gmane.editors.vim.devel

Re: [patch] v/V with a count

Bram Moolenaar — 2012-05-25T11:12:51

Christian Brabandt wrote:


Thanks, that's useful.  I will change it to take care of 'selectmode'.
And update the docs.

Re: Does it still make sense to have per-file/-type maintainers? [Was: Re: Added support for spell checking in runtime/syntax/ocaml.vim]

Bram Moolenaar — 2012-05-25T11:12:51

Ben Fritz wrote:


There was a specific library that some Linux versions compiled Vim with,
and this library was GPL.  A Vim built that way could not be
distributed, because there is a small incompatibility between GPL and
the Vim license.  To solve that the dual-license method was introduced.
Richard Stallman was involved in updating the license text, thus it
should be OK for everybody.

Patch 7.3.529

Bram Moolenaar — 2012-05-25T11:12:51

Patch 7.3.529
Problem:    Using a count before "v" and "V" does not work (Kikyous)
Solution:   Make the count select that many characters or lines. (Christian
    Brabandt)
Files:    src/normal.c


*** ../vim-7.3.528/src/normal.c2012-05-25 11:04:34.000000000 +0200
--- src/normal.c2012-05-25 13:12:06.000000000 +0200
***************
*** 7660,7672 ****
      else    /* start Visual mode */
      {
  check_visual_highlight();
! if (cap->count0)    /* use previously selected part */
  {
!     if (resel_VIsual_mode == NUL)   /* there is none */
!     {
! beep_flush();
! return;
!     }
      VIsual = curwin->w_cursor;
  
      VIsual_active = TRUE;
--- 7660,7668 ----
      else    /* start Visual mode */
      {
  check_visual_highlight();
! if (cap->count0 > 0 && resel_VIsual_mode != NUL)
  {
!     /* use previously selected part */
      VIsual = curwin->w_cursor;
  
      VIsual_active = TRUE;
***************
*** 7725,7730 ****
--- 7721,7736 ----
  /* start Select mode when 'selectmode' contains "cmd" */
  may_start_select('c');
      n_start_visual_mode(cap->cmdchar);
+     if (VIsual_mode != 'V' && *p_sel == 'e')
+ ++cap->count1;  /* include one more char */
+     if (cap->count0 > 0 && --cap->count1 > 0)
+     {
+ /* With a count select that many characters or lines. */
+ if (VIsual_mode == 'v' || VIsual_mode == Ctrl_V)
+     nv_right(cap);
+ else if (VIsual_mode == 'V')
+     nv_down(cap);
+     }
  }
      }
  }
*** ../vim-7.3.528/src/version.c2012-05-25 12:38:57.000000000 +0200
--- src/version.c2012-05-25 12:59:58.000000000 +0200
***************
*** 716,717 ****
--- 716,719 ----
  {   /* Add new patch number below this line */
+ /**/
+     529,
  /**/

Re: bug related to patch 7.3.449 (Autocommands caused command to abort)

Bram Moolenaar — 2012-05-25T10:39:38

Christian Brabandt wrote:


Great, thanks.

We can make the method static, since it's only used in window.c

Patch 7.3.528

Bram Moolenaar — 2012-05-25T10:39:38

Patch 7.3.528
Problem:    Crash when closing last window in a tab. (Alex Efros)
Solution:   Use common code in close_last_window_tabpage(). (Christian
    Brabandt)
Files:    src/window.c


*** ../vim-7.3.527/src/window.c2012-03-16 19:07:54.000000000 +0100
--- src/window.c2012-05-25 12:25:16.000000000 +0200
***************
*** 23,28 ****
--- 23,29 ----
  static void win_totop __ARGS((int size, int flags));
  static void win_equal_rec __ARGS((win_T *next_curwin, int current, frame_T *topfr, int dir, int col, int row, int width, int height));
  static int last_window __ARGS((void));
+ static int close_last_window_tabpage __ARGS((win_T *win, int free_buf, tabpage_T *prev_curtab));
  static win_T *win_free_mem __ARGS((win_T *win, int *dirp, tabpage_T *tp));
  static frame_T *win_altframe __ARGS((win_T *win, tabpage_T *tp));
  static tabpage_T *alt_tabpage __ARGS((void));
***************
*** 2105,2110 ****
--- 2106,2147 ----
  }
  
