ilya-biryukov added a comment. > Seems like this optimization is not worth (yet?). As soon as we get more > proximity paths (and hence more OR iterator children) that might make sense.
WDYT about storing **all** the elements with the minimal doc-id outside the heap? I.e. we can pop **all** elements with the minimum doc-id on 'advance' and iterator creation. That way we could potentially regain the improvements we've seen in the first version. https://reviews.llvm.org/D52083 _______________________________________________ cfe-commits mailing list cfe-commits@lists.llvm.org http://lists.llvm.org/cgi-bin/mailman/listinfo/cfe-commits
