Builtin

Nvim :help pages, generated from source using the tree-sitter-vimdoc parser.


Builtin functions builtin-functions
For functions grouped by what they are used for see function-list.
abs({expr}) abs() Return the absolute value of {expr}. When {expr} evaluates to a Float abs() returns a Float. When {expr} can be converted to a Number abs() returns a Number. Otherwise abs() gives an error message and returns -1. Examples:
echo abs(1.456)
1.456
echo abs(-5.456)
5.456
echo abs(-4)
4
acos({expr}) acos() Return the arc cosine of {expr} measured in radians, as a Float in the range of [0, pi]. {expr} must evaluate to a Float or a Number in the range [-1, 1]. Returns NaN if {expr} is outside the range [-1, 1]. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo acos(0)
1.570796
echo acos(-0.5)
2.094395
add({object}, {expr}) add() Append the item {expr} to List or Blob {object}. Returns the resulting List or Blob. Examples:
let alist = add([1, 2, 3], item)
call add(mylist, "woodstock")
Note that when {expr} is a List it is appended as a single item. Use extend() to concatenate Lists. When {object} is a Blob then {expr} must be a number. Use insert() to add an item at another position. Returns 1 if {object} is not a List or a Blob.
and({expr}, {expr}) and() Bitwise AND on the two arguments. The arguments are converted to a number. A List, Dict or Float argument causes an error. Also see or() and xor(). Example:
let flag = and(bits, 0x80)
api_info() api_info() Returns Dictionary of api-metadata.
View it in a nice human-readable format:
lua vim.print(vim.fn.api_info())
append({lnum}, {text}) append() When {text} is a List: Append each item of the List as a text line below line {lnum} in the current buffer. Otherwise append {text} as one text line below line {lnum} in the current buffer. Any type of item is accepted and converted to a String. {lnum} can be zero to insert a line before the first one. {lnum} is used like with getline(). Returns 1 for failure ({lnum} out of range or out of memory), 0 for success. When {text} is an empty list zero is returned, no matter the value of {lnum}. Example:
let failed = append(line('$'), "# THE END")
let failed = append(0, ["Chapter 1", "the beginning"])
appendbufline({buf}, {lnum}, {text}) appendbufline() Like append() but append the text in buffer {expr}.
This function works only for loaded buffers. First call bufload() if needed.
For the use of {buf}, see bufname().
{lnum} is the line number to append below. Note that using line() would use the current buffer, not the one appending to. Use "$" to append at the end of the buffer. Other string values are not supported.
On success 0 is returned, on failure 1 is returned.
If {buf} is not a valid buffer or {lnum} is not valid, an error message is given. Example:
let failed = appendbufline(13, 0, "# THE START")
However, when {text} is an empty list then no error is given for an invalid {lnum}, since {lnum} isn't actually used.
argc([{winid}]) argc() The result is the number of files in the argument list. See arglist. If {winid} is not supplied, the argument list of the current window is used. If {winid} is -1, the global argument list is used. Otherwise {winid} specifies the window of which the argument list is used: either the window number or the window ID. Returns -1 if the {winid} argument is invalid.
argidx() argidx() The result is the current index in the argument list. 0 is the first file. argc() - 1 is the last one. See arglist.
arglistid([{winnr} [, {tabnr}]]) arglistid() Return the argument list ID. This is a number which identifies the argument list being used. Zero is used for the global argument list. See arglist. Returns -1 if the arguments are invalid.
Without arguments use the current window. With {winnr} only use this window in the current tab page. With {winnr} and {tabnr} use the window in the specified tab page. {winnr} can be the window number or the window-ID.
argv([{nr} [, {winid}]]) argv() The result is the {nr}th file in the argument list. See arglist. "argv(0)" is the first one. Example:
let i = 0
while i < argc()
  let f = escape(fnameescape(argv(i)), '.')
  exe 'amenu Arg.' .. f .. ' :e ' .. f .. '<CR>'
  let i = i + 1
endwhile
Without the {nr} argument, or when {nr} is -1, a List with the whole arglist is returned.
The {winid} argument specifies the window ID, see argc(). For the Vim command line arguments see v:argv.
Returns an empty string if {nr}th argument is not present in the argument list. Returns an empty List if the {winid} argument is invalid.
asin({expr}) asin() Return the arc sine of {expr} measured in radians, as a Float in the range of [-pi/2, pi/2]. {expr} must evaluate to a Float or a Number in the range [-1, 1]. Returns NaN if {expr} is outside the range [-1, 1]. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo asin(0.8)
0.927295
echo asin(-0.5)
-0.523599
assert_beeps({cmd}) assert_beeps() Run {cmd} and add an error message to v:errors if it does NOT produce a beep or visual bell. Also see assert_fails(), assert_nobeep() and assert-return.
assert_equal({expected}, {actual} [, {msg}]) assert_equal() When {expected} and {actual} are not equal an error message is added to v:errors and 1 is returned. Otherwise zero is returned. assert-return The error is in the form "Expected {expected} but got {actual}". When {msg} is present it is prefixed to that.
There is no automatic conversion, the String "4" is different from the Number 4. And the number 4 is different from the Float 4.0. The value of 'ignorecase' is not used here, case always matters. Example:
assert_equal('foo', 'bar')
Will result in a string to be added to v:errors:
test.vim line 12: Expectedfoo but gotbar
assert_equalfile({fname-one}, {fname-two}) assert_equalfile() When the files {fname-one} and {fname-two} do not contain exactly the same text an error message is added to v:errors. Also see assert-return. When {fname-one} or {fname-two} does not exist the error will mention that.
assert_exception({error} [, {msg}]) assert_exception() When v:exception does not contain the string {error} an error message is added to v:errors. Also see assert-return. This can be used to assert that a command throws an exception. Using the error number, followed by a colon, avoids problems with translations:
try
  commandthatfails
  call assert_false(1, 'command should have failed')
catch
  call assert_exception('E492:')
endtry
assert_fails() assert_fails({cmd} [, {error} [, {msg} [, {lnum} [, {context}]]]]) Run {cmd} and add an error message to v:errors if it does NOT produce an error or when {error} is not found in the error message. Also see assert-return.
When {error} is a string it must be found literally in the first reported error. Most often this will be the error code, including the colon, e.g. "E123:".
assert_fails('bad cmd', 'E987:')
When {error} is a List with one or two strings, these are used as patterns. The first pattern is matched against the first reported error:
assert_fails('cmd', ['E987:.*expected bool'])
The second pattern, if present, is matched against the last reported error. To only match the last error use an empty string for the first error:
assert_fails('cmd', ['', 'E987:'])
If {msg} is empty then it is not used. Do this to get the default message when passing the {lnum} argument.
When {lnum} is present and not negative, and the {error} argument is present and matches, then this is compared with the line number at which the error was reported. That can be the line number in a function or in a script.
When {context} is present it is used as a pattern and matched against the context (script name or function name) where {lnum} is located in.
Note that beeping is not considered an error, and some failing commands only beep. Use assert_beeps() for those.
assert_false({actual} [, {msg}]) assert_false() When {actual} is not false an error message is added to v:errors, like with assert_equal(). The error is in the form "Expected False but got {actual}". When {msg} is present it is prepended to that. Also see assert-return.
A value is false when it is zero. When {actual} is not a number the assert fails.
assert_inrange({lower}, {upper}, {actual} [, {msg}]) assert_inrange() This asserts number and Float values. When {actual} is lower than {lower} or higher than {upper} an error message is added to v:errors. Also see assert-return. The error is in the form "Expected range {lower} - {upper}, but got {actual}". When {msg} is present it is prefixed to that.
assert_match({pattern}, {actual} [, {msg}]) assert_match() When {pattern} does not match {actual} an error message is added to v:errors. Also see assert-return. The error is in the form "Pattern {pattern} does not match {actual}". When {msg} is present it is prefixed to that.
{pattern} is used as with expr-=~: The matching is always done like 'magic' was set and 'cpoptions' is empty, no matter what the actual value of 'magic' or 'cpoptions' is.
{actual} is used as a string, automatic conversion applies. Use "^" and "$" to match with the start and end of the text. Use both to match the whole text.
Example:
assert_match('^f.*o$', 'foobar')
Will result in a string to be added to v:errors:
test.vim line 12: Pattern '^f.*o$' does not matchfoobar
assert_nobeep({cmd}) assert_nobeep() Run {cmd} and add an error message to v:errors if it produces a beep or visual bell. Also see assert_beeps().
assert_notequal({expected}, {actual} [, {msg}]) assert_notequal() The opposite of assert_equal(): add an error message to v:errors when {expected} and {actual} are equal. Also see assert-return.
assert_notmatch({pattern}, {actual} [, {msg}]) assert_notmatch() The opposite of assert_match(): add an error message to v:errors when {pattern} matches {actual}. Also see assert-return.
assert_report({msg}) assert_report() Report a test failure directly, using String {msg}. Always returns one.
assert_true({actual} [, {msg}]) assert_true() When {actual} is not true an error message is added to v:errors, like with assert_equal(). Also see assert-return. A value is TRUE when it is a non-zero number or v:true. When {actual} is not a number or v:true the assert fails. When {msg} is given it precedes the default message.
atan({expr}) atan() Return the principal value of the arc tangent of {expr}, in the range [-pi/2, +pi/2] radians, as a Float. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo atan(100)
1.560797
echo atan(-4.01)
-1.326405
atan2({expr1}, {expr2}) atan2() Return the arc tangent of {expr1} / {expr2}, measured in radians, as a Float in the range [-pi, pi]. {expr1} and {expr2} must evaluate to a Float or a Number. Returns 0.0 if {expr1} or {expr2} is not a Float or a Number. Examples:
echo atan2(-1, 1)
-0.785398
echo atan2(1, -1)
2.356194
blob2list({blob}) blob2list() Return a List containing the number value of each byte in Blob {blob}. Examples:
blob2list(0z0102.0304)        " returns [1, 2, 3, 4]
blob2list(0z)                " returns []
Returns an empty List on error. list2blob() does the opposite.
browse({save}, {title}, {initdir}, {default}) browse() Put up a file requester. This only works when "has("browse")" returns TRUE (only in some GUI versions). The input fields are: {save} when TRUE, select file to write {title} title for the requester {initdir} directory to start browsing in {default} default file name An empty string is returned when the "Cancel" button is hit, something went wrong, or browsing is not possible.
browsedir({title}, {initdir}) browsedir() Put up a directory requester. This only works when "has("browse")" returns TRUE (only in some GUI versions). On systems where a directory browser is not supported a file browser is used. In that case: select a file in the directory to be used. The input fields are: {title} title for the requester {initdir} directory to start browsing in When the "Cancel" button is hit, something went wrong, or browsing is not possible, an empty string is returned.
bufadd({name}) bufadd() Add a buffer to the buffer list with name {name} (must be a String). If a buffer for file {name} already exists, return that buffer number. Otherwise return the buffer number of the newly created buffer. When {name} is an empty string then a new buffer is always created. The buffer will not have 'buflisted' set and not be loaded yet. To add some text to the buffer use this:
let bufnr = bufadd('someName')
call bufload(bufnr)
call setbufline(bufnr, 1, ['some', 'text'])
Returns 0 on error.
bufexists({buf}) bufexists() The result is a Number, which is TRUE if a buffer called {buf} exists. If the {buf} argument is a number, buffer numbers are used. Number zero is the alternate buffer for the current window.
If the {buf} argument is a string it must match a buffer name exactly. The name can be:
Relative to the current directory.
A full path.
The name of a buffer with 'buftype' set to "nofile".
A URL name. Unlisted buffers will be found. Note that help files are listed by their short name in the output of :buffers, but bufexists() requires using their long name to be able to find them. bufexists() may report a buffer exists, but to use the name with a :buffer command you may need to use expand(). Esp for MS-Windows 8.3 names in the form "c:\DOCUME~1" Use "bufexists(0)" to test for the existence of an alternate file name.
buflisted({buf}) buflisted() The result is a Number, which is TRUE if a buffer called {buf} exists and is listed (has the 'buflisted' option set). The {buf} argument is used like with bufexists().
bufload({buf}) bufload() Ensure the buffer {buf} is loaded. When the buffer name refers to an existing file then the file is read. Otherwise the buffer will be empty. If the buffer was already loaded then there is no change. If the buffer is not related to a file then no file is read (e.g., when 'buftype' is "nofile"). If there is an existing swap file for the file of the buffer, there will be no dialog, the buffer will be loaded anyway. The {buf} argument is used like with bufexists().
bufloaded({buf}) bufloaded() The result is a Number, which is TRUE if a buffer called {buf} exists and is loaded (shown in a window or hidden). The {buf} argument is used like with bufexists().
bufname([{buf}]) bufname() The result is the name of a buffer. Mostly as it is displayed by the :ls command, but not using special names such as "[No Name]". If {buf} is omitted the current buffer is used. If {buf} is a Number, that buffer number's name is given. Number zero is the alternate buffer for the current window. If {buf} is a String, it is used as a file-pattern to match with the buffer names. This is always done like 'magic' is set and 'cpoptions' is empty. When there is more than one match an empty string is returned. "" or "%" can be used for the current buffer, "#" for the alternate buffer. A full match is preferred, otherwise a match at the start, end or middle of the buffer name is accepted. If you only want a full match then put "^" at the start and "$" at the end of the pattern. Listed buffers are found first. If there is a single match with a listed buffer, that one is returned. Next unlisted buffers are searched for. If the {buf} is a String, but you want to use it as a buffer number, force it to be a Number by adding zero to it:
echo bufname("3" + 0)
If the buffer doesn't exist, or doesn't have a name, an empty string is returned.
echo bufname("#")        " alternate buffer name
echo bufname(3)                " name of buffer 3
echo bufname("%")        " name of current buffer
echo bufname("file2")        " name of buffer where "file2" matches.
bufnr([{buf} [, {create}]]) bufnr() The result is the number of a buffer, as it is displayed by the :ls command. For the use of {buf}, see bufname() above. If the buffer doesn't exist, -1 is returned. Or, if the {create} argument is present and TRUE, a new, unlisted, buffer is created and its number is returned. bufnr("$") is the last buffer:
let last_buffer = bufnr("$")
The result is a Number, which is the highest buffer number of existing buffers. Note that not all buffers with a smaller number necessarily exist, because ":bwipeout" may have removed them. Use bufexists() to test for the existence of a buffer.
bufwinid({buf}) bufwinid() The result is a Number, which is the window-ID of the first window associated with buffer {buf}. For the use of {buf}, see bufname() above. If buffer {buf} doesn't exist or there is no such window, -1 is returned. Example:
echo "A window containing buffer 1 is " .. (bufwinid(1))
Only deals with the current tab page. See win_findbuf() for finding more.
bufwinnr({buf}) bufwinnr() Like bufwinid() but return the window number instead of the window-ID. If buffer {buf} doesn't exist or there is no such window, -1 is returned. Example:
echo "A window containing buffer 1 is " .. (bufwinnr(1))
The number can be used with CTRL-W_w and ":wincmd w" :wincmd.
byte2line({byte}) byte2line() Return the line number that contains the character at byte count {byte} in the current buffer. This includes the end-of-line character, depending on the 'fileformat' option for the current buffer. The first character has byte count one. Also see line2byte(), go and :goto.
Returns -1 if the {byte} value is invalid.
byteidx({expr}, {nr} [, {utf16}]) byteidx() Return byte index of the {nr}th character in the String {expr}. Use zero for the first character, it then returns zero. If there are no multibyte characters the returned value is equal to {nr}. Composing characters are not counted separately, their byte length is added to the preceding base character. See byteidxcomp() below for counting composing characters separately. When {utf16} is present and TRUE, {nr} is used as the UTF-16 index in the String {expr} instead of as the character index. The UTF-16 index is the index in the string when it is encoded with 16-bit words. If the specified UTF-16 index is in the middle of a character (e.g. in a 4-byte character), then the byte index of the first byte in the character is returned. Refer to string-offset-encoding for more information. Example :
echo matchstr(str, ".", byteidx(str, 3))
will display the fourth character. Another way to do the same:
let s = strpart(str, byteidx(str, 3))
echo strpart(s, 0, byteidx(s, 1))
If there are less than {nr} characters -1 is returned. If there are exactly {nr} characters the length of the string in bytes is returned. See charidx() and utf16idx() for getting the character and UTF-16 index respectively from the byte index. Examples:
echo byteidx('a😊😊', 2)        " returns 5
echo byteidx('a😊😊', 2, 1)        " returns 1
echo byteidx('a😊😊', 3, 1)        " returns 5
byteidxcomp({expr}, {nr} [, {utf16}]) byteidxcomp() Like byteidx(), except that a composing character is counted as a separate character. Example:
let s = 'e' .. nr2char(0x301)
echo byteidx(s, 1)
echo byteidxcomp(s, 1)
echo byteidxcomp(s, 2)
The first and third echo result in 3 ('e' plus composing character is 3 bytes), the second echo results in 1 ('e' is one byte).
call({func}, {arglist} [, {dict}]) call() E699 Call function {func} with the items in List {arglist} as arguments. {func} can either be a Funcref or the name of a function. a:firstline and a:lastline are set to the cursor line. Returns the return value of the called function. {dict} is for functions with the "dict" attribute. It will be used to set the local variable "self". Dictionary-function
ceil({expr}) ceil() Return the smallest integral value greater than or equal to {expr} as a Float (round up). {expr} must evaluate to a Float or a Number. Examples:
echo ceil(1.456)
2.0
echo ceil(-5.456)
-5.0
echo ceil(4.0)
4.0
Returns 0.0 if {expr} is not a Float or a Number.
chanclose({id} [, {stream}]) chanclose() Close a channel or a specific stream associated with it. For a job, {stream} can be one of "stdin", "stdout", "stderr" or "rpc" (closes stdin/stdout for a job started with "rpc":v:true) If {stream} is omitted, all streams are closed. If the channel is a pty, this will then close the pty master, sending SIGHUP to the job process. For a socket, there is only one stream, and {stream} should be omitted.
changenr() changenr() Return the number of the most recent change. This is the same number as what is displayed with :undolist and can be used with the :undo command. When a change was made it is the number of that change. After redo it is the number of the redone change. After undo it is one less than the number of the undone change. Returns 0 if the undo list is empty.
chansend({id}, {data}) chansend() Send data to channel {id}. For a job, it writes it to the stdin of the process. For the stdio channel channel-stdio, it writes to Nvim's stdout. Returns the number of bytes written if the write succeeded, 0 otherwise. See channel-bytes for more information.
{data} may be a string, string convertible, Blob, or a list. If {data} is a list, the items will be joined by newlines; any newlines in an item will be sent as NUL. To send a final newline, include a final empty string. Example:
call chansend(id, ["abc", "123\n456", ""])
will send "abc<NL>123<NUL>456<NL>".
chansend() writes raw data, not RPC messages. If the channel was created with "rpc":v:true then the channel expects RPC messages, use rpcnotify() and rpcrequest() instead.
char2nr({string} [, {utf8}]) char2nr() Return Number value of the first char in {string}. Examples:
echo char2nr(" ")        " returns 32
echo char2nr("ABC")        " returns 65
echo char2nr("Γ‘")        " returns 225
echo char2nr("Γ‘"[0])        " returns 195
echo char2nr("\<M-x>")        " returns 128
Non-ASCII characters are always treated as UTF-8 characters. {utf8} is ignored, it exists only for backwards-compatibility. A combining character is a separate character. nr2char() does the opposite.
Returns 0 if {string} is not a String.
charclass({string}) charclass() Return the character class of the first character in {string}. The character class is one of: 0 blank 1 punctuation 2 word character 3 emoji other specific Unicode class The class is used in patterns and word motions. Returns 0 if {string} is not a String.
charcol({expr} [, {winid}]) charcol() Same as col() but returns the character index of the column position given with {expr} instead of the byte position.
Example: With the cursor on 'μ„Έ' in line 5 with text "μ—¬λ³΄μ„Έμš”":
echo charcol('.')        " returns 3
echo col('.')                " returns 7
charidx({string}, {idx} [, {countcc} [, {utf16}]]) charidx() Return the character index of the byte at {idx} in {string}. The index of the first character is zero. If there are no multibyte characters the returned value is equal to {idx}.
When {countcc} is omitted or FALSE, then composing characters are not counted separately, their byte length is added to the preceding base character. When {countcc} is TRUE, then composing characters are counted as separate characters.
When {utf16} is present and TRUE, {idx} is used as the UTF-16 index in the String {expr} instead of as the byte index.
Returns -1 if the arguments are invalid or if there are less than {idx} bytes. If there are exactly {idx} bytes the length of the string in characters is returned.
An error is given and -1 is returned if the first argument is not a string, the second argument is not a number or when the third argument is present and is not zero or one.
See byteidx() and byteidxcomp() for getting the byte index from the character index and utf16idx() for getting the UTF-16 index from the character index. Refer to string-offset-encoding for more information. Examples:
echo charidx('áb́ć', 3)                " returns 1
echo charidx('áb́ć', 6, 1)        " returns 4
echo charidx('áb́ć', 16)                " returns -1
echo charidx('a😊😊', 4, 0, 1)        " returns 2
chdir({dir}) chdir() Change the current working directory to {dir}. The scope of the directory change depends on the directory of the current window:
If the current window has a window-local directory (:lcd), then changes the window local directory.
Otherwise, if the current tabpage has a local directory (:tcd) then changes the tabpage local directory.
Otherwise, changes the global directory. {dir} must be a String. If successful, returns the previous working directory. Pass this to another chdir() to restore the directory. On failure, returns an empty string.
Example:
let save_dir = chdir(newdir)
if save_dir != ""
   " ... do some work
   call chdir(save_dir)
endif
cindent({lnum}) cindent() Get the amount of indent for line {lnum} according the C indenting rules, as with 'cindent'. The indent is counted in spaces, the value of 'tabstop' is relevant. {lnum} is used just like in getline(). When {lnum} is invalid -1 is returned. See C-indenting.
clearmatches([{win}]) clearmatches() Clears all matches previously defined for the current window by matchadd() and the :match commands. If {win} is specified, use the window with this number or window ID instead of the current window.
col({expr} [, {winid}]) col() The result is a Number, which is the byte index of the column position given with {expr}. The accepted positions are: . the cursor position $ the end of the cursor line (the result is the number of bytes in the cursor line plus one) 'x position of mark x (if the mark is not set, 0 is returned) v In Visual mode: the start of the Visual area (the cursor is the end). When not in Visual mode returns the cursor position. Differs from '< in that it's updated right away. Additionally {expr} can be [lnum, col]: a List with the line and column number. Most useful when the column is "$", to get the last column of a specific line. When "lnum" or "col" is out of range then col() returns zero. With the optional {winid} argument the values are obtained for that window instead of the current window. To get the line number use line(). To get both use getpos(). For the screen column position use virtcol(). For the character position use charcol(). Note that only marks in the current file can be used. Examples:
echo col(".")                        " column of cursor
echo col("$")                        " length of cursor line plus one
echo col("'t")                        " column of mark t
echo col("'" .. markname)        " column of mark markname
The first column is 1. Returns 0 if {expr} is invalid or when the window with ID {winid} is not found. For an uppercase mark the column may actually be in another buffer. For the cursor position, when 'virtualedit' is active, the column is one higher if the cursor is after the end of the line. Also, when using a <Cmd> mapping the cursor isn't moved, this can be used to obtain the column in Insert mode:
imap <F2> <Cmd>echo col(".").."\n"<CR>
complete({startcol}, {matches}) complete() E785 Set the matches for Insert mode completion. Can only be used in Insert mode. You need to use a mapping with CTRL-R = (see i_CTRL-R). It does not work after CTRL-O or with an expression mapping. {startcol} is the byte offset in the line where the completed text start. The text up to the cursor is the original text that will be replaced by the matches. Use col('.') for an empty string. "col('.') - 1" will replace one character by a match. {matches} must be a List. Each List item is one match. See complete-items for the kind of items that are possible. "longest" in 'completeopt' is ignored. Note that the after calling this function you need to avoid inserting anything that would cause completion to stop. The match can be selected with CTRL-N and CTRL-P as usual with Insert mode completion. The popup menu will appear if specified, see ins-completion-menu. Example:
inoremap <F5> <C-R>=ListMonths()<CR>
func ListMonths()
  call complete(col('.'), ['January', 'February', 'March',
    \ 'April', 'May', 'June', 'July', 'August', 'September',
    \ 'October', 'November', 'December'])
  return ''
endfunc
This isn't very useful, but it shows how it works. Note that an empty string is returned to avoid a zero being inserted.
complete_add({expr}) complete_add() Add {expr} to the list of matches. Only to be used by the function specified with the 'completefunc' option. Returns 0 for failure (empty string or out of memory), 1 when the match was added, 2 when the match was already in the list. See complete-functions for an explanation of {expr}. It is the same as one item in the list that 'omnifunc' would return.
complete_check() complete_check() Check for a key typed while looking for completion matches. This is to be used when looking for matches takes some time. Returns TRUE when searching for matches is to be aborted, zero otherwise. Only to be used by the function specified with the 'completefunc' option.
complete_info([{what}]) complete_info() Returns a Dictionary with information about Insert mode completion. See ins-completion. The items are: mode Current completion mode name string. See complete_info_mode for the values. pum_visible TRUE if popup menu is visible. See pumvisible(). items List of completion matches. Each item is a dictionary containing the entries "word", "abbr", "menu", "kind", "info" and "user_data". See complete-items. selected Selected item index. First index is zero. Index is -1 if no item is selected (showing typed text only, or the last completion after no item is selected when using the <Up> or <Down> keys) inserted Inserted string. [NOT IMPLEMENTED YET] preview_winid Info floating preview window id. preview_bufnr Info floating preview buffer id.
complete_info_mode mode values are: "" Not in completion mode "keyword" Keyword completion i_CTRL-X_CTRL-N "ctrl_x" Just pressed CTRL-X i_CTRL-X "scroll" Scrolling with i_CTRL-X_CTRL-E or i_CTRL-X_CTRL-Y "whole_line" Whole lines i_CTRL-X_CTRL-L "files" File names i_CTRL-X_CTRL-F "tags" Tags i_CTRL-X_CTRL-] "path_defines" Definition completion i_CTRL-X_CTRL-D "path_patterns" Include completion i_CTRL-X_CTRL-I "dictionary" Dictionary i_CTRL-X_CTRL-K "thesaurus" Thesaurus i_CTRL-X_CTRL-T "cmdline" Vim Command line i_CTRL-X_CTRL-V "function" User defined completion i_CTRL-X_CTRL-U "omni" Omni completion i_CTRL-X_CTRL-O "spell" Spelling suggestions i_CTRL-X_s "eval" complete() completion "unknown" Other internal modes
If the optional {what} list argument is supplied, then only the items listed in {what} are returned. Unsupported items in {what} are silently ignored.
To get the position and size of the popup menu, see pum_getpos(). It's also available in v:event during the CompleteChanged event.
Returns an empty Dictionary on error.
Examples:
" Get all items
call complete_info()
" Get only 'mode'
call complete_info(['mode'])
" Get only 'mode' and 'pum_visible'
call complete_info(['mode', 'pum_visible'])
confirm({msg} [, {choices} [, {default} [, {type}]]]) confirm() confirm() offers the user a dialog, from which a choice can be made. It returns the number of the choice. For the first choice this is 1.
{msg} is displayed in a dialog with {choices} as the alternatives. When {choices} is missing or empty, "&OK" is used (and translated). {msg} is a String, use '\n' to include a newline. Only on some systems the string is wrapped when it doesn't fit.
{choices} is a String, with the individual choices separated by '\n', e.g.
confirm("Save changes?", "&Yes\n&No\n&Cancel")
The letter after the '&' is the shortcut key for that choice. Thus you can type 'c' to select "Cancel". The shortcut does not need to be the first letter:
confirm("file has been modified", "&Save\nSave &All")
For the console, the first letter of each choice is used as the default shortcut key. Case is ignored.
The optional {type} String argument gives the type of dialog. It can be one of these values: "Error", "Question", "Info", "Warning" or "Generic". Only the first character is relevant. When {type} is omitted, "Generic" is used.
The optional {type} argument gives the type of dialog. This is only used for the icon of the Win32 GUI. It can be one of these values: "Error", "Question", "Info", "Warning" or "Generic". Only the first character is relevant. When {type} is omitted, "Generic" is used.
If the user aborts the dialog by pressing <Esc>, CTRL-C, or another valid interrupt key, confirm() returns 0.
An example:
let choice = confirm("What do you want?",
                     \ "&Apples\n&Oranges\n&Bananas", 2)
if choice == 0
     echo "make up your mind!"
elseif choice == 3
     echo "tasteful"
else
     echo "I prefer bananas myself."
endif
In a GUI dialog, buttons are used. The layout of the buttons depends on the 'v' flag in 'guioptions'. If it is included, the buttons are always put vertically. Otherwise, confirm() tries to put the buttons in one horizontal line. If they don't fit, a vertical layout is used anyway. For some systems the horizontal layout is always used.
copy({expr}) copy() Make a copy of {expr}. For Numbers and Strings this isn't different from using {expr} directly. When {expr} is a List a shallow copy is created. This means that the original List can be changed without changing the copy, and vice versa. But the items are identical, thus changing an item changes the contents of both Lists. A Dictionary is copied in a similar way as a List. Also see deepcopy().
cos({expr}) cos() Return the cosine of {expr}, measured in radians, as a Float. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo cos(100)
0.862319
echo cos(-4.01)
-0.646043
cosh({expr}) cosh() Return the hyperbolic cosine of {expr} as a Float in the range [1, inf]. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo cosh(0.5)
1.127626
echo cosh(-0.5)
-1.127626
count({comp}, {expr} [, {ic} [, {start}]]) count() E706 Return the number of times an item with value {expr} appears in String, List or Dictionary {comp}.
If {start} is given then start with the item with this index. {start} can only be used with a List.
When {ic} is given and it's TRUE then case is ignored.
When {comp} is a string then the number of not overlapping occurrences of {expr} is returned. Zero is returned when {expr} is an empty string.
ctxget([{index}]) ctxget() Returns a Dictionary representing the context at {index} from the top of the context-stack (see context-dict). If {index} is not given, it is assumed to be 0 (i.e.: top).
ctxpop() ctxpop() Pops and restores the context at the top of the context-stack.
ctxpush([{types}]) ctxpush() Pushes the current editor state (context) on the context-stack. If {types} is given and is a List of Strings, it specifies which context-types to include in the pushed context. Otherwise, all context types are included.
ctxset({context} [, {index}]) ctxset() Sets the context at {index} from the top of the context-stack to that represented by {context}. {context} is a Dictionary with context data (context-dict). If {index} is not given, it is assumed to be 0 (i.e.: top).
ctxsize() ctxsize() Returns the size of the context-stack.
cursor({lnum}, {col} [, {off}]) cursor() cursor({list}) Positions the cursor at the column (byte count) {col} in the line {lnum}. The first column is one.
When there is one argument {list} this is used as a List with two, three or four item: [{lnum}, {col}] [{lnum}, {col}, {off}] [{lnum}, {col}, {off}, {curswant}] This is like the return value of getpos() or getcurpos(), but without the first item.
To position the cursor using {col} as the character count, use setcursorcharpos().
Does not change the jumplist. {lnum} is used like with getline(), except that if {lnum} is zero, the cursor will stay in the current line. If {lnum} is greater than the number of lines in the buffer, the cursor will be positioned at the last line in the buffer. If {col} is greater than the number of bytes in the line, the cursor will be positioned at the last character in the line. If {col} is zero, the cursor will stay in the current column. If {curswant} is given it is used to set the preferred column for vertical movement. Otherwise {col} is used.
When 'virtualedit' is used {off} specifies the offset in screen columns from the start of the character. E.g., a position within a <Tab> or after the last character. Returns 0 when the position could be set, -1 otherwise.
debugbreak({pid}) debugbreak() Specifically used to interrupt a program being debugged. It will cause process {pid} to get a SIGTRAP. Behavior for other processes is undefined. See terminal-debug. (Sends a SIGINT to a process {pid} other than MS-Windows)
Returns TRUE if successfully interrupted the program. Otherwise returns FALSE.
deepcopy({expr} [, {noref}]) deepcopy() E698 Make a copy of {expr}. For Numbers and Strings this isn't different from using {expr} directly. When {expr} is a List a full copy is created. This means that the original List can be changed without changing the copy, and vice versa. When an item is a List, a copy for it is made, recursively. Thus changing an item in the copy does not change the contents of the original List.
When {noref} is omitted or zero a contained List or Dictionary is only copied once. All references point to this single copy. With {noref} set to 1 every occurrence of a List or Dictionary results in a new copy. This also means that a cyclic reference causes deepcopy() to fail. E724 Nesting is possible up to 100 levels. When there is an item that refers back to a higher level making a deep copy with {noref} set to 1 will fail. Also see copy().
delete({fname} [, {flags}]) delete() Without {flags} or with {flags} empty: Deletes the file by the name {fname}.
This also works when {fname} is a symbolic link. The symbolic link itself is deleted, not what it points to.
When {flags} is "d": Deletes the directory by the name {fname}. This fails when directory {fname} is not empty.
When {flags} is "rf": Deletes the directory by the name {fname} and everything in it, recursively. BE CAREFUL! Note: on MS-Windows it is not possible to delete a directory that is being used.
The result is a Number, which is 0/false if the delete operation was successful and -1/true when the deletion failed or partly failed.
deletebufline({buf}, {first} [, {last}]) deletebufline() Delete lines {first} to {last} (inclusive) from buffer {buf}. If {last} is omitted then delete line {first} only. On success 0 is returned, on failure 1 is returned.
This function works only for loaded buffers. First call bufload() if needed.
For the use of {buf}, see bufname() above.
{first} and {last} are used like with getline(). Note that when using line() this refers to the current buffer. Use "$" to refer to the last line in buffer {buf}.
dictwatcheradd({dict}, {pattern}, {callback}) dictwatcheradd() Adds a watcher to a dictionary. A dictionary watcher is identified by three components:
A dictionary({dict});
A key pattern({pattern}).
A function({callback}).
After this is called, every change on {dict} and on keys matching {pattern} will result in {callback} being invoked.
For example, to watch all global variables:
silent! call dictwatcherdel(g:, '*', 'OnDictChanged')
function! OnDictChanged(d,k,z)
  echomsg string(a:k) string(a:z)
endfunction
call dictwatcheradd(g:, '*', 'OnDictChanged')
For now {pattern} only accepts very simple patterns that can contain a "*" at the end of the string, in which case it will match every key that begins with the substring before the "*". That means if "*" is not the last character of {pattern}, only keys that are exactly equal as {pattern} will be matched.
The {callback} receives three arguments:
The dictionary being watched.
The key which changed.
A dictionary containing the new and old values for the key.
The type of change can be determined by examining the keys present on the third argument:
If contains both old and new, the key was updated.
If it contains only new, the key was added.
If it contains only old, the key was deleted.
