Hi Stefan, (and other interested devs)
For additional testing I've created a directory/file structure like
1/
1 2/
2 3/
3 4/
ie a single file under a directory, the tree is 1038+ layers deep and
there is no difference between the original time taken by the
DirectoryScanner code and the modified code - if you want more
testing of this I'm willing to accommodate, but I'm pretty sure that
the Java reflection == slow is not always the case, and certainly
isn't in this case. I agree we must be careful to prevent new code
from causing the system to 'appear' slow (perceived performance), but
in this instance, I cannot prove that it is faster/slower, any
differences are minor (10s of milliseconds).
As an aside, my little recursive script to generate a mad dir
structure for testing breaks some kind of limit (unix or ruby not
sure) for file names, so any additional testing would require a
different script to generate the directory structure first :)
Kev
--
"In vain you tell me that Artificial Government is good, but that I
fall out with the Abuse. The Thing! The Thing itself is the Abuse!" -
Edmund Burke
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
