On Mon, Dec 6, 2010 at 7:29 PM, Chet Ramey <chet.ra...@case.edu> wrote: > On 12/6/10 8:22 PM, Chet Ramey wrote: > >> Yes. In this case, the documentation is less than clear. The count >> argument is actually set by the first call to yank-last-arg, and that >> argument sets the word to extract from the history entry. Subsequent >> calls to yank-last-arg without a different intervening editing command >> use that same count as the argument to extract while moving through the >> history, and any count given to those subsequent calls sets the number >> of history entries to move. A negative count on those calls changes the >> direction. > > Sorry, that was partly a future direction. The count to the second and > subsequent calls just changes the direction, not the number of history > entries to skip. > > Chet > > -- > ``The lyf so short, the craft so long to lerne.'' - Chaucer > ``Ars longa, vita brevis'' - Hippocrates > Chet Ramey, ITS, CWRU c...@case.edu http://cnswww.cns.cwru.edu/~chet/ >
As you may have been able to tell from the macros I included in my report, I was trying to create a macro analogous to yank-last-arg but instead of "last" it would be some fixed n or last-minus-n, such as "yank-last-1-arg" or "yank-3-arg" with a repeatable, unchanging keystroke akin to Meta-. Using a positive argument, this works exactly as I intended. It would have been nice to be able to do something like "yank-last-n-arg".