This function can be used by plugins to implement options with validation and parsing logic.
dictwatcherdel({dict}, {pattern}, {callback}) dictwatcherdel() Removes a watcher added with dictwatcheradd(). All three arguments must match the ones passed to dictwatcheradd() in order for the watcher to be successfully deleted.
did_filetype() did_filetype() Returns TRUE when autocommands are being executed and the FileType event has been triggered at least once. Can be used to avoid triggering the FileType event again in the scripts that detect the file type. FileType Returns FALSE when :setf FALLBACK was used. When editing another file, the counter is reset, thus this really checks if the FileType event has been triggered for the current buffer. This allows an autocommand that starts editing another buffer to set 'filetype' and load a syntax file.
diff_filler({lnum}) diff_filler() Returns the number of filler lines above line {lnum}. These are the lines that were inserted at this point in another diff'ed window. These filler lines are shown in the display but don't exist in the buffer. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc. Returns 0 if the current window is not in diff mode.
diff_hlID({lnum}, {col}) diff_hlID() Returns the highlight ID for diff mode at line {lnum} column {col} (byte index). When the current line does not have a diff change zero is returned. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc. {col} is 1 for the leftmost column, {lnum} is 1 for the first line. The highlight ID can be used with synIDattr() to obtain syntax information about the highlighting.
digraph_get({chars}) digraph_get() E1214 Return the digraph of {chars}. This should be a string with exactly two characters. If {chars} are not just two characters, or the digraph of {chars} does not exist, an error is given and an empty string is returned.
Examples:
" Get a built-in digraph
echo digraph_get('00')                " Returns '∞'
" Get a user-defined digraph
call digraph_set('aa', 'あ')
echo digraph_get('aa')                " Returns 'あ'
digraph_getlist([{listall}]) digraph_getlist() Return a list of digraphs. If the {listall} argument is given and it is TRUE, return all digraphs, including the default digraphs. Otherwise, return only user-defined digraphs.
Also see digraph_get().
Examples:
" Get user-defined digraphs
echo digraph_getlist()
" Get all the digraphs, including default digraphs
echo digraph_getlist(1)
digraph_set({chars}, {digraph}) digraph_set() Add digraph {chars} to the list. {chars} must be a string with two characters. {digraph} is a string with one UTF-8 encoded character. E1215 Be careful, composing characters are NOT ignored. This function is similar to :digraphs command, but useful to add digraphs start with a white space.
The function result is v:true if digraph is registered. If this fails an error message is given and v:false is returned.
If you want to define multiple digraphs at once, you can use digraph_setlist().
Example:
call digraph_set('  ', 'あ')
Can be used as a method:
GetString()->digraph_set('あ')
digraph_setlist({digraphlist}) digraph_setlist() Similar to digraph_set() but this function can add multiple digraphs at once. {digraphlist} is a list composed of lists, where each list contains two strings with {chars} and {digraph} as in digraph_set(). E1216 Example:
call digraph_setlist([['aa', 'あ'], ['ii', 'い']])
It is similar to the following:
for [chars, digraph] in [['aa', 'あ'], ['ii', 'い']]
      call digraph_set(chars, digraph)
endfor
Except that the function returns after the first error, following digraphs will not be added.
Can be used as a method:
GetList()->digraph_setlist()
empty({expr}) empty() Return the Number 1 if {expr} is empty, zero otherwise.
A List or Dictionary is empty when it does not have any items.
A String is empty when its length is zero.
A Number and Float are empty when their value is zero.
v:false and v:null are empty, v:true is not.
A Blob is empty when its length is zero.
environ() environ() Return all of environment variables as dictionary. You can check if an environment variable exists like this:
echo has_key(environ(), 'HOME')
Note that the variable name may be CamelCase; to ignore case use this:
echo index(keys(environ()), 'HOME', 0, 1) != -1
escape({string}, {chars}) escape() Escape the characters in {chars} that occur in {string} with a backslash. Example:
echo escape('c:\program files\vim', ' \')
results in:
c:\\program\ files\\vim
eval({string}) eval() Evaluate {string} and return the result. Especially useful to turn the result of string() back into the original value. This works for Numbers, Floats, Strings, Blobs and composites of them. Also works for Funcrefs that refer to existing functions.
eventhandler() eventhandler() Returns 1 when inside an event handler. That is that Vim got interrupted while waiting for the user to type a character, e.g., when dropping a file on Vim. This means interactive commands cannot be used. Otherwise zero is returned.
executable({expr}) executable() This function checks if an executable with the name {expr} exists. {expr} must be the name of the program without any arguments. executable() uses the value of $PATH and/or the normal searchpath for programs. PATHEXT On MS-Windows the ".exe", ".bat", etc. can optionally be included. Then the extensions in $PATHEXT are tried. Thus if "foo.exe" does not exist, "foo.exe.bat" can be found. If $PATHEXT is not set then ".exe;.com;.bat;.cmd" is used. A dot by itself can be used in $PATHEXT to try using the name without an extension. When 'shell' looks like a Unix shell, then the name is also tried without adding an extension. On MS-Windows it only checks if the file exists and is not a directory, not if it's really executable. On Windows an executable in the same directory as Vim is always found (it is added to $PATH at startup). The result is a Number: 1 exists 0 does not exist -1 not implemented on this system exepath() can be used to get the full path of an executable.
execute({command} [, {silent}]) execute() Execute {command} and capture its output. If {command} is a String, returns {command} output. If {command} is a List, returns concatenated outputs. Line continuations in {command} are not recognized. Examples:
echo execute('echon "foo"')
foo
echo execute(['echon "foo"', 'echon "bar"'])
foobar
The optional {silent} argument can have these values: "" no :silent used "silent" :silent used "silent!" :silent! used The default is "silent". Note that with "silent!", unlike :redir, error messages are dropped.
To get a list of lines use split() on the result:
execute('args')->split("\n")
This function is not available in the sandbox. Note: If nested, an outer execute() will not observe output of the inner calls. Note: Text attributes (highlights) are not captured. To execute a command in another window than the current one use win_execute().
exepath({expr}) exepath() Returns the full path of {expr} if it is an executable and given as a (partial or full) path or is found in $PATH. Returns empty string otherwise. If {expr} starts with "./" the current-directory is used.
exists({expr}) exists() The result is a Number, which is TRUE if {expr} is defined, zero otherwise.
For checking for a supported feature use has(). For checking if a file exists use filereadable().
The {expr} argument is a string, which contains one of these: varname internal variable (see dict.key internal-variables). Also works list[i] for curly-braces-names, Dictionary entries, List items, etc. Beware that evaluating an index may cause an error message for an invalid expression. E.g.:
let l = [1, 2, 3]
echo exists("l[5]")
0
echo exists("l[xx]")
E121: Undefined variable: xx 0 &option-name Vim option (only checks if it exists, not if it really works) +option-name Vim option that works. $ENVNAME environment variable (could also be done by comparing with an empty string) *funcname built-in function (see functions) or user defined function (see user-function). Also works for a variable that is a Funcref. :cmdname Ex command: built-in command, user command or command modifier :command. Returns: 1 for match with start of a command 2 full match with a command 3 matches several user commands To check for a supported command always check the return value to be 2. :2match The :2match command. :3match The :3match command (but you probably should not use it, it is reserved for internal usage) #event autocommand defined for this event #event#pattern autocommand defined for this event and pattern (the pattern is taken literally and compared to the autocommand patterns character by character) #group autocommand group exists #group#event autocommand defined for this group and event. #group#event#pattern autocommand defined for this group, event and pattern. ##event autocommand for this event is supported.
Examples:
echo exists("&mouse")
echo exists("$HOSTNAME")
echo exists("*strftime")
echo exists("*s:MyFunc")
echo exists("*MyFunc")
echo exists("*v:lua.Func")
echo exists("bufcount")
echo exists(":Make")
echo exists("#CursorHold")
echo exists("#BufReadPre#*.gz")
echo exists("#filetypeindent")
echo exists("#filetypeindent#FileType")
echo exists("#filetypeindent#FileType#*")
echo exists("##ColorScheme")
There must be no space between the symbol (&/$/*/#) and the name. There must be no extra characters after the name, although in a few cases this is ignored. That may become stricter in the future, thus don't count on it! Working example:
echo exists(":make")
NOT working example:
echo exists(":make install")
Note that the argument must be a string, not the name of the variable itself. For example:
echo exists(bufcount)
This doesn't check for existence of the "bufcount" variable, but gets the value of "bufcount", and checks if that exists.
exp({expr}) exp() Return the exponential of {expr} as a Float in the range [0, inf]. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo exp(2)
7.389056
echo exp(-1)
0.367879
expand({string} [, {nosuf} [, {list}]]) expand() Expand wildcards and the following special keywords in {string}. 'wildignorecase' applies.
If {list} is given and it is TRUE, a List will be returned. Otherwise the result is a String and when there are several matches, they are separated by <NL> characters.
If the expansion fails, the result is an empty string. A name for a non-existing file is not included, unless {string} does not start with '%', '#' or '<', see below.
When {string} starts with '%', '#' or '<', the expansion is done like for the cmdline-special variables with their associated modifiers. Here is a short overview:
% current file name # alternate file name #n alternate file name n <cfile> file name under the cursor <afile> autocmd file name <abuf> autocmd buffer number (as a String!) <amatch> autocmd matched name <cexpr> C expression under the cursor <sfile> sourced script file or function name <slnum> sourced script line number or function line number <sflnum> script file line number, also when in a function <SID> "<SNR>123_" where "123" is the current script ID <SID> <script> sourced script file, or script file where the current function was defined <stack> call stack <cword> word under the cursor <cWORD> WORD under the cursor <client> the {clientid} of the last received message Modifiers: :p expand to full path :h head (last path component removed) :t tail (last path component only) :r root (one extension removed) :e extension only
Example:
let &tags = expand("%:p:h") .. "/tags"
Note that when expanding a string that starts with '%', '#' or '<', any following text is ignored. This does NOT work:
let doesntwork = expand("%:h.bak")
Use this:
let doeswork = expand("%:h") .. ".bak"
Also note that expanding "<cfile>" and others only returns the referenced file name without further expansion. If "<cfile>" is "~/.cshrc", you need to do another expand() to have the "~/" expanded into the path of the home directory:
echo expand(expand("<cfile>"))
There cannot be white space between the variables and the following modifier. The fnamemodify() function can be used to modify normal file names.
When using '%' or '#', and the current or alternate file name is not defined, an empty string is used. Using "%:p" in a buffer with no name, results in the current directory, with a '/' added. When 'verbose' is set then expanding '%', '#' and <> items will result in an error message if the argument cannot be expanded.
When {string} does not start with '%', '#' or '<', it is expanded like a file name is expanded on the command line. 'suffixes' and 'wildignore' are used, unless the optional {nosuf} argument is given and it is TRUE. Names for non-existing files are included. The "**" item can be used to search in a directory tree. For example, to find all "README" files in the current directory and below:
echo expand("**/README")
expand() can also be used to expand variables and environment variables that are only known in a shell. But this can be slow, because a shell may be used to do the expansion. See expr-env-expand. The expanded variable is still handled like a list of file names. When an environment variable cannot be expanded, it is left unchanged. Thus ":echo expand('$FOOBAR')" results in "$FOOBAR".
See glob() for finding existing files. See system() for getting the raw output of an external command.
expandcmd({string} [, {options}]) expandcmd() Expand special items in String {string} like what is done for an Ex command such as :edit. This expands special keywords, like with expand(), and environment variables, anywhere in {string}. "~user" and "~/path" are only expanded at the start.
The following items are supported in the {options} Dict argument: errmsg If set to TRUE, error messages are displayed if an error is encountered during expansion. By default, error messages are not displayed.
Returns the expanded string. If an error is encountered during expansion, the unmodified {string} is returned.
Example:
echo expandcmd('make %<.o')
make /path/runtime/doc/builtin.o
echo expandcmd('make %<.o', {'errmsg': v:true})
extend({expr1}, {expr2} [, {expr3}]) extend() {expr1} and {expr2} must be both Lists or both Dictionaries.
If they are Lists: Append {expr2} to {expr1}. If {expr3} is given insert the items of {expr2} before the item with index {expr3} in {expr1}. When {expr3} is zero insert before the first item. When {expr3} is equal to len({expr1}) then {expr2} is appended. Examples:
echo sort(extend(mylist, [7, 5]))
call extend(mylist, [2, 3], 1)
When {expr1} is the same List as {expr2} then the number of items copied is equal to the original length of the List. E.g., when {expr3} is 1 you get N new copies of the first item (where N is the original length of the List). Use add() to concatenate one item to a list. To concatenate two lists into a new list use the + operator:
let newlist = [1, 2, 3] + [4, 5]
If they are Dictionaries: Add all entries from {expr2} to {expr1}. If a key exists in both {expr1} and {expr2} then {expr3} is used to decide what to do: {expr3} = "keep": keep the value of {expr1} {expr3} = "force": use the value of {expr2} {expr3} = "error": give an error message E737 When {expr3} is omitted then "force" is assumed.
{expr1} is changed when {expr2} is not empty. If necessary make a copy of {expr1} first. {expr2} remains unchanged. When {expr1} is locked and {expr2} is not empty the operation fails. Returns {expr1}. Returns 0 on error.
extendnew({expr1}, {expr2} [, {expr3}]) extendnew() Like extend() but instead of adding items to {expr1} a new List or Dictionary is created and returned. {expr1} remains unchanged.
feedkeys({string} [, {mode}]) feedkeys() Characters in {string} are queued for processing as if they come from a mapping or were typed by the user.
By default the string is added to the end of the typeahead buffer, thus if a mapping is still being executed the characters come after them. Use the 'i' flag to insert before other characters, they will be executed next, before any characters from a mapping.
The function does not wait for processing of keys contained in {string}.
To include special keys into {string}, use double-quotes and "\..." notation expr-quote. For example, feedkeys("\<CR>") simulates pressing of the <Enter> key. But feedkeys('\<CR>') pushes 5 characters. The <Ignore> keycode may be used to exit the wait-for-character without doing anything.
{mode} is a String, which can contain these character flags: 'm' Remap keys. This is default. If {mode} is absent, keys are remapped. 'n' Do not remap keys. 't' Handle keys as if typed; otherwise they are handled as if coming from a mapping. This matters for undo, opening folds, etc. 'i' Insert the string instead of appending (see above). 'x' Execute commands until typeahead is empty. This is similar to using ":normal!". You can call feedkeys() several times without 'x' and then one time with 'x' (possibly with an empty {string}) to execute all the typeahead. Note that when Vim ends in Insert mode it will behave as if <Esc> is typed, to avoid getting stuck, waiting for a character to be typed before the script continues. Note that if you manage to call feedkeys() while executing commands, thus calling it recursively, then all typeahead will be consumed by the last call. '!' When used with 'x' will not end Insert mode. Can be used in a test when a timer is set to exit Insert mode a little later. Useful for testing CursorHoldI.
Return value is always 0.
filereadable({file}) filereadable() The result is a Number, which is TRUE when a file with the name {file} exists, and can be read. If {file} doesn't exist, or is a directory, the result is FALSE. {file} is any expression, which is used as a String. If you don't care about the file being readable you can use glob(). {file} is used as-is, you may want to expand wildcards first:
echo filereadable('~/.vimrc')
0
echo filereadable(expand('~/.vimrc'))
1
filewritable({file}) filewritable() The result is a Number, which is 1 when a file with the name {file} exists, and can be written. If {file} doesn't exist, or is not writable, the result is 0. If {file} is a directory, and we can write to it, the result is 2.
filter({expr1}, {expr2}) filter() {expr1} must be a List, String, Blob or Dictionary. For each item in {expr1} evaluate {expr2} and when the result is zero or false remove the item from the List or Dictionary. Similarly for each byte in a Blob and each character in a String.
{expr2} must be a string or Funcref.
If {expr2} is a string, inside {expr2} v:val has the value of the current item. For a Dictionary v:key has the key of the current item and for a List v:key has the index of the current item. For a Blob v:key has the index of the current byte. For a String v:key has the index of the current character. Examples:
call filter(mylist, 'v:val !~ "OLD"')
Removes the items where "OLD" appears.
call filter(mydict, 'v:key >= 8')
Removes the items with a key below 8.
call filter(var, 0)
Removes all the items, thus clears the List or Dictionary.
Note that {expr2} is the result of expression and is then used as an expression again. Often it is good to use a literal-string to avoid having to double backslashes.
If {expr2} is a Funcref it must take two arguments: 1. the key or the index of the current item. 2. the value of the current item. The function must return TRUE if the item should be kept. Example that keeps the odd items of a list:
func Odd(idx, val)
  return a:idx % 2 == 1
endfunc
call filter(mylist, function('Odd'))
It is shorter when using a lambda:
call filter(myList, {idx, val -> idx * val <= 42})
If you do not use "val" you can leave it out:
call filter(myList, {idx -> idx % 2 == 1})
For a List and a Dictionary the operation is done in-place. If you want it to remain unmodified make a copy first:
let l = filter(copy(mylist), 'v:val =~ "KEEP"')
Returns {expr1}, the List or Dictionary that was filtered, or a new Blob or String. When an error is encountered while evaluating {expr2} no further items in {expr1} are processed. When {expr2} is a Funcref errors inside a function are ignored, unless it was defined with the "abort" flag.
finddir({name} [, {path} [, {count}]]) finddir() Find directory {name} in {path}. Supports both downwards and upwards recursive directory searches. See file-searching for the syntax of {path}.
Returns the path of the first found match. When the found directory is below the current directory a relative path is returned. Otherwise a full path is returned. If {path} is omitted or empty then 'path' is used.
If the optional {count} is given, find {count}'s occurrence of {name} in {path} instead of the first one. When {count} is negative return all the matches in a List.
Returns an empty string if the directory is not found.
This is quite similar to the ex-command :find.
findfile({name} [, {path} [, {count}]]) findfile() Just like finddir(), but find a file instead of a directory. Uses 'suffixesadd'. Example:
echo findfile("tags.vim", ".;")
Searches from the directory of the current file upwards until it finds the file "tags.vim".
flatten({list} [, {maxdepth}]) flatten() Flatten {list} up to {maxdepth} levels. Without {maxdepth} the result is a List without nesting, as if {maxdepth} is a very large number. The {list} is changed in place, use flattennew() if you do not want that. E900 {maxdepth} means how deep in nested lists changes are made. {list} is not modified when {maxdepth} is 0. {maxdepth} must be positive number.
If there is an error the number zero is returned.
Example:
echo flatten([1, [2, [3, 4]], 5])
[1, 2, 3, 4, 5]
echo flatten([1, [2, [3, 4]], 5], 1)
[1, 2, [3, 4], 5]
flattennew({list} [, {maxdepth}]) flattennew() Like flatten() but first make a copy of {list}.
float2nr({expr}) float2nr() Convert {expr} to a Number by omitting the part after the decimal point. {expr} must evaluate to a Float or a Number. Returns 0 if {expr} is not a Float or a Number. When the value of {expr} is out of range for a Number the result is truncated to 0x7fffffff or -0x7fffffff (or when 64-bit Number support is enabled, 0x7fffffffffffffff or -0x7fffffffffffffff). NaN results in -0x80000000 (or when 64-bit Number support is enabled, -0x8000000000000000). Examples:
echo float2nr(3.95)
3
echo float2nr(-23.45)
-23
echo float2nr(1.0e100)
2147483647 (or 9223372036854775807)
echo float2nr(-1.0e150)
-2147483647 (or -9223372036854775807)
echo float2nr(1.0e-100)
0
floor({expr}) floor() Return the largest integral value less than or equal to {expr} as a Float (round down). {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo floor(1.856)
1.0
echo floor(-5.456)
-6.0
echo floor(4.0)
4.0
fmod({expr1}, {expr2}) fmod() Return the remainder of {expr1} / {expr2}, even if the division is not representable. Returns {expr1} - i * {expr2} for some integer i such that if {expr2} is non-zero, the result has the same sign as {expr1} and magnitude less than the magnitude of {expr2}. If {expr2} is zero, the value returned is zero. The value returned is a Float. {expr1} and {expr2} must evaluate to a Float or a Number. Returns 0.0 if {expr1} or {expr2} is not a Float or a Number. Examples:
echo fmod(12.33, 1.22)
0.13
echo fmod(-12.33, 1.22)
-0.13
fnameescape({string}) fnameescape() Escape {string} for use as file name command argument. All characters that have a special meaning, such as '%' and '|' are escaped with a backslash. For most systems the characters escaped are " \t\n*?[{`$\\%#'\"|!<".} For systems where a backslash appears in a filename, it depends on the value of 'isfname'. A leading '+' and '>' is also escaped (special after :edit and :write). And a "-" by itself (special after :cd). Returns an empty string on error. Example:
let fname = '+some str%nge|name'
exe "edit " .. fnameescape(fname)
results in executing:
edit \+some\ str\%nge\|name
fnamemodify({fname}, {mods}) fnamemodify() Modify file name {fname} according to {mods}. {mods} is a string of characters like it is used for file names on the command line. See filename-modifiers. Example:
echo fnamemodify("main.c", ":p:h")
results in:
/home/user/vim/vim/src
If {mods} is empty or an unsupported modifier is used then {fname} is returned. When {fname} is empty then with {mods} ":h" returns ".", so that :cd can be used with it. This is different from expand('%:h') without a buffer name, which returns an empty string. Note: Environment variables don't work in {fname}, use expand() first then.
foldclosed({lnum}) foldclosed() The result is a Number. If the line {lnum} is in a closed fold, the result is the number of the first line in that fold. If the line {lnum} is not in a closed fold, -1 is returned. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc.
foldclosedend({lnum}) foldclosedend() The result is a Number. If the line {lnum} is in a closed fold, the result is the number of the last line in that fold. If the line {lnum} is not in a closed fold, -1 is returned. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc.
foldlevel({lnum}) foldlevel() The result is a Number, which is the foldlevel of line {lnum} in the current buffer. For nested folds the deepest level is returned. If there is no fold at line {lnum}, zero is returned. It doesn't matter if the folds are open or closed. When used while updating folds (from 'foldexpr') -1 is returned for lines where folds are still to be updated and the foldlevel is unknown. As a special case the level of the previous line is usually available. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc.
foldtext() foldtext() Returns a String, to be displayed for a closed fold. This is the default function used for the 'foldtext' option and should only be called from evaluating 'foldtext'. It uses the v:foldstart, v:foldend and v:folddashes variables. The returned string looks like this:
+-- 45 lines: abcdef
The number of leading dashes depends on the foldlevel. The "45" is the number of lines in the fold. "abcdef" is the text in the first non-blank line of the fold. Leading white space, "//" or "/*" and the text from the 'foldmarker' and 'commentstring' options is removed. When used to draw the actual foldtext, the rest of the line will be filled with the fold char from the 'fillchars' setting. Returns an empty string when there is no fold.
foldtextresult({lnum}) foldtextresult() Returns the text that is displayed for the closed fold at line {lnum}. Evaluates 'foldtext' in the appropriate context. When there is no closed fold at {lnum} an empty string is returned. {lnum} is used like with getline(). Thus "." is the current line, "'m" mark m, etc. Useful when exporting folded text, e.g., to HTML.
foreach({expr1}, {expr2}) foreach() {expr1} must be a List, String, Blob or Dictionary. For each item in {expr1} execute {expr2}. {expr1} is not modified; its values may be, as with :lockvar 1. E741 See map() and filter() to modify {expr1}.
{expr2} must be a string or Funcref.
If {expr2} is a string, inside {expr2} v:val has the value of the current item. For a Dictionary v:key has the key of the current item and for a List v:key has the index of the current item. For a Blob v:key has the index of the current byte. For a String v:key has the index of the current character. Examples:
call foreach(mylist, 'let used[v:val] = v:true')
This records the items that are in the {expr1} list.
Note that {expr2} is the result of expression and is then used as a command. Often it is good to use a literal-string to avoid having to double backslashes.
If {expr2} is a Funcref it must take two arguments: 1. the key or the index of the current item. 2. the value of the current item. With a lambda you don't get an error if it only accepts one argument. If the function returns a value, it is ignored.
Returns {expr1} in all cases. When an error is encountered while executing {expr2} no further items in {expr1} are processed. When {expr2} is a Funcref errors inside a function are ignored, unless it was defined with the "abort" flag.
fullcommand({name}) fullcommand() Get the full command name from a short abbreviated command name; see 20.2 for details on command abbreviations.
The string argument {name} may start with a : and can include a [range], these are skipped and not returned. Returns an empty string if a command doesn't exist or if it's ambiguous (for user-defined commands).
For example fullcommand('s'), fullcommand('sub'), fullcommand(':%substitute') all return "substitute".
funcref({name} [, {arglist}] [, {dict}]) funcref() Just like function(), but the returned Funcref will lookup the function by reference, not by name. This matters when the function {name} is redefined later.
Unlike function(), {name} must be an existing user function. It only works for an autoloaded function if it has already been loaded (to avoid mistakenly loading the autoload script when only intending to use the function name, use function() instead). {name} cannot be a builtin function. Returns 0 on error.
function({name} [, {arglist}] [, {dict}]) function() partial E700 E923 Return a Funcref variable that refers to function {name}. {name} can be the name of a user defined function or an internal function.
{name} can also be a Funcref or a partial. When it is a partial the dict stored in it will be used and the {dict} argument is not allowed. E.g.:
let FuncWithArg = function(dict.Func, [arg])
let Broken = function(dict.Func, [arg], dict)
When using the Funcref the function will be found by {name}, also when it was redefined later. Use funcref() to keep the same function.
When {arglist} or {dict} is present this creates a partial. That means the argument list and/or the dictionary is stored in the Funcref and will be used when the Funcref is called.
The arguments are passed to the function in front of other arguments, but after any argument from method. Example:
func Callback(arg1, arg2, name)
"...
endfunc
let Partial = function('Callback', ['one', 'two'])
"...
call Partial('name')
Invokes the function as with:
call Callback('one', 'two', 'name')
With a method:
func Callback(one, two, three)
"...
endfunc
let Partial = function('Callback', ['two'])
"...
eval 'one'->Partial('three')
Invokes the function as with:
call Callback('one', 'two', 'three')
The function() call can be nested to add more arguments to the Funcref. The extra arguments are appended to the list of arguments. Example:
func Callback(arg1, arg2, name)
"...
endfunc
let Func = function('Callback', ['one'])
let Func2 = function(Func, ['two'])
"...
call Func2('name')
Invokes the function as with:
call Callback('one', 'two', 'name')
The Dictionary is only useful when calling a "dict" function. In that case the {dict} is passed in as "self". Example:
function Callback() dict
   echo "called for " .. self.name
endfunction
"...
let context = {"name": "example"}
let Func = function('Callback', context)
"...
call Func()        " will echo: called for example
The use of function() is not needed when there are no extra arguments, these two are equivalent, if Callback() is defined as context.Callback():
let Func = function('Callback', context)
let Func = context.Callback
The argument list and the Dictionary can be combined:
function Callback(arg1, count) dict
"...
endfunction
let context = {"name": "example"}
let Func = function('Callback', ['one'], context)
"...
call Func(500)
Invokes the function as with:
call context.Callback('one', 500)
Returns 0 on error.
garbagecollect([{atexit}]) garbagecollect() Cleanup unused Lists and Dictionaries that have circular references.
There is hardly ever a need to invoke this function, as it is automatically done when Vim runs out of memory or is waiting for the user to press a key after 'updatetime'. Items without circular references are always freed when they become unused. This is useful if you have deleted a very big List and/or Dictionary with circular references in a script that runs for a long time.
When the optional {atexit} argument is one, garbage collection will also be done when exiting Vim, if it wasn't done before. This is useful when checking for memory leaks.
The garbage collection is not done immediately but only when it's safe to perform. This is when waiting for the user to type a character.
get({list}, {idx} [, {default}]) get() Get item {idx} from List {list}. When this item is not available return {default}. Return zero when {default} is omitted.
get({blob}, {idx} [, {default}]) Get byte {idx} from Blob {blob}. When this byte is not available return {default}. Return -1 when {default} is omitted.
get({dict}, {key} [, {default}]) Get item with key {key} from Dictionary {dict}. When this item is not available return {default}. Return zero when {default} is omitted. Useful example:
let val = get(g:, 'var_name', 'default')
This gets the value of g:var_name if it exists, and uses "default" when it does not exist.
get({func}, {what}) Get item {what} from Funcref {func}. Possible values for {what} are: "name" The function name "func" The function "dict" The dictionary "args" The list with arguments Returns zero on error.
getbufinfo([{buf}]) getbufinfo() getbufinfo([{dict}]) Get information about buffers as a List of Dictionaries.
Without an argument information about all the buffers is returned.
When the argument is a Dictionary only the buffers matching the specified criteria are returned. The following keys can be specified in {dict}: buflisted include only listed buffers. bufloaded include only loaded buffers. bufmodified include only modified buffers.
Otherwise, {buf} specifies a particular buffer to return information for. For the use of {buf}, see bufname() above. If the buffer is found the returned List has one item. Otherwise the result is an empty list.
Each returned List item is a dictionary with the following entries: bufnr Buffer number. changed TRUE if the buffer is modified. changedtick Number of changes made to the buffer. command TRUE if the buffer belongs to the command-line window cmdwin. hidden TRUE if the buffer is hidden. lastused Timestamp in seconds, like localtime(), when the buffer was last used. listed TRUE if the buffer is listed. lnum Line number used for the buffer when opened in the current window. Only valid if the buffer has been displayed in the window in the past. If you want the line number of the last known cursor position in a given window, use line():
echo line('.', {winid})
linecount Number of lines in the buffer (only valid when loaded) loaded TRUE if the buffer is loaded. name Full path to the file in the buffer. signs List of signs placed in the buffer. Each list item is a dictionary with the following fields: id sign identifier lnum line number name sign name variables A reference to the dictionary with buffer-local variables. windows List of window-IDs that display this buffer
Examples:
for buf in getbufinfo()
    echo buf.name
endfor
for buf in getbufinfo({'buflisted':1})
    if buf.changed
        " ....
    endif
endfor
To get buffer-local options use:
getbufvar({bufnr}, '&option_name')
getbufline({buf}, {lnum} [, {end}]) getbufline() Return a List with the lines starting from {lnum} to {end} (inclusive) in the buffer {buf}. If {end} is omitted, a List with only the line {lnum} is returned. See getbufoneline() for only getting the line.
For the use of {buf}, see bufname() above.
For {lnum} and {end} "$" can be used for the last line of the buffer. Otherwise a number must be used.
When {lnum} is smaller than 1 or bigger than the number of lines in the buffer, an empty List is returned.
When {end} is greater than the number of lines in the buffer, it is treated as {end} is set to the number of lines in the buffer. When {end} is before {lnum} an empty List is returned.
This function works only for loaded buffers. For unloaded and non-existing buffers, an empty List is returned.
Example:
let lines = getbufline(bufnr("myfile"), 1, "$")
getbufoneline({buf}, {lnum}) getbufoneline() Just like getbufline() but only get one line and return it as a string.
getbufvar({buf}, {varname} [, {def}]) getbufvar() The result is the value of option or local buffer variable {varname} in buffer {buf}. Note that the name without "b:" must be used. The {varname} argument is a string. When {varname} is empty returns a Dictionary with all the buffer-local variables. When {varname} is equal to "&" returns a Dictionary with all the buffer-local options. Otherwise, when {varname} starts with "&" returns the value of a buffer-local option. This also works for a global or buffer-local option, but it doesn't work for a global variable, window-local variable or window-local option. For the use of {buf}, see bufname() above. When the buffer or variable doesn't exist {def} or an empty string is returned, there is no error message. Examples:
let bufmodified = getbufvar(1, "&mod")
echo "todo myvar = " .. getbufvar("todo", "myvar")
getcellwidths() getcellwidths() Returns a List of cell widths of character ranges overridden by setcellwidths(). The format is equal to the argument of setcellwidths(). If no character ranges have their cell widths overridden, an empty List is returned.
getchangelist([{buf}]) getchangelist() Returns the changelist for the buffer {buf}. For the use of {buf}, see bufname() above. If buffer {buf} doesn't exist, an empty list is returned.
The returned list contains two entries: a list with the change locations and the current position in the list. Each entry in the change list is a dictionary with the following entries: col column number coladd column offset for 'virtualedit' lnum line number If buffer {buf} is the current buffer, then the current position refers to the position in the list. For other buffers, it is set to the length of the list.
getchar([expr]) getchar() Get a single character from the user or input stream. If [expr] is omitted, wait until a character is available. If [expr] is 0, only get a character when one is available. Return zero otherwise. If [expr] is 1, only check if a character is available, it is not consumed. Return zero if no character available. If you prefer always getting a string use getcharstr().
Without [expr] and when [expr] is 0 a whole character or special key is returned. If it is a single character, the result is a Number. Use nr2char() to convert it to a String. Otherwise a String is returned with the encoded character. For a special key it's a String with a sequence of bytes starting with 0x80 (decimal: 128). This is the same value as the String "\<Key>", e.g., "\<Left>". The returned value is also a String when a modifier (shift, control, alt) was used that is not included in the character.
When [expr] is 0 and Esc is typed, there will be a short delay while Vim waits to see if this is the start of an escape sequence.
When [expr] is 1 only the first byte is returned. For a one-byte character it is the character itself as a number. Use nr2char() to convert it to a String.
Use getcharmod() to obtain any additional modifiers.
When the user clicks a mouse button, the mouse event will be returned. The position can then be found in v:mouse_col, v:mouse_lnum, v:mouse_winid and v:mouse_win. getmousepos() can also be used. Mouse move events will be ignored. This example positions the mouse as it would normally happen:
let c = getchar()
if c == "\<LeftMouse>" && v:mouse_win > 0
  exe v:mouse_win .. "wincmd w"
  exe v:mouse_lnum
  exe "normal " .. v:mouse_col .. "|"
endif
There is no prompt, you will somehow have to make clear to the user that a character has to be typed. The screen is not redrawn, e.g. when resizing the window.
There is no mapping for the character. Key codes are replaced, thus when the user presses the <Del> key you get the code for the <Del> key, not the raw character sequence. Examples:
getchar() == "\<Del>"
getchar() == "\<S-Left>"
This example redefines "f" to ignore case:
nmap f :call FindChar()<CR>
function FindChar()
  let c = nr2char(getchar())
  while col('.') < col('$') - 1
    normal l
    if getline('.')[col('.') - 1] ==? c
      break
    endif
  endwhile
endfunction
getcharmod() getcharmod() The result is a Number which is the state of the modifiers for the last obtained character with getchar() or in another way. These values are added together: 2 shift 4 control 8 alt (meta) 16 meta (when it's different from ALT) 32 mouse double click 64 mouse triple click 96 mouse quadruple click (== 32 + 64) 128 command (Mac) or super Only the modifiers that have not been included in the character itself are obtained. Thus Shift-a results in "A" without a modifier. Returns 0 if no modifiers are used.
getcharpos({expr}) getcharpos() Get the position for String {expr}. Same as getpos() but the column number in the returned List is a character index instead of a byte index. If getpos() returns a very large column number, equal to v:maxcol, then getcharpos() will return the character index of the last character.