  /*
+  * Close the possibly last window in a tab page.
+  * Returns TRUE when the window was closed already.
+  */
+     static int
+ close_last_window_tabpage(win, free_buf, prev_curtab)
+     win_T*win;
+     intfree_buf;
+     tabpage_T   *prev_curtab;
+ {
+     if (firstwin == lastwin)
+     {
+ /*
+  * Closing the last window in a tab page.  First go to another tab
+  * page and then close the window and the tab page.  This avoids that
+  * curwin and curtab are invalid while we are freeing memory, they may
+  * be used in GUI events.
+  */
+ goto_tabpage_tp(alt_tabpage());
+ redraw_tabline = TRUE;
+ 
+ /* Safety check: Autocommands may have closed the window when jumping
+  * to the other tab page. */
+ if (valid_tabpage(prev_curtab) && prev_curtab->tp_firstwin == win)
+ {
+     int    h = tabline_height();
+ 
+     win_close_othertab(win, free_buf, prev_curtab);
+     if (h != tabline_height())
+ shell_new_rows();
+ }
+ return TRUE;
+     }
+     return FALSE;
+ }
+ 
+ /*
   * Close window "win".  Only works for the current tab page.
   * If "free_buf" is TRUE related buffer may be unloaded.
   *
***************
*** 2143,2171 ****
      }
  #endif
  
!     /*
!      * When closing the last window in a tab page first go to another tab
!      * page and then close the window and the tab page.  This avoids that
!      * curwin and curtab are not invalid while we are freeing memory, they may
!      * be used in GUI events.
!      */
!     if (firstwin == lastwin)
!     {
! goto_tabpage_tp(alt_tabpage());
! redraw_tabline = TRUE;
! 
! /* Safety check: Autocommands may have closed the window when jumping
!  * to the other tab page. */
! if (valid_tabpage(prev_curtab) && prev_curtab->tp_firstwin == win)
! {
!     int    h = tabline_height();
! 
!     win_close_othertab(win, free_buf, prev_curtab);
!     if (h != tabline_height())
! shell_new_rows();
! }
! return;
!     }
  
      /* When closing the help window, try restoring a snapshot after closing
       * the window.  Otherwise clear the snapshot, it's now invalid. */
--- 2180,2190 ----
      }
  #endif
  
!     /* When closing the last window in a tab page first go to another tab page
!      * and then close the window and the tab page to avoid that curwin and
!      * curtab are invalid while we are freeing memory. */
!     if (close_last_window_tabpage(win, free_buf, prev_curtab))
!       return;
  
      /* When closing the help window, try restoring a snapshot after closing
       * the window.  Otherwise clear the snapshot, it's now invalid. */
***************
*** 2225,2231 ****
  
      /* Autocommands may have closed the window already, or closed the only
       * other window or moved to another tab page. */
!     if (!win_valid(win) || last_window() || curtab != prev_curtab)
  return;
  
      /* Free the memory used for the window and get the window that received
--- 2244,2251 ----
  
      /* Autocommands may have closed the window already, or closed the only
       * other window or moved to another tab page. */
!     if (!win_valid(win) || last_window() || curtab != prev_curtab
!     || close_last_window_tabpage(win, free_buf, prev_curtab))
  return;
  
      /* Free the memory used for the window and get the window that received
***************
*** 2310,2316 ****
  
  /*
   * Close window "win" in tab page "tp", which is not the current tab page.
!  * This may be the last window ih that tab page and result in closing the tab,
   * thus "tp" may become invalid!
   * Caller must check if buffer is hidden and whether the tabline needs to be
   * updated.
--- 2330,2336 ----
  
  /*
   * Close window "win" in tab page "tp", which is not the current tab page.
!  * This may be the last window in that tab page and result in closing the tab,
   * thus "tp" may become invalid!
   * Caller must check if buffer is hidden and whether the tabline needs to be
   * updated.
*** ../vim-7.3.527/src/version.c2012-05-25 11:56:06.000000000 +0200
--- src/version.c2012-05-25 12:38:25.000000000 +0200
***************
*** 716,717 ****
--- 716,719 ----
  {   /* Add new patch number below this line */
+ /**/
+     528,
  /**/

Patch 7.3.527

Bram Moolenaar — 2012-05-25T09:56:47

Patch 7.3.527
Problem:    Clang complains about non-ASCII characters in a string.
Solution:   Change to \x88 form. (Dominique Pelle)
Files:    src/charset.c


*** ../vim-7.3.526/src/charset.c2012-01-26 13:40:04.000000000 +0100
--- src/charset.c2012-05-25 11:49:58.000000000 +0200
***************
*** 1602,1611 ****
  #define LATIN1LOWER 'l'
  #define LATIN1UPPER 'U'
  
