Jumplist oddity

Given two files A, and B, with the following jump lists: A: 3: Text A2 2: Text A1 1: > B: 3: Text in B2 2: Text in B1 1: > Starting from B, how would I ever be able to reach position "Text in B1" using ? And - as this is a rhetorical question - how could we improve this

Re: Setting the language

@Ken Thanks for the reply. set encoding=utf-8 is my default setting... However, in Windows: - how can I influence the values of v:lc_time, v:lang, and v:ctype? - how can I influence the way the internal sort command sorts (in Linux - having set $LC_ALL to "de_DE.UTF-8") sorting works with

Re: Setting the language

@Christian, Thanks for your reply. Is there any reason why anyone would not want UTF8? (just joking...) That's exactly the problem, I can't set the language to UTF-8 for the various types (messages, ctype, time) in Windows. language english_United States.1252 works (but is not what I want).

Setting the language

How would I set the language to en_US.UTF-8 via the "language" command in Windows gvim? -- -- You received this message from the "vim_use" 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

errorformat for make.gcc multiline output

I'd like to match the following output from gcc via make: ric.c:1703:18: error: line one line two ^~ line three (ignore) line four (ignore) Beside the standard file name, error type, line, and column information, the message

Re: buttype == quickfix plus WinEnter

@Ben Fritz Thanks for the answer. Using in place of - according to my findings - makes no difference. The *type is not available when entering the quickfix buffer for the first time. Shouldn't we consider this a bug and move the thread to vim_dev? -- -- You received this message from the

Re: buttype == quickfix plus WinEnter

@Tony Thanks for your reply. I already tried that first proposal of your's - to no avail... Tested the second one - with the same result... -- -- You received this message from the "vim_use" maillist. Do not top-post! Type your reply below the text you are replying to. For more information,

buttype == quickfix plus WinEnter

I would like to set the (unfortunately global) scrolloff option dynamically. To accomplish that, I use the following construct (simplified): autocmd BufEnter * call SetScrollOff() function SetScrollOff() let = == "quickfix" ? 3 : 5 endfunc This approach works fine for non-quickfix windows.

"<i{" - when issued from a function - is not repeatable via "."

Contrary to e.g. ":normal! gvim in.c :so in.vim /xxx iyy :normal!

Re: Vim as external editor for thunderbird

@Guido Milanese (wow, that's an italian name for a change ;-) 1) Just to mention it: there are four variables that can be configured in the advanced settings of Thunderbird (view_source.editor) which have not been working for me since Thunderbird 3.x (if I recall it correctly). As far as I

Re: Clear input prompt after input is entered?

I found this a (one possible) solution: let in = input("...") normal : -- -- You received this message from the "vim_use" 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

Re: Clear input prompt after input is entered?

Is there a way the is not so "invasive"? Redrawing the whole screen, clearing it first, seems to be a little bit much of overhead for just resetting the command line... -- -- You received this message from the "vim_use" maillist. Do not top-post! Type your reply below the text you are

Selecting multiple tags

Is there a way (w/o plugins or VIMLing) to select two successive tags (no children => siblings) analogously - but not similar to - using e.g. v2at or d2at? I'm not asking for a macro but for a native vim command (haven't found one...). Sample ('*' denotes the cursor position; the area between

Re: Using g://w to filter a buffer

Thanks for the answers. @DrChip: Hey, you know what I want ;-) -- -- You received this message from the "vim_use" 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

Using g://w to filter a buffer

Given the following description in the docs "The global commands work by first scanning through the [range] lines and marking each line where a match occurs (for a multi-line pattern, only the start of the match matters). In a second scan the [cmd] is executed for each marked line with its line

Re: Search for a mach in the last line

@ZyX O hell... I had seen this before but didn't make the connection. Sometimes you don't see the wood for the trees. Thanks a lot! -- -- You received this message from the "vim_use" maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Search for a mach in the last line

When I want to match a pattern in the first line of a document, I use "\%1l". However, no pattern for the last line is provided in the docs (my - somewhat naive - attempts using "\%$l", or "\%l" failed). Did I miss something? -- -- You received this message from the "vim_use" maillist.

