Closed llorllale closed 6 years ago
@llorllale but only @value
matters for us. Or you want our tests to be more detailed oriented. I didn't understand, what is broken?
@yegor256 For SCOM_minK
I need the tests to be more detail oriented.
SCOM_minK
is another number calculated for SCOM
separate from its @value
. The paper says that if SCOM < SCOM_minK then you should consider splitting the class into more cohesive classes.
@llorllale we don't need to test those supplementary variables. Each metric has them. We just have them in the XML report for visibility. That's enough.
(This may or may not be relevant if the proposal for #169 is approved.)
Problem: As per #142, I am implementing calculation of
SCOM_minK
value forSCOM
. I have the calculation ready. It will be stored in a newvar
in the metric.xml, butMetricsTest
currently only tests the//class/@value
of metrics.Proposal Ideally, enable tests of arbitrary xpaths in
MetricsTest
. Otherwise, enable tests of<var>
using an xpath similar to ://var[@id = 'SCOM_minK']
.