Crash when searching words with big number of occurrencies

[emadima]

Hi guys,

as I told in another topic, I'm testing this IDE and I find it GOOD!

BTW I have a problem when seerching words that have big number of occurrencies (i.e. "void").
If I start this kind of search, IDE crashes and I have to kill the application.

Searching is ok for words with few occurrencies.
My project is about 15000 files and I'm using Ubuntu 10.04.

Any idea on this issue?

Thanks ,
E.

[eranif]

I just tried it on windows and it seems to work fine with no problems (searching for 'void' under codelite's workspace)
I will need to see if it fails under linux (have you tried to attach debugger to codelite before performing the search and see where it crash?)

Eran

[emadima]

Hi Eran,

indeed the IDE don't crash because I tried to left it for an amount of time and after that the search was completed,
but

1) the amount of time was something high (tomorrow I will try to measure)

2)if I start to click somewhere, the IDE seems to frooze and I need to "Force quit"

of course this causes that if you start such kind of search you cannot control the IDE anymore ...
I would need some "kill the Search" tool to kill the searching process if I understand that it is too long...

BTW, very good good job!
Thanks
E.

[emadima]

I tried to execute a Search of "char"

1) after pressing some results on the output windows, the system froozes and you cannot change file or kill the Search
2) after 8 minutes the system wakes up, shows all the results, executes all the command you did during the 8 minutes and says "Mtches found:65752, elapsed time: 31 seconds"

I hope this can help you.
Bye,
E.

[eranif]

Mtches found:65752, elapsed time: 31 seconds"

From this line, I can say that this is a UI problem... the search did ended after 31 seconds, its just the UI can't print the huge output fast enough
I will see how this can be optimized (I believe that we are having here an event flooding that the system is queuing and processes, this explains why your commands were executed 8 minutes later, since all the command events were queued and only got processed after all the 'find-in-files' events

The way the find in files is working, is that an event is sent with the matches found for a given file, so if the searched word is very common (e.g. "char") and exists almost in every file and in your case you got 15K files, this means 15K events... which explains what you see

Eran

[emadima]

Hi Eran, thank you for your effort,

please let me know if you need some help or extra test I can do for you.
Meantime I will avoid to search very common words

bye,
E.