, but it seems to be related to certain includes or how these includes are structured. |
|
|
willdean
Tomato Guru
134 Posts |
Posted - Mar 19 2004 : 4:25:01 PM
|
Yes, I see that too - it's often more than just colouring though, because completion and case-correction etc doesn't seem to work for the uncoloured symbols.
|
|
|
Uniwares
Tomato Guru
Portugal
2322 Posts |
Posted - Mar 19 2004 : 5:01:14 PM
|
Exactly. Although when you switch to the file where the "unknown" symbols are defined (maybe some hidden 'StdUnknown.h' ??), all looks normal. Strange, very strange. |
|
|
gstelmack
Ketchup Master
USA
76 Posts |
Posted - Mar 22 2004 : 10:56:41 AM
|
Don't forget the whole "file is uncolored, start the debugger, now file is colored" problem and it's inverse (where the debugger uncolors a previously colored file).
I think parsing is working okay, I just think something is turning the coloring on/off. |
-- Greg Stelmack, Red Storm Entertainment |
|
|
Uniwares
Tomato Guru
Portugal
2322 Posts |
Posted - Mar 22 2004 : 12:21:00 PM
|
Sorry Greg, never seen what you describe. |
|
|
LarryLeonard
Tomato Guru
USA
1041 Posts |
Posted - Mar 22 2004 : 12:43:56 PM
|
quote: Don't forget the whole "file is uncolored, start the debugger, now file is colored" problem and it's inverse (where the debugger uncolors a previously colored file).
quote: Sorry Greg, never seen what you describe.
He's talking about this bit o'strangeness: http://forum.wholetomato.com/forum/topic.asp?TOPIC_ID=1852...
|
|
|
Uniwares
Tomato Guru
Portugal
2322 Posts |
Posted - Mar 22 2004 : 12:59:41 PM
|
Still, never seen this. |
|
|
|
Topic |
|