Currently, the merge function tries to match the entries using the UUIDs, and tries to match the groups using the name. My proposition is to split the merge function into a merge and a sync function.

The sync function would, among other things:

• Match and update groups using UUIDs;
• Match and update entries using UUIDs;
• Handle moved groups;
• Handle moved entries;
• Handle deleted objects (groups and entries)

The merge function would:

• Match and update groups using the names (current behaviour);
• Match and update entries using the titles (we currently use the UUIDs, thus the current merge is in part a sync function)

This would solve #818 and would constitute a starting point for #90.

Some of the work for the sync is in #807, but If that sounds good, I'll move the new sync stuff in a new synchronize function.

@keepassxreboot/core-developers I'd like your opinion on that feature.