How the merging of databases is done in EMMA?

[Nobody/Anonymous]

Hi all,

I am interested to know how this merging of databases is done in EMMA.

Consider , we have two coverage database c1 and c2 containing the merge information for the
Class Test. The database c1 has the information about the 50 % coverage of the Class Test and the database c2 has the information of 10% coverage of the Class Test.  We merge database c1 and database c2 to generate c (merged coverage database). In this case , what coverage information will the database C will have about the class Test.
     In other words, how much coverage we can see about the Test if we generate any report.

Thanks & regards,
Mihir Jha

[Vlad Roubtsov]

The merged runtime coverage data for class Test will be  a union of all merged profiles. EMMA keeps a boolean coverage flag for each basic block, with 'true' meaning that the basic block is covered.

When c1 and c2 are merged, the flag is set to ((value in c1) OR (value in c2)). A basic block will be considered covered if it was covered in either c1 or c2. In other words, coverage accumulates across different execution runs.