Really the only time that DM_CONVERT_CASE should not be considered a bug is when you are certain of the Locale(s) that code will be run in. For library developers at least, that is never the case.
In Hibernate, for example, this regression has continually crept into our codebase many times over the past few years
Reference https://hibernate.atlassian.net/browse/HHH-9733 and its linked issues
You can increase the rank of the warning, see
http://findbugs.sourceforge.net/api/edu/umd/cs/findbugs/BugRanker.html
Thanks Andrey! As far as Gradle integration anyway, the bigger hurdle is the confidence. For my concern, I'd like to say that these DM_CONVERT_CASE warnings have a high confidence of being a bug (I want to treat them as unconditionally a bug in our codebase). Is the a similar way to increase the confidence (priority) of a warning?
I will look into using the BugRanker as well, because to me the rank ought to be higher (err, lower :) than 20 which is what FindBugs reports atm:
FWIW, someone pointed out
-adjustPriority
in the FindBugs code which I tried to add support for in the Gradle integration. But FindBugs does not like that from the command line:So not sure from where
-adjustPriority
is actually considered valid.