-------------------------------------------- On Sun, 1/8/17, vim-dev ML <[email protected]> wrote:
Subject: Re: [vim/vim] bug in persistent undo -- `:earlier 100d` undoes one change instead of all (#1254) To: "vim/vim" <[email protected]> Cc: "vim-dev ML" <[email protected]>, "Your activity" <[email protected]> Date: Sunday, January 8, 2017, 4:06 AM 2016-11-17 16:49 GMT+03:00 Bram Moolenaar <[email protected]>: > > Christian Brabandt wrote: > >> I am quite sure it happens here in undo.c >> >> 2390 if (dosec) >> 2391 val = (long)(uhp->uh_time - starttime); >> >> val will get a negative value, because starttime is less than the time of >> the undo header. Then this condition here will only get hit once: >> >> >> 2403 if ((step < 0 ? uhp->uh_seq <= curbuf->b_u_seq_cur >> 2404 : uhp->uh_seq > curbuf->b_u_seq_cur) >> 2405 && ((dosec && val == closest) >> 2406 ? (step < 0 >> 2407 ? uhp->uh_seq < closest_seq >> 2408 : uhp->uh_seq > closest_seq) >> 2409 : closest == closest_start >> 2410 || (val > target >> 2411 ? (closest > target >> 2412 ? val - target <= closest - target >> 2413 : val - target <= target - closest) >> 2414 : (closest > target >> 2415 ? target - val <= closest - target >> 2416 : target - val <= target - closest)))) >> 2417 { >> 2418 closest = val; >> 2419 closest_seq = uhp->uh_seq; >> 2420 } >> >> This will only hit the first time, because only at the beginning closest >> == closest_start, but after the next iteration, it will never hit again. >> >> I think this patch will fix it, but I am not sure: >> >> ```patch >> diff --git a/src/undo.c b/src/undo.c >> index 57c3c20..07c5d3c 100644 >> --- a/src/undo.c >> +++ b/src/undo.c >> @@ -2406,7 +2406,7 @@ undo_time( >> ? (step < 0 >> ? uhp->uh_seq < closest_seq >> : uhp->uh_seq > closest_seq) >> - : closest == closest_start >> + : closest == closest_start || closest < 0 >> || (val > target >> ? (closest > target >> ? val - target <= closest - target >> ``` >> >> That condition is really hard to understand, I stared on it for about >> an hour and even now I am not quite sure I completly understand it.. > > I tried to write a test for this, but it passes without the fix: > > func Test_undofile_earlier() > " Issue #1254 > new Xfile > call feedkeys("ione\<Esc>", 'xt') > set ul=100 > call feedkeys("otwo\<Esc>", 'xt') > set ul=100 > call feedkeys("othree\<Esc>", 'xt') > set ul=100 > w > wundo Xundofile > bwipe! > new Xfile > rundo Xundofile > earlier 1m > call assert_equal('', getline(1)) > bwipe! > call delete('Xfile') > call delete('Xundofile') > endfunc > > And indeed, if I try the reproduce steps without exiting Vim then it > works as expected. So it seems that something gets reset when > restarting Vim, but not when reading the undo file. This is one of the reasons Neovim tests prefer to restart Neovim another time even during a single test rather then try to smash more then one test (note: not even a test module) in a single Neovim instance: you never know which state that is significant to the test was left from the previous test, additionally state may be just impossible to restore. Even if this makes tests an order of magnitude faster this is not worth the chance that dirty state will make the tests incorrect (also Neovim test suite does not require “sleep to avoid the xterm title being messed up”). > > -- > `The Guide says there is an art to flying,' said Ford, `or at least a > knack. The knack lies in learning how to throw yourself at the ground > and miss.' He smiled weakly. > -- Douglas Adams, "The Hitchhiker's Guide to the Galaxy" > > /// Bram Moolenaar -- [email protected] -- http://www.Moolenaar.net \\\ > /// sponsor Vim, vote for features -- http://www.Vim.org/sponsor/ \\\ > \\\ an exciting new programming language -- http://www.Zimbu.org /// > \\\ help me help AIDS victims -- http://ICCF-Holland.org /// > > — > You are receiving this because you are subscribed to this thread. > Reply to this email directly, view it on GitHub > > -- > -- > You received this message from the "vim_dev" maillist. > Do not top-post! Type your reply below the text you are replying to. > For more information, visit http://www.vim.org/maillist.php > > --- > You received this message because you are subscribed to the Google Groups > "vim_dev" group. > To unsubscribe from this group and stop receiving emails from it, send an > email to [email protected]. > For more options, visit https://groups.google.com/d/optout. — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub -- -- You received this message from the "vim_dev" maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit http://www.vim.org/maillist.php --- You received this message because you are subscribed to the Google Groups "vim_dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout. sadar anul 1944 s-a profilat ca hotarator pentru destinul Romaniei pe fondul cresterii nemultumirii populare al aruncarii Romaniei in sfera de influenta sovietica al intensificarii contactelor diplomatice pentru identificarea unor cai sigure de iesire din conflictul mondial. -- -- You received this message from the "vim_dev" maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit http://www.vim.org/maillist.php --- You received this message because you are subscribed to the Google Groups "vim_dev" group. To unsubscribe from this group and stop receiving emails from it, send an email to [email protected]. For more options, visit https://groups.google.com/d/optout.