Example: With the cursor on 'μ„Έ' in line 5 with text "μ—¬λ³΄μ„Έμš”":
getcharpos('.')                returns [0, 5, 3, 0]
getpos('.')                returns [0, 5, 7, 0]
getcharsearch() getcharsearch() Return the current character search information as a {dict} with the following entries:
char character previously used for a character search (t, f, T, or F); empty string if no character search has been performed forward direction of character search; 1 for forward, 0 for backward until type of character search; 1 for a t or T character search, 0 for an f or F character search
This can be useful to always have ; and , search forward/backward regardless of the direction of the previous character search:
nnoremap <expr> ; getcharsearch().forward ? ';' : ','
nnoremap <expr> , getcharsearch().forward ? ',' : ';'
Also see setcharsearch().
getcharstr([expr]) getcharstr() Get a single character from the user or input stream as a string. If [expr] is omitted, wait until a character is available. If [expr] is 0 or false, only get a character when one is available. Return an empty string otherwise. If [expr] is 1 or true, only check if a character is available, it is not consumed. Return an empty string if no character is available. Otherwise this works like getchar(), except that a number result is converted to a string.
getcmdcompltype() getcmdcompltype() Return the type of the current command-line completion. Only works when the command line is being edited, thus requires use of c_CTRL-\_e or c_CTRL-R_=. See :command-completion for the return string. Also see getcmdtype(), setcmdpos(), getcmdline() and setcmdline(). Returns an empty string when completion is not defined.
getcmdline() getcmdline() Return the current command-line. Only works when the command line is being edited, thus requires use of c_CTRL-\_e or c_CTRL-R_=. Example:
cmap <F7> <C-\>eescape(getcmdline(), ' \')<CR>
Also see getcmdtype(), getcmdpos(), setcmdpos() and setcmdline(). Returns an empty string when entering a password or using inputsecret().
getcmdpos() getcmdpos() Return the position of the cursor in the command line as a byte count. The first column is 1. Only works when editing the command line, thus requires use of c_CTRL-\_e or c_CTRL-R_= or an expression mapping. Returns 0 otherwise. Also see getcmdtype(), setcmdpos(), getcmdline() and setcmdline().
getcmdscreenpos() getcmdscreenpos() Return the screen position of the cursor in the command line as a byte count. The first column is 1. Instead of getcmdpos(), it adds the prompt position. Only works when editing the command line, thus requires use of c_CTRL-\_e or c_CTRL-R_= or an expression mapping. Returns 0 otherwise. Also see getcmdpos(), setcmdpos(), getcmdline() and setcmdline().
getcmdtype() getcmdtype() Return the current command-line type. Possible return values are: : normal Ex command > debug mode command debug-mode / forward search command ? backward search command @ input() command - :insert or :append command = i_CTRL-R_= Only works when editing the command line, thus requires use of c_CTRL-\_e or c_CTRL-R_= or an expression mapping. Returns an empty string otherwise. Also see getcmdpos(), setcmdpos() and getcmdline().
getcmdwintype() getcmdwintype() Return the current command-line-window type. Possible return values are the same as getcmdtype(). Returns an empty string when not in the command-line window.
getcompletion({pat}, {type} [, {filtered}]) getcompletion() Return a list of command-line completion matches. The String {type} argument specifies what for. The following completion types are supported:
arglist file names in argument list augroup autocmd groups buffer buffer names breakpoint :breakadd and :breakdel suboptions cmdline cmdline-completion result color color schemes command Ex command compiler compilers custom,{func} custom completion, defined via {func} customlist,{func} custom completion, defined via {func} diff_buffer :diffget and :diffput completion dir directory names environment environment variable names event autocommand events expression Vim expression file file and directory names file_in_path file and directory names in 'path' filetype filetype names 'filetype' function function name help help subjects highlight highlight groups history :history suboptions keymap keyboard mappings locale locale names (as output of locale -a) mapclear buffer argument mapping mapping name menu menus messages :messages suboptions option options packadd optional package pack-add names runtime :runtime completion scriptnames sourced script names :scriptnames shellcmd Shell command sign :sign suboptions syntax syntax file names 'syntax' syntime :syntime suboptions tag tags tag_listfiles tags, file names user user names var user variables
If {pat} is an empty string, then all the matches are returned. Otherwise only items matching {pat} are returned. See wildcards for the use of special characters in {pat}.
If the optional {filtered} flag is set to 1, then 'wildignore' is applied to filter the results. Otherwise all the matches are returned. The 'wildignorecase' option always applies.
If the 'wildoptions' option contains "fuzzy", then fuzzy matching is used to get the completion matches. Otherwise regular expression matching is used. Thus this function follows the user preference, what happens on the command line. If you do not want this you can make 'wildoptions' empty before calling getcompletion() and restore it afterwards.
If {type} is "cmdline", then the cmdline-completion result is returned. For example, to complete the possible values after a ":call" command:
echo getcompletion('call ', 'cmdline')
If there are no matches, an empty list is returned. An invalid value for {type} produces an error.
getcurpos([{winid}]) getcurpos() Get the position of the cursor. This is like getpos('.'), but includes an extra "curswant" item in the list:
[0, lnum, col, off, curswant]
The "curswant" number is the preferred column when moving the cursor vertically. After $ command it will be a very large number equal to v:maxcol. Also see getcursorcharpos() and getpos(). The first "bufnum" item is always zero. The byte position of the cursor is returned in "col". To get the character position, use getcursorcharpos().
The optional {winid} argument can specify the window. It can be the window number or the window-ID. The last known cursor position is returned, this may be invalid for the current value of the buffer if it is not the current window. If {winid} is invalid a list with zeroes is returned.
This can be used to save and restore the cursor position:
let save_cursor = getcurpos()
MoveTheCursorAround
call setpos('.', save_cursor)
Note that this only works within the window. See winrestview() for restoring more state.
getcursorcharpos([{winid}]) getcursorcharpos() Same as getcurpos() but the column number in the returned List is a character index instead of a byte index.
Example: With the cursor on '보' in line 3 with text "μ—¬λ³΄μ„Έμš”":
getcursorcharpos()        " returns [0, 3, 2, 0, 3]
getcurpos()                " returns [0, 3, 4, 0, 3]
getcwd([{winnr} [, {tabnr}]]) getcwd() With no arguments, returns the name of the effective current-directory. With {winnr} or {tabnr} the working directory of that scope is returned, and 'autochdir' is ignored. Tabs and windows are identified by their respective numbers, 0 means current tab or window. Missing tab number implies 0. Thus the following are equivalent:
getcwd(0)
getcwd(0, 0)
If {winnr} is -1 it is ignored, only the tab is resolved. {winnr} can be the window number or the window-ID. If both {winnr} and {tabnr} are -1 the global working directory is returned. Throw error if the arguments are invalid. E5000 E5001 E5002
getenv({name}) getenv() Return the value of environment variable {name}. The {name} argument is a string, without a leading '$'. Example:
myHome = getenv('HOME')
When the variable does not exist v:null is returned. That is different from a variable set to an empty string. See also expr-env.
getfontname([{name}]) getfontname() Without an argument returns the name of the normal font being used. Like what is used for the Normal highlight group hl-Normal. With an argument a check is done whether String {name} is a valid font name. If not then an empty string is returned. Otherwise the actual font name is returned, or {name} if the GUI does not support obtaining the real name. Only works when the GUI is running, thus not in your vimrc or gvimrc file. Use the GUIEnter autocommand to use this function just after the GUI has started.
getfperm({fname}) getfperm() The result is a String, which is the read, write, and execute permissions of the given file {fname}. If {fname} does not exist or its directory cannot be read, an empty string is returned. The result is of the form "rwxrwxrwx", where each group of "rwx" flags represent, in turn, the permissions of the owner of the file, the group the file belongs to, and other users. If a user does not have a given permission the flag for this is replaced with the string "-". Examples:
echo getfperm("/etc/passwd")
echo getfperm(expand("~/.config/nvim/init.vim"))
This will hopefully (from a security point of view) display the string "rw-r--r--" or even "rw-------".
For setting permissions use setfperm().
getfsize({fname}) getfsize() The result is a Number, which is the size in bytes of the given file {fname}. If {fname} is a directory, 0 is returned. If the file {fname} can't be found, -1 is returned. If the size of {fname} is too big to fit in a Number then -2 is returned.
getftime({fname}) getftime() The result is a Number, which is the last modification time of the given file {fname}. The value is measured as seconds since 1st Jan 1970, and may be passed to strftime(). See also localtime() and strftime(). If the file {fname} can't be found -1 is returned.
getftype({fname}) getftype() The result is a String, which is a description of the kind of file of the given file {fname}. If {fname} does not exist an empty string is returned. Here is a table over different kinds of files and their results: Normal file "file" Directory "dir" Symbolic link "link" Block device "bdev" Character device "cdev" Socket "socket" FIFO "fifo" All other "other" Example:
getftype("/home")
Note that a type such as "link" will only be returned on systems that support it. On some systems only "dir" and "file" are returned.
getjumplist([{winnr} [, {tabnr}]]) getjumplist() Returns the jumplist for the specified window.
Without arguments use the current window. With {winnr} only use this window in the current tab page. {winnr} can also be a window-ID. With {winnr} and {tabnr} use the window in the specified tab page. If {winnr} or {tabnr} is invalid, an empty list is returned.
The returned list contains two entries: a list with the jump locations and the last used jump position number in the list. Each entry in the jump location list is a dictionary with the following entries: bufnr buffer number col column number coladd column offset for 'virtualedit' filename filename if available lnum line number
getline({lnum} [, {end}]) getline() Without {end} the result is a String, which is line {lnum} from the current buffer. Example:
getline(1)
When {lnum} is a String that doesn't start with a digit, line() is called to translate the String into a Number. To get the line under the cursor:
getline(".")
When {lnum} is a number smaller than 1 or bigger than the number of lines in the buffer, an empty string is returned.
When {end} is given the result is a List where each item is a line from the current buffer in the range {lnum} to {end}, including line {end}. {end} is used in the same way as {lnum}. Non-existing lines are silently omitted. When {end} is before {lnum} an empty List is returned. Example:
let start = line('.')
let end = search("^$") - 1
let lines = getline(start, end)
To get lines from another buffer see getbufline() and getbufoneline()
getloclist({nr} [, {what}]) getloclist() Returns a List with all the entries in the location list for window {nr}. {nr} can be the window number or the window-ID. When {nr} is zero the current window is used.
For a location list window, the displayed location list is returned. For an invalid window number {nr}, an empty list is returned. Otherwise, same as getqflist().
If the optional {what} dictionary argument is supplied, then returns the items listed in {what} as a dictionary. Refer to getqflist() for the supported items in {what}.
In addition to the items supported by getqflist() in {what}, the following item is supported by getloclist():
filewinid id of the window used to display files from the location list. This field is applicable only when called from a location list window. See location-list-file-window for more details.
Returns a Dictionary with default values if there is no location list for the window {nr}. Returns an empty Dictionary if window {nr} does not exist.
Examples (See also getqflist-examples):
echo getloclist(3, {'all': 0})
echo getloclist(5, {'filewinid': 0})
getmarklist([{buf}]) getmarklist() Without the {buf} argument returns a List with information about all the global marks. mark
If the optional {buf} argument is specified, returns the local marks defined in buffer {buf}. For the use of {buf}, see bufname(). If {buf} is invalid, an empty list is returned.
Each item in the returned List is a Dict with the following: mark name of the mark prefixed by "'" pos a List with the position of the mark: [bufnum, lnum, col, off] Refer to getpos() for more information. file file name
Refer to getpos() for getting information about a specific mark.
getmatches([{win}]) getmatches() Returns a List with all matches previously defined for the current window by matchadd() and the :match commands. getmatches() is useful in combination with setmatches(), as setmatches() can restore a list of matches saved by getmatches(). If {win} is specified, use the window with this number or window ID instead of the current window. If {win} is invalid, an empty list is returned. Example:
echo getmatches()
[{"group": "MyGroup1", "pattern": "TODO",
"priority": 10, "id": 1}, {"group": "MyGroup2",
"pattern": "FIXME", "priority": 10, "id": 2}]
let m = getmatches()
call clearmatches()
echo getmatches()
[]
call setmatches(m)
echo getmatches()
[{"group": "MyGroup1", "pattern": "TODO",
"priority": 10, "id": 1}, {"group": "MyGroup2",
"pattern": "FIXME", "priority": 10, "id": 2}]
unlet m
getmousepos() getmousepos() Returns a Dictionary with the last known position of the mouse. This can be used in a mapping for a mouse click. The items are: screenrow screen row screencol screen column winid Window ID of the click winrow row inside "winid" wincol column inside "winid" line text line inside "winid" column text column inside "winid" coladd offset (in screen columns) from the start of the clicked char All numbers are 1-based.
If not over a window, e.g. when in the command line, then only "screenrow" and "screencol" are valid, the others are zero.
When on the status line below a window or the vertical separator right of a window, the "line" and "column" values are zero.
When the position is after the text then "column" is the length of the text in bytes plus one.
If the mouse is over a focusable floating window then that window is used.
When using getchar() the Vim variables v:mouse_lnum, v:mouse_col and v:mouse_winid also provide these values.
getpid() getpid() Return a Number which is the process ID of the Vim process. This is a unique number, until Vim exits.
getpos({expr}) getpos() Get the position for String {expr}. For possible values of {expr} see line(). For getting the cursor position see getcurpos(). The result is a List with four numbers: [bufnum, lnum, col, off] "bufnum" is zero, unless a mark like '0 or 'A is used, then it is the buffer number of the mark. "lnum" and "col" are the position in the buffer. The first column is 1. The "off" number is zero, unless 'virtualedit' is used. Then it is the offset in screen columns from the start of the character. E.g., a position within a <Tab> or after the last character. Note that for '< and '> Visual mode matters: when it is "V" (visual line mode) the column of '< is zero and the column of '> is a large number equal to v:maxcol. The column number in the returned List is the byte position within the line. To get the character position in the line, use getcharpos(). A very large column number equal to v:maxcol can be returned, in which case it means "after the end of the line". If {expr} is invalid, returns a list with all zeros. This can be used to save and restore the position of a mark:
let save_a_mark = getpos("'a")
" ...
call setpos("'a", save_a_mark)
getqflist([{what}]) getqflist() Returns a List with all the current quickfix errors. Each list item is a dictionary with these entries: bufnr number of buffer that has the file name, use bufname() to get the name module module name lnum line number in the buffer (first line is 1) end_lnum end of line number if the item is multiline col column number (first column is 1) end_col end of column number if the item has range vcol TRUE: "col" is visual column FALSE: "col" is byte index nr error number pattern search pattern used to locate the error text description of the error type type of the error, 'E', '1', etc. valid TRUE: recognized error message user_data custom data associated with the item, can be any type.
When there is no error list or it's empty, an empty list is returned. Quickfix list entries with a non-existing buffer number are returned with "bufnr" set to zero (Note: some functions accept buffer number zero for the alternate buffer, you may need to explicitly check for zero).
Useful application: Find pattern matches in multiple files and do something with them:
vimgrep /theword/jg *.c
for d in getqflist()
   echo bufname(d.bufnr) ':' d.lnum '=' d.text
endfor
If the optional {what} dictionary argument is supplied, then returns only the items listed in {what} as a dictionary. The following string items are supported in {what}: changedtick get the total number of changes made to the list quickfix-changedtick context get the quickfix-context efm errorformat to use when parsing "lines". If not present, then the 'errorformat' option value is used. id get information for the quickfix list with quickfix-ID; zero means the id for the current list or the list specified by "nr" idx get information for the quickfix entry at this index in the list specified by "id" or "nr". If set to zero, then uses the current entry. See quickfix-index items quickfix list entries lines parse a list of lines using 'efm' and return the resulting entries. Only a List type is accepted. The current quickfix list is not modified. See quickfix-parse. nr get information for this quickfix list; zero means the current quickfix list and "$" means the last quickfix list qfbufnr number of the buffer displayed in the quickfix window. Returns 0 if the quickfix buffer is not present. See quickfix-buffer. size number of entries in the quickfix list title get the list title quickfix-title winid get the quickfix window-ID all all of the above quickfix properties Non-string items in {what} are ignored. To get the value of a particular item, set it to zero. If "nr" is not present then the current quickfix list is used. If both "nr" and a non-zero "id" are specified, then the list specified by "id" is used. To get the number of lists in the quickfix stack, set "nr" to "$" in {what}. The "nr" value in the returned dictionary contains the quickfix stack size. When "lines" is specified, all the other items except "efm" are ignored. The returned dictionary contains the entry "items" with the list of entries.
The returned dictionary contains the following entries: changedtick total number of changes made to the list quickfix-changedtick context quickfix list context. See quickfix-context If not present, set to "". id quickfix list ID quickfix-ID. If not present, set to 0. idx index of the quickfix entry in the list. If not present, set to 0. items quickfix list entries. If not present, set to an empty list. nr quickfix list number. If not present, set to 0 qfbufnr number of the buffer displayed in the quickfix window. If not present, set to 0. size number of entries in the quickfix list. If not present, set to 0. title quickfix list title text. If not present, set to "". winid quickfix window-ID. If not present, set to 0
Examples (See also getqflist-examples):
echo getqflist({'all': 1})
echo getqflist({'nr': 2, 'title': 1})
echo getqflist({'lines' : ["F1:10:L10"]})
getreg([{regname} [, 1 [, {list}]]]) getreg() The result is a String, which is the contents of register {regname}. Example:
let cliptext = getreg('*')
When register {regname} was not set the result is an empty string. The {regname} argument must be a string.
getreg('=') returns the last evaluated value of the expression register. (For use in maps.) getreg('=', 1) returns the expression itself, so that it can be restored with setreg(). For other registers the extra argument is ignored, thus you can always give it.
If {list} is present and TRUE, the result type is changed to List. Each list item is one text line. Use it if you care about zero bytes possibly present inside register: without third argument both NLs and zero bytes are represented as NLs (see NL-used-for-Nul). When the register was not set an empty list is returned.
If {regname} is not specified, v:register is used.
getreginfo([{regname}]) getreginfo() Returns detailed information about register {regname} as a Dictionary with the following entries: regcontents List of lines contained in register {regname}, like getreg({regname}, 1, 1). regtype the type of register {regname}, as in getregtype(). isunnamed Boolean flag, v:true if this register is currently pointed to by the unnamed register. points_to for the unnamed register, gives the single letter name of the register currently pointed to (see quotequote). For example, after deleting a line with dd, this field will be "1", which is the register that got the deleted text.
The {regname} argument is a string. If {regname} is invalid or not set, an empty Dictionary will be returned. If {regname} is not specified, v:register is used. The returned Dictionary can be passed to setreg().
getregion({pos1}, {pos2} [, {opts}]) getregion() Returns the list of strings from {pos1} to {pos2} from a buffer.
{pos1} and {pos2} must both be Lists with four numbers. See getpos() for the format of the list. It's possible to specify positions from a different buffer, but please note the limitations at getregion-notes.
The optional argument {opts} is a Dict and supports the following items:
type Specify the region's selection type (default: "v"): "v" for charwise mode "V" for linewise mode "<CTRL-V>" for blockwise-visual mode
exclusive If TRUE, use exclusive selection for the end position (default: follow 'selection')
You can get the last selection type by visualmode(). If Visual mode is active, use mode() to get the Visual mode (e.g., in a :vmap). This function is useful to get text starting and ending in different columns, such as a charwise-visual selection.
getregion-notes Note that:
Order of {pos1} and {pos2} doesn't matter, it will always return content from the upper left position to the lower right position.
If 'virtualedit' is enabled and the region is past the end of the lines, resulting lines are padded with spaces.
If the region is blockwise and it starts or ends in the middle of a multi-cell character, it is not included but its selected part is substituted with spaces.
If {pos1} and {pos2} are not in the same buffer, an empty list is returned.
{pos1} and {pos2} must belong to a bufloaded() buffer.
It is evaluated in current window context, which makes a difference if the buffer is displayed in a window with different 'virtualedit' or 'list' values.
Examples:
:xnoremap <CR>
\ <Cmd>echom getregion(
\ getpos('v'), getpos('.'), #{ type: mode() })<CR>
getregtype([{regname}]) getregtype() The result is a String, which is type of register {regname}. The value will be one of: "v" for charwise text "V" for linewise text "<CTRL-V>{width}" for blockwise-visual text "" for an empty or unknown register <CTRL-V> is one character with value 0x16. The {regname} argument is a string. If {regname} is not specified, v:register is used.
getscriptinfo([{opts}]) getscriptinfo() Returns a List with information about all the sourced Vim scripts in the order they were sourced, like what :scriptnames shows.
The optional Dict argument {opts} supports the following optional items: name Script name match pattern. If specified, and "sid" is not specified, information about scripts with a name that match the pattern "name" are returned. sid Script ID <SID>. If specified, only information about the script with ID "sid" is returned and "name" is ignored.
Each item in the returned List is a Dict with the following items: autoload Always set to FALSE. functions List of script-local function names defined in the script. Present only when a particular script is specified using the "sid" item in {opts}. name Vim script file name. sid Script ID <SID>. variables A dictionary with the script-local variables. Present only when a particular script is specified using the "sid" item in {opts}. Note that this is a copy, the value of script-local variables cannot be changed using this dictionary. version Vim script version, always 1
Examples:
echo getscriptinfo({'name': 'myscript'})
echo getscriptinfo({'sid': 15}).variables
gettabinfo([{tabnr}]) gettabinfo() If {tabnr} is not specified, then information about all the tab pages is returned as a List. Each List item is a Dictionary. Otherwise, {tabnr} specifies the tab page number and information about that one is returned. If the tab page does not exist an empty List is returned.
Each List item is a Dictionary with the following entries: tabnr tab page number. variables a reference to the dictionary with tabpage-local variables windows List of window-IDs in the tab page.
gettabvar({tabnr}, {varname} [, {def}]) gettabvar() Get the value of a tab-local variable {varname} in tab page {tabnr}. t:var Tabs are numbered starting with one. The {varname} argument is a string. When {varname} is empty a dictionary with all tab-local variables is returned. Note that the name without "t:" must be used. When the tab or variable doesn't exist {def} or an empty string is returned, there is no error message.
gettabwinvar({tabnr}, {winnr}, {varname} [, {def}]) gettabwinvar() Get the value of window-local variable {varname} in window {winnr} in tab page {tabnr}. The {varname} argument is a string. When {varname} is empty a dictionary with all window-local variables is returned. When {varname} is equal to "&" get the values of all window-local options in a Dictionary. Otherwise, when {varname} starts with "&" get the value of a window-local option. Note that {varname} must be the name without "w:". Tabs are numbered starting with one. For the current tabpage use getwinvar(). {winnr} can be the window number or the window-ID. When {winnr} is zero the current window is used. This also works for a global option, buffer-local option and window-local option, but it doesn't work for a global variable or buffer-local variable. When the tab, window or variable doesn't exist {def} or an empty string is returned, there is no error message. Examples:
let list_is_on = gettabwinvar(1, 2, '&list')
echo "myvar = " .. gettabwinvar(3, 1, 'myvar')
To obtain all window-local variables use:
gettabwinvar({tabnr}, {winnr}, '&')
gettagstack([{winnr}]) gettagstack() The result is a Dict, which is the tag stack of window {winnr}. {winnr} can be the window number or the window-ID. When {winnr} is not specified, the current window is used. When window {winnr} doesn't exist, an empty Dict is returned.
The returned dictionary contains the following entries: curidx Current index in the stack. When at top of the stack, set to (length + 1). Index of bottom of the stack is 1. items List of items in the stack. Each item is a dictionary containing the entries described below. length Number of entries in the stack.
Each item in the stack is a dictionary with the following entries: bufnr buffer number of the current jump from cursor position before the tag jump. See getpos() for the format of the returned list. matchnr current matching tag number. Used when multiple matching tags are found for a name. tagname name of the tag
See tagstack for more information about the tag stack.
gettext({text}) gettext() Translate String {text} if possible. This is mainly for use in the distributed Vim scripts. When generating message translations the {text} is extracted by xgettext, the translator can add the translated message in the .po file and Vim will lookup the translation when gettext() is called. For {text} double quoted strings are preferred, because xgettext does not understand escaping in single quoted strings.
getwininfo([{winid}]) getwininfo() Returns information about windows as a List with Dictionaries.
If {winid} is given Information about the window with that ID is returned, as a List with one item. If the window does not exist the result is an empty list.
Without {winid} information about all the windows in all the tab pages is returned.
Each List item is a Dictionary with the following entries: botline last complete displayed buffer line bufnr number of buffer in the window height window height (excluding winbar) loclist 1 if showing a location list quickfix 1 if quickfix or location list window terminal 1 if a terminal window tabnr tab page number topline first displayed buffer line variables a reference to the dictionary with window-local variables width window width winbar 1 if the window has a toolbar, 0 otherwise wincol leftmost screen column of the window; "col" from win_screenpos() textoff number of columns occupied by any 'foldcolumn', 'signcolumn' and line number in front of the text winid window-ID winnr window number winrow topmost screen line of the window; "row" from win_screenpos()
getwinpos([{timeout}]) getwinpos() The result is a List with two numbers, the result of getwinposx() and getwinposy() combined: [x-pos, y-pos] {timeout} can be used to specify how long to wait in msec for a response from the terminal. When omitted 100 msec is used.
Use a longer time for a remote terminal. When using a value less than 10 and no response is received within that time, a previously reported position is returned, if available. This can be used to poll for the position and do some work in the meantime:
while 1
  let res = getwinpos(1)
  if res[0] >= 0
    break
  endif
  " Do some work here
endwhile
getwinposx() getwinposx() The result is a Number, which is the X coordinate in pixels of the left hand side of the GUI Vim window. The result will be -1 if the information is not available. The value can be used with :winpos.
getwinposy() getwinposy() The result is a Number, which is the Y coordinate in pixels of the top of the GUI Vim window. The result will be -1 if the information is not available. The value can be used with :winpos.
getwinvar({winnr}, {varname} [, {def}]) getwinvar() Like gettabwinvar() for the current tabpage. Examples:
let list_is_on = getwinvar(2, '&list')
echo "myvar = " .. getwinvar(1, 'myvar')
glob({expr} [, {nosuf} [, {list} [, {alllinks}]]]) glob() Expand the file wildcards in {expr}. See wildcards for the use of special characters.
Unless the optional {nosuf} argument is given and is TRUE, the 'suffixes' and 'wildignore' options apply: Names matching one of the patterns in 'wildignore' will be skipped and 'suffixes' affect the ordering of matches. 'wildignorecase' always applies.
When {list} is present and it is TRUE the result is a List with all matching files. The advantage of using a List is, you also get filenames containing newlines correctly. Otherwise the result is a String and when there are several matches, they are separated by <NL> characters.
If the expansion fails, the result is an empty String or List.
You can also use readdir() if you need to do complicated things, such as limiting the number of matches.
A name for a non-existing file is not included. A symbolic link is only included if it points to an existing file. However, when the {alllinks} argument is present and it is TRUE then all symbolic links are included.
For most systems backticks can be used to get files names from any external command. Example:
let tagfiles = glob("`find . -name tags -print`")
let &tags = substitute(tagfiles, "\n", ",", "g")
The result of the program inside the backticks should be one item per line. Spaces inside an item are allowed.
See expand() for expanding special Vim variables. See system() for getting the raw output of an external command.
glob2regpat({string}) glob2regpat() Convert a file pattern, as used by glob(), into a search pattern. The result can be used to match with a string that is a file name. E.g.
if filename =~ glob2regpat('Make*.mak')
  " ...
endif
This is equivalent to:
if filename =~ '^Make.*\.mak$'
  " ...
endif
When {string} is an empty string the result is "^$", match an empty string. Note that the result depends on the system. On MS-Windows a backslash usually means a path separator.
globpath({path}, {expr} [, {nosuf} [, {list} [, {allinks}]]]) globpath() Perform glob() for String {expr} on all directories in {path} and concatenate the results. Example:
echo globpath(&rtp, "syntax/c.vim")
{path} is a comma-separated list of directory names. Each directory name is prepended to {expr} and expanded like with glob(). A path separator is inserted when needed. To add a comma inside a directory name escape it with a backslash. Note that on MS-Windows a directory may have a trailing backslash, remove it if you put a comma after it. If the expansion fails for one of the directories, there is no error message.
Unless the optional {nosuf} argument is given and is TRUE, the 'suffixes' and 'wildignore' options apply: Names matching one of the patterns in 'wildignore' will be skipped and 'suffixes' affect the ordering of matches.
When {list} is present and it is TRUE the result is a List with all matching files. The advantage of using a List is, you also get filenames containing newlines correctly. Otherwise the result is a String and when there are several matches, they are separated by <NL> characters. Example:
echo globpath(&rtp, "syntax/c.vim", 0, 1)
{allinks} is used as with glob().
The "**" item can be used to search in a directory tree. For example, to find all "README.txt" files in the directories in 'runtimepath' and below:
echo globpath(&rtp, "**/README.txt")
Upwards search and limiting the depth of "**" is not supported, thus using 'path' will not always work properly.
has({feature}) has() Returns 1 if {feature} is supported, 0 otherwise. The {feature} argument is a feature name like "nvim-0.2.1" or "win32", see below. See also exists().
To get the system name use vim.uv.os_uname() in Lua:
print(vim.uv.os_uname().sysname)
If the code has a syntax error then Vimscript may skip the rest of the line. Put :if and :endif on separate lines to avoid the syntax error:
if has('feature')
  let x = this_breaks_without_the_feature()
endif
Vim's compile-time feature-names (prefixed with "+") are not recognized because Nvim is always compiled with all possible features. feature-compile
Feature names can be: 1. Nvim version. For example the "nvim-0.2.1" feature means that Nvim is version 0.2.1 or later:
if has("nvim-0.2.1")
  " ...
endif
2. Runtime condition or other pseudo-feature. For example the "win32" feature checks if the current system is Windows:
if has("win32")
  " ...
endif
feature-list List of supported pseudo-feature names: acl ACL support. bsd BSD system (not macOS, use "mac" for that). clipboard clipboard provider is available. fname_case Case in file names matters (for Darwin and MS-Windows this is not present). gui_running Nvim has a GUI. iconv Can use iconv() for conversion. linux Linux system. mac MacOS system. nvim This is Nvim. python3 Legacy Vim python3 interface. has-python pythonx Legacy Vim python_x interface. has-pythonx sun SunOS system. ttyin input is a terminal (tty). ttyout output is a terminal (tty). unix Unix system. vim_starting True during startup. win32 Windows system (32 or 64 bit). win64 Windows system (64 bit). wsl WSL (Windows Subsystem for Linux) system.
has-patch 3. Vim patch. For example the "patch123" feature means that Vim patch 123 at the current v:version was included:
if v:version > 602 || v:version == 602 && has("patch148")
  " ...
endif
4. Vim version. For example the "patch-7.4.237" feature means that Nvim is Vim-compatible to version 7.4.237 or later.
if has("patch-7.4.237")
  " ...
endif
has_key({dict}, {key}) has_key() The result is a Number, which is TRUE if Dictionary {dict} has an entry with key {key}. FALSE otherwise. The {key} argument is a string.
haslocaldir([{winnr} [, {tabnr}]]) haslocaldir() The result is a Number, which is 1 when the window has set a local path via :lcd or when {winnr} is -1 and the tabpage has set a local path via :tcd, otherwise 0.
Tabs and windows are identified by their respective numbers, 0 means current tab or window. Missing argument implies 0. Thus the following are equivalent:
echo haslocaldir()
echo haslocaldir(0)
echo haslocaldir(0, 0)
With {winnr} use that window in the current tabpage. With {winnr} and {tabnr} use the window in that tabpage. {winnr} can be the window number or the window-ID. If {winnr} is -1 it is ignored, only the tab is resolved. Throw error if the arguments are invalid. E5000 E5001 E5002
hasmapto({what} [, {mode} [, {abbr}]]) hasmapto() The result is a Number, which is TRUE if there is a mapping that contains {what} in somewhere in the rhs (what it is mapped to) and this mapping exists in one of the modes indicated by {mode}. The arguments {what} and {mode} are strings. When {abbr} is there and it is TRUE use abbreviations instead of mappings. Don't forget to specify Insert and/or Command-line mode. Both the global mappings and the mappings local to the current buffer are checked for a match. If no matching mapping is found FALSE is returned. The following characters are recognized in {mode}: n Normal mode v Visual and Select mode x Visual mode s Select mode o Operator-pending mode i Insert mode l Language-Argument ("r", "f", "t", etc.) c Command-line mode When {mode} is omitted, "nvo" is used.
This function is useful to check if a mapping already exists to a function in a Vim script. Example:
if !hasmapto('\ABCdoit')
   map <Leader>d \ABCdoit
endif
This installs the mapping to "\ABCdoit" only if there isn't already a mapping to "\ABCdoit".
histadd({history}, {item}) histadd() Add the String {item} to the history {history} which can be one of: hist-names "cmd" or ":" command line history "search" or "/" search pattern history "expr" or "=" typed expression history "input" or "@" input line history "debug" or ">" debug command history empty the current or last used history The {history} string does not need to be the whole name, one character is sufficient. If {item} does already exist in the history, it will be shifted to become the newest entry. The result is a Number: TRUE if the operation was successful, otherwise FALSE is returned.
Example:
call histadd("input", strftime("%Y %b %d"))
let date=input("Enter date: ")
This function is not available in the sandbox.
histdel({history} [, {item}]) histdel() Clear {history}, i.e. delete all its entries. See hist-names for the possible values of {history}.
If the parameter {item} evaluates to a String, it is used as a regular expression. All entries matching that expression will be removed from the history (if there are any). Upper/lowercase must match, unless "\c" is used /\c. If {item} evaluates to a Number, it will be interpreted as an index, see :history-indexing. The respective entry will be removed if it exists.
The result is TRUE for a successful operation, otherwise FALSE is returned.
Examples: Clear expression register history:
call histdel("expr")
Remove all entries starting with "*" from the search history:
call histdel("/", '^\*')
The following three are equivalent:
call histdel("search", histnr("search"))
call histdel("search", -1)
call histdel("search", '^' .. histget("search", -1) .. '$')
To delete the last search pattern and use the last-but-one for the "n" command and 'hlsearch':
call histdel("search", -1)
let @/ = histget("search", -1)
histget({history} [, {index}]) histget() The result is a String, the entry with Number {index} from {history}. See hist-names for the possible values of {history}, and :history-indexing for {index}. If there is no such entry, an empty String is returned. When {index} is omitted, the most recent item from the history is used.
Examples: Redo the second last search from history.
execute '/' .. histget("search", -2)
Define an Ex command ":H {num}" that supports re-execution of the {num}th entry from the output of :history.
command -nargs=1 H execute histget("cmd", 0+<args>)
histnr({history}) histnr() The result is the Number of the current entry in {history}. See hist-names for the possible values of {history}. If an error occurred, -1 is returned.
Example:
let inp_index = histnr("expr")
hlID({name}) hlID() The result is a Number, which is the ID of the highlight group with name {name}. When the highlight group doesn't exist, zero is returned. This can be used to retrieve information about the highlight group. For example, to get the background color of the "Comment" group:
echo synIDattr(synIDtrans(hlID("Comment")), "bg")
hlexists({name}) hlexists() The result is a Number, which is TRUE if a highlight group called {name} exists. This is when the group has been defined in some way. Not necessarily when highlighting has been defined for it, it may also have been used for a syntax item.
hostname() hostname() The result is a String, which is the name of the machine on which Vim is currently running. Machine names greater than 256 characters long are truncated.
iconv({string}, {from}, {to}) iconv() The result is a String, which is the text {string} converted from encoding {from} to encoding {to}. When the conversion completely fails an empty string is returned. When some characters could not be converted they are replaced with "?". The encoding names are whatever the iconv() library function can accept, see ":!man 3 iconv". Note that Vim uses UTF-8 for all Unicode encodings, conversion from/to UCS-2 is automatically changed to use UTF-8. You cannot use UCS-2 in a string anyway, because of the NUL bytes.
id({expr}) id() Returns a String which is a unique identifier of the container type (List, Dict, Blob and Partial). It is guaranteed that for the mentioned types id(v1) ==# id(v2) returns true iff type(v1) == type(v2) && v1 is v2. Note that v:_null_string, v:_null_list, v:_null_dict and v:_null_blob have the same id() with different types because they are internally represented as NULL pointers. id() returns a hexadecimal representanion of the pointers to the containers (i.e. like 0x994a40), same asprintf("%p", {expr})`, but it is advised against counting on the exact format of the return value.
