Faster Find in Files

kokoglen · Post by **kokoglen** » Thu May 03, 2007 6:47 pm

Find in Files speed used to be fine for me, but recently someone dropped alot of files into the project repository (SVN) exactly where I need to search.

I am not sure the best way to improve speed, but some possibilities include:
1. Actively indexing the content of a project.
2. Using Microsoft's Indexing Service (if is available)
3. Limiting the folders to search and saving the set. Rather than "all subdirectories", I could choose specific folders to search through and save that set for repeated searching.

I am searching PHP, CSS and JS files through 21.5 megs of stuff (3,599 files in 1,639 folders)

Thanks. Looking forward to 2007 version!

Glen Lipka

Post by **syrupcore** » Thu May 03, 2007 9:15 pm

I wonder if ignore .svn folders would help?

kokoglen · Post by **kokoglen** » Fri May 04, 2007 6:32 pm

Is there a way of doing that currently?

The problem also might be related to some large files that were added. Like a 1 meg js file.

The thing is, using Aptana, the search is very fast, which makes me think it is indexing the files in advance.

Post by **syrupcore** » Fri May 04, 2007 10:59 pm

1 meg js file? holy macarel! Can I ask, what it is? NASA celestial body tracking software with automated satellite trajectory realignment?

No, there's no way to ignore SVN files at the moment. Just came to mind when I read your post and I said it out loud for the dev team's sake, in case it would help. I have no idea though.

WereCougar · Post by **WereCougar** » Mon Nov 05, 2007 3:09 pm

I too am having lots of problems with the search times. I have to use a separate file-searching utilility now.

I have *.php specified as a filter, and that should search about 250 files (there are 70,000+ files in all subdirectories, mostly images). But it takes several minutes and I give up.

Would it be valid to guess that RapidPHP is first making a master list of all files (the 70,000), then filtering out the ones to search (the 250)?

Post by **Karlis** » Sat Nov 24, 2007 12:23 am

I guess our editor is simply not designed to cope with 70 000+ files. It is insane. We could improve it by 20% maybe but not make it 10 times faster, it is just too complicated for us tight now. Indexing would be required and that is a big task, so a special file-search utility could be a solution for now.

There should be no problems for projects with a couple of thousands of files.

Post by **syrupcore** » Sun Nov 25, 2007 6:51 am

Karlis wrote:... It is insane...

haha... it totally is. 70,000 files = time for a database.

I'm not a systems fella but I imagine webuilder isn't the problem - it's the file system. When I search through thousands and thousands of files in windows directly, it takes a while as well.