Patch 8.1.1737
Bram Moolenaar
Patch 8.1.1737
Problem: :args command that outputs one line gives more prompt.
Solution: Only output line break if needed. (Daniel Hahler, closes #4715)
Files: src/version.c, src/testdir/test_arglist.vim
*** ../vim-8.1.1736/src/version.c 2019-07-23 22:15:21.311518857 +0200
--- src/version.c 2019-07-23 22:55:05.615818414 +0200
***************
*** 4351,4356 ****
--- 4353,4359 ----
int i;
int ncol;
int nrow;
+ int cur_row = 1;
int item_count = 0;
int width = 0;
#ifdef FEAT_SYN_HL
***************
*** 4381,4392 ****
return;
}
! /* The rightmost column doesn't need a separator.
! * Sacrifice it to fit in one more column if possible. */
ncol = (int) (Columns + 1) / width;
nrow = item_count / ncol + (item_count % ncol ? 1 : 0);
! /* i counts columns then rows. idx counts rows then columns. */
for (i = 0; !got_int && i < nrow * ncol; ++i)
{
int idx = (i / ncol) + (i % ncol) * nrow;
--- 4384,4395 ----
return;
}
! // The rightmost column doesn't need a separator.
! // Sacrifice it to fit in one more column if possible.
ncol = (int) (Columns + 1) / width;
nrow = item_count / ncol + (item_count % ncol ? 1 : 0);
! // "i" counts columns then rows. idx counts rows then columns.
for (i = 0; !got_int && i < nrow * ncol; ++i)
{
int idx = (i / ncol) + (i % ncol) * nrow;
***************
*** 4407,4414 ****
msg_putchar(']');
if (last_col)
{
! if (msg_col > 0)
msg_putchar('\n');
}
else
{
--- 4410,4418 ----
msg_putchar(']');
if (last_col)
{
! if (msg_col > 0 && cur_row < nrow)
msg_putchar('\n');
+ ++cur_row;
}
else
{
***************
*** 4416,4426 ****
msg_putchar(' ');
}
}
- else
- {
- if (msg_col > 0)
- msg_putchar('\n');
- }
}
}
--- 4420,4425 ----
*** ../vim-8.1.1736/src/testdir/test_arglist.vim 2018-12-28 19:29:31.743633958 +0100
--- src/testdir/test_arglist.vim 2019-07-23 22:52:04.144231816 +0200
***************
*** 140,149 ****
call assert_equal(['d', 'c', 'b', 'a', 'c'], g:buffers)
! redir => result
! args
! redir END
! call assert_equal('a b [c] d', trim(result))
.argd
call assert_equal(['a', 'b', 'd'], argv())
--- 140,146 ----
call assert_equal(['d', 'c', 'b', 'a', 'c'], g:buffers)
! call assert_equal("\na b [c] d ", execute(':args'))
.argd
call assert_equal(['a', 'b', 'd'], argv())
*** ../vim-8.1.1736/src/version.c 2019-07-23 22:15:21.311518857 +0200
--- src/version.c 2019-07-23 22:55:05.615818414 +0200
***************
*** 779,780 ****
--- 779,782 ----
{ /* Add new patch number below this line */
+ /**/
+ 1737,
/**/
--
Clothes make the man. Naked people have little or no influence on society.
-- Mark Twain (Samuel Clemens) (1835-1910)
/// Bram Moolenaar -- Br...@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 ///
Tom M
> Patch 8.1.1737
> Problem: :args command that outputs one line gives more prompt.
> Solution: Only output line break if needed. (Daniel Hahler, closes #4715)
> Files: src/version.c, src/testdir/test_arglist.vim
>
>
This change introduced a bug. Start executing the following (after starting "vim --clean"):
:let width = &columns / 3 - 4
:let farg = repeat('f', width)
:let garg = repeat('g', width)
:exe 'args a b c d e' farg garg
:args
The arguments will be printed on 2 lines with the 2nd line too long so it'll be wrapped:
[a] d gggggggggggggggggggggg
b e c fffffffffff
fffffffffff
The proper way to display the args (the way previous vim versions did) is to print them on 3 lines:
[a] d gggggggggggggggggggggg
b e
c ffffffffffffffffffffff
The problem lies in this part:
> - else
> - {
> - if (msg_col > 0)
> - msg_putchar('\n');
> - }
> }
> }
I suppose this was axed to avoid sending a '\n' at the end of the last line. But it skips '\n' even for the second to last line - after printing args "b" and "e". Or even more often. (In fact any time when idx >= item_count.) So I suggest introducing the "else" branch back, like in the following patch:
@@ -4438,6 +4438,12 @@ list_in_columns(char_u **items, int size, int current)
msg_putchar(' ');
}
}
+ else
+ {
+ if (msg_col > 0 && cur_row < nrow)
+ msg_putchar('\n');
+ ++cur_row;
+ }
}
}
This issue is related to other 2 topics - to the ones which toothpik mentioned with subjects "version display horked" and "script stopped working". In the latter, he mentioned:
> i still maintain the --version output is horked
After the fix, the number of columns with features will be uniform once again. It might even help to get rid of the hang. Though I am not sure I run vim in the same way as toothpik as this is depends on &columns.
Tom M
Bram Moolenaar
test, but I don't want to delay sending out the fix.
> This issue is related to other 2 topics - to the ones which toothpik
> mentioned with subjects "version display horked" and "script stopped
> working". In the latter, he mentioned:
>
> > i still maintain the --version output is horked
>
> After the fix, the number of columns with features will be uniform
> once again. It might even help to get rid of the hang. Though I am not
> sure I run vim in the same way as toothpik as this is depends on
> &columns.
DENNIS: Oh, very nice. King, eh! I expect you've got a palace and fine
clothes and courtiers and plenty of food. And how d'you get that? By
exploiting the workers! By hanging on to outdated imperialist dogma
which perpetuates the social and economic differences in our society!
"Monty Python and the Holy Grail" PYTHON (MONTY) PICTURES LTD