It is not guaranteed that id(no_longer_existing_container) will not be equal to some other id(): new containers may reuse identifiers of the garbage-collected ones.
indent({lnum}) indent() The result is a Number, which is indent of line {lnum} in the current buffer. The indent is counted in spaces, the value of 'tabstop' is relevant. {lnum} is used just like in getline(). When {lnum} is invalid -1 is returned.
index({object}, {expr} [, {start} [, {ic}]]) index() Find {expr} in {object} and return its index. See indexof() for using a lambda to select the item.
If {object} is a List return the lowest index where the item has a value equal to {expr}. There is no automatic conversion, so the String "4" is different from the Number 4. And the Number 4 is different from the Float 4.0. The value of 'ignorecase' is not used here, case matters as indicated by the {ic} argument.
If {object} is a Blob return the lowest index where the byte value is equal to {expr}.
If {start} is given then start looking at the item with index {start} (may be negative for an item relative to the end).
When {ic} is given and it is TRUE, ignore case. Otherwise case must match.
-1 is returned when {expr} is not found in {object}. Example:
let idx = index(words, "the")
if index(numbers, 123) >= 0
  " ...
endif
indexof({object}, {expr} [, {opts}]) indexof() Returns the index of an item in {object} where {expr} is v:true. {object} must be a List or a Blob.
If {object} is a List, evaluate {expr} for each item in the List until the expression is v:true and return the index of this item.
If {object} is a Blob evaluate {expr} for each byte in the Blob until the expression is v:true and return the index of this byte.
{expr} must be a string or Funcref.
If {expr} is a string: If {object} is a List, inside {expr} v:key has the index of the current List item and v:val has the value of the item. If {object} is a Blob, inside {expr} v:key has the index of the current byte and v:val has the byte value.
If {expr} is a Funcref it must take two arguments: 1. the key or the index of the current item. 2. the value of the current item. The function must return TRUE if the item is found and the search should stop.
The optional argument {opts} is a Dict and supports the following items: startidx start evaluating {expr} at the item with this index; may be negative for an item relative to the end Returns -1 when {expr} evaluates to v:false for all the items. Example:
let l = [#{n: 10}, #{n: 20}, #{n: 30}]
echo indexof(l, "v:val.n == 20")
echo indexof(l, {i, v -> v.n == 30})
echo indexof(l, "v:val.n == 20", #{startidx: 1})
input({prompt} [, {text} [, {completion}]]) input()
input({opts}) The result is a String, which is whatever the user typed on the command-line. The {prompt} argument is either a prompt string, or a blank string (for no prompt). A '\n' can be used in the prompt to start a new line.
In the second form it accepts a single dictionary with the following keys, any of which may be omitted:
Key Default Description
prompt "" Same as {prompt} in the first form. default "" Same as {text} in the first form. completion nothing Same as {completion} in the first form. cancelreturn "" The value returned when the dialog is cancelled. highlight nothing Highlight handler: Funcref.
The highlighting set with :echohl is used for the prompt. The input is entered just like a command-line, with the same editing commands and mappings. There is a separate history for lines typed for input(). Example:
if input("Coffee or beer? ") == "beer"
  echo "Cheers!"
endif
If the optional {text} argument is present and not empty, this is used for the default reply, as if the user typed this. Example:
let color = input("Color? ", "white")
The optional {completion} argument specifies the type of completion supported for the input. Without it completion is not performed. The supported completion types are the same as that can be supplied to a user-defined command using the "-complete=" argument. Refer to :command-completion for more information. Example:
let fname = input("File: ", "", "file")
input()-highlight E5400 E5402 The optional highlight key allows specifying function which will be used for highlighting user input. This function receives user input as its only argument and must return a list of 3-tuples [hl_start_col, hl_end_col + 1, hl_group] where hl_start_col is the first highlighted column, hl_end_col is the last highlighted column (+ 1!), hl_group is :hi group used for highlighting. E5403 E5404 E5405 E5406 Both hl_start_col and hl_end_col + 1 must point to the start of the multibyte character (highlighting must not break multibyte characters), hl_end_col + 1 may be equal to the input length. Start column must be in range [0, len(input)), end column must be in range (hl_start_col, len(input)], sections must be ordered so that next hl_start_col is greater then or equal to previous hl_end_col.
Example (try some input with parentheses):
highlight RBP1 guibg=Red ctermbg=red
highlight RBP2 guibg=Yellow ctermbg=yellow
highlight RBP3 guibg=Green ctermbg=green
highlight RBP4 guibg=Blue ctermbg=blue
let g:rainbow_levels = 4
function! RainbowParens(cmdline)
  let ret = []
  let i = 0
  let lvl = 0
  while i < len(a:cmdline)
    if a:cmdline[i] is# '('
      call add(ret, [i, i + 1, 'RBP' .. ((lvl % g:rainbow_levels) + 1)])
      let lvl += 1
    elseif a:cmdline[i] is# ')'
      let lvl -= 1
      call add(ret, [i, i + 1, 'RBP' .. ((lvl % g:rainbow_levels) + 1)])
    endif
    let i += 1
  endwhile
  return ret
endfunction
call input({'prompt':'>','highlight':'RainbowParens'})
Highlight function is called at least once for each new displayed input string, before command-line is redrawn. It is expected that function is pure for the duration of one input() call, i.e. it produces the same output for the same input, so output may be memoized. Function is run like under :silent modifier. If the function causes any errors, it will be skipped for the duration of the current input() call.
Highlighting is disabled if command-line contains arabic characters.
NOTE: This function must not be used in a startup file, for the versions that only run in GUI mode (e.g., the Win32 GUI). Note: When input() is called from within a mapping it will consume remaining characters from that mapping, because a mapping is handled like the characters were typed. Use inputsave() before input() and inputrestore() after input() to avoid that. Another solution is to avoid that further characters follow in the mapping, e.g., by using :execute or :normal.
Example with a mapping:
nmap \x :call GetFoo()<CR>:exe "/" .. Foo<CR>
function GetFoo()
  call inputsave()
  let g:Foo = input("enter search pattern: ")
  call inputrestore()
endfunction
inputlist({textlist}) inputlist() {textlist} must be a List of strings. This List is displayed, one string per line. The user will be prompted to enter a number, which is returned. The user can also select an item by clicking on it with the mouse, if the mouse is enabled in the command line ('mouse' is "a" or includes "c"). For the first string 0 is returned. When clicking above the first item a negative number is returned. When clicking on the prompt one more than the length of {textlist} is returned. Make sure {textlist} has less than 'lines' entries, otherwise it won't work. It's a good idea to put the entry number at the start of the string. And put a prompt in the first item. Example:
let color = inputlist(['Select color:', '1. red',
        \ '2. green', '3. blue'])
inputrestore() inputrestore() Restore typeahead that was saved with a previous inputsave(). Should be called the same number of times inputsave() is called. Calling it more often is harmless though. Returns TRUE when there is nothing to restore, FALSE otherwise.
inputsave() inputsave() Preserve typeahead (also from mappings) and clear it, so that a following prompt gets input from the user. Should be followed by a matching inputrestore() after the prompt. Can be used several times, in which case there must be just as many inputrestore() calls. Returns TRUE when out of memory, FALSE otherwise.
inputsecret({prompt} [, {text}]) inputsecret() This function acts much like the input() function with but two exceptions: a) the user's response will be displayed as a sequence of asterisks ("*") thereby keeping the entry secret, and b) the user's response will not be recorded on the input history stack. The result is a String, which is whatever the user actually typed on the command-line in response to the issued prompt. NOTE: Command-line completion is not supported.
insert({object}, {item} [, {idx}]) insert() When {object} is a List or a Blob insert {item} at the start of it.
If {idx} is specified insert {item} before the item with index {idx}. If {idx} is zero it goes before the first item, just like omitting {idx}. A negative {idx} is also possible, see list-index. -1 inserts just before the last item.
Returns the resulting List or Blob. Examples:
let mylist = insert([2, 3, 5], 1)
call insert(mylist, 4, -1)
call insert(mylist, 6, len(mylist))
The last example can be done simpler with add(). Note that when {item} is a List it is inserted as a single item. Use extend() to concatenate Lists.
interrupt() interrupt() Interrupt script execution. It works more or less like the user typing CTRL-C, most commands won't execute and control returns to the user. This is useful to abort execution from lower down, e.g. in an autocommand. Example:
function s:check_typoname(file)
   if fnamemodify(a:file, ':t') == '['
       echomsg 'Maybe typo'
       call interrupt()
   endif
endfunction
au BufWritePre * call s:check_typoname(expand('<amatch>'))
invert({expr}) invert() Bitwise invert. The argument is converted to a number. A List, Dict or Float argument causes an error. Example:
let bits = invert(bits)
isdirectory({directory}) isdirectory() The result is a Number, which is TRUE when a directory with the name {directory} exists. If {directory} doesn't exist, or isn't a directory, the result is FALSE. {directory} is any expression, which is used as a String.
isinf({expr}) isinf() Return 1 if {expr} is a positive infinity, or -1 a negative infinity, otherwise 0.
echo isinf(1.0 / 0.0)
1
echo isinf(-1.0 / 0.0)
-1
islocked({expr}) islocked() E786 The result is a Number, which is TRUE when {expr} is the name of a locked variable. The string argument {expr} must be the name of a variable, List item or Dictionary entry, not the variable itself! Example:
let alist = [0, ['a', 'b'], 2, 3]
lockvar 1 alist
echo islocked('alist')                " 1
echo islocked('alist[1]')        " 0
When {expr} is a variable that does not exist you get an error message. Use exists() to check for existence.
isnan({expr}) isnan() Return TRUE if {expr} is a float with value NaN.
echo isnan(0.0 / 0.0)
1
items({dict}) items() Return a List with all the key-value pairs of {dict}. Each List item is a list with two items: the key of a {dict} entry and the value of this entry. The List is in arbitrary order. Also see keys() and values(). Example:
for [key, value] in items(mydict)
   echo key .. ': ' .. value
endfor
jobpid({job}) jobpid() Return the PID (process id) of job-id {job}.
jobresize({job}, {width}, {height}) jobresize() Resize the pseudo terminal window of job-id {job} to {width} columns and {height} rows. Fails if the job was not started with "pty":v:true.
jobstart({cmd} [, {opts}]) jobstart() Note: Prefer vim.system() in Lua (unless using the pty option).
Spawns {cmd} as a job. If {cmd} is a List it runs directly (no 'shell'). If {cmd} is a String it runs in the 'shell', like this:
call jobstart(split(&shell) + split(&shellcmdflag) + ['{cmd}'])
(See shell-unquoting for details.)
Example:
call jobstart('nvim -h', {'on_stdout':{j,d,e->append(line('.'),d)}})
Returns job-id on success, 0 on invalid arguments (or job table is full), -1 if {cmd}[0] or 'shell' is not executable. The returned job-id is a valid channel-id representing the job's stdio streams. Use chansend() (or rpcnotify() and rpcrequest() if "rpc" was enabled) to send data to stdin and chanclose() to close the streams without stopping the job.
See job-control and RPC.
NOTE: on Windows if {cmd} is a List:
cmd[0] must be an executable (not a "built-in"). If it is in $PATH it can be called by name, without an extension:
call jobstart(['ping', 'neovim.io'])
If it is a full or partial path, extension is required:
call jobstart(['System32\ping.exe', 'neovim.io'])
{cmd} is collapsed to a string of quoted args as expected by CommandLineToArgvW https://msdn.microsoft.com/bb776391 unless cmd[0] is some form of "cmd.exe".
jobstart-env The job environment is initialized as follows: $NVIM is set to v:servername of the parent Nvim $NVIM_LISTEN_ADDRESS is unset $NVIM_LOG_FILE is unset $VIM is unset $VIMRUNTIME is unset You can set these with the env option.
jobstart-options {opts} is a dictionary with these keys: clear_env: (boolean) env defines the job environment exactly, instead of merging current environment. cwd: (string, default=|current-directory|) Working directory of the job. detach: (boolean) Detach the job process: it will not be killed when Nvim exits. If the process exits before Nvim, on_exit will be invoked. env: (dict) Map of environment variable name:value pairs extending (or replace with "clear_env") the current environment. jobstart-env height: (number) Height of the pty terminal. on_exit: (function) Callback invoked when the job exits. on_stdout: (function) Callback invoked when the job emits stdout data. on_stderr: (function) Callback invoked when the job emits stderr data. overlapped: (boolean) Sets FILE_FLAG_OVERLAPPED for the stdio passed to the child process. Only on MS-Windows; ignored on other platforms. pty: (boolean) Connect the job to a new pseudo terminal, and its streams to the master file descriptor. on_stdout receives all output, on_stderr is ignored. terminal-start rpc: (boolean) Use msgpack-rpc to communicate with the job over stdio. Then on_stdout is ignored, but on_stderr can still be used. stderr_buffered: (boolean) Collect data until EOF (stream closed) before invoking on_stderr. channel-buffered stdout_buffered: (boolean) Collect data until EOF (stream closed) before invoking on_stdout. channel-buffered stdin: (string) Either "pipe" (default) to connect the job's stdin to a channel or "null" to disconnect stdin. width: (number) Width of the pty terminal.
{opts} is passed as self dictionary to the callback; the caller may set other keys to pass application-specific data.
Returns:
channel-id on success
0 on invalid arguments
-1 if {cmd}[0] is not executable. See also job-control, channel, msgpack-rpc.
jobstop({id}) jobstop() Stop job-id {id} by sending SIGTERM to the job process. If the process does not terminate after a timeout then SIGKILL will be sent. When the job terminates its on_exit handler (if any) will be invoked. See job-control.
Returns 1 for valid job id, 0 for invalid id, including jobs have exited or stopped.
jobwait({jobs} [, {timeout}]) jobwait() Waits for jobs and their on_exit handlers to complete.
{jobs} is a List of job-ids to wait for. {timeout} is the maximum waiting time in milliseconds. If omitted or -1, wait forever.
Timeout of 0 can be used to check the status of a job:
let running = jobwait([{job-id}], 0)[0] == -1
During jobwait() callbacks for jobs not in the {jobs} list may be invoked. The screen will not redraw unless :redraw is invoked by a callback.
Returns a list of len({jobs}) integers, where each integer is the status of the corresponding job: Exit-code, if the job exited -1 if the timeout was exceeded -2 if the job was interrupted (by CTRL-C) -3 if the job-id is invalid
join({list} [, {sep}]) join() Join the items in {list} together into one String. When {sep} is specified it is put in between the items. If {sep} is omitted a single space is used. Note that {sep} is not added at the end. You might want to add it there too:
let lines = join(mylist, "\n") .. "\n"
String items are used as-is. Lists and Dictionaries are converted into a string like with string(). The opposite function is split().
json_decode({expr}) json_decode() Convert {expr} from JSON object. Accepts readfile()-style list as the input, as well as regular string. May output any Vim value. In the following cases it will output msgpack-special-dict: 1. Dictionary contains duplicate key. 2. String contains NUL byte. Two special dictionaries: for dictionary and for string will be emitted in case string with NUL byte was a dictionary key.
Note: function treats its input as UTF-8 always. The JSON standard allows only a few encodings, of which UTF-8 is recommended and the only one required to be supported. Non-UTF-8 characters are an error.
json_encode({expr}) json_encode() Convert {expr} into a JSON string. Accepts msgpack-special-dict as the input. Will not convert Funcrefs, mappings with non-string keys (can be created as msgpack-special-dict), values with self-referencing containers, strings which contain non-UTF-8 characters, pseudo-UTF-8 strings which contain codepoints reserved for surrogate pairs (such strings are not valid UTF-8 strings). Non-printable characters are converted into "\u1234" escapes or special escapes like "\t", other are dumped as-is. Blobs are converted to arrays of the individual bytes.
keys({dict}) keys() Return a List with all the keys of {dict}. The List is in arbitrary order. Also see items() and values().
keytrans({string}) keytrans() Turn the internal byte representation of keys into a form that can be used for :map. E.g.
let xx = "\<C-Home>"
echo keytrans(xx)
<C-Home>
len({expr}) len() E701 The result is a Number, which is the length of the argument. When {expr} is a String or a Number the length in bytes is used, as with strlen(). When {expr} is a List the number of items in the List is returned. When {expr} is a Blob the number of bytes is returned. When {expr} is a Dictionary the number of entries in the Dictionary is returned. Otherwise an error is given and returns zero.
libcall({libname}, {funcname}, {argument}) libcall() E364 E368 Call function {funcname} in the run-time library {libname} with single argument {argument}. This is useful to call functions in a library that you especially made to be used with Vim. Since only one argument is possible, calling standard library functions is rather limited. The result is the String returned by the function. If the function returns NULL, this will appear as an empty string "" to Vim. If the function returns a number, use libcallnr()! If {argument} is a number, it is passed to the function as an int; if {argument} is a string, it is passed as a null-terminated string.
libcall() allows you to write your own 'plug-in' extensions to Vim without having to recompile the program. It is NOT a means to call system functions! If you try to do so Vim will very probably crash.
For Win32, the functions you write must be placed in a DLL and use the normal C calling convention (NOT Pascal which is used in Windows System DLLs). The function must take exactly one parameter, either a character pointer or a long integer, and must return a character pointer or NULL. The character pointer returned must point to memory that will remain valid after the function has returned (e.g. in static data in the DLL). If it points to allocated memory, that memory will leak away. Using a static buffer in the function should work, it's then freed when the DLL is unloaded.
WARNING: If the function returns a non-valid pointer, Vim may crash! This also happens if the function returns a number, because Vim thinks it's a pointer. For Win32 systems, {libname} should be the filename of the DLL without the ".DLL" suffix. A full path is only required if the DLL is not in the usual places. For Unix: When compiling your own plugins, remember that the object code must be compiled as position-independent ('PIC'). Examples:
echo libcall("libc.so", "getenv", "HOME")
libcallnr({libname}, {funcname}, {argument}) libcallnr() Just like libcall(), but used for a function that returns an int instead of a string. Examples:
echo libcallnr("/usr/lib/libc.so", "getpid", "")
call libcallnr("libc.so", "printf", "Hello World!\n")
call libcallnr("libc.so", "sleep", 10)
line({expr} [, {winid}]) line() The result is a Number, which is the line number of the file position given with {expr}. The {expr} argument is a string. The accepted positions are: . the cursor position $ the last line in the current buffer 'x position of mark x (if the mark is not set, 0 is returned) w0 first line visible in current window (one if the display isn't updated, e.g. in silent Ex mode) w$ last line visible in current window (this is one less than "w0" if no lines are visible) v In Visual mode: the start of the Visual area (the cursor is the end). When not in Visual mode returns the cursor position. Differs from '< in that it's updated right away. Note that a mark in another file can be used. The line number then applies to another buffer. To get the column number use col(). To get both use getpos(). With the optional {winid} argument the values are obtained for that window instead of the current window. Returns 0 for invalid values of {expr} and {winid}. Examples:
echo line(".")                        " line number of the cursor
echo line(".", winid)                " idem, in window "winid"
echo line("'t")                        " line number of mark t
echo line("'" .. marker)        " line number of mark marker
To jump to the last known position when opening a file see last-position-jump.
line2byte({lnum}) line2byte() Return the byte count from the start of the buffer for line {lnum}. This includes the end-of-line character, depending on the 'fileformat' option for the current buffer. The first line returns 1. UTF-8 encoding is used, 'fileencoding' is ignored. This can also be used to get the byte count for the line just below the last line:
echo line2byte(line("$") + 1)
This is the buffer size plus one. If 'fileencoding' is empty it is the file size plus one. {lnum} is used like with getline(). When {lnum} is invalid -1 is returned. Also see byte2line(), go and :goto.
lispindent({lnum}) lispindent() Get the amount of indent for line {lnum} according the lisp indenting rules, as with 'lisp'. The indent is counted in spaces, the value of 'tabstop' is relevant. {lnum} is used just like in getline(). When {lnum} is invalid, -1 is returned.
list2blob({list}) list2blob() Return a Blob concatenating all the number values in {list}. Examples:
echo list2blob([1, 2, 3, 4])        " returns 0z01020304
echo list2blob([])                " returns 0z
Returns an empty Blob on error. If one of the numbers is negative or more than 255 error E1239 is given.
blob2list() does the opposite.
list2str({list} [, {utf8}]) list2str() Convert each number in {list} to a character string can concatenate them all. Examples:
echo list2str([32])                " returns " "
echo list2str([65, 66, 67])        " returns "ABC"
The same can be done (slowly) with:
echo join(map(list, {nr, val -> nr2char(val)}), '')
str2list() does the opposite.
UTF-8 encoding is always used, {utf8} option has no effect, and exists only for backwards-compatibility. With UTF-8 composing characters work as expected:
echo list2str([97, 769])        " returns "á"
Returns an empty string on error.
localtime() localtime() Return the current time, measured as seconds since 1st Jan 1970. See also strftime(), strptime() and getftime().
log({expr}) log() Return the natural logarithm (base e) of {expr} as a Float. {expr} must evaluate to a Float or a Number in the range (0, inf]. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo log(10)
2.302585
echo log(exp(5))
5.0
log10({expr}) log10() Return the logarithm of Float {expr} to base 10 as a Float. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo log10(1000)
3.0
echo log10(0.01)
-2.0
luaeval({expr} [, {expr}]) luaeval() Evaluate Lua expression {expr} and return its result converted to Vim data structures. See lua-eval for more details.
map({expr1}, {expr2}) map() {expr1} must be a List, String, Blob or Dictionary. When {expr1} is a List| or Dictionary, replace each item in {expr1} with the result of evaluating {expr2}. For a Blob each byte is replaced. For a String, each character, including composing characters, is replaced. If the item type changes you may want to use mapnew() to create a new List or Dictionary.
{expr2} must be a String or Funcref.
If {expr2} is a String, inside {expr2} v:val has the value of the current item. For a Dictionary v:key has the key of the current item and for a List v:key has the index of the current item. For a Blob v:key has the index of the current byte. For a String v:key has the index of the current character. Example:
call map(mylist, '"> " .. v:val .. " <"')
This puts "> " before and " <" after each item in "mylist".
Note that {expr2} is the result of an expression and is then used as an expression again. Often it is good to use a literal-string to avoid having to double backslashes. You still have to double ' quotes
If {expr2} is a Funcref it is called with two arguments: 1. The key or the index of the current item. 2. the value of the current item. The function must return the new value of the item. Example that changes each value by "key-value":
func KeyValue(key, val)
  return a:key .. '-' .. a:val
endfunc
call map(myDict, function('KeyValue'))
It is shorter when using a lambda:
call map(myDict, {key, val -> key .. '-' .. val})
If you do not use "val" you can leave it out:
call map(myDict, {key -> 'item: ' .. key})
If you do not use "key" you can use a short name:
call map(myDict, {_, val -> 'item: ' .. val})
The operation is done in-place for a List and Dictionary. If you want it to remain unmodified make a copy first:
let tlist = map(copy(mylist), ' v:val .. "\t"')
Returns {expr1}, the List or Dictionary that was filtered, or a new Blob or String. When an error is encountered while evaluating {expr2} no further items in {expr1} are processed. When {expr2} is a Funcref errors inside a function are ignored, unless it was defined with the "abort" flag.
maparg({name} [, {mode} [, {abbr} [, {dict}]]]) maparg() When {dict} is omitted or zero: Return the rhs of mapping {name} in mode {mode}. The returned String has special characters translated like in the output of the ":map" command listing. When {dict} is TRUE a dictionary is returned, see below. To get a list of all mappings see maplist().
When there is no mapping for {name}, an empty String is returned if {dict} is FALSE, otherwise returns an empty Dict. When the mapping for {name} is empty, then "<Nop>" is returned.
The {name} can have special key names, like in the ":map" command.
{mode} can be one of these strings: "n" Normal "v" Visual (including Select) "o" Operator-pending "i" Insert "c" Cmd-line "s" Select "x" Visual "l" langmap language-mapping "t" Terminal "" Normal, Visual and Operator-pending When {mode} is omitted, the modes for "" are used.
When {abbr} is there and it is TRUE use abbreviations instead of mappings.
When {dict} is there and it is TRUE return a dictionary containing all the information of the mapping with the following items: mapping-dict "lhs" The {lhs} of the mapping as it would be typed "lhsraw" The {lhs} of the mapping as raw bytes "lhsrawalt" The {lhs} of the mapping as raw bytes, alternate form, only present when it differs from "lhsraw" "rhs" The {rhs} of the mapping as typed. "silent" 1 for a :map-silent mapping, else 0. "noremap" 1 if the {rhs} of the mapping is not remappable. "script" 1 if mapping was defined with <script>. "expr" 1 for an expression mapping (:map-<expr>). "buffer" 1 for a buffer local mapping (:map-local). "mode" Modes for which the mapping is defined. In addition to the modes mentioned above, these characters will be used: " " Normal, Visual and Operator-pending "!" Insert and Commandline mode (mapmode-ic) "sid" The script local ID, used for <sid> mappings (<SID>). Negative for special contexts. "scriptversion" The version of the script, always 1. "lnum" The line number in "sid", zero if unknown. "nowait" Do not wait for other, longer mappings. (:map-<nowait>). "abbr" True if this is an abbreviation. "mode_bits" Nvim's internal binary representation of "mode". mapset() ignores this; only "mode" is used. See maplist() for usage examples. The values are from src/nvim/state_defs.h and may change in the future.
The dictionary can be used to restore a mapping with mapset().
The mappings local to the current buffer are checked first, then the global mappings. This function can be used to map a key even when it's already mapped, and have it do the original mapping too. Sketch:
exe 'nnoremap <Tab> ==' .. maparg('<Tab>', 'n')
mapcheck({name} [, {mode} [, {abbr}]]) mapcheck() Check if there is a mapping that matches with {name} in mode {mode}. See maparg() for {mode} and special names in {name}. When {abbr} is there and it is non-zero use abbreviations instead of mappings. A match happens with a mapping that starts with {name} and with a mapping which is equal to the start of {name}.
matches mapping "a" "ab" "abc"
mapcheck("a") yes yes yes mapcheck("abc") yes yes yes mapcheck("ax") yes no no mapcheck("b") no no no
The difference with maparg() is that mapcheck() finds a mapping that matches with {name}, while maparg() only finds a mapping for {name} exactly. When there is no mapping that starts with {name}, an empty String is returned. If there is one, the RHS of that mapping is returned. If there are several mappings that start with {name}, the RHS of one of them is returned. This will be "<Nop>" if the RHS is empty. The mappings local to the current buffer are checked first, then the global mappings. This function can be used to check if a mapping can be added without being ambiguous. Example:
if mapcheck("_vv") == ""
   map _vv :set guifont=7x13<CR>
endif
This avoids adding the "_vv" mapping when there already is a mapping for "_v" or for "_vvv".
maplist([{abbr}]) maplist() Returns a List of all mappings. Each List item is a Dict, the same as what is returned by maparg(), see mapping-dict. When {abbr} is there and it is TRUE use abbreviations instead of mappings.
Example to show all mappings with "MultiMatch" in rhs:
echo maplist()->filter({_, m ->
        \ match(get(m, 'rhs', ''), 'MultiMatch') >= 0
        \ })
It can be tricky to find mappings for particular :map-modes. mapping-dict's "mode_bits" can simplify this. For example, the mode_bits for Normal, Insert or Command-line modes are 0x19. To find all the mappings available in those modes you can do:
let saved_maps = []
for m in maplist()
    if and(m.mode_bits, 0x19) != 0
        eval saved_maps->add(m)
    endif
endfor
echo saved_maps->mapnew({_, m -> m.lhs})
The values of the mode_bits are defined in Nvim's src/nvim/state_defs.h file and they can be discovered at runtime using :map-commands and "maplist()". Example:
omap xyzzy <Nop>
let op_bit = maplist()->filter(
    \ {_, m -> m.lhs == 'xyzzy'})[0].mode_bits
ounmap xyzzy
echo printf("Operator-pending mode bit: 0x%x", op_bit)
mapnew({expr1}, {expr2}) mapnew() Like map() but instead of replacing items in {expr1} a new List or Dictionary is created and returned. {expr1} remains unchanged. Items can still be changed by {expr2}, if you don't want that use deepcopy() first.
mapset({mode}, {abbr}, {dict}) mapset() mapset({dict}) Restore a mapping from a dictionary, possibly returned by maparg() or maplist(). A buffer mapping, when dict.buffer is true, is set on the current buffer; it is up to the caller to ensure that the intended buffer is the current buffer. This feature allows copying mappings from one buffer to another. The dict.mode value may restore a single mapping that covers more than one mode, like with mode values of '!', ' ', "nox", or 'v'. E1276
In the first form, {mode} and {abbr} should be the same as for the call to maparg(). E460 {mode} is used to define the mode in which the mapping is set, not the "mode" entry in {dict}. Example for saving and restoring a mapping:
let save_map = maparg('K', 'n', 0, 1)
nnoremap K somethingelse
" ...
call mapset('n', 0, save_map)
Note that if you are going to replace a map in several modes, e.g. with :map!, you need to save/restore the mapping for all of them, when they might differ.
In the second form, with {dict} as the only argument, mode and abbr are taken from the dict. Example:
let save_maps = maplist()->filter(
                        \ {_, m -> m.lhs == 'K'})
nnoremap K somethingelse
cnoremap K somethingelse2
" ...
unmap K
for d in save_maps
    call mapset(d)
endfor
match({expr}, {pat} [, {start} [, {count}]]) match() When {expr} is a List then this returns the index of the first item where {pat} matches. Each item is used as a String, Lists and Dictionaries are used as echoed.
Otherwise, {expr} is used as a String. The result is a Number, which gives the index (byte offset) in {expr} where {pat} matches.
A match at the first character or List item returns zero. If there is no match -1 is returned.
For getting submatches see matchlist(). Example:
echo match("testing", "ing")        " results in 4
echo match([1, 'x'], '\a')        " results in 1
See string-match for how {pat} is used. strpbrk() Vim doesn't have a strpbrk() function. But you can do:
let sepidx = match(line, '[.,;: \t]')
strcasestr() Vim doesn't have a strcasestr() function. But you can add "\c" to the pattern to ignore case:
let idx = match(haystack, '\cneedle')
If {start} is given, the search starts from byte index {start} in a String or item {start} in a List. The result, however, is still the index counted from the first character/item. Example:
echo match("testing", "ing", 2)
result is again "4".
echo match("testing", "ing", 4)
result is again "4".
echo match("testing", "t", 2)
result is "3". For a String, if {start} > 0 then it is like the string starts {start} bytes later, thus "^" will match at {start}. Except when {count} is given, then it's like matches before the {start} byte are ignored (this is a bit complicated to keep it backwards compatible). For a String, if {start} < 0, it will be set to 0. For a list the index is counted from the end. If {start} is out of range ({start} > strlen({expr}) for a String or {start} > len({expr}) for a List) -1 is returned.
When {count} is given use the {count}th match. When a match is found in a String the search for the next one starts one character further. Thus this example results in 1:
echo match("testing", "..", 0, 2)
In a List the search continues in the next item. Note that when {count} is added the way {start} works changes, see above.
match-pattern See pattern for the patterns that are accepted. The 'ignorecase' option is used to set the ignore-caseness of the pattern. 'smartcase' is NOT used. The matching is always done like 'magic' is set and 'cpoptions' is empty. Note that a match at the start is preferred, thus when the pattern is using "*" (any number of matches) it tends to find zero matches at the start instead of a number of matches further down in the text.
matchadd() E798 E799 E801 E957 matchadd({group}, {pattern} [, {priority} [, {id} [, {dict}]]]) Defines a pattern to be highlighted in the current window (a "match"). It will be highlighted with {group}. Returns an identification number (ID), which can be used to delete the match using matchdelete(). The ID is bound to the window. Matching is case sensitive and magic, unless case sensitivity or magicness are explicitly overridden in {pattern}. The 'magic', 'smartcase' and 'ignorecase' options are not used. The "Conceal" value is special, it causes the match to be concealed.
The optional {priority} argument assigns a priority to the match. A match with a high priority will have its highlighting overrule that of a match with a lower priority. A priority is specified as an integer (negative numbers are no exception). If the {priority} argument is not specified, the default priority is 10. The priority of 'hlsearch' is zero, hence all matches with a priority greater than zero will overrule it. Syntax highlighting (see 'syntax') is a separate mechanism, and regardless of the chosen priority a match will always overrule syntax highlighting.
The optional {id} argument allows the request for a specific match ID. If a specified ID is already taken, an error message will appear and the match will not be added. An ID is specified as a positive integer (zero excluded). IDs 1, 2 and 3 are reserved for :match, :2match and :3match, respectively. 3 is reserved for use by the matchparen plugin. If the {id} argument is not specified or -1, matchadd() automatically chooses a free ID, which is at least 1000.
The optional {dict} argument allows for further custom values. Currently this is used to specify a match specific conceal character that will be shown for hl-Conceal highlighted matches. The dict can have the following members:
conceal Special character to show instead of the match (only for hl-Conceal highlighted matches, see :syn-cchar) window Instead of the current window use the window with this number or window ID.
The number of matches is not limited, as it is the case with the :match commands.
Returns -1 on error.
Example:
highlight MyGroup ctermbg=green guibg=green
let m = matchadd("MyGroup", "TODO")
Deletion of the pattern:
call matchdelete(m)
A list of matches defined by matchadd() and :match are available from getmatches(). All matches can be deleted in one operation by clearmatches().
matchaddpos({group}, {pos} [, {priority} [, {id} [, {dict}]]]) matchaddpos() Same as matchadd(), but requires a list of positions {pos} instead of a pattern. This command is faster than matchadd() because it does not require to handle regular expressions and sets buffer line boundaries to redraw screen. It is supposed to be used when fast match additions and deletions are required, for example to highlight matching parentheses. E5030 E5031 {pos} is a list of positions. Each position can be one of these:
A number. This whole line will be highlighted. The first line has number 1.
A list with one number, e.g., [23]. The whole line with this number will be highlighted.
A list with two numbers, e.g., [23, 11]. The first number is the line number, the second one is the column number (first column is 1, the value must correspond to the byte index as col() would return). The character at this position will be highlighted.
A list with three numbers, e.g., [23, 11, 3]. As above, but the third number gives the length of the highlight in bytes.
Entries with zero and negative line numbers are silently ignored, as well as entries with negative column numbers and lengths.
Returns -1 on error.
Example:
highlight MyGroup ctermbg=green guibg=green
let m = matchaddpos("MyGroup", [[23, 24], 34])
Deletion of the pattern:
call matchdelete(m)
Matches added by matchaddpos() are returned by getmatches().
matcharg({nr}) matcharg() Selects the {nr} match item, as set with a :match, :2match or :3match command. Return a List with two elements: The name of the highlight group used The pattern used. When {nr} is not 1, 2 or 3 returns an empty List. When there is no match item set returns ['', '']. This is useful to save and restore a :match. Highlighting matches using the :match commands are limited to three matches. matchadd() does not have this limitation.
matchbufline({buf}, {pat}, {lnum}, {end}, [, {dict}]) matchbufline() Returns the List of matches in lines from {lnum} to {end} in buffer {buf} where {pat} matches.
{lnum} and {end} can either be a line number or the string "$" to refer to the last line in {buf}.