Re: [Q] Confine CTRL-o to the current file

Hello Christian, thanks for the info which (unfortunately) implies that there's no native solution to the problem... Greetings -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

[Q] Confine CTRL-o to the current file

Is it possible to confine CTRL-o (jumping backwards in the jump list) to the current file (i.e. not having the command jump to another file in the buffer list)? -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For

Re: Capturing setting ro/noro

On Thursday, February 26, 2015 at 9:11:35 AM UTC+1, Axel Bender wrote: I had to dismiss a solution based on CursorHold as it isn't immediate. Sadly there is no ReadonlyChanged event... Where in the sources would I start btw. if I wanted to implement a new event? -- -- You received

Re: Capturing setting ro/noro

I had to dismiss a solution based on CursorHold as it isn't immediate. Sadly there is no ReadonlyChanged event... -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: Capturing setting ro/noro

@Christian Correct me, if I get sth wrong here, but autocommands are based on events? How would I implement e.g. a ReadonlyChanged with other events? -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more

Re: Capturing setting ro/noro

I'd like to set the editor's background whenever I switch to/read a RO file, or when I set the readonly option. Right now I'm abusing statusline for my needs; it works pretty good but it is also pretty dirty... -- -- You received this message from the vim_use maillist. Do not top-post! Type

Capturing setting ro/noro

Is there a way to capture the setting of readonly? I.e. I'd like to be able to inject a function when the readonly setting is changed. -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Use calculations on sub-expressions

Given the following file: abc3 abc4 bcd1 bcd3 I'd like to find the line containing bcd1 having a gap (bcd2) between it and the next line. Searching like this: /^\(...\)\(\d\)\n\1\=\2+2 /^\(...\)\(\d\)\n\1\=submatch(2)+2 /^\(...\)\(\d\)\n\1\=eval(submatch(2)+2) seemingly doesn't work. I

Separate buffer lists for split windows

I'm looking for a possibility to have different split windows (say two) have different buffer lists (via plugin or settings). E.g. having the following screen layout --- |||A: file01, file02 | Win1 | Win2 |B: file03, file04, file05 |||

Re: Separate buffer lists for split windows

Hm, nice try that one ;-) However not exactly what was I was hoping to hear from you... -- -- You received this message from the vim_use 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

Re: polish characters

Czesc Mikolaj, no problems here (see attached pics); I'm on Windows 7 (64 bit), and use DejaVuSansMono.ttf from SourceForge. Encoding is set to UTF-8. Can you provide a sample - non working - text? -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply

Re: Determine the position we left off before moving to the command line

Hm, somehow my last (yesterday evening) post got removed... Anyway. I attached a VimL file which shows the problem (at least here: version 7.3.903, Windows 7, 64 bit). Please try to execute :Hi w/o a visual block, the cursor being in line 2. -- -- You received this message from the vim_use

Re: Determine the position we left off before moving to the command line

Thanks Christian, that did it! -- -- You received this message from the vim_use 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

Determine the position we left off before moving to the command line

I'm searching a way to determine the last cursor position before entering the command line to execute a command defined with :command -range=% aaa call line1,line2func(). If I start from a visual block, I'm able to retrieve that position with line(.) or a:firstline in func(). However, if I

\ze position

Is there any *native* (i.e. not programmed) way to access the column position of the last match's \zs or \ze? -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: \ze position

@ben Sorry, I was too unspecific. I want to use the \z[es] from the last match() to prevent searching for the same - complicated - expression two times (once with match() the second time with matchend()...). matchlist() might come to the rescue, but I expect that having access to these column

Re: \ze position

Thanks for the feedback! I had hoped that - like in Perl - there is a [(possibly) undocumented] feature or a trick that lets you extract the start and the end positions (columns) after a *single* call to match(), using both \zs and \ze. I would prefer such a solution over having to determine

Re: Is there a way to see autocommand events fire?

@all Not seeing my own post either... As for the problem: I know this classic ;.-) approach. I was hoping that there's some switch around that I had missed. Thanks! -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying

Is there a way to see autocommand events fire?

I'm looking for a way to see the order in which autocommand events fire. Setting verbose doesn't seem to do... -- -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: How to determine the start of a visual block (C-V) when it starts in the void?

Hi John, unfortunately this approach doesn't work in my situation. I'm using the same function in a mapping operating either on visuals, or on motions (:map-operator). Also, I think that if a visual block is displayed (set virtualedit=block, upper left corner in the void), then when switching

Re: How to determine the start of a visual block (C-V) when it starts in the void?

Hi Ben, try this: xnoremap silent _C :C-Ucall SomeFunc()CRMy current mapping xnoremap silentexpr_D SomeFunc()CR John's proposal xnoremap silent _E SomeFunc()CR same w/o expr fun! SomeFunc() let col = virtcol(') - 1 let ecol = virtcol(')

Re: How to determine the start of a visual block (C-V) when it starts in the void?

Please try it again w/o commenting out the stuff. -- -- You received this message from the vim_use 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

Re: How to determine the start of a visual block (C-V) when it starts in the void?

Hi Christian, thanks for the answer. I completely go with you concerning expr, but this was not my original problem. In fact, I already have a solution for it (temporarily setting ve to all), which doesn't require me to change any mapping. However the inconsistency remains: when - with

Re: How to determine the start of a visual block (C-V) when it starts in the void?

