Hello gri@golang.org (cc: golang-codereviews@googlegroups.com), I'd like you to review this change to https://code.google.com/p/go.tools
LGTM + suggestion for unrelated change https://codereview.appspot.com/99310043/diff/10002/container/intsets/util_tes... File container/intsets/util_test.go (right): https://codereview.appspot.com/99310043/diff/10002/container/intsets/util_tes... container/intsets/util_test.go:14: n := nlz(word(v)) I think popcount and nlz should return an int (or perhaps even a uint). It might even remove some conversions where you use the nlz to compute shifts, etc.
https://codereview.appspot.com/99310043/diff/10002/container/intsets/util_tes... File container/intsets/util_test.go (right): https://codereview.appspot.com/99310043/diff/10002/container/intsets/util_tes... container/intsets/util_test.go:14: n := nlz(word(v)) On 2014/05/15 17:58:02, gri wrote: > I think popcount and nlz should return an int (or perhaps even a uint). It might > even remove some conversions where you use the nlz to compute shifts, etc. Done.
*** Submitted as https://code.google.com/p/go/source/detail?r=58c7b2933d04&repo=tools *** go.tools/container/intsets: fix compile error on 32-bit platforms. Also, fix typo. LGTM=gri R=gri CC=golang-codereviews https://codereview.appspot.com/99310043
LGTM but now you can remove the unnecessary int conversions at sparse.go:134, 147