The {dict} argument supports following items: submatches include submatch information (/\()
For each match, a Dict with the following items is returned: byteidx starting byte index of the match lnum line number where there is a match text matched string Note that there can be multiple matches in a single line.
This function works only for loaded buffers. First call bufload() if needed.
See match-pattern for information about the effect of some option settings on the pattern.
When {buf} is not a valid buffer, the buffer is not loaded or {lnum} or {end} is not valid then an error is given and an empty List is returned.
Examples:
" Assuming line 3 in buffer 5 contains "a"
:echo matchbufline(5, '\<\k\+\>', 3, 3)
[{'lnum': 3, 'byteidx': 0, 'text': 'a'}]
" Assuming line 4 in buffer 10 contains "tik tok"
:echo matchbufline(10, '\<\k\+\>', 1, 4)
[{'lnum': 4, 'byteidx': 0, 'text': 'tik'}, {'lnum': 4, 'byteidx': 4, 'text': 'tok'}]
If {submatch} is present and is v:true, then submatches like "\1", "\2", etc. are also returned. Example:
" Assuming line 2 in buffer 2 contains "acd"
:echo matchbufline(2, '\(a\)\?\(b\)\?\(c\)\?\(.*\)', 2, 2
                            \ {'submatches': v:true})
[{'lnum': 2, 'byteidx': 0, 'text': 'acd', 'submatches': ['a', '', 'c', 'd', '', '', '', '', '']}]
The "submatches" List always contains 9 items. If a submatch is not found, then an empty string is returned for that submatch.
matchdelete({id} [, {win}]) matchdelete() E802 E803 Deletes a match with ID {id} previously defined by matchadd() or one of the :match commands. Returns 0 if successful, otherwise -1. See example for matchadd(). All matches can be deleted in one operation by clearmatches(). If {win} is specified, use the window with this number or window ID instead of the current window.
matchend({expr}, {pat} [, {start} [, {count}]]) matchend() Same as match(), but return the index of first character after the match. Example:
echo matchend("testing", "ing")
results in "7". strspn() strcspn() Vim doesn't have a strspn() or strcspn() function, but you can do it with matchend():
let span = matchend(line, '[a-zA-Z]')
let span = matchend(line, '[^a-zA-Z]')
Except that -1 is returned when there are no matches.
The {start}, if given, has the same meaning as for match().
echo matchend("testing", "ing", 2)
results in "7".
echo matchend("testing", "ing", 5)
result is "-1". When {expr} is a List the result is equal to match().
matchfuzzy({list}, {str} [, {dict}]) matchfuzzy() If {list} is a list of strings, then returns a List with all the strings in {list} that fuzzy match {str}. The strings in the returned list are sorted based on the matching score.
The optional {dict} argument always supports the following items: matchseq When this item is present return only matches that contain the characters in {str} in the given sequence. limit Maximum number of matches in {list} to be returned. Zero means no limit.
If {list} is a list of dictionaries, then the optional {dict} argument supports the following additional items: key Key of the item which is fuzzy matched against {str}. The value of this item should be a string. text_cb Funcref that will be called for every item in {list} to get the text for fuzzy matching. This should accept a dictionary item as the argument and return the text for that item to use for fuzzy matching.
{str} is treated as a literal string and regular expression matching is NOT supported. The maximum supported {str} length is 256.
When {str} has multiple words each separated by white space, then the list of strings that have all the words is returned.
If there are no matching strings or there is an error, then an empty list is returned. If length of {str} is greater than 256, then returns an empty list.
When {limit} is given, matchfuzzy() will find up to this number of matches in {list} and return them in sorted order.
Refer to fuzzy-matching for more information about fuzzy matching strings.
Example:
echo matchfuzzy(["clay", "crow"], "cay")
results in ["clay"].
echo getbufinfo()->map({_, v -> v.name})->matchfuzzy("ndl")
results in a list of buffer names fuzzy matching "ndl".
echo getbufinfo()->matchfuzzy("ndl", {'key' : 'name'})
results in a list of buffer information dicts with buffer names fuzzy matching "ndl".
echo getbufinfo()->matchfuzzy("spl",
                             \ {'text_cb' : {v -> v.name}})
results in a list of buffer information dicts with buffer names fuzzy matching "spl".
echo v:oldfiles->matchfuzzy("test")
results in a list of file names fuzzy matching "test".
let l = readfile("buffer.c")->matchfuzzy("str")
results in a list of lines in "buffer.c" fuzzy matching "str".
echo ['one two', 'two one']->matchfuzzy('two one')
results in ['two one', 'one two'] .
echo ['one two', 'two one']->matchfuzzy('two one',
                             \ {'matchseq': 1})
results in ['two one'].
matchfuzzypos({list}, {str} [, {dict}]) matchfuzzypos() Same as matchfuzzy(), but returns the list of matched strings, the list of character positions where characters in {str} matches and a list of matching scores. You can use byteidx() to convert a character position to a byte position.
If {str} matches multiple times in a string, then only the positions for the best match is returned.
If there are no matching strings or there is an error, then a list with three empty list items is returned.
Example:
echo matchfuzzypos(['testing'], 'tsg')
results in [["testing"], [[0, 2, 6]], [99]]
echo matchfuzzypos(['clay', 'lacy'], 'la')
results in [["lacy", "clay"], [[0, 1], [1, 2]], [153, 133]]
echo [{'text': 'hello', 'id' : 10}]
        \ ->matchfuzzypos('ll', {'key' : 'text'})
results in [[{"id": 10, "text": "hello"}], [[2, 3]], [127]]
matchlist({expr}, {pat} [, {start} [, {count}]]) matchlist() Same as match(), but return a List. The first item in the list is the matched string, same as what matchstr() would return. Following items are submatches, like "\1", "\2", etc. in :substitute. When an optional submatch didn't match an empty string is used. Example:
echo matchlist('acd', '\(a\)\?\(b\)\?\(c\)\?\(.*\)')
Results in: ['acd', 'a', '', 'c', 'd', '', '', '', '', ''] When there is no match an empty list is returned.
You can pass in a List, but that is not very useful.
matchstr({expr}, {pat} [, {start} [, {count}]]) matchstr() Same as match(), but return the matched string. Example:
echo matchstr("testing", "ing")
results in "ing". When there is no match "" is returned. The {start}, if given, has the same meaning as for match().
echo matchstr("testing", "ing", 2)
results in "ing".
echo matchstr("testing", "ing", 5)
result is "". When {expr} is a List then the matching item is returned. The type isn't changed, it's not necessarily a String.
matchstrlist({list}, {pat} [, {dict}]) matchstrlist() Returns the List of matches in {list} where {pat} matches. {list} is a List of strings. {pat} is matched against each string in {list}.
The {dict} argument supports following items: submatches include submatch information (/\()
For each match, a Dict with the following items is returned: byteidx starting byte index of the match. idx index in {list} of the match. text matched string submatches a List of submatches. Present only if "submatches" is set to v:true in {dict}.
See match-pattern for information about the effect of some option settings on the pattern.
Example:
:echo matchstrlist(['tik tok'], '\<\k\+\>')
[{'idx': 0, 'byteidx': 0, 'text': 'tik'}, {'idx': 0, 'byteidx': 4, 'text': 'tok'}]
:echo matchstrlist(['a', 'b'], '\<\k\+\>')
[{'idx': 0, 'byteidx': 0, 'text': 'a'}, {'idx': 1, 'byteidx': 0, 'text': 'b'}]
If "submatches" is present and is v:true, then submatches like "\1", "\2", etc. are also returned. Example:
:echo matchstrlist(['acd'], '\(a\)\?\(b\)\?\(c\)\?\(.*\)',
                            \ #{submatches: v:true})
[{'idx': 0, 'byteidx': 0, 'text': 'acd', 'submatches': ['a', '', 'c', 'd', '', '', '', '', '']}]
The "submatches" List always contains 9 items. If a submatch is not found, then an empty string is returned for that submatch.
matchstrpos({expr}, {pat} [, {start} [, {count}]]) matchstrpos() Same as matchstr(), but return the matched string, the start position and the end position of the match. Example:
echo matchstrpos("testing", "ing")
results in ["ing", 4, 7]. When there is no match ["", -1, -1] is returned. The {start}, if given, has the same meaning as for match().
echo matchstrpos("testing", "ing", 2)
results in ["ing", 4, 7].
echo matchstrpos("testing", "ing", 5)
result is ["", -1, -1]. When {expr} is a List then the matching item, the index of first item where {pat} matches, the start position and the end position of the match are returned.
echo matchstrpos([1, '__x'], '\a')
result is ["x", 1, 2, 3]. The type isn't changed, it's not necessarily a String.
max({expr}) max() Return the maximum value of all items in {expr}. Example:
echo max([apples, pears, oranges])
{expr} can be a List or a Dictionary. For a Dictionary, it returns the maximum of all values in the Dictionary. If {expr} is neither a List nor a Dictionary, or one of the items in {expr} cannot be used as a Number this results in an error. An empty List or Dictionary results in zero.
menu_get({path} [, {modes}]) menu_get() Returns a List of Dictionaries describing menus (defined by :menu, :amenu, …), including hidden-menus.
{path} matches a menu by name, or all menus if {path} is an empty string. Example:
echo menu_get('File','')
echo menu_get('')
{modes} is a string of zero or more modes (see maparg() or creating-menus for the list of modes). "a" means "all".
Example:
nnoremenu &Test.Test inormal
inoremenu Test.Test insert
vnoremenu Test.Test x
echo menu_get("")
returns something like this:
[ {
  "hidden": 0,
  "name": "Test",
  "priority": 500,
  "shortcut": 84,
  "submenus": [ {
    "hidden": 0,
    "mappings": {
      i": {
        "enabled": 1,
        "noremap": 1,
        "rhs": "insert",
        "sid": 1,
        "silent": 0
      },
      n": { ... },
      s": { ... },
      v": { ... }
    },
    "name": "Test",
    "priority": 500,
    "shortcut": 0
  } ]
} ]
menu_info({name} [, {mode}]) menu_info() Return information about the specified menu {name} in mode {mode}. The menu name should be specified without the shortcut character ('&'). If {name} is "", then the top-level menu names are returned.
{mode} can be one of these strings: "n" Normal "v" Visual (including Select) "o" Operator-pending "i" Insert "c" Cmd-line "s" Select "x" Visual "t" Terminal-Job "" Normal, Visual and Operator-pending "!" Insert and Cmd-line When {mode} is omitted, the modes for "" are used.
Returns a Dictionary containing the following items: accel menu item accelerator text menu-text display display name (name without '&') enabled v:true if this menu item is enabled Refer to :menu-enable icon name of the icon file (for toolbar) toolbar-icon iconidx index of a built-in icon modes modes for which the menu is defined. In addition to the modes mentioned above, these characters will be used: " " Normal, Visual and Operator-pending name menu item name. noremenu v:true if the {rhs} of the menu item is not remappable else v:false. priority menu order priority menu-priority rhs right-hand-side of the menu item. The returned string has special characters translated like in the output of the ":menu" command listing. When the {rhs} of a menu item is empty, then "<Nop>" is returned. script v:true if script-local remapping of {rhs} is allowed else v:false. See :menu-script. shortcut shortcut key (character after '&' in the menu name) menu-shortcut silent v:true if the menu item is created with <silent> argument :menu-silent submenus List containing the names of all the submenus. Present only if the menu item has submenus.
Returns an empty dictionary if the menu item is not found.
Examples:
echo menu_info('Edit.Cut')
echo menu_info('File.Save', 'n')
" Display the entire menu hierarchy in a buffer
func ShowMenu(name, pfx)
  let m = menu_info(a:name)
  call append(line('$'), a:pfx .. m.display)
  for child in m->get('submenus', [])
    call ShowMenu(a:name .. '.' .. escape(child, '.'),
                                \ a:pfx .. '    ')
  endfor
endfunc
new
for topmenu in menu_info('').submenus
  call ShowMenu(topmenu, '')
endfor
min({expr}) min() Return the minimum value of all items in {expr}. Example:
echo min([apples, pears, oranges])
{expr} can be a List or a Dictionary. For a Dictionary, it returns the minimum of all values in the Dictionary. If {expr} is neither a List nor a Dictionary, or one of the items in {expr} cannot be used as a Number this results in an error. An empty List or Dictionary results in zero.
mkdir({name} [, {flags} [, {prot}]]) mkdir() E739 Create directory {name}.
When {flags} is present it must be a string. An empty string has no effect.
If {flags} contains "p" then intermediate directories are created as necessary.
If {flags} contains "D" then {name} is deleted at the end of the current function, as with:
defer delete({name}, 'd')
If {flags} contains "R" then {name} is deleted recursively at the end of the current function, as with:
defer delete({name}, 'rf')
Note that when {name} has more than one part and "p" is used some directories may already exist. Only the first one that is created and what it contains is scheduled to be deleted. E.g. when using:
call mkdir('subdir/tmp/autoload', 'pR')
and "subdir" already exists then "subdir/tmp" will be scheduled for deletion, like with:
defer delete('subdir/tmp', 'rf')
If {prot} is given it is used to set the protection bits of the new directory. The default is 0o755 (rwxr-xr-x: r/w for the user, readable for others). Use 0o700 to make it unreadable for others.
{prot} is applied for all parts of {name}. Thus if you create /tmp/foo/bar then /tmp/foo will be created with 0o700. Example:
call mkdir($HOME .. "/tmp/foo/bar", "p", 0o700)
This function is not available in the sandbox.
If you try to create an existing directory with {flags} set to "p" mkdir() will silently exit.
The function result is a Number, which is TRUE if the call was successful or FALSE if the directory creation failed or partly failed.
mode([expr]) mode() Return a string that indicates the current mode. If [expr] is supplied and it evaluates to a non-zero Number or a non-empty String (non-zero-arg), then the full mode is returned, otherwise only the first letter is returned. Also see state().
n Normal no Operator-pending nov Operator-pending (forced charwise o_v) noV Operator-pending (forced linewise o_V) noCTRL-V Operator-pending (forced blockwise o_CTRL-V) CTRL-V is one character niI Normal using i_CTRL-O in Insert-mode niR Normal using i_CTRL-O in Replace-mode niV Normal using i_CTRL-O in Virtual-Replace-mode nt Normal in terminal-emulator (insert goes to Terminal mode) ntT Normal using t_CTRL-\_CTRL-O in Terminal-mode v Visual by character vs Visual by character using v_CTRL-O in Select mode V Visual by line Vs Visual by line using v_CTRL-O in Select mode CTRL-V Visual blockwise CTRL-Vs Visual blockwise using v_CTRL-O in Select mode s Select by character S Select by line CTRL-S Select blockwise i Insert ic Insert mode completion compl-generic ix Insert mode i_CTRL-X completion R Replace R Rc Replace mode completion compl-generic Rx Replace mode i_CTRL-X completion Rv Virtual Replace gR Rvc Virtual Replace mode completion compl-generic Rvx Virtual Replace mode i_CTRL-X completion c Command-line editing cr Command-line editing overstrike mode c_<Insert> cv Vim Ex mode gQ cvr Vim Ex mode while in overstrike mode c_<Insert> r Hit-enter prompt rm The -- more -- prompt r? A :confirm query of some sort ! Shell or external command is executing t Terminal mode: keys go to the job
This is useful in the 'statusline' option or RPC calls. In most other places it always returns "c" or "n". Note that in the future more modes and more specific modes may be added. It's better not to compare the whole string but only the leading character(s). Also see visualmode().
msgpackdump({list} [, {type}]) msgpackdump() Convert a list of Vimscript objects to msgpack. Returned value is a readfile()-style list. When {type} contains "B", a Blob is returned instead. Example:
call writefile(msgpackdump([{}]), 'fname.mpack', 'b')
or, using a Blob:
call writefile(msgpackdump([{}], 'B'), 'fname.mpack')
This will write the single 0x80 byte to a fname.mpack file (dictionary with zero items is represented by 0x80 byte in messagepack).
Limitations: E5004 E5005 1. Funcrefs cannot be dumped. 2. Containers that reference themselves cannot be dumped. 3. Dictionary keys are always dumped as STR strings. 4. Other strings and Blobs are always dumped as BIN strings. 5. Points 3. and 4. do not apply to msgpack-special-dicts.
msgpackparse({data}) msgpackparse() Convert a readfile()-style list or a Blob to a list of Vimscript objects. Example:
let fname = expand('~/.config/nvim/shada/main.shada')
let mpack = readfile(fname, 'b')
let shada_objects = msgpackparse(mpack)
This will read ~/.config/nvim/shada/main.shada file to shada_objects list.
Limitations: 1. Mapping ordering is not preserved unless messagepack mapping is dumped using generic mapping (msgpack-special-map). 2. Since the parser aims to preserve all data untouched (except for 1.) some strings are parsed to msgpack-special-dict format which is not convenient to use. msgpack-special-dict Some messagepack strings may be parsed to special dictionaries. Special dictionaries are dictionaries which
1. Contain exactly two keys: _TYPE and _VAL. 2. _TYPE key is one of the types found in v:msgpack_types variable. 3. Value for _VAL has the following format (Key column contains name of the key from v:msgpack_types):
Key Value
nil Zero, ignored when dumping. Not returned by msgpackparse() since v:null was introduced. boolean One or zero. When dumping it is only checked that value is a Number. Not returned by msgpackparse() since v:true and v:false were introduced. integer List with four numbers: sign (-1 or 1), highest two bits, number with bits from 62nd to 31st, lowest 31 bits. I.e. to get actual number one will need to use code like
_VAL[0] * ((_VAL[1] << 62)
           & (_VAL[2] << 31)
           & _VAL[3])
Special dictionary with this type will appear in msgpackparse() output under one of the following circumstances: 1. Number is 32-bit and value is either above INT32_MAX or below INT32_MIN. 2. Number is 64-bit and value is above INT64_MAX. It cannot possibly be below INT64_MIN because msgpack C parser does not support such values. float Float. This value cannot possibly appear in msgpackparse() output. string readfile()-style list of strings. This value will appear in msgpackparse() output if string contains zero byte or if string is a mapping key and mapping is being represented as special dictionary for other reasons. binary String, or Blob if binary string contains zero byte. This value cannot appear in msgpackparse() output since blobs were introduced. array List. This value cannot appear in msgpackparse() output. msgpack-special-map map List of Lists with two items (key and value) each. This value will appear in msgpackparse() output if parsed mapping contains one of the following keys: 1. Any key that is not a string (including keys which are binary strings). 2. String with NUL byte inside. 3. Duplicate key. ext List with two values: first is a signed integer representing extension type. Second is readfile()-style list of strings.
nextnonblank({lnum}) nextnonblank() Return the line number of the first line at or below {lnum} that is not blank. Example:
if getline(nextnonblank(1)) =~ "Java" | endif
When {lnum} is invalid or there is no non-blank line at or below it, zero is returned. {lnum} is used like with getline(). See also prevnonblank().
nr2char({expr} [, {utf8}]) nr2char() Return a string with a single character, which has the number value {expr}. Examples:
echo nr2char(64)                " returns '@'
echo nr2char(32)                " returns ' '
Example for "utf-8":
echo nr2char(300)                " returns I with bow character
UTF-8 encoding is always used, {utf8} option has no effect, and exists only for backwards-compatibility. Note that a NUL character in the file is specified with nr2char(10), because NULs are represented with newline characters. nr2char(0) is a real NUL and terminates the string, thus results in an empty string.
nvim_...({...}) nvim_...() E5555 eval-api Call nvim api functions. The type checking of arguments will be stricter than for most other builtins. For instance, if Integer is expected, a Number must be passed in, a String will not be autoconverted. Buffer numbers, as returned by bufnr() could be used as first argument to nvim_buf_... functions. All functions expecting an object (buffer, window or tabpage) can also take the numerical value 0 to indicate the current (focused) object.
or({expr}, {expr}) or() Bitwise OR on the two arguments. The arguments are converted to a number. A List, Dict or Float argument causes an error. Also see and() and xor(). Example:
let bits = or(bits, 0x80)
Rationale: The reason this is a function and not using the "|" character like many languages, is that Vi has always used "|" to separate commands. In many places it would not be clear if "|" is an operator or a command separator.
pathshorten({path} [, {len}]) pathshorten() Shorten directory names in the path {path} and return the result. The tail, the file name, is kept as-is. The other components in the path are reduced to {len} letters in length. If {len} is omitted or smaller than 1 then 1 is used (single letters). Leading '~' and '.' characters are kept. Examples:
echo pathshorten('~/.config/nvim/autoload/file1.vim')
~/.c/n/a/file1.vim
echo pathshorten('~/.config/nvim/autoload/file2.vim', 2)
~/.co/nv/au/file2.vim
It doesn't matter if the path exists or not. Returns an empty string on error.
perleval({expr}) perleval() Evaluate perl expression {expr} and return its result converted to Vim data structures. Numbers and strings are returned as they are (strings are copied though). Lists are represented as Vim List type. Dictionaries are represented as Vim Dictionary type, non-string keys result in error.
Note: If you want an array or hash, {expr} must return a reference to it. Example:
echo perleval('[1 .. 4]')
[1, 2, 3, 4]
pow({x}, {y}) pow() Return the power of {x} to the exponent {y} as a Float. {x} and {y} must evaluate to a Float or a Number. Returns 0.0 if {x} or {y} is not a Float or a Number. Examples:
echo pow(3, 3)
27.0
echo pow(2, 16)
65536.0
echo pow(32, 0.20)
2.0
prevnonblank({lnum}) prevnonblank() Return the line number of the first line at or above {lnum} that is not blank. Example:
let ind = indent(prevnonblank(v:lnum - 1))
When {lnum} is invalid or there is no non-blank line at or above it, zero is returned. {lnum} is used like with getline(). Also see nextnonblank().
printf({fmt}, {expr1} ...) printf() Return a String with {fmt}, where "%" items are replaced by the formatted form of their respective arguments. Example:
echo printf("%4d: E%d %.30s", lnum, errno, msg)
May result in:
" 99: E42 asdfasdfasdfasdfasdfasdfasdfas"
When used as a method the base is passed as the second argument:
Compute()->printf("result: %d")
You can use call() to pass the items as a list.
Often used items are: %s string %6S string right-aligned in 6 display cells %6s string right-aligned in 6 bytes %.9s string truncated to 9 bytes %c single byte %d decimal number %5d decimal number padded with spaces to 5 characters %b binary number %08b binary number padded with zeros to at least 8 characters %B binary number using upper case letters %x hex number %04x hex number padded with zeros to at least 4 characters %X hex number using upper case letters %o octal number %f floating point number as 12.23, inf, -inf or nan %F floating point number as 12.23, INF, -INF or NAN %e floating point number as 1.23e3, inf, -inf or nan %E floating point number as 1.23E3, INF, -INF or NAN %g floating point number, as %f or %e depending on value %G floating point number, as %F or %E depending on value %% the % character itself %p representation of the pointer to the container
Conversion specifications start with '%' and end with the conversion type. All other characters are copied unchanged to the result.
The "%" starts a conversion specification. The following arguments appear in sequence:
% [pos-argument] [flags] [field-width] [.precision] type
pos-argument At most one positional argument specifier. These take the form {n$}, where n is >= 1.
flags Zero or more of the following flags:
# The value should be converted to an "alternate form". For c, d, and s conversions, this option has no effect. For o conversions, the precision of the number is increased to force the first character of the output string to a zero (except if a zero value is printed with an explicit precision of zero). For x and X conversions, a non-zero result has the string "0x" (or "0X" for X conversions) prepended to it.
0 (zero) Zero padding. For all conversions the converted value is padded on the left with zeros rather than blanks. If a precision is given with a numeric conversion (d, o, x, and X), the 0 flag is ignored.
- A negative field width flag; the converted value is to be left adjusted on the field boundary. The converted value is padded on the right with blanks, rather than on the left with blanks or zeros. A - overrides a 0 if both are given.
' ' (space) A blank should be left before a positive number produced by a signed conversion (d).
+ A sign must always be placed before a number produced by a signed conversion. A + overrides a space if both are used.
field-width An optional decimal digit string specifying a minimum field width. If the converted value has fewer bytes than the field width, it will be padded with spaces on the left (or right, if the left-adjustment flag has been given) to fill out the field width. For the S conversion the count is in cells.
.precision An optional precision, in the form of a period '.' followed by an optional digit string. If the digit string is omitted, the precision is taken as zero. This gives the minimum number of digits to appear for d, o, x, and X conversions, the maximum number of bytes to be printed from a string for s conversions, or the maximum number of cells to be printed from a string for S conversions. For floating point it is the number of digits after the decimal point.
type A character that specifies the type of conversion to be applied, see below.
A field width or precision, or both, may be indicated by an asterisk "*" instead of a digit string. In this case, a Number argument supplies the field width or precision. A negative field width is treated as a left adjustment flag followed by a positive field width; a negative precision is treated as though it were missing. Example:
echo printf("%d: %.*s", nr, width, line)
This limits the length of the text used from "line" to "width" bytes.
If the argument to be formatted is specified using a positional argument specifier, and a '*' is used to indicate that a number argument is to be used to specify the width or precision, the argument(s) to be used must also be specified using a {n$} positional argument specifier. See printf-$.
The conversion specifiers and their meanings are:
printf-d printf-b printf-B printf-o printf-x printf-X dbBoxX The Number argument is converted to signed decimal (d), unsigned binary (b and B), unsigned octal (o), or unsigned hexadecimal (x and X) notation. The letters "abcdef" are used for x conversions; the letters "ABCDEF" are used for X conversions. The precision, if any, gives the minimum number of digits that must appear; if the converted value requires fewer digits, it is padded on the left with zeros. In no case does a non-existent or small field width cause truncation of a numeric field; if the result of a conversion is wider than the field width, the field is expanded to contain the conversion result. The 'h' modifier indicates the argument is 16 bits. The 'l' modifier indicates the argument is a long integer. The size will be 32 bits or 64 bits depending on your platform. The "ll" modifier indicates the argument is 64 bits. The b and B conversion specifiers never take a width modifier and always assume their argument is a 64 bit integer. Generally, these modifiers are not useful. They are ignored when type is known from the argument.
i alias for d D alias for ld U alias for lu O alias for lo
printf-c c The Number argument is converted to a byte, and the resulting character is written.
printf-s s The text of the String argument is used. If a precision is specified, no more bytes than the number specified are used. If the argument is not a String type, it is automatically converted to text with the same format as ":echo". printf-S S The text of the String argument is used. If a precision is specified, no more display cells than the number specified are used.
printf-f E807 f F The Float argument is converted into a string of the form 123.456. The precision specifies the number of digits after the decimal point. When the precision is zero the decimal point is omitted. When the precision is not specified 6 is used. A really big number (out of range or dividing by zero) results in "inf" or "-inf" with %f (INF or -INF with %F). "0.0 / 0.0" results in "nan" with %f (NAN with %F). Example:
echo printf("%.2f", 12.115)
12.12 Note that roundoff depends on the system libraries. Use round() when in doubt.
printf-e printf-E e E The Float argument is converted into a string of the form 1.234e+03 or 1.234E+03 when using 'E'. The precision specifies the number of digits after the decimal point, like with 'f'.
printf-g printf-G g G The Float argument is converted like with 'f' if the value is between 0.001 (inclusive) and 10000000.0 (exclusive). Otherwise 'e' is used for 'g' and 'E' for 'G'. When no precision is specified superfluous zeroes and '+' signs are removed, except for the zero immediately after the decimal point. Thus 10000000.0 results in 1.0e7.
printf-% % A '%' is written. No argument is converted. The complete conversion specification is "%%".
When a Number argument is expected a String argument is also accepted and automatically converted. When a Float or String argument is expected a Number argument is also accepted and automatically converted. Any other argument type results in an error message.
E766 E767 The number of {exprN} arguments must exactly match the number of "%" items. If there are not sufficient or too many arguments an error is given. Up to 18 arguments can be used.
printf-$ In certain languages, error and informative messages are more readable when the order of words is different from the corresponding message in English. To accommodate translations having a different word order, positional arguments may be used to indicate this. For instance:
#, c-format
msgid "%s returning %s"
msgstr "waarde %2$s komt terug van %1$s"
In this example, the sentence has its 2 string arguments reversed in the output.
echo printf(
    "In The Netherlands, vim's creator's name is: %1$s %2$s",
    "Bram", "Moolenaar")
In The Netherlands, vim's creator's name is: Bram Moolenaar
echo printf(
    "In Belgium, vim's creator's name is: %2$s %1$s",
    "Bram", "Moolenaar")
In Belgium, vim's creator's name is: Moolenaar Bram
Width (and precision) can be specified using the '*' specifier. In this case, you must specify the field width position in the argument list.
echo printf("%1$*2$.*3$d", 1, 2, 3)
001
echo printf("%2$*3$.*1$d", 1, 2, 3)
2
echo printf("%3$*1$.*2$d", 1, 2, 3)
03
echo printf("%1$*2$.*3$g", 1.4142, 2, 3)
1.414
You can mix specifying the width and/or precision directly and via positional arguments:
echo printf("%1$4.*2$f", 1.4142135, 6)
1.414214
echo printf("%1$*2$.4f", 1.4142135, 6)
1.4142
echo printf("%1$*2$.*3$f", 1.4142135, 6, 2)
1.41
You will get an overflow error E1510, when the field-width or precision will result in a string longer than 6400 chars.
E1500 You cannot mix positional and non-positional arguments:
echo printf("%s%1$s", "One", "Two")
E1500: Cannot mix positional and non-positional arguments: %s%1$s
E1501 You cannot skip a positional argument in a format string:
echo printf("%3$s%1$s", "One", "Two", "Three")
E1501: format argument 2 unused in $-style format: %3$s%1$s
E1502 You can re-use a [field-width] (or [precision]) argument:
echo printf("%1$d at width %2$d is: %01$*2$d", 1, 2)
1 at width 2 is: 01
However, you can't use it as a different type:
echo printf("%1$d at width %2$ld is: %01$*2$d", 1, 2)
E1502: Positional argument 2 used as field width reused as different type: long int/int
E1503 When a positional argument is used, but not the correct number or arguments is given, an error is raised:
echo printf("%1$d at width %2$d is: %01$*2$.*3$d", 1, 2)
E1503: Positional argument 3 out of bounds: %1$d at width %2$d is: %01$*2$.*3$d
Only the first error is reported:
echo printf("%01$*2$.*3$d %4$d", 1, 2)
E1503: Positional argument 3 out of bounds: %01$*2$.*3$d %4$d
E1504 A positional argument can be used more than once:
echo printf("%1$s %2$s %1$s", "One", "Two")
One Two One
However, you can't use a different type the second time:
echo printf("%1$s %2$s %1$d", "One", "Two")
E1504: Positional argument 1 type used inconsistently: int/string
E1505 Various other errors that lead to a format string being wrongly formatted lead to:
echo printf("%1$d at width %2$d is: %01$*2$.3$d", 1, 2)
E1505: Invalid format specifier: %1$d at width %2$d is: %01$*2$.3$d
E1507 This internal error indicates that the logic to parse a positional format argument ran into a problem that couldn't be otherwise reported. Please file a bug against Vim if you run into this, copying the exact format string and parameters that were used.
prompt_getprompt({buf}) prompt_getprompt() Returns the effective prompt text for buffer {buf}. {buf} can be a buffer name or number. See prompt-buffer.
If the buffer doesn't exist or isn't a prompt buffer, an empty string is returned.
prompt_setcallback({buf}, {expr}) prompt_setcallback() Set prompt callback for buffer {buf} to {expr}. When {expr} is an empty string the callback is removed. This has only effect if {buf} has 'buftype' set to "prompt".
The callback is invoked when pressing Enter. The current buffer will always be the prompt buffer. A new line for a prompt is added before invoking the callback, thus the prompt for which the callback was invoked will be in the last but one line. If the callback wants to add text to the buffer, it must insert it above the last line, since that is where the current prompt is. This can also be done asynchronously. The callback is invoked with one argument, which is the text that was entered at the prompt. This can be an empty string if the user only typed Enter. Example:
func s:TextEntered(text)
  if a:text == 'exit' || a:text == 'quit'
    stopinsert
    " Reset 'modified' to allow the buffer to be closed.
    " We assume there is nothing useful to be saved.
    set nomodified
    close
  else
    " Do something useful with "a:text".  In this example
    " we just repeat it.
    call append(line('$') - 1, 'Entered: "' .. a:text .. '"')
  endif
endfunc
call prompt_setcallback(bufnr(), function('s:TextEntered'))
prompt_setinterrupt({buf}, {expr}) prompt_setinterrupt() Set a callback for buffer {buf} to {expr}. When {expr} is an empty string the callback is removed. This has only effect if {buf} has 'buftype' set to "prompt".
This callback will be invoked when pressing CTRL-C in Insert mode. Without setting a callback Vim will exit Insert mode, as in any buffer.
prompt_setprompt({buf}, {text}) prompt_setprompt() Set prompt for buffer {buf} to {text}. You most likely want {text} to end in a space. The result is only visible if {buf} has 'buftype' set to "prompt". Example:
call prompt_setprompt(bufnr(''), 'command: ')
pum_getpos() pum_getpos() If the popup menu (see ins-completion-menu) is not visible, returns an empty Dictionary, otherwise, returns a Dictionary with the following keys: height nr of items visible width screen cells row top screen row (0 first row) col leftmost screen column (0 first col) size total nr of items scrollbar TRUE if scrollbar is visible
The values are the same as in v:event during CompleteChanged.
pumvisible() pumvisible() Returns non-zero when the popup menu is visible, zero otherwise. See ins-completion-menu. This can be used to avoid some things that would remove the popup menu.
py3eval({expr}) py3eval() Evaluate Python expression {expr} and return its result converted to Vim data structures. Numbers and strings are returned as they are (strings are copied though, Unicode strings are additionally converted to UTF-8). Lists are represented as Vim List type. Dictionaries are represented as Vim Dictionary type with keys converted to strings.
pyeval({expr}) pyeval() E858 E859 Evaluate Python expression {expr} and return its result converted to Vim data structures. Numbers and strings are returned as they are (strings are copied though). Lists are represented as Vim List type. Dictionaries are represented as Vim Dictionary type, non-string keys result in error.
pyxeval({expr}) pyxeval() Evaluate Python expression {expr} and return its result converted to Vim data structures. Uses Python 2 or 3, see python_x and 'pyxversion'. See also: pyeval(), py3eval()
rand([{expr}]) rand() Return a pseudo-random Number generated with an xoshiro128** algorithm using seed {expr}. The returned number is 32 bits, also on 64 bits systems, for consistency. {expr} can be initialized by srand() and will be updated by rand(). If {expr} is omitted, an internal seed value is used and updated. Returns -1 if {expr} is invalid.
Examples:
echo rand()
let seed = srand()
echo rand(seed)
echo rand(seed) % 16  " random number 0 - 15
range({expr} [, {max} [, {stride}]]) range() E726 E727 Returns a List with Numbers:
If only {expr} is specified: [0, 1, ..., {expr} - 1]
If {max} is specified: [{expr}, {expr} + 1, ..., {max}]
If {stride} is specified: [{expr}, {expr} + {stride}, ..., {max}] (increasing {expr} with {stride} each time, not producing a value past {max}). When the maximum is one before the start the result is an empty list. When the maximum is more than one before the start this is an error. Examples:
echo range(4)                " [0, 1, 2, 3]
echo range(2, 4)        " [2, 3, 4]
echo range(2, 9, 3)        " [2, 5, 8]
echo range(2, -2, -1)        " [2, 1, 0, -1, -2]
echo range(0)                " []
echo range(2, 0)        " error!
readblob({fname} [, {offset} [, {size}]]) readblob() Read file {fname} in binary mode and return a Blob. If {offset} is specified, read the file from the specified offset. If it is a negative value, it is used as an offset from the end of the file. E.g., to read the last 12 bytes:
echo readblob('file.bin', -12)
If {size} is specified, only the specified size will be read. E.g. to read the first 100 bytes of a file:
echo readblob('file.bin', 0, 100)
If {size} is -1 or omitted, the whole data starting from {offset} will be read. This can be also used to read the data from a character device on Unix when {size} is explicitly set. Only if the device supports seeking {offset} can be used. Otherwise it should be zero. E.g. to read 10 bytes from a serial console:
echo readblob('/dev/ttyS0', 0, 10)
When the file can't be opened an error message is given and the result is an empty Blob. When the offset is beyond the end of the file the result is an empty blob. When trying to read more bytes than are available the result is truncated. Also see readfile() and writefile().
readdir({directory} [, {expr}]) readdir() Return a list with file and directory names in {directory}. You can also use glob() if you don't need to do complicated things, such as limiting the number of matches.
When {expr} is omitted all entries are included. When {expr} is given, it is evaluated to check what to do: If {expr} results in -1 then no further entries will be handled. If {expr} results in 0 then this entry will not be added to the list. If {expr} results in 1 then this entry will be added to the list. Each time {expr} is evaluated v:val is set to the entry name. When {expr} is a function the name is passed as the argument. For example, to get a list of files ending in ".txt":
echo readdir(dirname, {n -> n =~ '.txt$'})
To skip hidden and backup files:
echo readdir(dirname, {n -> n !~ '^\.\|\~$'})
If you want to get a directory tree:
function! s:tree(dir)
    return {a:dir : map(readdir(a:dir),
    \ {_, x -> isdirectory(x) ?
    \          {x : s:tree(a:dir .. '/' .. x)} : x})}
endfunction
echo s:tree(".")
Returns an empty List on error.
readfile({fname} [, {type} [, {max}]]) readfile() Read file {fname} and return a List, each line of the file as an item. Lines are broken at NL characters. Macintosh files separated with CR will result in a single long line (unless a NL appears somewhere). All NUL characters are replaced with a NL character. When {type} contains "b" binary mode is used:
When the last line ends in a NL an extra empty list item is added.
No CR characters are removed. Otherwise:
CR characters that appear before a NL are removed.
Whether the last line ends in a NL or not does not matter.
Any UTF-8 byte order mark is removed from the text. When {max} is given this specifies the maximum number of lines to be read. Useful if you only want to check the first ten lines of a file:
for line in readfile(fname, '', 10)
  if line =~ 'Date' | echo line | endif
endfor
When {max} is negative -{max} lines from the end of the file are returned, or as many as there are. When {max} is zero the result is an empty list. Note that without {max} the whole file is read into memory. Also note that there is no recognition of encoding. Read a file into a buffer if you need to. Deprecated (use readblob() instead): When {type} contains "B" a Blob is returned with the binary data of the file unmodified. When the file can't be opened an error message is given and the result is an empty list. Also see writefile().
reduce({object}, {func} [, {initial}]) reduce() E998 {func} is called for every item in {object}, which can be a String, List or a Blob. {func} is called with two arguments: the result so far and current item. After processing all items the result is returned.
{initial} is the initial result. When omitted, the first item in {object} is used and {func} is first called for the second item. If {initial} is not given and {object} is empty no result can be computed, an E998 error is given.
Examples:
echo reduce([1, 3, 5], { acc, val -> acc + val })
echo reduce(['x', 'y'], { acc, val -> acc .. val }, 'a')
echo reduce(0z1122, { acc, val -> 2 * acc + val })
echo reduce('xyz', { acc, val -> acc .. ',' .. val })
reg_executing() reg_executing() Returns the single letter name of the register being executed. Returns an empty string when no register is being executed. See @.
reg_recorded() reg_recorded() Returns the single letter name of the last recorded register. Returns an empty string when nothing was recorded yet. See q and Q.
reg_recording() reg_recording() Returns the single letter name of the register being recorded. Returns an empty string when not recording. See q.
reltime() reltime() reltime({start}) reltime({start}, {end}) Return an item that represents a time value. The item is a list with items that depend on the system. The item can be passed to reltimestr() to convert it to a string or reltimefloat() to convert to a Float.
Without an argument it returns the current "relative time", an implementation-defined value meaningful only when used as an argument to reltime(), reltimestr() and reltimefloat().
With one argument it returns the time passed since the time specified in the argument. With two arguments it returns the time passed between {start} and {end}.
The {start} and {end} arguments must be values returned by reltime(). Returns zero on error.
Note: localtime() returns the current (non-relative) time.
reltimefloat({time}) reltimefloat() Return a Float that represents the time value of {time}. Unit of time is seconds. Example: let start = reltime() call MyFunction() let seconds = reltimefloat(reltime(start)) See the note of reltimestr() about overhead. Also see profiling. If there is an error an empty string is returned
reltimestr({time}) reltimestr() Return a String that represents the time value of {time}. This is the number of seconds, a dot and the number of microseconds. Example:
let start = reltime()
call MyFunction()
echo reltimestr(reltime(start))
Note that overhead for the commands will be added to the time. Leading spaces are used to make the string align nicely. You can use split() to remove it.
echo split(reltimestr(reltime(start)))[0]
Also see profiling. If there is an error an empty string is returned
remove({list}, {idx}) remove() remove({list}, {idx}, {end}) Without {end}: Remove the item at {idx} from List {list} and return the item. With {end}: Remove items from {idx} to {end} (inclusive) and return a List with these items. When {idx} points to the same item as {end} a list with one item is returned. When {end} points to an item before {idx} this is an error. See list-index for possible values of {idx} and {end}. Returns zero on error. Example:
echo "last item: " .. remove(mylist, -1)
call remove(mylist, 0, 9)
Use delete() to remove a file.
remove({blob}, {idx}) remove({blob}, {idx}, {end}) Without {end}: Remove the byte at {idx} from Blob {blob} and return the byte. With {end}: Remove bytes from {idx} to {end} (inclusive) and return a Blob with these bytes. When {idx} points to the same byte as {end} a Blob with one byte is returned. When {end} points to a byte before {idx} this is an error. Returns zero on error. Example:
echo "last byte: " .. remove(myblob, -1)
call remove(mylist, 0, 9)
remove({dict}, {key}) Remove the entry from {dict} with key {key} and return it. Example:
echo "removed " .. remove(dict, "one")
If there is no {key} in {dict} this is an error. Returns zero on error.
rename({from}, {to}) rename() Rename the file by the name {from} to the name {to}. This should also work to move files across file systems. The result is a Number, which is 0 if the file was renamed successfully, and non-zero when the renaming failed. NOTE: If {to} exists it is overwritten without warning. This function is not available in the sandbox.
repeat({expr}, {count}) repeat() Repeat {expr} {count} times and return the concatenated result. Example:
let separator = repeat('-', 80)
When {count} is zero or negative the result is empty. When {expr} is a List or a Blob the result is {expr} concatenated {count} times. Example:
let longlist = repeat(['a', 'b'], 3)
Results in ['a', 'b', 'a', 'b', 'a', 'b'].
resolve({filename}) resolve() E655 On MS-Windows, when {filename} is a shortcut (a .lnk file), returns the path the shortcut points to in a simplified form. On Unix, repeat resolving symbolic links in all path components of {filename} and return the simplified result. To cope with link cycles, resolving of symbolic links is stopped after 100 iterations. On other systems, return the simplified {filename}. The simplification step is done as by simplify(). resolve() keeps a leading path component specifying the current directory (provided the result is still a relative path name) and also keeps a trailing path separator.
reverse({object}) reverse() Reverse the order of items in {object}. {object} can be a List, a Blob or a String. For a List and a Blob the items are reversed in-place and {object} is returned. For a String a new String is returned. Returns zero if {object} is not a List, Blob or a String. If you want a List or Blob to remain unmodified make a copy first:
let revlist = reverse(copy(mylist))
round({expr}) round() Round off {expr} to the nearest integral value and return it as a Float. If {expr} lies halfway between two integral values, then use the larger one (away from zero). {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo round(0.456)
0.0
echo round(4.5)
5.0
echo round(-4.5)
-5.0
rpcnotify({channel}, {event} [, {args}...]) rpcnotify() Sends {event} to {channel} via RPC and returns immediately. If {channel} is 0, the event is broadcast to all channels. Example:
au VimLeave call rpcnotify(0, "leaving")
rpcrequest({channel}, {method} [, {args}...]) rpcrequest() Sends a request to {channel} to invoke {method} via RPC and blocks until a response is received. Example:
let result = rpcrequest(rpc_chan, "func", 1, 2, 3)
rpcstart({prog} [, {argv}]) rpcstart() Deprecated. Replace
let id = rpcstart('prog', ['arg1', 'arg2'])
with
let id = jobstart(['prog', 'arg1', 'arg2'], {'rpc': v:true})
rubyeval({expr}) rubyeval() Evaluate Ruby expression {expr} and return its result converted to Vim data structures. Numbers, floats and strings are returned as they are (strings are copied though). Arrays are represented as Vim List type. Hashes are represented as Vim Dictionary type. Other objects are represented as strings resulted from their "Object#to_s" method.
screenattr({row}, {col}) screenattr() Like screenchar(), but return the attribute. This is a rather arbitrary number that can only be used to compare to the attribute at other positions. Returns -1 when row or col is out of range.
screenchar({row}, {col}) screenchar() The result is a Number, which is the character at position [row, col] on the screen. This works for every possible screen position, also status lines, window separators and the command line. The top left position is row one, column one The character excludes composing characters. For double-byte encodings it may only be the first byte. This is mainly to be used for testing. Returns -1 when row or col is out of range.
screenchars({row}, {col}) screenchars() The result is a List of Numbers. The first number is the same as what screenchar() returns. Further numbers are composing characters on top of the base character. This is mainly to be used for testing. Returns an empty List when row or col is out of range.
screencol() screencol() The result is a Number, which is the current screen column of the cursor. The leftmost column has number 1. This function is mainly used for testing.
Note: Always returns the current screen column, thus if used in a command (e.g. ":echo screencol()") it will return the column inside the command line, which is 1 when the command is executed. To get the cursor position in the file use one of the following mappings:
nnoremap <expr> GG ":echom " .. screencol() .. "\n"
nnoremap <silent> GG :echom screencol()<CR>
noremap GG <Cmd>echom screencol()<Cr>
screenpos({winid}, {lnum}, {col}) screenpos() The result is a Dict with the screen position of the text character in window {winid} at buffer line {lnum} and column {col}. {col} is a one-based byte index. The Dict has these members: row screen row col first screen column endcol last screen column curscol cursor screen column If the specified position is not visible, all values are zero. The "endcol" value differs from "col" when the character occupies more than one screen cell. E.g. for a Tab "col" can be 1 and "endcol" can be 8. The "curscol" value is where the cursor would be placed. For a Tab it would be the same as "endcol", while for a double width character it would be the same as "col". The conceal feature is ignored here, the column numbers are as if 'conceallevel' is zero. You can set the cursor to the right position and use screencol() to get the value with conceal taken into account. If the position is in a closed fold the screen position of the first character is returned, {col} is not used. Returns an empty Dict if {winid} is invalid.
screenrow() screenrow() The result is a Number, which is the current screen row of the cursor. The top line has number one. This function is mainly used for testing. Alternatively you can use winline().
Note: Same restrictions as with screencol().
screenstring({row}, {col}) screenstring() The result is a String that contains the base character and any composing characters at position [row, col] on the screen. This is like screenchars() but returning a String with the characters. This is mainly to be used for testing. Returns an empty String when row or col is out of range.
search({pattern} [, {flags} [, {stopline} [, {timeout} [, {skip}]]]]) search() Search for regexp pattern {pattern}. The search starts at the cursor position (you can use cursor() to set it).
When a match has been found its line number is returned. If there is no match a 0 is returned and the cursor doesn't move. No error message is given. To get the matched string, use matchbufline().
{flags} is a String, which can contain these character flags: 'b' search Backward instead of forward 'c' accept a match at the Cursor position 'e' move to the End of the match 'n' do Not move the cursor 'p' return number of matching sub-Pattern (see below) 's' Set the ' mark at the previous location of the cursor 'w' Wrap around the end of the file 'W' don't Wrap around the end of the file 'z' start searching at the cursor column instead of Zero If neither 'w' or 'W' is given, the 'wrapscan' option applies.
If the 's' flag is supplied, the ' mark is set, only if the cursor is moved. The 's' flag cannot be combined with the 'n' flag.
When the 'z' flag is not given, forward searching always starts in column zero and then matches before the cursor are skipped. When the 'c' flag is present in 'cpo' the next search starts after the match. Without the 'c' flag the next search starts one column after the start of the match. This matters for overlapping matches. See cpo-c. You can also insert "\ze" to change where the match ends, see /\ze.
When searching backwards and the 'z' flag is given then the search starts in column zero, thus no match in the current line will be found (unless wrapping around the end of the file).
When the {stopline} argument is given then the search stops after searching this line. This is useful to restrict the search to a range of lines. Examples:
let match = search('(', 'b', line("w0"))
let end = search('END', '', line("w$"))
When {stopline} is used and it is not zero this also implies that the search does not wrap around the end of the file. A zero value is equal to not giving the argument.
When the {timeout} argument is given the search stops when more than this many milliseconds have passed. Thus when {timeout} is 500 the search stops after half a second. The value must not be negative. A zero value is like not giving the argument.
If the {skip} expression is given it is evaluated with the cursor positioned on the start of a match. If it evaluates to non-zero this match is skipped. This can be used, for example, to skip a match in a comment or a string. {skip} can be a string, which is evaluated as an expression, a function reference or a lambda. When {skip} is omitted or empty, every match is accepted. When evaluating {skip} causes an error the search is aborted and -1 returned. search()-sub-match With the 'p' flag the returned value is one more than the first sub-match in \(\). One if none of them matched but the whole pattern did match. To get the column number too use searchpos().
The cursor will be positioned at the match, unless the 'n' flag is used.
Example (goes over all files in the argument list):
let n = 1
while n <= argc()            " loop over all files in arglist
  exe "argument " .. n
  " start at the last char in the file and wrap for the
  " first search to find match at start of file
  normal G$
  let flags = "w"
  while search("foo", flags) > 0
    s/foo/bar/g
    let flags = "W"
  endwhile
  update                    " write the file if modified
  let n = n + 1
endwhile
Example for using some flags:
echo search('\<if\|\(else\)\|\(endif\)', 'ncpe')
This will search for the keywords "if", "else", and "endif" under or after the cursor. Because of the 'p' flag, it returns 1, 2, or 3 depending on which keyword is found, or 0 if the search fails. With the cursor on the first word of the line:
if (foo == 0) | let foo = foo + 1 | endif
the function returns 1. Without the 'c' flag, the function finds the "endif" and returns 3. The same thing happens without the 'e' flag if the cursor is on the "f" of "if". The 'n' flag tells the function not to move the cursor.
searchcount([{options}]) searchcount() Get or update the last search count, like what is displayed without the "S" flag in 'shortmess'. This works even if 'shortmess' does contain the "S" flag.
This returns a Dictionary. The dictionary is empty if the previous pattern was not set and "pattern" was not specified.
key type meaning
current Number current position of match; 0 if the cursor position is before the first match exact_match Boolean 1 if "current" is matched on "pos", otherwise 0 total Number total count of matches found incomplete Number 0: search was fully completed 1: recomputing was timed out 2: max count exceeded
For {options} see further down.
To get the last search count when n or N was pressed, call this function with recompute: 0 . This sometimes returns wrong information because n and N's maximum count is 99. If it exceeded 99 the result must be max count + 1 (100). If you want to get correct information, specify recompute: 1:
" result == maxcount + 1 (100) when many matches
let result = searchcount(#{recompute: 0})
" Below returns correct result (recompute defaults
" to 1)
let result = searchcount()
The function is useful to add the count to 'statusline':
function! LastSearchCount() abort
  let result = searchcount(#{recompute: 0})
  if empty(result)
    return ''
  endif
  if result.incomplete ==# 1     " timed out
    return printf(' /%s [?/??]', @/)
  elseif result.incomplete ==# 2 " max count exceeded
    if result.total > result.maxcount &&
    \  result.current > result.maxcount
      return printf(' /%s [>%d/>%d]', @/,
      \             result.current, result.total)
    elseif result.total > result.maxcount
      return printf(' /%s [%d/>%d]', @/,
      \             result.current, result.total)
    endif
  endif
  return printf(' /%s [%d/%d]', @/,
  \             result.current, result.total)
endfunction
let &statusline ..= '%{LastSearchCount()}'
" Or if you want to show the count only when
" 'hlsearch' was on
" let &statusline ..=
" \   '%{v:hlsearch ? LastSearchCount() : ""}'
You can also update the search count, which can be useful in a CursorMoved or CursorMovedI autocommand:
autocmd CursorMoved,CursorMovedI *
  \ let s:searchcount_timer = timer_start(
  \   200, function('s:update_searchcount'))
function! s:update_searchcount(timer) abort
  if a:timer ==# s:searchcount_timer
    call searchcount(#{
    \ recompute: 1, maxcount: 0, timeout: 100})
    redrawstatus
  endif
endfunction
This can also be used to count matched texts with specified pattern in the current buffer using "pattern":
" Count '\<foo\>' in this buffer
" (Note that it also updates search count)
let result = searchcount(#{pattern: '\<foo\>'})
" To restore old search count by old pattern,
" search again
call searchcount()
{options} must be a Dictionary. It can contain:
key type meaning
recompute Boolean if TRUE, recompute the count like n or N was executed. otherwise returns the last computed result (when n or N was used when "S" is not in 'shortmess', or this function was called). (default: TRUE) pattern String recompute if this was given and different with @/. this works as same as the below command is executed before calling this function
let @/ = pattern
(default: @/) timeout Number 0 or negative number is no timeout. timeout milliseconds for recomputing the result (default: 0) maxcount Number 0 or negative number is no limit. max count of matched text while recomputing the result. if search exceeded total count, "total" value becomes maxcount + 1 (default: 0) pos List [lnum, col, off] value when recomputing the result. this changes "current" result value. see cursor(), getpos() (default: cursor's position)
searchdecl({name} [, {global} [, {thisblock}]]) searchdecl() Search for the declaration of {name}.
With a non-zero {global} argument it works like gD, find first match in the file. Otherwise it works like gd, find first match in the function.
With a non-zero {thisblock} argument matches in a {} block that ends before the cursor position are ignored. Avoids finding variable declarations only valid in another scope.
Moves the cursor to the found match. Returns zero for success, non-zero for failure. Example:
if searchdecl('myvar') == 0
   echo getline('.')
endif
searchpair() searchpair({start}, {middle}, {end} [, {flags} [, {skip} [, {stopline} [, {timeout}]]]]) Search for the match of a nested start-end pair. This can be used to find the "endif" that matches an "if", while other if/endif pairs in between are ignored. The search starts at the cursor. The default is to search forward, include 'b' in {flags} to search backward. If a match is found, the cursor is positioned at it and the line number is returned. If no match is found 0 or -1 is returned and the cursor doesn't move. No error message is given.
{start}, {middle} and {end} are patterns, see pattern. They must not contain \( \) pairs. Use of \%( \) is allowed. When {middle} is not empty, it is found when searching from either direction, but only when not in a nested start-end pair. A typical use is:
echo searchpair('\<if\>', '\<else\>', '\<endif\>')
By leaving {middle} empty the "else" is skipped.
{flags} 'b', 'c', 'n', 's', 'w' and 'W' are used like with search(). Additionally: 'r' Repeat until no more matches found; will find the outer pair. Implies the 'W' flag. 'm' Return number of matches instead of line number with the match; will be > 1 when 'r' is used. Note: it's nearly always a good idea to use the 'W' flag, to avoid wrapping around the end of the file.
When a match for {start}, {middle} or {end} is found, the {skip} expression is evaluated with the cursor positioned on the start of the match. It should return non-zero if this match is to be skipped. E.g., because it is inside a comment or a string. When {skip} is omitted or empty, every match is accepted. When evaluating {skip} causes an error the search is aborted and -1 returned. {skip} can be a string, a lambda, a funcref or a partial. Anything else makes the function fail.
For {stopline} and {timeout} see search().
The value of 'ignorecase' is used. 'magic' is ignored, the patterns are used like it's on.
The search starts exactly at the cursor. A match with {start}, {middle} or {end} at the next character, in the direction of searching, is the first one found. Example:
if 1
  if 2
  endif 2
endif 1
When starting at the "if 2", with the cursor on the "i", and searching forwards, the "endif 2" is found. When starting on the character just before the "if 2", the "endif 1" will be found. That's because the "if 2" will be found first, and then this is considered to be a nested if/endif from "if 2" to "endif 2". When searching backwards and {end} is more than one character, it may be useful to put "\zs" at the end of the pattern, so that when the cursor is inside a match with the end it finds the matching start.
Example, to find the "endif" command in a Vim script:
echo searchpair('\<if\>', '\<el\%[seif]\>', '\<en\%[dif]\>', 'W',
\ 'getline(".") =~ "^\\s*\""')
The cursor must be at or after the "if" for which a match is to be found. Note that single-quote strings are used to avoid having to double the backslashes. The skip expression only catches comments at the start of a line, not after a command. Also, a word "en" or "if" halfway through a line is considered a match. Another example, to search for the matching "{" of a "}":
echo searchpair('{', '', '}', 'bW')
This works when the cursor is at or before the "}" for which a match is to be found. To reject matches that syntax highlighting recognized as strings:
echo searchpair('{', '', '}', 'bW',
     \ 'synIDattr(synID(line("."), col("."), 0), "name") =~? "string"')
searchpairpos() searchpairpos({start}, {middle}, {end} [, {flags} [, {skip} [, {stopline} [, {timeout}]]]]) Same as searchpair(), but returns a List with the line and column position of the match. The first element of the List is the line number and the second element is the byte index of the column position of the match. If no match is found, returns [0, 0].
let [lnum,col] = searchpairpos('{', '', '}', 'n')
See match-parens for a bigger and more useful example.
searchpos() searchpos({pattern} [, {flags} [, {stopline} [, {timeout} [, {skip}]]]]) Same as search(), but returns a List with the line and column position of the match. The first element of the List is the line number and the second element is the byte index of the column position of the match. If no match is found, returns [0, 0]. Example:
let [lnum, col] = searchpos('mypattern', 'n')
When the 'p' flag is given then there is an extra item with the sub-pattern match number search()-sub-match. Example:
let [lnum, col, submatch] = searchpos('\(\l\)\|\(\u\)', 'np')
In this example "submatch" is 2 when a lowercase letter is found /\l, 3 when an uppercase letter is found /\u.
serverlist() serverlist() Returns a list of server addresses, or empty if all servers were stopped. serverstart() serverstop() Example:
echo serverlist()
serverstart([{address}]) serverstart() Opens a socket or named pipe at {address} and listens for RPC messages. Clients can send API commands to the returned address to control Nvim.
Returns the address string (which may differ from the {address} argument, see below).
If {address} has a colon (":") it is a TCP/IPv4/IPv6 address where the last ":" separates host and port (empty or zero assigns a random port).
Else {address} is the path to a named pipe (except on Windows).
If {address} has no slashes ("/") it is treated as the "name" part of a generated path in this format:
stdpath("run").."/{name}.{pid}.{counter}"
If {address} is omitted the name is "nvim".
echo serverstart()
=> /tmp/nvim.bram/oknANW/nvim.15430.5
Example bash command to list all Nvim servers:
ls ${XDG_RUNTIME_DIR:-${TMPDIR}nvim.${USER}}/*/nvim.*.0
Example named pipe:
if has('win32')
  echo serverstart('\\.\pipe\nvim-pipe-1234')
else
  echo serverstart('nvim.sock')
endif
Example TCP/IP address:
echo serverstart('::1:12345')
serverstop({address}) serverstop() Closes the pipe or socket at {address}. Returns TRUE if {address} is valid, else FALSE. If v:servername is stopped it is set to the next available address in serverlist().
setbufline({buf}, {lnum}, {text}) setbufline() Set line {lnum} to {text} in buffer {buf}. This works like setline() for the specified buffer.
This function works only for loaded buffers. First call bufload() if needed.
To insert lines use appendbufline().
{text} can be a string to set one line, or a List of strings to set multiple lines. If the List extends below the last line then those lines are added. If the List is empty then nothing is changed and zero is returned.
For the use of {buf}, see bufname() above.
{lnum} is used like with setline(). Use "$" to refer to the last line in buffer {buf}. When {lnum} is just below the last line the {text} will be added below the last line. On success 0 is returned, on failure 1 is returned.
If {buf} is not a valid buffer or {lnum} is not valid, an error message is given.
setbufvar({buf}, {varname}, {val}) setbufvar() Set option or local variable {varname} in buffer {buf} to {val}. This also works for a global or local window option, but it doesn't work for a global or local window variable. For a local window option the global value is unchanged. For the use of {buf}, see bufname() above. The {varname} argument is a string. Note that the variable name without "b:" must be used. Examples:
call setbufvar(1, "&mod", 1)
call setbufvar("todo", "myvar", "foobar")
This function is not available in the sandbox.
setcellwidths({list}) setcellwidths() Specify overrides for cell widths of character ranges. This tells Vim how wide characters are when displayed in the terminal, counted in screen cells. The values override 'ambiwidth'. Example:
call setcellwidths([
             \ [0x111, 0x111, 1],
             \ [0x2194, 0x2199, 2],
             \ ])
The {list} argument is a List of Lists with each three numbers: [{low}, {high}, {width}]. E1109 E1110 {low} and {high} can be the same, in which case this refers to one character. Otherwise it is the range of characters from {low} to {high} (inclusive). E1111 E1114 Only characters with value 0x80 and higher can be used.
{width} must be either 1 or 2, indicating the character width in screen cells. E1112 An error is given if the argument is invalid, also when a range overlaps with another. E1113
If the new value causes 'fillchars' or 'listchars' to become invalid it is rejected and an error is given.
To clear the overrides pass an empty {list}:
call setcellwidths([])
You can use the script $VIMRUNTIME/tools/emoji_list.vim to see the effect for known emoji characters. Move the cursor through the text to check if the cell widths of your terminal match with what Vim knows about each emoji. If it doesn't look right you need to adjust the {list} argument.
setcharpos({expr}, {list}) setcharpos() Same as setpos() but uses the specified column number as the character index instead of the byte index in the line.
Example: With the text "μ—¬λ³΄μ„Έμš”" in line 8:
call setcharpos('.', [0, 8, 4, 0])
positions the cursor on the fourth character 'μš”'.
call setpos('.', [0, 8, 4, 0])
positions the cursor on the second character '보'.
setcharsearch({dict}) setcharsearch() Set the current character search information to {dict}, which contains one or more of the following entries:
char character which will be used for a subsequent , or ; command; an empty string clears the character search forward direction of character search; 1 for forward, 0 for backward until type of character search; 1 for a t or T character search, 0 for an f or F character search
This can be useful to save/restore a user's character search from a script:
let prevsearch = getcharsearch()
" Perform a command which clobbers user's search
call setcharsearch(prevsearch)
Also see getcharsearch().
setcmdline({str} [, {pos}]) setcmdline() Set the command line to {str} and set the cursor position to {pos}. If {pos} is omitted, the cursor is positioned after the text. Returns 0 when successful, 1 when not editing the command line.
setcmdpos({pos}) setcmdpos() Set the cursor position in the command line to byte position {pos}. The first position is 1. Use getcmdpos() to obtain the current position. Only works while editing the command line, thus you must use c_CTRL-\_e, c_CTRL-R_= or c_CTRL-R_CTRL-R with '='. For c_CTRL-\_e and c_CTRL-R_CTRL-R with '=' the position is set after the command line is set to the expression. For c_CTRL-R_= it is set after evaluating the expression but before inserting the resulting text. When the number is too big the cursor is put at the end of the line. A number smaller than one has undefined results. Returns 0 when successful, 1 when not editing the command line.
setcursorcharpos({lnum}, {col} [, {off}]) setcursorcharpos() setcursorcharpos({list}) Same as cursor() but uses the specified column number as the character index instead of the byte index in the line.
Example: With the text "μ—¬λ³΄μ„Έμš”" in line 4:
call setcursorcharpos(4, 3)
positions the cursor on the third character 'μ„Έ'.
call cursor(4, 3)
positions the cursor on the first character 'μ—¬'.
setenv({name}, {val}) setenv() Set environment variable {name} to {val}. Example:
call setenv('HOME', '/home/myhome')
When {val} is v:null the environment variable is deleted. See also expr-env.
setfperm({fname}, {mode}) setfperm() chmod Set the file permissions for {fname} to {mode}. {mode} must be a string with 9 characters. It is of the form "rwxrwxrwx", where each group of "rwx" flags represent, in turn, the permissions of the owner of the file, the group the file belongs to, and other users. A '-' character means the permission is off, any other character means on. Multi-byte characters are not supported.
For example "rw-r-----" means read-write for the user, readable by the group, not accessible by others. "xx-x-----" would do the same thing.
Returns non-zero for success, zero for failure.
To read permissions see getfperm().
setline({lnum}, {text}) setline() Set line {lnum} of the current buffer to {text}. To insert lines use append(). To set lines in another buffer use setbufline().
{lnum} is used like with getline(). When {lnum} is just below the last line the {text} will be added below the last line. {text} can be any type or a List of any type, each item is converted to a String. When {text} is an empty List then nothing is changed and FALSE is returned.
If this succeeds, FALSE is returned. If this fails (most likely because {lnum} is invalid) TRUE is returned.
Example:
call setline(5, strftime("%c"))
When {text} is a List then line {lnum} and following lines will be set to the items in the list. Example:
call setline(5, ['aaa', 'bbb', 'ccc'])
This is equivalent to:
for [n, l] in [[5, 'aaa'], [6, 'bbb'], [7, 'ccc']]
  call setline(n, l)
endfor
Note: The '[ and '] marks are not set.
setloclist({nr}, {list} [, {action} [, {what}]]) setloclist() Create or replace or add to the location list for window {nr}. {nr} can be the window number or the window-ID. When {nr} is zero the current window is used.
For a location list window, the displayed location list is modified. For an invalid window number {nr}, -1 is returned. Otherwise, same as setqflist(). Also see location-list.
For {action} see setqflist-action.
If the optional {what} dictionary argument is supplied, then only the items listed in {what} are set. Refer to setqflist() for the list of supported keys in {what}.
setmatches({list} [, {win}]) setmatches() Restores a list of matches saved by getmatches() for the current window. Returns 0 if successful, otherwise -1. All current matches are cleared before the list is restored. See example for getmatches(). If {win} is specified, use the window with this number or window ID instead of the current window.
setpos({expr}, {list}) setpos() Set the position for String {expr}. Possible values: . the cursor 'x mark x
{list} must be a List with four or five numbers: [bufnum, lnum, col, off] [bufnum, lnum, col, off, curswant]
"bufnum" is the buffer number. Zero can be used for the current buffer. When setting an uppercase mark "bufnum" is used for the mark position. For other marks it specifies the buffer to set the mark in. You can use the bufnr() function to turn a file name into a buffer number. For setting the cursor and the ' mark "bufnum" is ignored, since these are associated with a window, not a buffer. Does not change the jumplist.
"lnum" and "col" are the position in the buffer. The first column is 1. Use a zero "lnum" to delete a mark. If "col" is smaller than 1 then 1 is used. To use the character count instead of the byte count, use setcharpos().
The "off" number is only used when 'virtualedit' is set. Then it is the offset in screen columns from the start of the character. E.g., a position within a <Tab> or after the last character.
The "curswant" number is only used when setting the cursor position. It sets the preferred column for when moving the cursor vertically. When the "curswant" number is missing the preferred column is not set. When it is present and setting a mark position it is not used.
Note that for '< and '> changing the line number may result in the marks to be effectively be swapped, so that '< is always before '>.
Returns 0 when the position could be set, -1 otherwise. An error message is given if {expr} is invalid.
This does not restore the preferred column for moving vertically; if you set the cursor position with this, j and k motions will jump to previous columns! Use cursor() to also set the preferred column. Also see the "curswant" key in winrestview().
setqflist({list} [, {action} [, {what}]]) setqflist() Create or replace or add to the quickfix list.
If the optional {what} dictionary argument is supplied, then only the items listed in {what} are set. The first {list} argument is ignored. See below for the supported items in {what}. setqflist-what When {what} is not present, the items in {list} are used. Each item must be a dictionary. Non-dictionary items in {list} are ignored. Each dictionary item can contain the following entries:
bufnr buffer number; must be the number of a valid buffer filename name of a file; only used when "bufnr" is not present or it is invalid. module name of a module; if given it will be used in quickfix error window instead of the filename. lnum line number in the file end_lnum end of lines, if the item spans multiple lines pattern search pattern used to locate the error col column number vcol when non-zero: "col" is visual column when zero: "col" is byte index end_col end column, if the item spans multiple columns nr error number text description of the error type single-character error type, 'E', 'W', etc. valid recognized error message user_data custom data associated with the item, can be any type.
The "col", "vcol", "nr", "type" and "text" entries are optional. Either "lnum" or "pattern" entry can be used to locate a matching error line. If the "filename" and "bufnr" entries are not present or neither the "lnum" or "pattern" entries are present, then the item will not be handled as an error line. If both "pattern" and "lnum" are present then "pattern" will be used. If the "valid" entry is not supplied, then the valid flag is set when "bufnr" is a valid buffer or "filename" exists. If you supply an empty {list}, the quickfix list will be cleared. Note that the list is not exactly the same as what getqflist() returns.
{action} values: setqflist-action E927 'a' The items from {list} are added to the existing quickfix list. If there is no existing list, then a new list is created.
'r' The items from the current quickfix list are replaced with the items from {list}. This can also be used to clear the list:
call setqflist([], 'r')
'f' All the quickfix lists in the quickfix stack are freed.
If {action} is not present or is set to ' ', then a new list is created. The new quickfix list is added after the current quickfix list in the stack and all the following lists are freed. To add a new quickfix list at the end of the stack, set "nr" in {what} to "$".
The following items can be specified in dictionary {what}: context quickfix list context. See quickfix-context efm errorformat to use when parsing text from "lines". If this is not present, then the 'errorformat' option value is used. See quickfix-parse id quickfix list identifier quickfix-ID idx index of the current entry in the quickfix list specified by "id" or "nr". If set to '$', then the last entry in the list is set as the current entry. See quickfix-index items list of quickfix entries. Same as the {list} argument. lines use 'errorformat' to parse a list of lines and add the resulting entries to the quickfix list {nr} or {id}. Only a List value is supported. See quickfix-parse nr list number in the quickfix stack; zero means the current quickfix list and "$" means the last quickfix list. quickfixtextfunc function to get the text to display in the quickfix window. The value can be the name of a function or a funcref or a lambda. Refer to quickfix-window-function for an explanation of how to write the function and an example. title quickfix list title text. See quickfix-title Unsupported keys in {what} are ignored. If the "nr" item is not present, then the current quickfix list is modified. When creating a new quickfix list, "nr" can be set to a value one greater than the quickfix stack size. When modifying a quickfix list, to guarantee that the correct list is modified, "id" should be used instead of "nr" to specify the list.