@John Thanks for the answer. You're right, I've actually set 'virtualedit' to 'block'. Nevertheless I would have assumed (hoped), that - whatever mode I'm in - 'block', or 'all' - virtcol(') would reflect the actual start column. One can - of course - argue in favor or against that, but I

Re: How to determine the start of a visual block (C-V) when it starts in the void?

Addendum: It's obviously sufficient to set virtualedit to 'block' just before retrieving from virtcol(), and resetting it thereafter. I'm using this in a plugin mapping that operates on a visual area or on a motion. -- -- You received this message from the vim_use maillist. Do not top-post!

How to determine the start of a visual block (C-V) when it starts in the void?

Vim 7.3.785@Win7-64 Please, consider the following content of a buffer: $ ...p$ Then start a visual block (C-V) at the left angle of p. Extend the block to the row above. Then use :C-Uecho col(') (or virtcol('), or getpos('), or v, or .) to determine the start column of the visual block. All

Re: How to determine the start of a visual block (C-V) when it starts in the void?

virtcol(...) doesn't work in the example given. Please note that the end/start - depending from where you start to select - of the block is in the void, i.e. after the line's end (as indicated by $). Here even virtcol(...) returns 1. -- -- You received this message from the vim_use maillist.

Re: Strange behavior using caret or backtick

@Ben Yes, that's the exact bahavior, and yes I used the switches you specified. Also, the described behavior is not appearing in 7.3.046. -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: Strange behavior using caret or backtick

Sorry for opening another thread. After having been away for a week, I was not aware of having posted this before (also had not found it in a quick search). I think, the thread should continue here. @Tony yes, I'm using the German keyboard using ^ as a combining key. The exact order of input

dab on escaped backslashes

Is there any possibility to make dib (or dab for that matter) ignore escaped backslashes, e.g. Current behavior: ( \(a[cursor_here]bc) ) - dib - () ) Desired behavior: ( \(a[cursor_here]bc) ) - dib - ( () ) -- You received this message from the vim_use maillist. Do not top-post! Type your

Re: dab on escaped backslashes

Addendum: I'm not looking for a macro/function but for a setting. -- You received this message from the vim_use 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

Re: dab on escaped backslashes

I believe to remember that this wasn't so in earlier versions of GVim. Do you by chance know, why this was changed? It - usually - comes relatively unhandy when editing, e.g. Windows file names. -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the

Strange behavior using caret or backtick

Since some patches I experience the following behavior ([] denotes the cursor position, current patch level: 7.2.709): a[a]aa - a aa[]aa - spc aa []aa - ^ aa []aa - spcShould result in aa ^[]aa aa []aa - esc aa[ ]aa - sgA shows 0x20 []$a aa Should result

Re: dab on escaped backslashes

In fact I don't see any influence of cpo+=M either. However I don't know if there should be a change... On the other hand there seems to be no other place at which we might toggle this behavior. -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the

Re: Strange behavior using caret or backtick

Erratum: The current patch level here is 7.3.709 (not 7.2.709). -- You received this message from the vim_use 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

Re: C-indentation not working?

Thanks for pointing me in the right direction, Ben. Sometimes you don't see the forest because of the trees... -- You received this message from the vim_use 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

C-indentation not working?

After starting vim (7.3.608) via gvim -U NONE -u NONE test.c and setting nocompatible cindent indentexpr=cindent autoindent smartindent filetype=c vim wouldn't autoindent a after { in the first line (stays on first column). What am I doing wrong? -- You received this message from the

Determine the buffer that was last closed

Is there a way to determine the number/name of the buffer that was last closed with :bd? -- You received this message from the vim_use 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

Change the command line

Is there a way to change the content of the command line from within a function? -- You received this message from the vim_use 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

Re: Change the command line

Thanks Andy. This one is on me... I want to change the command line in the way - a not existing function - setcmdline() would do. The function would keep running (a kind of command parser if you like) until the user hits e.g. CR. -- You received this message from the vim_use maillist. Do not

Re: Determine the buffer that was last closed

So there's no builtin function that returns that value... Ok, thanks! Greetings Axel -- You received this message from the vim_use 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

German spell correction ignores ß

When using the spell correction feature (set spelllang=de_de spell) a word containing a german ß (0xDF) is displayed incorrectly if it is found to be misspelled: Wir wohnen nicht in der Georgenkirchstraße, sondern in der Hauptstraße. In this sample Georgenkirchstraße is flagged as misspelled

Re: start cmd doesn't work anymore with vim 7.3.446

Same problem here: execute !start regedit (just a sample) works fine with 7.3.441 (last version I updated to); it's no longer working in 7.3.446: 441: !start regedit.exe = regedit starts 446: !start regedit.exe = E371: Command not found -- You received this message from the vim_use maillist.