- /*                                                                 !"#$%&'()*+,-./0123456789:;<=>?< at >ABCDEFGHIJKLMNOPQRSTUVWXYZ[\]%_'abcdefghijklmnopqrstuvwxyz{|}~                                  ¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ */
  static char_u latin1flags[257] = "                                                                 UUUUUUUUUUUUUUUUUUUUUUUUUU      llllllllllllllllllllllllll                                                                     UUUUUUUUUUUUUUUUUUUUUUU UUUUUUUllllllllllllllllllllllll llllllll";
! static char_u latin1upper[257] = "                                 !\"#$%&'()*+,-./0123456789:;<=>?< at >ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`ABCDEFGHIJKLMNOPQRSTUVWXYZ{|}~ € ‚ƒ„…†‡ˆ‰Š‹Œ Ž  ‘’“”•–—˜™š›œ žŸ ¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞßÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ÷ØÙÚÛÜÝÞÿ";
! static char_u latin1lower[257] = "                                 !\"#$%&'()*+,-./0123456789:;<=>?< at >abcdefghijklmnopqrstuvwxyz[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~ € ‚ƒ„…†‡ˆ‰Š‹Œ Ž  ‘’“”•–—˜™š›œ žŸ ¡¢£¤¥¦§¨©ª«¬®¯°±²³´µ¶·¸¹º»¼½¾¿àáâãäåæçèéêëìíîïðñòóôõö×øùúûüýþßàáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ";
  
      int
  vim_islower(c)
--- 1602,1610 ----
  #define LATIN1LOWER 'l'
  #define LATIN1UPPER 'U'
  
  static char_u latin1flags[257] = "                                                                 UUUUUUUUUUUUUUUUUUUUUUUUUU      llllllllllllllllllllllllll                                                                     UUUUUUUUUUUUUUUUUUUUUUU UUUUUUUllllllllllllllllllllllll llllllll";
! static char_u latin1upper[257] = "                                 !\"#$%&'()*+,-./0123456789:;<=>?< at >ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_`ABCDEFGHIJKLMNOPQRSTUVWXYZ{|}~\x7f\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xd7\xd8\xd9\xda\xdb\xdc\xdd\xde\xdf\xc0\xc1\xc2\xc3\xc4\xc5\xc6\xc7\xc8\xc9\xca\xcb\xcc\xcd\xce\xcf\xd0\xd1\xd2\xd3\xd4\xd5\xd6\xf7\xd8\xd9\xda\xdb\xdc\xdd\xde\xff";
! static char_u latin1lower[257] = "                                 !\"#$%&'()*+,-./0123456789:;<=>?< at >abcdefghijklmnopqrstuvwxyz[\\]^_`abcdefghijklmnopqrstuvwxyz{|}~\x7f\x80\x81\x82\x83\x84\x85\x86\x87\x88\x89\x8a\x8b\x8c\x8d\x8e\x8f\x90\x91\x92\x93\x94\x95\x96\x97\x98\x99\x9a\x9b\x9c\x9d\x9e\x9f\xa0\xa1\xa2\xa3\xa4\xa5\xa6\xa7\xa8\xa9\xaa\xab\xac\xad\xae\xaf\xb0\xb1\xb2\xb3\xb4\xb5\xb6\xb7\xb8\xb9\xba\xbb\xbc\xbd\xbe\xbf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xd7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xdf\xe0\xe1\xe2\xe3\xe4\xe5\xe6\xe7\xe8\xe9\xea\xeb\xec\xed\xee\xef\xf0\xf1\xf2\xf3\xf4\xf5\xf6\xf7\xf8\xf9\xfa\xfb\xfc\xfd\xfe\xff";
  
      int
  vim_islower(c)
*** ../vim-7.3.526/src/version.c2012-05-25 11:04:34.000000000 +0200
--- src/version.c2012-05-25 11:52:06.000000000 +0200
***************
*** 716,717 ****
--- 716,719 ----
  {   /* Add new patch number below this line */
+ /**/
+     527,
  /**/

Re: [patch] fixed clang-3.1 warnings: src/charset.c uses latin1 char in strings

Bram Moolenaar — 2012-05-25T09:56:47

Dominique Pelle wrote:


Thanks, I'll include it.  Patch applied without a warning, so I assume
there are no encoding problems.  And I also compared the relevant part
of charset.o.

Re: Does it still make sense to have per-file/-type maintainers?

