Closed icsfy closed 5 months ago
The first issue was an accounting bug - fixed in eb5bd65591. (It also turns out that at least one of the lcov regression tests exposed the bug - but was not sensitive to the difference and did not notice. Sigh.)
The second appears to be a misunderstanding or an inconsistency in your environment/configuration.
exception
(and orphan
) filters were added in SHA e5e6355ccc42d4 - which was pushed sometime around 24 April.
The feature is not part of the 2.1-1 release - and I do not think it is mentioned in the man pages of the release.
I'm not sure which combination of code and doc you have or how you got them, but you may want to start fresh with a clean clone of TOT (or the 2.1 release - whichever you prefer) - to ensure consistency.In any event: please try your testcase again with this update - and close the issue if it appears fixed (or update the description to explain the new problem and how to see it).
Thanks
Henry
lcov: LCOV version 2.1-1 test.cpp:
compile and run:
use
--rc no_exception_branch=1
and then use--summary
to show the output .info:The branch coverage summary first gives
branches....: 60.0% (3 of 5 branches)
, but later isbranches....: 50.0% (2 of 4 branches)
And
man lcovrc
tells:but
--filter
does not recognize 'exception':geninfo: ERROR: unknown argument for --filter: 'exception'
. And there is aexception
argument which belongs to--ignore-errors
option inman genhtml
: