Yes, extending the key with a sequence number makes it unique again. I have missed a 'diff' stage quite a few times (or xcompare if you want). After http://rvdheij.wordpress.com/2014/05/01/writing-stages-in-c-using-gcc-part-1/ it looks like I should be able to pick up portions of the Linux source to do this. I suppose it would show such a group as one deleted and one added. Wonder if it would help your case.
Rob On 19 August 2014 22:57, Glenn Knickerbocker <[email protected]> wrote: > I have two lists of files that are mostly copies of each other, but > the old and new filenames have nothing to do with each other. I can > *nearly* interleave them by merging on size and timestamp, but there are > some groups that are the same size and written in the same second. I'd > like to interleave those groups alternating one from each (to the extent > possible). > > Aha, the answer occurred to me as I typed this, so I'll go ahead and > post it in case anybody else has the same problem: Use UNIQUE COUNT > to number the repeats, and add the count to the merge key. > > (end /) ... | uniq1: unique count 52-* first | mult1: faninany > | leave: merge 62-* d 1.10 | not chop 10 | ... > / uniq1: | mult1: > / ... | uniq2: unique count 52-* first | mult2: faninany | leave: > / uniq2: | mult2: > > ¬R >