hg Rollback

Being new to Mercurial a quick question: How would I roll back a source code version e.g. from 7.3.446 to 7.3.441? -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: hg Rollback

Thanks for your help, it worked right out of the box! Is there a way to find out the syntax of the revision number/the available revisions? -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: start cmd doesn't work anymore with vim 7.3.446

Reverting to the versions below 7.3.446, I found that the first version the command is no longer working in, is version 7.3.445. -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: :language command

Thanks for your answer Tony. UTF-8 corresponds (afaik) to CodePage 65001. Setting language to this value (e.g. :language English_United States.65001) is possible, however, what are its effects? I'll try to grok through the sources... -- You received this message from the vim_use maillist. Do

Re: :language command

Hi Tony! Oh, you have the sources of Windows? There was a time when I did have them... But I think - for now - vim's sources will suffice ;-) Greetings -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more

:language command

Why isn't it possible (in Windows?) to use :lanugage to set the language to e.g. ge-GE.UTF-8 (while ge-GE.1252 would work)? How - for that matter - does :language relate to the 'encoding' setting? -- You received this message from the vim_use maillist. Do not top-post! Type your reply below

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

@Andy thanks for the input! However - knowing about escape(cword) - I don't want to use it because it for the reasons given in one of my previous posts (e.g. ~ always returns a value and thus hides the info I need). Thanks again! -- You received this message from the vim_use maillist. Do

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

@Tim The answer I found (cmp. above) is my replacement for escape(cword), which I didn't find helpful in all situations. I wanted to include fewer/more characters in vim's notion of what a cword is, and I wanted to be able to define a left boundary different from the right one. Adding to

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

@Andy Thanks for the answer! The split on '\zs' was new to me. Made some speed tests and found that for lines with more than 180 chars the first solution is faster, and for shorter lines the solution below, which is based on the split() approach. Opted for the new solution as short lines

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

Thanks for the replies! The problem however isn't solved... Having obtained the current byte index using col(.), I can't simply add n to/subtract n from the return value of col(.) to find the beginning of the next/previous multi-byte character (the -1 offset is ok because there's a base

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

Addendum. Heureka, I've found it! The solution is based on a combination of virtcol(.) and byteidx(...): function! CWord(string, class_start, class_end) let l:len = len(substitute(a:string, ., x, g)) let l:pos = virtcol(.) - 1 let l:start = l:pos while l:start = 0

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

Addendum: Is there a UTF-8-safe version? The following doesn't work for a UTF-8 file (both enc and fenc): :echo getline(.)[col(.) ] =~ '\k' when over an ü (i.e. an character that - if in Latin-1 - would be in the default iskeyword option range). Note: The inspected character is the *second*

[Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

Sometimes it's necessary to determine whether a given character belongs to one of the options whose names are reflected by the above (suggested) function names. Though it's possible to write those functions in vim itself, one would either have to call such a function manually a) after a change

Re: [Suggestion] New functions iskeywordchar(), isfnamechar(), isprintchar(), and isidentchar()

@Andy, Tim thanks! Sometimes one doesn't see the forrest for the trees. -- You received this message from the vim_use 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

Re: Vim as default email (thunderbird) editor

Not working here (Windows 2003, Thunderbird 7.0, exteditor 1.0). Same error in JS Console. This seems to be an error in the plugin; even when using notepad as the text editor, it would not get called (verified with procmon). = Faulty plugin; nothing to do with GVim. This returns us to the

Re: Vim as default email (thunderbird) editor

Thanks for your input, Calon. This in fact does the trick (still an error in the plugin); I would never have tried this myself... Have a nice weekend -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more

Re: Assign the result of a COMMAND to a variable

Thanks for your input XyZ! Your improvment works for me if I replace bufexists() with buflisted() (I don't care about hidden buffers). Axel -- You received this message from the vim_use maillist. Do not top-post! Type your reply below the text you are replying to. For more information, visit

Re: Assign the result of a COMMAND to a variable

Thanks for all your input! Used it to create a function that quits vi when the last buffer is closed (didn't find one that worked out of the box with my settings): function! QuitIfLast(wipe) Delete/wipe-out the current buffer if a:wipe == 1 bwipeout! else bdelete endif

Re: Function/variable containing the number of current buffers

Thanks for your help, Tony! Meanwhile I created the missing function with the functionality I need. Search for Assign the result of a COMMAND to a variable if you'd like to comment on it. Best regards Axel -- You received this message from the vim_use maillist. Do not top-post! Type your

Assign the result of a COMMAND to a variable

I wonder if it is possible to assign the result of a command to a variable like so (not working): let [var] = [command] (where [command] stands for an arbitrary command producing output, like e.g. ls]. -- You received this message from the vim_use maillist. Do not top-post! Type your reply