Bram,
This was a faster turn around than I expected, thanks! Unfortunately there's
another issue that comes from this: undo items are overwritten. If you place
text on the line, then press `<esc>`, the undo entry should be created for the
text on that line (e.g. "aaa"). If you jump to the end of the line and insert
with `A`, trigger the completion, then select an item in the menu, the text on
the line is completely replaced with the selected menu entry (because
complete() starts at column 1). This causes an undo to erase the line instead
of reverting to "aaa".
Here's the steps to make the problem clearer:
1. Enter on first line: "aaa", press `<esc>o`
2. Enter text on second line: "aaa", press `<esc>a<right><c-n><c-y><esc>o`
3. Enter text on third line: "aaa"
4. Press `u` twice.
On the last step, I would expect there to be two lines with "aaa" on them, but
instead there's only one.
On Sunday, October 16, 2016 at 9:36:30 AM UTC-4, Bram Moolenaar wrote:
> Patch 8.0.0041
> Problem: When using Insert mode completion but not actually inserting
> anything an undo item is still created. (Tommy Allen)
> Solution: Do not call stop_arrow() when not inserting anything.
> Files: src/edit.c, src/testdir/test_popup.vim
>
>
> *** ../vim-8.0.0040/src/edit.c 2016-10-15 17:06:42.086912756 +0200
> --- src/edit.c 2016-10-16 15:34:26.985734970 +0200
> ***************
> *** 2799,2807 ****
> ins_compl_prep(' ');
> ins_compl_clear();
>
> - if (stop_arrow() == FAIL)
> - return;
> -
> compl_direction = FORWARD;
> if (startcol > curwin->w_cursor.col)
> startcol = curwin->w_cursor.col;
> --- 2799,2804 ----
> ***************
> *** 3876,3882 ****
> /* put the cursor on the last char, for 'tw' formatting */
> if (prev_col > 0)
> dec_cursor();
> ! if (stop_arrow() == OK)
> insertchar(NUL, 0, -1);
> if (prev_col > 0
> && ml_get_curline()[curwin->w_cursor.col] != NUL)
> --- 3873,3880 ----
> /* put the cursor on the last char, for 'tw' formatting */
> if (prev_col > 0)
> dec_cursor();
> ! /* only format when something was inserted */
> ! if (!arrow_used && !ins_need_undo)
> insertchar(NUL, 0, -1);
> if (prev_col > 0
> && ml_get_curline()[curwin->w_cursor.col] != NUL)
> *** ../vim-8.0.0040/src/testdir/test_popup.vim 2016-10-15
> 17:06:42.094912699 +0200
> --- src/testdir/test_popup.vim 2016-10-16 15:33:41.586069579 +0200
> ***************
> *** 378,384 ****
> endif
> endfunc
>
> ! :"Test that 'completefunc' works when it's OK.
> func Test_omnifunc_with_check()
> new
> setlocal omnifunc=DummyCompleteFour
> --- 378,384 ----
> endif
> endfunc
>
> ! " Test that 'completefunc' works when it's OK.
> func Test_omnifunc_with_check()
> new
> setlocal omnifunc=DummyCompleteFour
> ***************
> *** 400,403 ****
> --- 400,429 ----
> q!
> endfunc
>
> + function UndoComplete()
> + call complete(1, ['January', 'February', 'March',
> + \ 'April', 'May', 'June', 'July', 'August', 'September',
> + \ 'October', 'November', 'December'])
> + return ''
> + endfunc
> +
> + " Test that no undo item is created when no completion is inserted
> + func Test_complete_no_undo()
> + set completeopt=menu,preview,noinsert,noselect
> + inoremap <Right> <C-R>=UndoComplete()<CR>
> + new
> + call feedkeys("ixxx\<CR>\<CR>yyy\<Esc>k", 'xt')
> + call feedkeys("iaaa\<Esc>0", 'xt')
> + call assert_equal('aaa', getline(2))
> + call feedkeys("i\<Right>\<Esc>", 'xt')
> + call assert_equal('aaa', getline(2))
> + call feedkeys("u", 'xt')
> + call assert_equal('', getline(2))
> +
> + iunmap <Right>
> + set completeopt&
> + q!
> + endfunc
> +
> +
> " vim: shiftwidth=2 sts=2 expandtab
> *** ../vim-8.0.0040/src/version.c 2016-10-16 14:35:44.547696415 +0200
> --- src/version.c 2016-10-16 15:35:09.805419375 +0200
> ***************
> *** 766,767 ****
> --- 766,769 ----
> { /* Add new patch number below this line */
> + /**/
> + 41,
> /**/
>
> --
> Sometimes I think the surest sign that intelligent life exists elsewhere
> in the universe is that none of it has tried to contact us. (Calvin)
>
> /// Bram Moolenaar -- b...@moolenaar.net -- 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 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 vim_dev+unsubscr...@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.
