2016-08-03 10:59 GMT+03:00 gpe atmpro <
gpe....@gmail.com>:
> First of all thanks for everything :)
You are welcome. =)
> --RemoveKeywords works perfectly for the log.html. That already helps a lot
> on the memory side, but as written in the doc, this filtering only happens
> on parsing the output.xml. It leaves this xml with all the useless keywords.
> Is it there a way to get it out of the xml all together. Something like a
> annotation [NOOUTPUT] on the keyword would be very convenient.
Unfortunately it isn't possible to remove information from output.xml
during test execution. You can, however, do that after execution using
the same option with Rebot and specifying a new output file:
rebot --removekeywords <pattern> --output new.xml original.xml
If you want to reduce the overall time for running tests and
processing results, you can disable log and report generation when
running tests using `--log NONE --report NONE`.
Depending on your needs, you may also want to consider using
--FlattenKeywords. It drops unnecessary elements already when
output.xml is parsed and thus makes the overall processing faster.
> Having so much output from a for loop, our xml is big on memory. Writing
> these megabytes of data certainly also slows down the whole testing process,
> which is unfortunate useless data.
Sure. The best way to avoid unnecessary data and to control exactly
what is logged, is moving the logic from the for loop in Robot data to
a test library.