When using just a committed .mailmap
the aggregation is done nicely, but when --people-dict=...
is added the mailmap is not used at all. This is due to the branching in internal/plumbing/identity/identity.go
and lack of handling of mailmap in the IdentityDetector.LoadPeopleDict(...)
.
Another issue is that if you do use input to --people-dict
which has multiple emails mapping to the same name, those results will not be aggregated. I believe this is down to this method which doesn't check for duplicate ids[0]
before appending. A workaround (or intended but not clearly documented way) for this is to roll any duplicates into the same line e.g.
To work around the mailmap not being used, you can convert your mailmap into a --people-dict
and pass that.