Examples (See also setqflist-examples):
call setqflist([], 'r', {'title': 'My search'})
call setqflist([], 'r', {'nr': 2, 'title': 'Errors'})
call setqflist([], 'a', {'id':qfid, 'lines':["F1:10:L10"]})
Returns zero for success, -1 for failure.
This function can be used to create a quickfix list independent of the 'errorformat' setting. Use a command like :cc 1 to jump to the first position.
setreg({regname}, {value} [, {options}]) setreg() Set the register {regname} to {value}. If {regname} is "" or "@", the unnamed register '"' is used. The {regname} argument is a string.
{value} may be any value returned by getreg() or getreginfo(), including a List or Dict. If {options} contains "a" or {regname} is upper case, then the value is appended.
{options} can also contain a register type specification: "c" or "v" charwise mode "l" or "V" linewise mode "b" or "<CTRL-V>" blockwise-visual mode If a number immediately follows "b" or "<CTRL-V>" then this is used as the width of the selection - if it is not specified then the width of the block is set to the number of characters in the longest line (counting a <Tab> as 1 character). If {options} contains "u" or '"', then the unnamed register is set to point to register {regname}.
If {options} contains no register settings, then the default is to use character mode unless {value} ends in a <NL> for string {value} and linewise mode for list {value}. Blockwise mode is never selected automatically. Returns zero for success, non-zero for failure.
E883 Note: you may not use List containing more than one item to set search and expression registers. Lists containing no items act like empty strings.
Examples:
call setreg(v:register, @*)
call setreg('*', @%, 'ac')
call setreg('a', "1\n2\n3", 'b5')
call setreg('"', { 'points_to': 'a'})
This example shows using the functions to save and restore a register:
let var_a = getreginfo()
call setreg('a', var_a)
or:
let var_a = getreg('a', 1, 1)
let var_amode = getregtype('a')
" ....
call setreg('a', var_a, var_amode)
Note: you may not reliably restore register value without using the third argument to getreg() as without it newlines are represented as newlines AND Nul bytes are represented as newlines as well, see NL-used-for-Nul.
You can also change the type of a register by appending nothing:
call setreg('a', '', 'al')
settabvar({tabnr}, {varname}, {val}) settabvar() Set tab-local variable {varname} to {val} in tab page {tabnr}. t:var The {varname} argument is a string. Note that the variable name without "t:" must be used. Tabs are numbered starting with one. This function is not available in the sandbox.
settabwinvar({tabnr}, {winnr}, {varname}, {val}) settabwinvar() Set option or local variable {varname} in window {winnr} to {val}. Tabs are numbered starting with one. For the current tabpage use setwinvar(). {winnr} can be the window number or the window-ID. When {winnr} is zero the current window is used. This also works for a global or local buffer option, but it doesn't work for a global or local buffer variable. For a local buffer option the global value is unchanged. Note that the variable name without "w:" must be used. Examples:
call settabwinvar(1, 1, "&list", 0)
call settabwinvar(3, 2, "myvar", "foobar")
This function is not available in the sandbox.
settagstack({nr}, {dict} [, {action}]) settagstack() Modify the tag stack of the window {nr} using {dict}. {nr} can be the window number or the window-ID.
For a list of supported items in {dict}, refer to gettagstack(). "curidx" takes effect before changing the tag stack. E962 How the tag stack is modified depends on the {action} argument:
If {action} is not present or is set to 'r', then the tag stack is replaced.
If {action} is set to 'a', then new entries from {dict} are pushed (added) onto the tag stack.
If {action} is set to 't', then all the entries from the current entry in the tag stack or "curidx" in {dict} are removed and then new entries are pushed to the stack.
The current index is set to one after the length of the tag stack after the modification.
Returns zero for success, -1 for failure.
Examples (for more examples see tagstack-examples): Empty the tag stack of window 3:
call settagstack(3, {'items' : []})
Save and restore the tag stack:
let stack = gettagstack(1003)
" do something else
call settagstack(1003, stack)
unlet stack
setwinvar({nr}, {varname}, {val}) setwinvar() Like settabwinvar() for the current tab page. Examples:
call setwinvar(1, "&list", 0)
call setwinvar(2, "myvar", "foobar")
sha256({string}) sha256() Returns a String with 64 hex characters, which is the SHA256 checksum of {string}.
shellescape({string} [, {special}]) shellescape() Escape {string} for use as a shell command argument.
On Windows when 'shellslash' is not set, encloses {string} in double-quotes and doubles all double-quotes within {string}. Otherwise encloses {string} in single-quotes and replaces all "'" with "'\''".
If {special} is a non-zero-arg:
Special items such as "!", "%", "#" and "<cword>" will be preceded by a backslash. The backslash will be removed again by the :! command.
The <NL> character is escaped.
If 'shell' contains "csh" in the tail:
The "!" character will be escaped. This is because csh and tcsh use "!" for history replacement even in single-quotes.
The <NL> character is escaped (twice if {special} is a non-zero-arg).
If 'shell' contains "fish" in the tail, the "\" character will be escaped because in fish it is used as an escape character inside single quotes.
Example of use with a :! command:
exe '!dir ' .. shellescape(expand('<cfile>'), 1)
This results in a directory listing for the file under the cursor. Example of use with system():
call system("chmod +w -- " .. shellescape(expand("%")))
See also ::S.
shiftwidth([{col}]) shiftwidth() Returns the effective value of 'shiftwidth'. This is the 'shiftwidth' value unless it is zero, in which case it is the 'tabstop' value. To be backwards compatible in indent plugins, use this:
if exists('*shiftwidth')
  func s:sw()
    return shiftwidth()
  endfunc
else
  func s:sw()
    return &sw
  endfunc
endif
And then use s:sw() instead of &sw.
When there is one argument {col} this is used as column number for which to return the 'shiftwidth' value. This matters for the 'vartabstop' feature. If no {col} argument is given, column 1 will be assumed.
sign_define({name} [, {dict}]) sign_define() sign_define({list}) Define a new sign named {name} or modify the attributes of an existing sign. This is similar to the :sign-define command.
Prefix {name} with a unique text to avoid name collisions. There is no {group} like with placing signs.
The {name} can be a String or a Number. The optional {dict} argument specifies the sign attributes. The following values are supported: icon full path to the bitmap file for the sign. linehl highlight group used for the whole line the sign is placed in. numhl highlight group used for the line number where the sign is placed. text text that is displayed when there is no icon or the GUI is not being used. texthl highlight group used for the text item culhl highlight group used for the text item when the cursor is on the same line as the sign and 'cursorline' is enabled.
If the sign named {name} already exists, then the attributes of the sign are updated.
The one argument {list} can be used to define a list of signs. Each list item is a dictionary with the above items in {dict} and a "name" item for the sign name.
Returns 0 on success and -1 on failure. When the one argument {list} is used, then returns a List of values one for each defined sign.
Examples:
call sign_define("mySign", {
        \ "text" : "=>",
        \ "texthl" : "Error",
        \ "linehl" : "Search"})
call sign_define([
        \ {'name' : 'sign1',
        \  'text' : '=>'},
        \ {'name' : 'sign2',
        \  'text' : '!!'}
        \ ])
sign_getdefined([{name}]) sign_getdefined() Get a list of defined signs and their attributes. This is similar to the :sign-list command.
If the {name} is not supplied, then a list of all the defined signs is returned. Otherwise the attribute of the specified sign is returned.
Each list item in the returned value is a dictionary with the following entries: icon full path to the bitmap file of the sign linehl highlight group used for the whole line the sign is placed in; not present if not set. name name of the sign numhl highlight group used for the line number where the sign is placed; not present if not set. text text that is displayed when there is no icon or the GUI is not being used. texthl highlight group used for the text item; not present if not set. culhl highlight group used for the text item when the cursor is on the same line as the sign and 'cursorline' is enabled; not present if not set.
Returns an empty List if there are no signs and when {name} is not found.
Examples:
" Get a list of all the defined signs
echo sign_getdefined()
" Get the attribute of the sign named mySign
echo sign_getdefined("mySign")
sign_getplaced([{buf} [, {dict}]]) sign_getplaced() Return a list of signs placed in a buffer or all the buffers. This is similar to the :sign-place-list command.
If the optional buffer name {buf} is specified, then only the list of signs placed in that buffer is returned. For the use of {buf}, see bufname(). The optional {dict} can contain the following entries: group select only signs in this group id select sign with this identifier lnum select signs placed in this line. For the use of {lnum}, see line(). If {group} is "*", then signs in all the groups including the global group are returned. If {group} is not supplied or is an empty string, then only signs in the global group are returned. If no arguments are supplied, then signs in the global group placed in all the buffers are returned. See sign-group.
Each list item in the returned value is a dictionary with the following entries: bufnr number of the buffer with the sign signs list of signs placed in {bufnr}. Each list item is a dictionary with the below listed entries
The dictionary for each sign contains the following entries: group sign group. Set to '' for the global group. id identifier of the sign lnum line number where the sign is placed name name of the defined sign priority sign priority
The returned signs in a buffer are ordered by their line number and priority.
Returns an empty list on failure or if there are no placed signs.
Examples:
" Get a List of signs placed in eval.c in the
" global group
echo sign_getplaced("eval.c")
" Get a List of signs in group 'g1' placed in eval.c
echo sign_getplaced("eval.c", {'group' : 'g1'})
" Get a List of signs placed at line 10 in eval.c
echo sign_getplaced("eval.c", {'lnum' : 10})
" Get sign with identifier 10 placed in a.py
echo sign_getplaced("a.py", {'id' : 10})
" Get sign with id 20 in group 'g1' placed in a.py
echo sign_getplaced("a.py", {'group' : 'g1',
                                \  'id' : 20})
" Get a List of all the placed signs
echo sign_getplaced()
sign_jump({id}, {group}, {buf}) sign_jump() Open the buffer {buf} or jump to the window that contains {buf} and position the cursor at sign {id} in group {group}. This is similar to the :sign-jump command.
If {group} is an empty string, then the global group is used. For the use of {buf}, see bufname().
Returns the line number of the sign. Returns -1 if the arguments are invalid.
Example:
" Jump to sign 10 in the current buffer
call sign_jump(10, '', '')
sign_place({id}, {group}, {name}, {buf} [, {dict}]) sign_place() Place the sign defined as {name} at line {lnum} in file or buffer {buf} and assign {id} and {group} to sign. This is similar to the :sign-place command.
If the sign identifier {id} is zero, then a new identifier is allocated. Otherwise the specified number is used. {group} is the sign group name. To use the global sign group, use an empty string. {group} functions as a namespace for {id}, thus two groups can use the same IDs. Refer to sign-identifier and sign-group for more information.
{name} refers to a defined sign. {buf} refers to a buffer name or number. For the accepted values, see bufname().
The optional {dict} argument supports the following entries: lnum line number in the file or buffer {buf} where the sign is to be placed. For the accepted values, see line(). priority priority of the sign. See sign-priority for more information.
If the optional {dict} is not specified, then it modifies the placed sign {id} in group {group} to use the defined sign {name}.
Returns the sign identifier on success and -1 on failure.
Examples:
" Place a sign named sign1 with id 5 at line 20 in
" buffer json.c
call sign_place(5, '', 'sign1', 'json.c',
                                \ {'lnum' : 20})
" Updates sign 5 in buffer json.c to use sign2
call sign_place(5, '', 'sign2', 'json.c')
" Place a sign named sign3 at line 30 in
" buffer json.c with a new identifier
let id = sign_place(0, '', 'sign3', 'json.c',
                                \ {'lnum' : 30})
" Place a sign named sign4 with id 10 in group 'g3'
" at line 40 in buffer json.c with priority 90
call sign_place(10, 'g3', 'sign4', 'json.c',
                \ {'lnum' : 40, 'priority' : 90})
sign_placelist({list}) sign_placelist() Place one or more signs. This is similar to the sign_place() function. The {list} argument specifies the List of signs to place. Each list item is a dict with the following sign attributes: buffer Buffer name or number. For the accepted values, see bufname(). group Sign group. {group} functions as a namespace for {id}, thus two groups can use the same IDs. If not specified or set to an empty string, then the global group is used. See sign-group for more information. id Sign identifier. If not specified or zero, then a new unique identifier is allocated. Otherwise the specified number is used. See sign-identifier for more information. lnum Line number in the buffer where the sign is to be placed. For the accepted values, see line(). name Name of the sign to place. See sign_define() for more information. priority Priority of the sign. When multiple signs are placed on a line, the sign with the highest priority is used. If not specified, the default value of 10 is used. See sign-priority for more information.
If {id} refers to an existing sign, then the existing sign is modified to use the specified {name} and/or {priority}.
Returns a List of sign identifiers. If failed to place a sign, the corresponding list item is set to -1.
Examples:
" Place sign s1 with id 5 at line 20 and id 10 at line
" 30 in buffer a.c
let [n1, n2] = sign_placelist([
        \ {'id' : 5,
        \  'name' : 's1',
        \  'buffer' : 'a.c',
        \  'lnum' : 20},
        \ {'id' : 10,
        \  'name' : 's1',
        \  'buffer' : 'a.c',
        \  'lnum' : 30}
        \ ])
" Place sign s1 in buffer a.c at line 40 and 50
" with auto-generated identifiers
let [n1, n2] = sign_placelist([
        \ {'name' : 's1',
        \  'buffer' : 'a.c',
        \  'lnum' : 40},
        \ {'name' : 's1',
        \  'buffer' : 'a.c',
        \  'lnum' : 50}
        \ ])
sign_undefine([{name}]) sign_undefine() sign_undefine({list}) Deletes a previously defined sign {name}. This is similar to the :sign-undefine command. If {name} is not supplied, then deletes all the defined signs.
The one argument {list} can be used to undefine a list of signs. Each list item is the name of a sign.
Returns 0 on success and -1 on failure. For the one argument {list} call, returns a list of values one for each undefined sign.
Examples:
" Delete a sign named mySign
call sign_undefine("mySign")
" Delete signs 'sign1' and 'sign2'
call sign_undefine(["sign1", "sign2"])
" Delete all the signs
call sign_undefine()
sign_unplace({group} [, {dict}]) sign_unplace() Remove a previously placed sign in one or more buffers. This is similar to the :sign-unplace command.
{group} is the sign group name. To use the global sign group, use an empty string. If {group} is set to "*", then all the groups including the global group are used. The signs in {group} are selected based on the entries in {dict}. The following optional entries in {dict} are supported: buffer buffer name or number. See bufname(). id sign identifier If {dict} is not supplied, then all the signs in {group} are removed.
Returns 0 on success and -1 on failure.
Examples:
" Remove sign 10 from buffer a.vim
call sign_unplace('', {'buffer' : "a.vim", 'id' : 10})
" Remove sign 20 in group 'g1' from buffer 3
call sign_unplace('g1', {'buffer' : 3, 'id' : 20})
" Remove all the signs in group 'g2' from buffer 10
call sign_unplace('g2', {'buffer' : 10})
" Remove sign 30 in group 'g3' from all the buffers
call sign_unplace('g3', {'id' : 30})
" Remove all the signs placed in buffer 5
call sign_unplace('*', {'buffer' : 5})
" Remove the signs in group 'g4' from all the buffers
call sign_unplace('g4')
" Remove sign 40 from all the buffers
call sign_unplace('*', {'id' : 40})
" Remove all the placed signs from all the buffers
call sign_unplace('*')
sign_unplacelist({list}) sign_unplacelist() Remove previously placed signs from one or more buffers. This is similar to the sign_unplace() function.
The {list} argument specifies the List of signs to remove. Each list item is a dict with the following sign attributes: buffer buffer name or number. For the accepted values, see bufname(). If not specified, then the specified sign is removed from all the buffers. group sign group name. If not specified or set to an empty string, then the global sign group is used. If set to "*", then all the groups including the global group are used. id sign identifier. If not specified, then all the signs in the specified group are removed.
Returns a List where an entry is set to 0 if the corresponding sign was successfully removed or -1 on failure.
Example:
" Remove sign with id 10 from buffer a.vim and sign
" with id 20 from buffer b.vim
call sign_unplacelist([
        \ {'id' : 10, 'buffer' : "a.vim"},
        \ {'id' : 20, 'buffer' : 'b.vim'},
        \ ])
simplify({filename}) simplify() Simplify the file name as much as possible without changing the meaning. Shortcuts (on MS-Windows) or symbolic links (on Unix) are not resolved. If the first path component in {filename} designates the current directory, this will be valid for the result as well. A trailing path separator is not removed either. On Unix "//path" is unchanged, but "///path" is simplified to "/path" (this follows the Posix standard). Example:
simplify("./dir/.././/file/") == "./file/"
Note: The combination "dir/.." is only removed if "dir" is a searchable directory or does not exist. On Unix, it is also removed when "dir" is a symbolic link within the same directory. In order to resolve all the involved symbolic links before simplifying the path name, use resolve().
sin({expr}) sin() Return the sine of {expr}, measured in radians, as a Float. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo sin(100)
-0.506366
echo sin(-4.01)
0.763301
sinh({expr}) sinh() Return the hyperbolic sine of {expr} as a Float in the range [-inf, inf]. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo sinh(0.5)
0.521095
echo sinh(-0.9)
-1.026517
slice({expr}, {start} [, {end}]) slice() Similar to using a slice "expr[start : end]", but "end" is used exclusive. And for a string the indexes are used as character indexes instead of byte indexes. Also, composing characters are treated as a part of the preceding base character. When {end} is omitted the slice continues to the last item. When {end} is -1 the last item is omitted. Returns an empty value if {start} or {end} are invalid.
sockconnect({mode}, {address} [, {opts}]) sockconnect() Connect a socket to an address. If {mode} is "pipe" then {address} should be the path of a local domain socket (on unix) or named pipe (on Windows). If {mode} is "tcp" then {address} should be of the form "host:port" where the host should be an ip address or host name, and port the port number.
For "pipe" mode, see luv-pipe-handle. For "tcp" mode, see luv-tcp-handle.
Returns a channel ID. Close the socket with chanclose(). Use chansend() to send data over a bytes socket, and rpcrequest() and rpcnotify() to communicate with a RPC socket.
{opts} is an optional dictionary with these keys: on_data : callback invoked when data was read from socket data_buffered : read socket data in channel-buffered mode. rpc : If set, msgpack-rpc will be used to communicate over the socket. Returns:
The channel ID on success (greater than zero)
0 on invalid arguments or connection failure.
sort({list} [, {how} [, {dict}]]) sort() E702 Sort the items in {list} in-place. Returns {list}.
If you want a list to remain unmodified make a copy first:
let sortedlist = sort(copy(mylist))
When {how} is omitted or is a string, then sort() uses the string representation of each item to sort on. Numbers sort after Strings, Lists after Numbers. For sorting text in the current buffer use :sort.
When {how} is given and it is 'i' then case is ignored. For backwards compatibility, the value one can be used to ignore case. Zero means to not ignore case.
When {how} is given and it is 'l' then the current collation locale is used for ordering. Implementation details: strcoll() is used to compare strings. See :language check or set the collation locale. v:collate can also be used to check the current locale. Sorting using the locale typically ignores case. Example:
" ΓΆ is sorted similarly to o with English locale.
language collate en_US.UTF8
echo sort(['n', 'o', 'O', 'ΓΆ', 'p', 'z'], 'l')
['n', 'o', 'O', 'ΓΆ', 'p', 'z']
" ΓΆ is sorted after z with Swedish locale.
language collate sv_SE.UTF8
echo sort(['n', 'o', 'O', 'ΓΆ', 'p', 'z'], 'l')
['n', 'o', 'O', 'p', 'z', 'ΓΆ']
This does not work properly on Mac.
When {how} is given and it is 'n' then all items will be sorted numerical (Implementation detail: this uses the strtod() function to parse numbers, Strings, Lists, Dicts and Funcrefs will be considered as being 0).
When {how} is given and it is 'N' then all items will be sorted numerical. This is like 'n' but a string containing digits will be used as the number they represent.
When {how} is given and it is 'f' then all items will be sorted numerical. All values must be a Number or a Float.
When {how} is a Funcref or a function name, this function is called to compare items. The function is invoked with two items as argument and must return zero if they are equal, 1 or bigger if the first one sorts after the second one, -1 or smaller if the first one sorts before the second one.
{dict} is for functions with the "dict" attribute. It will be used to set the local variable "self". Dictionary-function
The sort is stable, items which compare equal (as number or as string) will keep their relative position. E.g., when sorting on numbers, text strings will sort next to each other, in the same order as they were originally.
Example:
func MyCompare(i1, i2)
   return a:i1 == a:i2 ? 0 : a:i1 > a:i2 ? 1 : -1
endfunc
eval mylist->sort("MyCompare")
A shorter compare version for this specific simple case, which ignores overflow:
func MyCompare(i1, i2)
   return a:i1 - a:i2
endfunc
For a simple expression you can use a lambda:
eval mylist->sort({i1, i2 -> i1 - i2})
soundfold({word}) soundfold() Return the sound-folded equivalent of {word}. Uses the first language in 'spelllang' for the current window that supports soundfolding. 'spell' must be set. When no sound folding is possible the {word} is returned unmodified. This can be used for making spelling suggestions. Note that the method can be quite slow.
spellbadword([{sentence}]) spellbadword() Without argument: The result is the badly spelled word under or after the cursor. The cursor is moved to the start of the bad word. When no bad word is found in the cursor line the result is an empty string and the cursor doesn't move.
With argument: The result is the first word in {sentence} that is badly spelled. If there are no spelling mistakes the result is an empty string.
The return value is a list with two items:
The badly spelled word or an empty string.
The type of the spelling error: "bad" spelling mistake "rare" rare word "local" word only valid in another region "caps" word should start with Capital Example:
echo spellbadword("the quik brown fox")
[quik,bad]
The spelling information for the current window and the value of 'spelllang' are used.
spellsuggest({word} [, {max} [, {capital}]]) spellsuggest() Return a List with spelling suggestions to replace {word}. When {max} is given up to this number of suggestions are returned. Otherwise up to 25 suggestions are returned.
When the {capital} argument is given and it's non-zero only suggestions with a leading capital will be given. Use this after a match with 'spellcapcheck'.
{word} can be a badly spelled word followed by other text. This allows for joining two words that were split. The suggestions also include the following text, thus you can replace a line.
{word} may also be a good word. Similar words will then be returned. {word} itself is not included in the suggestions, although it may appear capitalized.
The spelling information for the current window is used. The values of 'spelllang' and 'spellsuggest' are used.
split({string} [, {pattern} [, {keepempty}]]) split() Make a List out of {string}. When {pattern} is omitted or empty each white-separated sequence of characters becomes an item. Otherwise the string is split where {pattern} matches, removing the matched characters. 'ignorecase' is not used here, add \c to ignore case. /\c When the first or last item is empty it is omitted, unless the {keepempty} argument is given and it's non-zero. Other empty items are kept when {pattern} matches at least one character or when {keepempty} is non-zero. Example:
let words = split(getline('.'), '\W\+')
To split a string in individual characters:
for c in split(mystring, '\zs') | endfor
If you want to keep the separator you can also use '\zs' at the end of the pattern:
echo split('abc:def:ghi', ':\zs')
['abc:', 'def:', 'ghi']
Splitting a table where the first element can be empty:
let items = split(line, ':', 1)
The opposite function is join().
sqrt({expr}) sqrt() Return the non-negative square root of Float {expr} as a Float. {expr} must evaluate to a Float or a Number. When {expr} is negative the result is NaN (Not a Number). Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo sqrt(100)
10.0
echo sqrt(-4.01)
str2float("nan") NaN may be different, it depends on system libraries.
srand([{expr}]) srand() Initialize seed used by rand():
If {expr} is not given, seed values are initialized by reading from /dev/urandom, if possible, or using time(NULL) a.k.a. epoch time otherwise; this only has second accuracy.
If {expr} is given it must be a Number. It is used to initialize the seed values. This is useful for testing or when a predictable sequence is intended.
Examples:
let seed = srand()
let seed = srand(userinput)
echo rand(seed)
state([{what}]) state() Return a string which contains characters indicating the current state. Mostly useful in callbacks that want to do work that may not always be safe. Roughly this works like:
callback uses state() to check if work is safe to do. Yes: then do it right away. No: add to work queue and add a SafeState autocommand.
When SafeState is triggered and executes your autocommand, check with state() if the work can be done now, and if yes remove it from the queue and execute. Remove the autocommand if the queue is now empty. Also see mode().
When {what} is given only characters in this string will be added. E.g, this checks if the screen has scrolled:
if state('s') == ''
   " screen has not scrolled
These characters indicate the state, generally indicating that something is busy: m halfway a mapping, :normal command, feedkeys() or stuffed command o operator pending, e.g. after d a Insert mode autocomplete active x executing an autocommand S not triggering SafeState, e.g. after f or a count c callback invoked, including timer (repeats for recursiveness up to "ccc") s screen has scrolled for messages
stdioopen({opts}) stdioopen() With --headless this opens stdin and stdout as a channel. May be called only once. See channel-stdio. stderr is not handled by this function, see v:stderr.
Close the stdio handles with chanclose(). Use chansend() to send data to stdout, and rpcrequest() and rpcnotify() to communicate over RPC.
{opts} is a dictionary with these keys: on_stdin : callback invoked when stdin is written to. on_print : callback invoked when Nvim needs to print a message, with the message (whose type is string) as sole argument. stdin_buffered : read stdin in channel-buffered mode. rpc : If set, msgpack-rpc will be used to communicate over stdio Returns:
channel-id on success (value is always 1)
0 on invalid arguments
stdpath({what}) stdpath() E6100 Returns standard-path locations of various default files and directories.
{what} Type Description
cache String Cache directory: arbitrary temporary storage for plugins, etc. config String User configuration directory. init.vim is stored here. config_dirs List Other configuration directories. data String User data directory. data_dirs List Other data directories. log String Logs directory (for use by plugins too). run String Run directory: temporary, local storage for sockets, named pipes, etc. state String Session state directory: storage for file drafts, swap, undo, shada.
Example:
echo stdpath("config")
str2float({string} [, {quoted}]) str2float() Convert String {string} to a Float. This mostly works the same as when using a floating point number in an expression, see floating-point-format. But it's a bit more permissive. E.g., "1e40" is accepted, while in an expression you need to write "1.0e40". The hexadecimal form "0x123" is also accepted, but not others, like binary or octal. When {quoted} is present and non-zero then embedded single quotes before the dot are ignored, thus "1'000.0" is a thousand. Text after the number is silently ignored. The decimal point is always '.', no matter what the locale is set to. A comma ends the number: "12,345.67" is converted to 12.0. You can strip out thousands separators with substitute():
let f = str2float(substitute(text, ',', '', 'g'))
Returns 0.0 if the conversion fails.
str2list({string} [, {utf8}]) str2list() Return a list containing the number values which represent each character in String {string}. Examples:
echo str2list(" ")                " returns [32]
echo str2list("ABC")                " returns [65, 66, 67]
list2str() does the opposite.
UTF-8 encoding is always used, {utf8} option has no effect, and exists only for backwards-compatibility. With UTF-8 composing characters are handled properly:
echo str2list("á")                " returns [97, 769]
str2nr({string} [, {base}]) str2nr() Convert string {string} to a number. {base} is the conversion base, it can be 2, 8, 10 or 16. When {quoted} is present and non-zero then embedded single quotes are ignored, thus "1'000'000" is a million.
When {base} is omitted base 10 is used. This also means that a leading zero doesn't cause octal conversion to be used, as with the default String to Number conversion. Example:
let nr = str2nr('0123')
When {base} is 16 a leading "0x" or "0X" is ignored. With a different base the result will be zero. Similarly, when {base} is 8 a leading "0", "0o" or "0O" is ignored, and when {base} is 2 a leading "0b" or "0B" is ignored. Text after the number is silently ignored.
Returns 0 if {string} is empty or on error.
strcharlen({string}) strcharlen() The result is a Number, which is the number of characters in String {string}. Composing characters are ignored. strchars() can count the number of characters, counting composing characters separately.
Returns 0 if {string} is empty or on error.
strcharpart({src}, {start} [, {len} [, {skipcc}]]) strcharpart() Like strpart() but using character index and length instead of byte index and length. When {skipcc} is omitted or zero, composing characters are counted separately. When {skipcc} set to 1, composing characters are treated as a part of the preceding base character, similar to slice(). When a character index is used where a character does not exist it is omitted and counted as one character. For example:
echo strcharpart('abc', -1, 2)
results in 'a'.
Returns an empty string on error.
strchars({string} [, {skipcc}]) strchars() The result is a Number, which is the number of characters in String {string}. When {skipcc} is omitted or zero, composing characters are counted separately. When {skipcc} set to 1, composing characters are ignored. strcharlen() always does this.
Returns zero on error.
{skipcc} is only available after 7.4.755. For backward compatibility, you can define a wrapper function:
if has("patch-7.4.755")
  function s:strchars(str, skipcc)
    return strchars(a:str, a:skipcc)
  endfunction
else
  function s:strchars(str, skipcc)
    if a:skipcc
      return strlen(substitute(a:str, ".", "x", "g"))
    else
      return strchars(a:str)
    endif
  endfunction
endif
strdisplaywidth({string} [, {col}]) strdisplaywidth() The result is a Number, which is the number of display cells String {string} occupies on the screen when it starts at {col} (first column is zero). When {col} is omitted zero is used. Otherwise it is the screen column where to start. This matters for Tab characters. The option settings of the current window are used. This matters for anything that's displayed differently, such as 'tabstop' and 'display'. When {string} contains characters with East Asian Width Class Ambiguous, this function's return value depends on 'ambiwidth'. Returns zero on error. Also see strlen(), strwidth() and strchars().
strftime({format} [, {time}]) strftime() The result is a String, which is a formatted date and time, as specified by the {format} string. The given {time} is used, or the current time if no time is given. The accepted {format} depends on your system, thus this is not portable! See the manual page of the C function strftime() for the format. The maximum length of the result is 80 characters. See also localtime(), getftime() and strptime(). The language can be changed with the :language command. Examples:
echo strftime("%c")                   " Sun Apr 27 11:49:23 1997
echo strftime("%Y %b %d %X")           " 1997 Apr 27 11:53:25
echo strftime("%y%m%d %T")           " 970427 11:53:55
echo strftime("%H:%M")                   " 11:55
echo strftime("%c", getftime("file.c"))
                                 " Show mod time of file.c.
strgetchar({str}, {index}) strgetchar() Get a Number corresponding to the character at {index} in {str}. This uses a zero-based character index, not a byte index. Composing characters are considered separate characters here. Use nr2char() to convert the Number to a String. Returns -1 if {index} is invalid. Also see strcharpart() and strchars().
stridx({haystack}, {needle} [, {start}]) stridx() The result is a Number, which gives the byte index in {haystack} of the first occurrence of the String {needle}. If {start} is specified, the search starts at index {start}. This can be used to find a second match:
let colon1 = stridx(line, ":")
let colon2 = stridx(line, ":", colon1 + 1)
The search is done case-sensitive. For pattern searches use match(). -1 is returned if the {needle} does not occur in {haystack}. See also strridx(). Examples:
echo stridx("An Example", "Example")     " 3
echo stridx("Starting point", "Start")   " 0
echo stridx("Starting point", "start")   " -1
strstr() strchr() stridx() works similar to the C function strstr(). When used with a single character it works similar to strchr().
string({expr}) string() Return {expr} converted to a String. If {expr} is a Number, Float, String, Blob or a composition of them, then the result can be parsed back with eval().
{expr} type result
Stringstring Number 123 Float 123.123456 or 1.123456e8 or str2float('inf') Funcref function('name') Blob 0z00112233.44556677.8899 List [item, item] Dictionary {key: value, key: value} Note that in String values the ' character is doubled. Also see strtrans(). Note 2: Output format is mostly compatible with YAML, except for infinite and NaN floating-point values representations which use str2float(). Strings are also dumped literally, only single quote is escaped, which does not allow using YAML for parsing back binary strings. eval() should always work for strings and floats though and this is the only official method, use msgpackdump() or json_encode() if you need to share data with other application.
strlen({string}) strlen() The result is a Number, which is the length of the String {string} in bytes. If the argument is a Number it is first converted to a String. For other types an error is given and zero is returned. If you want to count the number of multibyte characters use strchars(). Also see len(), strdisplaywidth() and strwidth().
strpart({src}, {start} [, {len} [, {chars}]]) strpart() The result is a String, which is part of {src}, starting from byte {start}, with the byte length {len}. When {chars} is present and TRUE then {len} is the number of characters positions (composing characters are not counted separately, thus "1" means one base character and any following composing characters). To count {start} as characters instead of bytes use strcharpart().
When bytes are selected which do not exist, this doesn't result in an error, the bytes are simply omitted. If {len} is missing, the copy continues from {start} till the end of the {src}.
echo strpart("abcdefg", 3, 2)    " returns 'de'
echo strpart("abcdefg", -2, 4)   " returns 'ab'
echo strpart("abcdefg", 5, 4)    " returns 'fg'
echo strpart("abcdefg", 3)         " returns 'defg'
Note: To get the first character, {start} must be 0. For example, to get the character under the cursor:
strpart(getline("."), col(".") - 1, 1, v:true)
Returns an empty string on error.
strptime({format}, {timestring}) strptime() The result is a Number, which is a unix timestamp representing the date and time in {timestring}, which is expected to match the format specified in {format}.
The accepted {format} depends on your system, thus this is not portable! See the manual page of the C function strptime() for the format. Especially avoid "%c". The value of $TZ also matters.
If the {timestring} cannot be parsed with {format} zero is returned. If you do not know the format of {timestring} you can try different {format} values until you get a non-zero result.
See also strftime(). Examples:
echo strptime("%Y %b %d %X", "1997 Apr 27 11:49:23")
862156163
echo strftime("%c", strptime("%y%m%d %T", "970427 11:53:55"))
Sun Apr 27 11:53:55 1997
echo strftime("%c", strptime("%Y%m%d%H%M%S", "19970427115355") + 3600)
Sun Apr 27 12:53:55 1997
strridx({haystack}, {needle} [, {start}]) strridx() The result is a Number, which gives the byte index in {haystack} of the last occurrence of the String {needle}. When {start} is specified, matches beyond this index are ignored. This can be used to find a match before a previous match:
let lastcomma = strridx(line, ",")
let comma2 = strridx(line, ",", lastcomma - 1)
The search is done case-sensitive. For pattern searches use match(). -1 is returned if the {needle} does not occur in {haystack}. If the {needle} is empty the length of {haystack} is returned. See also stridx(). Examples:
echo strridx("an angry armadillo", "an")             3
strrchr() When used with a single character it works similar to the C function strrchr().
strtrans({string}) strtrans() The result is a String, which is {string} with all unprintable characters translated into printable characters 'isprint'. Like they are shown in a window. Example:
echo strtrans(@a)
This displays a newline in register a as "^@" instead of starting a new line.
Returns an empty string on error.
strutf16len({string} [, {countcc}]) strutf16len() The result is a Number, which is the number of UTF-16 code units in String {string} (after converting it to UTF-16).
When {countcc} is TRUE, composing characters are counted separately. When {countcc} is omitted or FALSE, composing characters are ignored.
Returns zero on error.