Thilo Six — 2012-05-25T09:09:57

Hello Tony and Ben,


Excerpt from Ben Fritz:

Patch 7.3.526

Bram Moolenaar — 2012-05-25T09:04:50

Patch 7.3.526
Problem:    Confusing indenting for #ifdef.
Solution:   Remove and add indent. (Elias Diem)
Files:    src/normal.c


*** ../vim-7.3.525/src/normal.c2012-04-30 18:18:43.000000000 +0200
--- src/normal.c2012-05-23 20:35:13.000000000 +0200
***************
*** 29,37 ****
  static voidset_vcount_ca __ARGS((cmdarg_T *cap, int *set_prevcount));
  #endif
  static int
! # ifdef __BORLANDC__
! _RTLENTRYF
! # endif
  nv_compare __ARGS((const void *s1, const void *s2));
  static intfind_command __ARGS((int cmdchar));
  static voidop_colon __ARGS((oparg_T *oap));
--- 29,37 ----
  static voidset_vcount_ca __ARGS((cmdarg_T *cap, int *set_prevcount));
  #endif
  static int
! #ifdef __BORLANDC__
!     _RTLENTRYF
! #endif
  nv_compare __ARGS((const void *s1, const void *s2));
  static intfind_command __ARGS((int cmdchar));
  static voidop_colon __ARGS((oparg_T *oap));
*** ../vim-7.3.525/src/version.c2012-05-25 11:02:34.000000000 +0200
--- src/version.c2012-05-25 11:03:37.000000000 +0200
***************
*** 716,717 ****
--- 716,719 ----
  {   /* Add new patch number below this line */
+ /**/
+     526,
  /**/

Patch 7.3.525

Bram Moolenaar — 2012-05-25T09:03:01

Patch 7.3.525
Problem:    Compiler warning on 64 bit MS-Windows.
Solution:   Add type cast. (Mike Williams)
Files:    src/ex_getln.c


*** ../vim-7.3.524/src/ex_getln.c2012-04-30 18:48:38.000000000 +0200
--- src/ex_getln.c2012-05-23 20:33:16.000000000 +0200
***************
*** 5263,5269 ****
  {
      static char_u compl[2] = { NUL, NUL };
      char *short_names = ":=< at >>?/";
!     int short_names_count = STRLEN(short_names);
      int history_name_count = sizeof(history_names) / sizeof(char *) - 1;
  
      if (idx < short_names_count)
--- 5263,5269 ----
  {
      static char_u compl[2] = { NUL, NUL };
      char *short_names = ":=< at >>?/";
!     int short_names_count = (int)STRLEN(short_names);
      int history_name_count = sizeof(history_names) / sizeof(char *) - 1;
  
      if (idx < short_names_count)
*** ../vim-7.3.524/src/version.c2012-05-18 21:53:29.000000000 +0200
--- src/version.c2012-05-25 11:01:51.000000000 +0200
***************
*** 716,717 ****
--- 716,719 ----
  {   /* Add new patch number below this line */
+ /**/
+     525,
  /**/

Re: bug related to patch 7.3.449 (Autocommands caused command to abort)

Christian Brabandt — 2012-05-25T04:34:01

Yes, indeed. Thanks.

regards,
Christian

Re: bug related to patch 7.3.449 (Autocommands caused command to abort)

James McCoy — 2012-05-24T22:20:04

                                ^^^
The "not" should be removed.  You fixed this in the original version of
the comment, but not when you copied it here.

Re: Updated fasm vim syntax file

Charles Campbell — 2012-05-24T20:42:22

Hello!

The fasm.vim file's maintainer is Ron Aaron, whose email you may find in 
the first few lines of the syntax/fasm.vim script. He may or may not pay 
attention to this list. May I suggest that you forward a copy directly 
to him.

Regards,
Charles Campbell

Re: Substitution of metacharacters

Christian Brabandt — 2012-05-24T20:14:35

Hi Bob!

On Do, 24 Mai 2012, Bob von Knobloch wrote:


Not possible, the replace text is escaped:

,----[ gui.c ]-
|5276  ga_concat(&ga, (char_u *)"/");
|5277                                         /* escape / and \ */
|5278  p = vim_strsave_escaped(repl_text, (char_u *)"/\\");
`----

Here is an experimental patch against the gtk gui, that adds an extra 
flag, and allows to replace with special chars ,e.g. \t for Tab

Looking at the dialog, it could possibly also get some more 
possibilities, e.g. a confirm/undo button, perhaps even more.

regards,
Christian

Re: bug related to patch 7.3.449 (Autocommands caused command to abort)

Christian Brabandt — 2012-05-24T20:14:41

Hi Alex!

