I know find in project is in for a revamp some time in the future, but in the mean time, if I run it on a project with many many many files or with a few really really big files, it goes spinning beachball of death for eternity and killing is the only thing left to do -- which kinda sucks when you have unsaved changes.
Is there anything that can be done about this?
On Nov 20, 2005, at 2:47 PM, Caio Chassot wrote:
I know find in project is in for a revamp some time in the future, but in the mean time, if I run it on a project with many many many files or with a few really really big files, it goes spinning beachball of death for eternity and killing is the only thing left to do -- which kinda sucks when you have unsaved changes.
Does this happen regardless of what you search for? I often find that if I search for some complicated regexp, it might take much longer, and for a while I'd be seeing the beachball. It often takes a while until I regain control, so long in fact that I've considered killing it many times. Are we talking here about files with very long lines, or just with too many lines? Are they local or remote?
Is there anything that can be done about this?
Haris
Does this happen regardless of what you search for? I often find that if I search for some complicated regexp, it might take much longer, and for a while I'd be seeing the beachball. It often takes a while until I regain control, so long in fact that I've considered killing it many times. Are we talking here about files with very long lines, or just with too many lines? Are they local or remote?
Can be any simple search string
The files are either a 20-100mb mysql dump, which likely means very long lines, or a hundred word doc files. neither is even relevant to the project, but because i just open the projects with "e .", they end up in it. (where e is an alias for mate or open -a textmate)
They're local
If I wait a few aeons, sometimes I can regain control, but sometimes it goes so resource hungry I just kill it.
On Nov 20, 2005, at 3:18 PM, Caio Chassot wrote:
Can be any simple search string
The files are either a 20-100mb mysql dump, which likely means very long lines, or a hundred word doc files. neither is even relevant to the project, but because i just open the projects with "e .", they end up in it. (where e is an alias for mate or open -a textmate)
I'm pretty certain this is the problem then. TM doesn't take too kindly to long lines at the moment. They only workaround I can think of for now is to create a TMproject file with all the other relevant files. It would be nice if one could have a project variable that would be used as a regexp for the filename of files to omit from any search. I don't know how easy it would be for Allan to incorporate this into the current search system, and given that he is planning to rewrite the entire project system anyway, I would guess it might not be too high up on the priority list ATM.
Haris
On 20/11/2005, at 22:28, Charilaos Skiadas wrote:
[...] The only workaround I can think of for now is to create a TMproject file with all the other relevant files.
Possibly go to Preferences -> Advanced -> Folder References and modify the default patterns to exclude word documents and SQL dumps.
[...] and given that he is planning to rewrite the entire project system anyway, I would guess it might not be too high up on the priority list ATM.
Indeed -- current public build is one or two releases from being the 1.1 final release.
On Nov 20, 2005, at 3:42 PM, Allan Odgaard wrote:
On 20/11/2005, at 22:28, Charilaos Skiadas wrote: Possibly go to Preferences -> Advanced -> Folder References and modify the default patterns to exclude word documents and SQL dumps.
Aah, that's what those things are for. I have to say I didn't find the explanation very clarifying when I first read it, or when I read it for the second time either. Maybe a note that these are used to filter which files will appear? Finally, I don't have to keep seeing the aux and log files that latex generates. :-)
[...] and given that he is planning to rewrite the entire project system anyway, I would guess it might not be too high up on the priority list ATM.
Indeed -- current public build is one or two releases from being the 1.1 final release.
This is great news! Is there some big celebration party scheduled? :-)
Haris
On 20/11/2005, at 22:50, Charilaos Skiadas wrote:
Possibly go to Preferences -> Advanced -> Folder References and modify the default patterns to exclude word documents and SQL dumps.
Aah, that's what those things are for. I have to say I didn't find the explanation very clarifying when I first read it, or when I read it for the second time either. Maybe a note that these are used to filter which files will appear?
I'm likely to change it to just one list of “hide patterns” where a pattern is instead a (shell) glob pattern, e.g. *.aux. Improvements in this area is on the 1.2 to-do (I also want it out of the general preferences and into “project attributes”).
Indeed -- current public build is one or two releases from being the 1.1 final release.
This is great news! Is there some big celebration party scheduled? :-)
Of course -- I'll keep port 6600 open on my local machine for anyone who wants to join it! ;)
It would be nice if one could have a project variable that would be used as a regexp for the filename of files to omit from any search.
I guess what I could do is change the folder references pattern, as the files are not only not supposed to be searched, but not to be added to the project at all. Doesn't keep all problem files from slipping in, though.
I don't know how easy it would be for Allan to incorporate this into the current search system, and given that he is planning to rewrite the entire project system anyway, I would guess it might not be too high up on the priority list ATM.
If he could just implement a "press esc to halt find" feature, or anything, just to save me from having to kill tm, I'd be more than glad while I wait for the project rewrite.
-
Allan Odgaard -
Caio Chassot -
Charilaos Skiadas