Hi,
Iulian's response in the other thread prompted me to spend a little bit more time on my little AspectJ experiment. Here's the result: https://github.com/gkossakowski/scalac-aspects#typing-timings-typingtimingsaj
I also created a tool for outputting the raw data in HTML: https://gist.github.com/4543164
The good news is that it's already usable. If you are curious whether you have pathological code snippets that take exceptionally long to compile you can run this tool to collect some useful data. Once you get results you can sort them by second column to see what takes the longest time.
The tool turned out to be already usable enough that I managed to find a problem in our own code base, see: https://issues.scala-lang.org/browse/SI-6979
-- Francois ARMAND http://fanf42.blogspot.com http://www.normation.com
Hi,
You can try "mvn compile -DdisplayCmd=true"
The plugin will print the command used to run scalac (a big part of
the command is to stored in a file). I think you can reuse those info
to create a script to call scalac-aspects.
scalac-aspects require asjectj and other stuff (script, .aj) to run.
Currently, I don't think it's a good idea to include it in the
scala-maven-plugin, but it can provide usefull info.
Any info, idea are welcome. Tell me how to integrate them.
try with
<configuration>
<args>
<arg>-nobootcp</arg>
<arg>-toolcp</arg>
<arg>/DIR/workspaces/rudder-project/scalac-aspects/aspects.jar</arg>
</args>
You can tried with scala-maven-plugin 3.1.3-SNAPSHOT (repo : https://oss.sonatype.org/content/repositories/snapshots)
Nice you find a workaround.
Yes, please open a ticket, then i'll not forgot. I'll try to provide support for toolcp with dedicated config (next week).
If you have time, open an other ticket with info to integrate this metric in the plugin, or write a wiki page with instructions and your feedback for other users.
Thanks
Nice you find a workaround.
Yes, please open a ticket, then i'll not forgot. I'll try to provide support for toolcp with dedicated config (next week).
If you have time, open an other ticket with info to integrate this metric in the plugin, or write a wiki page with instructions and your feedback for other users.