On Do, 24 Mai 2012, Alex Efros wrote:


I have noticed something similar in my NrrwRgn plugin.
I am not sure, this is a bug, because at the time you issue :q the 
location list ist still open and in case another window is open, :q does 
not exit Vim.


Attached patch fixes this issue.

regards,
Christian

[patch] v/V with a count

Christian Brabandt — 2012-05-24T20:14:23

Bram,
attached patch fixes this issue from the todo list:

,----
| Use a count before "v" and "V" to select that many characters or lines?
| (Kikyous)
`----

regards,
Christian

Updated fasm vim syntax file

Антон Кочков — 2012-05-24T18:42:13

Hello!
I've updated fasm vim syntax file to match fasm-1.70.02 (latest
stable) version syntax and support 64bit, MMX, SSE1-SSE4, AVX-AVX2
instructions.
Hope it can be useful.

Best regards,
Anton Kochkov.

[patch] fixed clang-3.1 warnings: src/charset.c uses latin1 char in strings

Dominique Pellé — 2012-05-24T18:48:24

Hi

Compiling Vim-7.3.524 with clang-3.1 gives these warnings:

charset.c:1607:34: warning: illegal character encoding in string
literal [-Winvalid-source-encoding]
charset.c:1608:34: warning: illegal character encoding in string
literal [-Winvalid-source-encoding]

The charset.c source file uses strings in latin1 encoding.

This seems dangerous since compilers often use utf-8 for
input source file or the locale settings according
to man gcc (see -finput-charset=...).  So I wonder how
portable it is to use latin1 strings in source files.

Attached patch fixes the warning by making charset.c
source use ASCII characters only.

Since the fix can be error prone, I verified with
"objdump -s objects/charset.o" that the generated
object files are identical before and after fix (when
stripping debug info which was different).

Regards

Re: Does it still make sense to have per-file/-type maintainers? [Was: Re: Added support for spell checking in runtime/syntax/ocaml.vim]

Ben Fritz — 2012-05-24T16:59:59

Right, but one that is more relaxed than the Vim license, like the MIT license for example, would not work. The Vim license requires anyone making changes to send those changes to Bram (perhaps if requested, I don't remember off the top of my head). Licensing with the MIT or Berkely-style, this requirement could not be enforced, whereas the GPL requires that the code be made available to anyone, accomplishing the goal of getting the change back to Bram.

Re: Does it still make sense to have per-file/-type maintainers? [Was: Re: Added support for spell checking in runtime/syntax/ocaml.vim]

Thilo Six — 2012-05-24T16:37:33

Hello Tony,


Excerpt from Tony Mechelynck:

bug related to patch 7.3.449 (Autocommands caused command to abort)

Alex Efros — 2012-05-24T10:28:55

Hi!

I've discover this bug when trying to ':wq' in file with syntax errors using
syntastic plugin with
    let g:syntastic_auto_loc_list=1
Instead of exiting from vim no matter there are syntax errors I got either
    E855: Autocommands caused command to abort
or vim segfault.


To reproduce this bug it's enough to:

    $ vi -u /dev/null --noplugin
    :autocmd BufWinLeave * if empty(&bt) | lclose | endif
    :lexpr system('echo :1:some')
    :lopen
    :wincmd p
    :q
    (location list closes; got message E855: Autocommands caused command to
    abort; vim doesn't exit)
    :q
    (now vim exit)

Or, with tabs and segfault:

    $ vi -u /dev/null --noplugin
    :autocmd BufWinLeave * if empty(&bt) | lclose | endif
    :tabnew
    :lexpr system('echo :1:some')
    :lopen
    :wincmd p
    :q
    Vim: Caught deadly signal SEGV
    Vim: Finished.
    Segmentation fault

This happens on 7.3.515. Versions before 7.3.449 behave differently
without tabs, but with tabs they all crash.


There is probably bug in that autocmd -
    &bt
should be replaced with
    getbufvar(0+expand('<abuf>'), '&bt')
but it fix this issue only partially, in these cases:
- :lclose in location-list window
- :lclose in file window (with opened location-list)
- :q in location-list window
but we still got error (or segfault when tabs used) in this case:
- :q in file window (with opened location-list)

In last case expand('<abuf>') return file window's number, which is
correct IMO. In this situation doing :lclose looks correct and shouldn't
result in 'Autocommands caused command to abort'.


I've implemented workaround for syntastic:
https://github.com/scrooloose/syntastic/pull/256
but it's bug in vim and should be fixed in vim.