yuja added a comment.
> > I also expect hg grep --all-files -r0+1 foo will show matches from both rev 0 and 1. > > Suppose there are ten hits in 0 and the same 10 hits in 1, do you mean we print out all the 20 results, What purpose that would serve? Yes. I think that's the least surprising behavior for --all-files, which is the option to ignore file status. Say a file is modified at both rev 0 and 1, which revisions should be grepped? a. only rev 0 b. rev 0 and rev 1 (because a file is changed at rev 1) c. rev 0 and rev 1 (no matter if a file is changed or not) d. --all-files for rev 0, and --diff for rev 1 (a) is the awkward behavior of the current "hg grep" which we're trying to fix. (b) might sound sensible, but why are unchanged lines in rev 1 displayed again? (c) shows redundant matches, but is consistent. (d) seems a bit tricky, but will be useful. So my proposal was (c). If we take (d), which I think is also good, we'll need to find more appropriate option name than --all-files. REPOSITORY rHG Mercurial REVISION DETAIL https://phab.mercurial-scm.org/D3976 To: sangeet259, #hg-reviewers Cc: yuja, mercurial-devel _______________________________________________ Mercurial-devel mailing list Mercurial-devel@mercurial-scm.org https://www.mercurial-scm.org/mailman/listinfo/mercurial-devel