Performance of JavadocTagContinuationIndentation, AtclauseOrder, NonEmptyAtclauseDescription etc.

47 views
Skip to first unread message

Csaba Kozák

unread,
May 10, 2015, 4:01:11 AM5/10/15
to check...@googlegroups.com
Hi!

We just switched to CS 6.5 from 5.8. I tried to enable the:

JavadocTagContinuationIndentation, AtclauseOrder, NonEmptyAtclauseDescription and SingleLineJavadoc

checks.

However i noticed that the CS execution increased from 4 seconds to 44 seconds on this codebase.
Is this normal? Should these checks need this much execution time?
We already have JavadocStyle, JavadocVariable, JavadocMethod and JavadocType and lots of other non-JavaDoc checks
and they running within the 4 seconds execution time.

Regards,
Csaba

Michal Kordas

unread,
May 10, 2015, 6:08:17 AM5/10/15
to Csaba Kozák, check...@googlegroups.com

Hi Csaba,

After investigation I've found that AtclauseOrder is the check with performance problem. Ticket created: https://github.com/checkstyle/checkstyle/issues/1064

Thanks for reporting,
Michal Kordas

--
You received this message because you are subscribed to the Google Groups "checkstyle" group.
To unsubscribe from this group and stop receiving emails from it, send an email to checkstyle+...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Csaba Kozák

unread,
May 10, 2015, 9:54:31 AM5/10/15
to check...@googlegroups.com, kozak...@gmail.com
Hi Michał,

Thanks for the quick answer!

I also enabled these checks individually, and i think not just AtclauseOrder is slow:
with AtclauseOrder: 44 sec
with JavadocTagContinuationIndentation: 60 sec
with NonEmptyAtclauseDescription: 52 sec
with SingleLineJavadoc: 53 sec

Instead of the normal 4 sec check time. I guess both of these are using a same routine which is responsible for the performance problem.

Regards,
Csaba

Roman Ivanov

unread,
May 13, 2015, 12:11:58 PM5/13/15
to Csaba Kozák, check...@googlegroups.com
Here is explanation of the performance issue: https://github.com/checkstyle/checkstyle/issues/1064#issuecomment-100697131
Reply all
Reply to author
Forward
0 new messages