Also see strlen() and strcharlen(). Examples:
echo strutf16len('a')                " returns 1
echo strutf16len('Β©')                " returns 1
echo strutf16len('😊')                " returns 2
echo strutf16len('ą́')                " returns 1
echo strutf16len('ą́', v:true)        " returns 3
strwidth({string}) strwidth() The result is a Number, which is the number of display cells String {string} occupies. A Tab character is counted as one cell, alternatively use strdisplaywidth(). When {string} contains characters with East Asian Width Class Ambiguous, this function's return value depends on 'ambiwidth'. Returns zero on error. Also see strlen(), strdisplaywidth() and strchars().
submatch({nr} [, {list}]) submatch() E935 Only for an expression in a :substitute command or substitute() function. Returns the {nr}th submatch of the matched text. When {nr} is 0 the whole matched text is returned. Note that a NL in the string can stand for a line break of a multi-line match or a NUL character in the text. Also see sub-replace-expression.
If {list} is present and non-zero then submatch() returns a list of strings, similar to getline() with two arguments. NL characters in the text represent NUL characters in the text. Only returns more than one item for :substitute, inside substitute() this list will always contain one or zero items, since there are no real line breaks.
When substitute() is used recursively only the submatches in the current (deepest) call can be obtained.
Returns an empty string or list on error.
Examples:
s/\d\+/\=submatch(0) + 1/
echo substitute(text, '\d\+', '\=submatch(0) + 1', '')
This finds the first number in the line and adds one to it. A line break is included as a newline character.
substitute({string}, {pat}, {sub}, {flags}) substitute() The result is a String, which is a copy of {string}, in which the first match of {pat} is replaced with {sub}. When {flags} is "g", all matches of {pat} in {string} are replaced. Otherwise {flags} should be "".
This works like the ":substitute" command (without any flags). But the matching with {pat} is always done like the 'magic' option is set and 'cpoptions' is empty (to make scripts portable). 'ignorecase' is still relevant, use /\c or /\C if you want to ignore or match case and ignore 'ignorecase'. 'smartcase' is not used. See string-match for how {pat} is used.
A "~" in {sub} is not replaced with the previous {sub}. Note that some codes in {sub} have a special meaning sub-replace-special. For example, to replace something with "\n" (two characters), use "\\\\n" or '\\n'.
When {pat} does not match in {string}, {string} is returned unmodified.
Example:
let &path = substitute(&path, ",\\=[^,]*$", "", "")
This removes the last component of the 'path' option.
echo substitute("testing", ".*", "\\U\\0", "")
results in "TESTING".
When {sub} starts with "\=", the remainder is interpreted as an expression. See sub-replace-expression. Example:
echo substitute(s, '%\(\x\x\)',
   \ '\=nr2char("0x" .. submatch(1))', 'g')
When {sub} is a Funcref that function is called, with one optional argument. Example:
echo substitute(s, '%\(\x\x\)', SubNr, 'g')
The optional argument is a list which contains the whole matched string and up to nine submatches, like what submatch() returns. Example:
echo substitute(s, '%\(\x\x\)', {m -> '0x' .. m[1]}, 'g')
Returns an empty string on error.
swapfilelist() swapfilelist() Returns a list of swap file names, like what "vim -r" shows. See the -r command argument. The 'directory' option is used for the directories to inspect. If you only want to get a list of swap files in the current directory then temporarily set 'directory' to a dot:
let save_dir = &directory
let &directory = '.'
let swapfiles = swapfilelist()
let &directory = save_dir
swapinfo({fname}) swapinfo() The result is a dictionary, which holds information about the swapfile {fname}. The available fields are: version Vim version user user name host host name fname original file name pid PID of the Nvim process that created the swap file, or zero if not running. mtime last modification time in seconds inode Optional: INODE number of the file dirty 1 if file was modified, 0 if not In case of failure an "error" item is added with the reason: Cannot open file: file not found or in accessible Cannot read file: cannot read first block Not a swap file: does not contain correct block ID Magic number mismatch: Info in first block is invalid
swapname({buf}) swapname() The result is the swap file path of the buffer {buf}. For the use of {buf}, see bufname() above. If buffer {buf} is the current buffer, the result is equal to :swapname (unless there is no swap file). If buffer {buf} has no swap file, returns an empty string.
synID({lnum}, {col}, {trans}) synID() The result is a Number, which is the syntax ID at the position {lnum} and {col} in the current window. The syntax ID can be used with synIDattr() and synIDtrans() to obtain syntax information about text.
{col} is 1 for the leftmost column, {lnum} is 1 for the first line. 'synmaxcol' applies, in a longer line zero is returned. Note that when the position is after the last character, that's where the cursor can be in Insert mode, synID() returns zero. {lnum} is used like with getline().
When {trans} is TRUE, transparent items are reduced to the item that they reveal. This is useful when wanting to know the effective color. When {trans} is FALSE, the transparent item is returned. This is useful when wanting to know which syntax item is effective (e.g. inside parens). Warning: This function can be very slow. Best speed is obtained by going through the file in forward direction.
Returns zero on error.
Example (echoes the name of the syntax item under the cursor):
echo synIDattr(synID(line("."), col("."), 1), "name")
synIDattr({synID}, {what} [, {mode}]) synIDattr() The result is a String, which is the {what} attribute of syntax ID {synID}. This can be used to obtain information about a syntax item. {mode} can be "gui" or "cterm", to get the attributes for that mode. When {mode} is omitted, or an invalid value is used, the attributes for the currently active highlighting are used (GUI or cterm). Use synIDtrans() to follow linked highlight groups. {what} result "name" the name of the syntax item "fg" foreground color (GUI: color name used to set the color, cterm: color number as a string, term: empty string) "bg" background color (as with "fg") "font" font name (only available in the GUI) highlight-font "sp" special color (as with "fg") guisp "fg#" like "fg", but for the GUI and the GUI is running the name in "#RRGGBB" form "bg#" like "fg#" for "bg" "sp#" like "fg#" for "sp" "bold" "1" if bold "italic" "1" if italic "reverse" "1" if reverse "inverse" "1" if inverse (= reverse) "standout" "1" if standout "underline" "1" if underlined "undercurl" "1" if undercurled "underdouble" "1" if double underlined "underdotted" "1" if dotted underlined "underdashed" "1" if dashed underlined "strikethrough" "1" if struckthrough "altfont" "1" if alternative font "nocombine" "1" if nocombine
Returns an empty string on error.
Example (echoes the color of the syntax item under the cursor):
echo synIDattr(synIDtrans(synID(line("."), col("."), 1)), "fg")
Can also be used as a method:
echo synID(line("."), col("."), 1)->synIDtrans()->synIDattr("fg")
synIDtrans({synID}) synIDtrans() The result is a Number, which is the translated syntax ID of {synID}. This is the syntax group ID of what is being used to highlight the character. Highlight links given with ":highlight link" are followed.
Returns zero on error.
synconcealed({lnum}, {col}) synconcealed() The result is a List with currently three items: 1. The first item in the list is 0 if the character at the position {lnum} and {col} is not part of a concealable region, 1 if it is. {lnum} is used like with getline(). 2. The second item in the list is a string. If the first item is 1, the second item contains the text which will be displayed in place of the concealed text, depending on the current setting of 'conceallevel' and 'listchars'. 3. The third and final item in the list is a number representing the specific syntax region matched in the line. When the character is not concealed the value is zero. This allows detection of the beginning of a new concealable region if there are two consecutive regions with the same replacement character. For an example, if the text is "123456" and both "23" and "45" are concealed and replaced by the character "X", then:
call returns
synconcealed(lnum, 1) [0, '', 0] synconcealed(lnum, 2) [1, 'X', 1] synconcealed(lnum, 3) [1, 'X', 1] synconcealed(lnum, 4) [1, 'X', 2] synconcealed(lnum, 5) [1, 'X', 2] synconcealed(lnum, 6) [0, '', 0]
synstack({lnum}, {col}) synstack() Return a List, which is the stack of syntax items at the position {lnum} and {col} in the current window. {lnum} is used like with getline(). Each item in the List is an ID like what synID() returns. The first item in the List is the outer region, following are items contained in that one. The last one is what synID() returns, unless not the whole item is highlighted or it is a transparent item. This function is useful for debugging a syntax file. Example that shows the syntax stack under the cursor:
for id in synstack(line("."), col("."))
   echo synIDattr(id, "name")
endfor
When the position specified with {lnum} and {col} is invalid an empty list is returned. The position just after the last character in a line and the first column in an empty line are valid positions.
system({cmd} [, {input}]) system() E677 Note: Prefer vim.system() in Lua.
Gets the output of {cmd} as a string (systemlist() returns a List) and sets v:shell_error to the error code. {cmd} is treated as in jobstart(): If {cmd} is a List it runs directly (no 'shell'). If {cmd} is a String it runs in the 'shell', like this:
call jobstart(split(&shell) + split(&shellcmdflag) + ['{cmd}'])
Not to be used for interactive commands.
Result is a String, filtered to avoid platform-specific quirks:
<CR><NL> is replaced with <NL>
NUL characters are replaced with SOH (0x01)
Example:
echo system(['ls', expand('%:h')])
If {input} is a string it is written to a pipe and passed as stdin to the command. The string is written as-is, line separators are not changed. If {input} is a List it is written to the pipe as writefile() does with {binary} set to "b" (i.e. with a newline between each list item, and newlines inside list items converted to NULs). When {input} is given and is a valid buffer id, the content of the buffer is written to the file line by line, each line terminated by NL (and NUL where the text has NL). E5677 Note: system() cannot write to or read from backgrounded ("&") shell commands, e.g.:
echo system("cat - &", "foo")
which is equivalent to:
$ echo foo | bash -c 'cat - &'
The pipes are disconnected (unless overridden by shell redirection syntax) before input can reach it. Use jobstart() instead.
Note: Use shellescape() or ::S with expand() or fnamemodify() to escape special characters in a command argument. 'shellquote' and 'shellxquote' must be properly configured. Example:
echo system('ls '..shellescape(expand('%:h')))
echo system('ls '..expand('%:h:S'))
Unlike ":!cmd" there is no automatic check for changed files. Use :checktime to force a check.
systemlist({cmd} [, {input} [, {keepempty}]]) systemlist() Same as system(), but returns a List with lines (parts of output separated by NL) with NULs transformed into NLs. Output is the same as readfile() will output with {binary} argument set to "b", except that a final newline is not preserved, unless {keepempty} is non-zero. Note that on MS-Windows you may get trailing CR characters.
To see the difference between "echo hello" and "echo -n hello" use system() and split():
echo split(system('echo hello'), '\n', 1)
Returns an empty string on error.
tabpagebuflist([{arg}]) tabpagebuflist() The result is a List, where each item is the number of the buffer associated with each window in the current tab page. {arg} specifies the number of the tab page to be used. When omitted the current tab page is used. When {arg} is invalid the number zero is returned. To get a list of all buffers in all tabs use this:
let buflist = []
for i in range(tabpagenr('$'))
   call extend(buflist, tabpagebuflist(i + 1))
endfor
Note that a buffer may appear in more than one window.
tabpagenr([{arg}]) tabpagenr() The result is a Number, which is the number of the current tab page. The first tab page has number 1.
The optional argument {arg} supports the following values: $ the number of the last tab page (the tab page count). # the number of the last accessed tab page (where g<Tab> goes to). If there is no previous tab page, 0 is returned. The number can be used with the :tab command.
Returns zero on error.
tabpagewinnr({tabarg} [, {arg}]) tabpagewinnr() Like winnr() but for tab page {tabarg}. {tabarg} specifies the number of tab page to be used. {arg} is used like with winnr():
When omitted the current window number is returned. This is the window which will be used when going to this tab page.
When "$" the number of windows is returned.
When "#" the previous window nr is returned. Useful examples:
tabpagewinnr(1)            " current window of tab page 1
tabpagewinnr(4, '$')    " number of windows in tab page 4
When {tabarg} is invalid zero is returned.
tagfiles() tagfiles() Returns a List with the file names used to search for tags for the current buffer. This is the 'tags' option expanded.
taglist({expr} [, {filename}]) taglist() Returns a List of tags matching the regular expression {expr}.
If {filename} is passed it is used to prioritize the results in the same way that :tselect does. See tag-priority. {filename} should be the full path of the file.
Each list item is a dictionary with at least the following entries: name Name of the tag. filename Name of the file where the tag is defined. It is either relative to the current directory or a full path. cmd Ex command used to locate the tag in the file. kind Type of the tag. The value for this entry depends on the language specific kind values. Only available when using a tags file generated by Universal/Exuberant ctags or hdrtag. static A file specific tag. Refer to static-tag for more information. More entries may be present, depending on the content of the tags file: access, implementation, inherits and signature. Refer to the ctags documentation for information about these fields. For C code the fields "struct", "class" and "enum" may appear, they give the name of the entity the tag is contained in.
The ex-command "cmd" can be either an ex search pattern, a line number or a line number followed by a byte number.
If there are no matching tags, then an empty list is returned.
To get an exact tag match, the anchors '^' and '$' should be used in {expr}. This also make the function work faster. Refer to tag-regexp for more information about the tag search regular expression pattern.
Refer to 'tags' for information about how the tags file is located by Vim. Refer to tags-file-format for the format of the tags file generated by the different ctags tools.
tan({expr}) tan() Return the tangent of {expr}, measured in radians, as a Float in the range [-inf, inf]. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo tan(10)
0.648361
echo tan(-4.01)
-1.181502
tanh({expr}) tanh() Return the hyperbolic tangent of {expr} as a Float in the range [-1, 1]. {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo tanh(0.5)
0.462117
echo tanh(-1)
-0.761594
tempname() tempname() Generates a (non-existent) filename located in the Nvim root tempdir. Scripts can use the filename as a temporary file. Example:
let tmpfile = tempname()
exe "redir > " .. tmpfile
termopen({cmd} [, {opts}]) termopen() Spawns {cmd} in a new pseudo-terminal session connected to the current (unmodified) buffer. Parameters and behavior are the same as jobstart() except "pty", "width", "height", and "TERM" are ignored: "height" and "width" are taken from the current window. Note that termopen() implies a "pty" arg to jobstart(), and thus has the implications documented at jobstart().
Returns the same values as jobstart().
Terminal environment is initialized as in jobstart-env, except $TERM is set to "xterm-256color". Full behavior is described in terminal.
test_garbagecollect_now() test_garbagecollect_now() Like garbagecollect(), but executed right away. This must only be called directly to avoid any structure to exist internally, and v:testing must have been set before calling any function.
timer_info([{id}]) timer_info() Return a list with information about timers. When {id} is given only information about this timer is returned. When timer {id} does not exist an empty list is returned. When {id} is omitted information about all timers is returned.
For each timer the information is stored in a Dictionary with these items: "id" the timer ID "time" time the timer was started with "repeat" number of times the timer will still fire; -1 means forever "callback" the callback
timer_pause({timer}, {paused}) timer_pause() Pause or unpause a timer. A paused timer does not invoke its callback when its time expires. Unpausing a timer may cause the callback to be invoked almost immediately if enough time has passed.
Pausing a timer is useful to avoid the callback to be called for a short time.
If {paused} evaluates to a non-zero Number or a non-empty String, then the timer is paused, otherwise it is unpaused. See non-zero-arg.
timer_start({time}, {callback} [, {options}]) timer_start() timer Create a timer and return the timer ID.
{time} is the waiting time in milliseconds. This is the minimum time before invoking the callback. When the system is busy or Vim is not waiting for input the time will be longer. Zero can be used to execute the callback when Vim is back in the main loop.
{callback} is the function to call. It can be the name of a function or a Funcref. It is called with one argument, which is the timer ID. The callback is only invoked when Vim is waiting for input.
{options} is a dictionary. Supported entries: "repeat" Number of times to repeat the callback. -1 means forever. Default is 1. If the timer causes an error three times in a row the repeat is cancelled.
Returns -1 on error.
Example:
func MyHandler(timer)
  echo 'Handler called'
endfunc
let timer = timer_start(500, 'MyHandler',
        \ {'repeat': 3})
This invokes MyHandler() three times at 500 msec intervals.
timer_stop({timer}) timer_stop() Stop a timer. The timer callback will no longer be invoked. {timer} is an ID returned by timer_start(), thus it must be a Number. If {timer} does not exist there is no error.
timer_stopall() timer_stopall() Stop all timers. The timer callbacks will no longer be invoked. Useful if some timers is misbehaving. If there are no timers there is no error.
tolower({expr}) tolower() The result is a copy of the String given, with all uppercase characters turned into lowercase (just like applying gu to the string). Returns an empty string on error.
toupper({expr}) toupper() The result is a copy of the String given, with all lowercase characters turned into uppercase (just like applying gU to the string). Returns an empty string on error.
tr({src}, {fromstr}, {tostr}) tr() The result is a copy of the {src} string with all characters which appear in {fromstr} replaced by the character in that position in the {tostr} string. Thus the first character in {fromstr} is translated into the first character in {tostr} and so on. Exactly like the unix "tr" command. This code also deals with multibyte characters properly.
Returns an empty string on error.
Examples:
echo tr("hello there", "ht", "HT")
returns "Hello THere"
echo tr("<blob>", "<>", "{}")
returns "{blob}"
trim({text} [, {mask} [, {dir}]]) trim() Return {text} as a String where any character in {mask} is removed from the beginning and/or end of {text}.
If {mask} is not given, or is an empty string, {mask} is all characters up to 0x20, which includes Tab, space, NL and CR, plus the non-breaking space character 0xa0.
The optional {dir} argument specifies where to remove the characters: 0 remove from the beginning and end of {text} 1 remove only at the beginning of {text} 2 remove only at the end of {text} When omitted both ends are trimmed.
This function deals with multibyte characters properly. Returns an empty string on error.
Examples:
echo trim("   some text ")
returns "some text"
echo trim("  \r\t\t\r RESERVE \t\n\x0B\xA0") .. "_TAIL"
returns "RESERVE_TAIL"
echo trim("rm<Xrm<>X>rrm", "rm<>")
returns "Xrm<>X" (characters in the middle are not removed)
echo trim("  vim  ", " ", 2)
returns " vim"
trunc({expr}) trunc() Return the largest integral value with magnitude less than or equal to {expr} as a Float (truncate towards zero). {expr} must evaluate to a Float or a Number. Returns 0.0 if {expr} is not a Float or a Number. Examples:
echo trunc(1.456)
1.0
echo trunc(-5.456)
-5.0
echo trunc(4.0)
4.0
type({expr}) type() The result is a Number representing the type of {expr}. Instead of using the number directly, it is better to use the v:t_ variable that has the value: Number: 0 v:t_number String: 1 v:t_string Funcref: 2 v:t_func List: 3 v:t_list Dictionary: 4 v:t_dict Float: 5 v:t_float Boolean: 6 v:t_bool (v:false and v:true) Null: 7 (v:null) Blob: 10 v:t_blob For backward compatibility, this method can be used:
if type(myvar) == type(0) | endif
if type(myvar) == type("") | endif
if type(myvar) == type(function("tr")) | endif
if type(myvar) == type([]) | endif
if type(myvar) == type({}) | endif
if type(myvar) == type(0.0) | endif
if type(myvar) == type(v:true) | endif
In place of checking for v:null type it is better to check for v:null directly as it is the only value of this type:
if myvar is v:null | endif
To check if the v:t_ variables exist use this:
if exists('v:t_number') | endif
undofile({name}) undofile() Return the name of the undo file that would be used for a file with name {name} when writing. This uses the 'undodir' option, finding directories that exist. It does not check if the undo file exists. {name} is always expanded to the full path, since that is what is used internally. If {name} is empty undofile() returns an empty string, since a buffer without a file name will not write an undo file. Useful in combination with :wundo and :rundo.
undotree([{buf}]) undotree() Return the current state of the undo tree for the current buffer, or for a specific buffer if {buf} is given. The result is a dictionary with the following items: "seq_last" The highest undo sequence number used. "seq_cur" The sequence number of the current position in the undo tree. This differs from "seq_last" when some changes were undone. "time_cur" Time last used for :earlier and related commands. Use strftime() to convert to something readable. "save_last" Number of the last file write. Zero when no write yet. "save_cur" Number of the current position in the undo tree. "synced" Non-zero when the last undo block was synced. This happens when waiting from input from the user. See undo-blocks. "entries" A list of dictionaries with information about undo blocks.
The first item in the "entries" list is the oldest undo item. Each List item is a Dictionary with these items: "seq" Undo sequence number. Same as what appears in :undolist. "time" Timestamp when the change happened. Use strftime() to convert to something readable. "newhead" Only appears in the item that is the last one that was added. This marks the last change and where further changes will be added. "curhead" Only appears in the item that is the last one that was undone. This marks the current position in the undo tree, the block that will be used by a redo command. When nothing was undone after the last change this item will not appear anywhere. "save" Only appears on the last block before a file write. The number is the write count. The first write has number 1, the last one the "save_last" mentioned above. "alt" Alternate entry. This is again a List of undo blocks. Each item may again have an "alt" item.
uniq({list} [, {func} [, {dict}]]) uniq() E882 Remove second and succeeding copies of repeated adjacent {list} items in-place. Returns {list}. If you want a list to remain unmodified make a copy first:
let newlist = uniq(copy(mylist))
The default compare function uses the string representation of each item. For the use of {func} and {dict} see sort().
Returns zero if {list} is not a List.
utf16idx({string}, {idx} [, {countcc} [, {charidx}]]) utf16idx() Same as charidx() but returns the UTF-16 code unit index of the byte at {idx} in {string} (after converting it to UTF-16).
When {charidx} is present and TRUE, {idx} is used as the character index in the String {string} instead of as the byte index. An {idx} in the middle of a UTF-8 sequence is rounded downwards to the beginning of that sequence.
Returns -1 if the arguments are invalid or if there are less than {idx} bytes in {string}. If there are exactly {idx} bytes the length of the string in UTF-16 code units is returned.
See byteidx() and byteidxcomp() for getting the byte index from the UTF-16 index and charidx() for getting the character index from the UTF-16 index. Refer to string-offset-encoding for more information. Examples:
echo utf16idx('a😊😊', 3)        " returns 2
echo utf16idx('a😊😊', 7)        " returns 4
echo utf16idx('a😊😊', 1, 0, 1)        " returns 2
echo utf16idx('a😊😊', 2, 0, 1)        " returns 4
echo utf16idx('aą́c', 6)                " returns 2
echo utf16idx('aą́c', 6, 1)        " returns 4
echo utf16idx('a😊😊', 9)        " returns -1
values({dict}) values() Return a List with all the values of {dict}. The List is in arbitrary order. Also see items() and keys(). Returns zero if {dict} is not a Dict.
virtcol({expr} [, {list} [, {winid}]]) virtcol() The result is a Number, which is the screen column of the file position given with {expr}. That is, the last screen position occupied by the character at that position, when the screen would be of unlimited width. When there is a <Tab> at the position, the returned Number will be the column at the end of the <Tab>. For example, for a <Tab> in column 1, with 'ts' set to 8, it returns 8. conceal is ignored. For the byte position use col().
For the use of {expr} see col().
When 'virtualedit' is used {expr} can be [lnum, col, off], where "off" is the offset in screen columns from the start of the character. E.g., a position within a <Tab> or after the last character. When "off" is omitted zero is used. When Virtual editing is active in the current mode, a position beyond the end of the line can be returned. Also see 'virtualedit'
The accepted positions are: . the cursor position $ the end of the cursor line (the result is the number of displayed characters in the cursor line plus one) 'x position of mark x (if the mark is not set, 0 is returned) v In Visual mode: the start of the Visual area (the cursor is the end). When not in Visual mode returns the cursor position. Differs from '< in that it's updated right away.
If {list} is present and non-zero then virtcol() returns a List with the first and last screen position occupied by the character.
With the optional {winid} argument the values are obtained for that window instead of the current window.
Note that only marks in the current file can be used. Examples:
" With text "foo^Lbar" and cursor on the "^L":
echo virtcol(".")        " returns 5
echo virtcol(".", 1)        " returns [4, 5]
echo virtcol("$")        " returns 9
" With text "          there", with 't at 'h':
echo virtcol("'t")        " returns 6
The first column is 1. 0 or [0, 0] is returned for an error. A more advanced example that echoes the maximum length of all lines:
echo max(map(range(1, line('$')), "virtcol([v:val, '$'])"))
virtcol2col({winid}, {lnum}, {col}) virtcol2col() The result is a Number, which is the byte index of the character in window {winid} at buffer line {lnum} and virtual column {col}.
If buffer line {lnum} is an empty line, 0 is returned.
If {col} is greater than the last virtual column in line {lnum}, then the byte index of the character at the last virtual column is returned.
For a multi-byte character, the column number of the first byte in the character is returned.
The {winid} argument can be the window number or the window-ID. If this is zero, then the current window is used.
Returns -1 if the window {winid} doesn't exist or the buffer line {lnum} or virtual column {col} is invalid.
See also screenpos(), virtcol() and col().
visualmode([{expr}]) visualmode() The result is a String, which describes the last Visual mode used in the current buffer. Initially it returns an empty string, but once Visual mode has been used, it returns "v", "V", or "<CTRL-V>" (a single CTRL-V character) for character-wise, line-wise, or block-wise Visual mode respectively. Example:
exe "normal " .. visualmode()
This enters the same Visual mode as before. It is also useful in scripts if you wish to act differently depending on the Visual mode that was used. If Visual mode is active, use mode() to get the Visual mode (e.g., in a :vmap). If {expr} is supplied and it evaluates to a non-zero Number or a non-empty String, then the Visual mode will be cleared and the old value is returned. See non-zero-arg.
wait({timeout}, {condition} [, {interval}]) wait() Waits until {condition} evaluates to TRUE, where {condition} is a Funcref or string containing an expression.
{timeout} is the maximum waiting time in milliseconds, -1 means forever.
Condition is evaluated on user events, internal events, and every {interval} milliseconds (default: 200).
Returns a status integer: 0 if the condition was satisfied before timeout -1 if the timeout was exceeded -2 if the function was interrupted (by CTRL-C) -3 if an error occurred
wildmenumode() wildmenumode() Returns TRUE when the wildmenu is active and FALSE otherwise. See 'wildmenu' and 'wildmode'. This can be used in mappings to handle the 'wildcharm' option gracefully. (Makes only sense with mapmode-c mappings).
For example to make <c-j> work like <down> in wildmode, use:
cnoremap <expr> <C-j> wildmenumode() ? "\<Down>\<Tab>" : "\<c-j>"
(Note, this needs the 'wildcharm' option set appropriately).
win_execute({id}, {command} [, {silent}]) win_execute() Like execute() but in the context of window {id}. The window will temporarily be made the current window, without triggering autocommands or changing directory. When executing {command} autocommands will be triggered, this may have unexpected side effects. Use :noautocmd if needed. Example:
call win_execute(winid, 'syntax enable')
Doing the same with setwinvar() would not trigger autocommands and not actually show syntax highlighting.
When window {id} does not exist then no error is given and an empty string is returned.
win_findbuf({bufnr}) win_findbuf() Returns a List with window-IDs for windows that contain buffer {bufnr}. When there is none the list is empty.
win_getid([{win} [, {tab}]]) win_getid() Get the window-ID for the specified window. When {win} is missing use the current window. With {win} this is the window number. The top window has number 1. Without {tab} use the current tab, otherwise the tab with number {tab}. The first tab has number one. Return zero if the window cannot be found.
win_gettype([{nr}]) win_gettype() Return the type of the window: "autocmd" autocommand window. Temporary window used to execute autocommands. "command" command-line window cmdwin (empty) normal window "loclist" location-list-window "popup" floating window api-floatwin "preview" preview window preview-window "quickfix" quickfix-window "unknown" window {nr} not found
When {nr} is omitted return the type of the current window. When {nr} is given return the type of this window by number or window-ID.
Also see the 'buftype' option.
win_gotoid({expr}) win_gotoid() Go to window with ID {expr}. This may also change the current tabpage. Return TRUE if successful, FALSE if the window cannot be found.
win_id2tabwin({expr}) win_id2tabwin() Return a list with the tab number and window number of window with ID {expr}: [tabnr, winnr]. Return [0, 0] if the window cannot be found.
win_id2win({expr}) win_id2win() Return the window number of window with ID {expr}. Return 0 if the window cannot be found in the current tabpage.
win_move_separator({nr}, {offset}) win_move_separator() Move window {nr}'s vertical separator (i.e., the right border) by {offset} columns, as if being dragged by the mouse. {nr} can be a window number or window-ID. A positive {offset} moves right and a negative {offset} moves left. Moving a window's vertical separator will change the width of the window and the width of other windows adjacent to the vertical separator. The magnitude of movement may be smaller than specified (e.g., as a consequence of maintaining 'winminwidth'). Returns TRUE if the window can be found and FALSE otherwise. This will fail for the rightmost window and a full-width window, since it has no separator on the right. Only works for the current tab page. E1308
win_move_statusline({nr}, {offset}) win_move_statusline() Move window {nr}'s status line (i.e., the bottom border) by {offset} rows, as if being dragged by the mouse. {nr} can be a window number or window-ID. A positive {offset} moves down and a negative {offset} moves up. Moving a window's status line will change the height of the window and the height of other windows adjacent to the status line. The magnitude of movement may be smaller than specified (e.g., as a consequence of maintaining 'winminheight'). Returns TRUE if the window can be found and FALSE otherwise. Only works for the current tab page.
win_screenpos({nr}) win_screenpos() Return the screen position of window {nr} as a list with two numbers: [row, col]. The first window always has position [1, 1], unless there is a tabline, then it is [2, 1]. {nr} can be the window number or the window-ID. Use zero for the current window. Returns [0, 0] if the window cannot be found.
win_splitmove({nr}, {target} [, {options}]) win_splitmove() Temporarily switch to window {target}, then move window {nr} to a new split adjacent to {target}. Unlike commands such as :split, no new windows are created (the window-ID of window {nr} is unchanged after the move).
Both {nr} and {target} can be window numbers or window-IDs. Both must be in the current tab page.
Returns zero for success, non-zero for failure.
{options} is a Dictionary with the following optional entries: "vertical" When TRUE, the split is created vertically, like with :vsplit. "rightbelow" When TRUE, the split is made below or to the right (if vertical). When FALSE, it is done above or to the left (if vertical). When not present, the values of 'splitbelow' and 'splitright' are used.
winbufnr({nr}) winbufnr() The result is a Number, which is the number of the buffer associated with window {nr}. {nr} can be the window number or the window-ID. When {nr} is zero, the number of the buffer in the current window is returned. When window {nr} doesn't exist, -1 is returned. Example:
echo "The file in the current window is " .. bufname(winbufnr(0))
wincol() wincol() The result is a Number, which is the virtual column of the cursor in the window. This is counting screen cells from the left side of the window. The leftmost column is one.
windowsversion() windowsversion() The result is a String. For MS-Windows it indicates the OS version. E.g, Windows 10 is "10.0", Windows 8 is "6.2", Windows XP is "5.1". For non-MS-Windows systems the result is an empty string.
winheight({nr}) winheight() The result is a Number, which is the height of window {nr}. {nr} can be the window number or the window-ID. When {nr} is zero, the height of the current window is returned. When window {nr} doesn't exist, -1 is returned. An existing window always has a height of zero or more. This excludes any window toolbar line. Examples:
echo "The current window has " .. winheight(0) .. " lines."
winlayout([{tabnr}]) winlayout() The result is a nested List containing the layout of windows in a tabpage.
Without {tabnr} use the current tabpage, otherwise the tabpage with number {tabnr}. If the tabpage {tabnr} is not found, returns an empty list.
For a leaf window, it returns:
["leaf", {winid}]
For horizontally split windows, which form a column, it returns:
["col", [{nested list of windows}]]
For vertically split windows, which form a row, it returns:
["row", [{nested list of windows}]]
Example:
" Only one window in the tab page
echo winlayout()
['leaf', 1000]
" Two horizontally split windows
echo winlayout()
['col', [['leaf', 1000], ['leaf', 1001]]]
" The second tab page, with three horizontally split
" windows, with two vertically split windows in the
" middle window
echo winlayout(2)
['col', [['leaf', 1002], ['row', [['leaf', 1003],
                    ['leaf', 1001]]], ['leaf', 1000]]]
winline() winline() The result is a Number, which is the screen line of the cursor in the window. This is counting screen lines from the top of the window. The first line is one. If the cursor was moved the view on the file will be updated first, this may cause a scroll.
winnr([{arg}]) winnr() The result is a Number, which is the number of the current window. The top window has number 1. Returns zero for a popup window.
The optional argument {arg} supports the following values: $ the number of the last window (the window count). # the number of the last accessed window (where CTRL-W_p goes to). If there is no previous window or it is in another tab page 0 is returned. May refer to the current window in some cases (e.g. when evaluating 'statusline' expressions). {N}j the number of the Nth window below the current window (where CTRL-W_j goes to). {N}k the number of the Nth window above the current window (where CTRL-W_k goes to). {N}h the number of the Nth window left of the current window (where CTRL-W_h goes to). {N}l the number of the Nth window right of the current window (where CTRL-W_l goes to). The number can be used with CTRL-W_w and ":wincmd w" :wincmd. When {arg} is invalid an error is given and zero is returned. Also see tabpagewinnr() and win_getid(). Examples:
let window_count = winnr('$')
let prev_window = winnr('#')
let wnum = winnr('3k')
winrestcmd() winrestcmd() Returns a sequence of :resize commands that should restore the current window sizes. Only works properly when no windows are opened or closed and the current window and tab page is unchanged. Example:
let cmd = winrestcmd()
call MessWithWindowSizes()
exe cmd
winrestview({dict}) winrestview() Uses the Dictionary returned by winsaveview() to restore the view of the current window. Note: The {dict} does not have to contain all values, that are returned by winsaveview(). If values are missing, those settings won't be restored. So you can use:
call winrestview({'curswant': 4})
This will only set the curswant value (the column the cursor wants to move on vertical movements) of the cursor to column 5 (yes, that is 5), while all other settings will remain the same. This is useful, if you set the cursor position manually.
If you have changed the values the result is unpredictable. If the window size changed the result won't be the same.
winsaveview() winsaveview() Returns a Dictionary that contains information to restore the view of the current window. Use winrestview() to restore the view. This is useful if you have a mapping that jumps around in the buffer and you want to go back to the original view. This does not save fold information. Use the 'foldenable' option to temporarily switch off folding, so that folds are not opened when moving around. This may have side effects. The return value includes: lnum cursor line number col cursor column (Note: the first column zero, as opposed to what getcurpos() returns) coladd cursor column offset for 'virtualedit' curswant column for vertical movement (Note: the first column is zero, as opposed to what getcurpos() returns). After $ command it will be a very large number equal to v:maxcol. topline first line in the window topfill filler lines, only in diff mode leftcol first column displayed; only used when 'wrap' is off skipcol columns skipped Note that no option values are saved.
winwidth({nr}) winwidth() The result is a Number, which is the width of window {nr}. {nr} can be the window number or the window-ID. When {nr} is zero, the width of the current window is returned. When window {nr} doesn't exist, -1 is returned. An existing window always has a width of zero or more. Examples:
echo "The current window has " .. winwidth(0) .. " columns."
if winwidth(0) <= 50
  50 wincmd |
endif
For getting the terminal or screen size, see the 'columns' option.
wordcount() wordcount() The result is a dictionary of byte/chars/word statistics for the current buffer. This is the same info as provided by g_CTRL-G The return value includes: bytes Number of bytes in the buffer chars Number of chars in the buffer words Number of words in the buffer cursor_bytes Number of bytes before cursor position (not in Visual mode) cursor_chars Number of chars before cursor position (not in Visual mode) cursor_words Number of words before cursor position (not in Visual mode) visual_bytes Number of bytes visually selected (only in Visual mode) visual_chars Number of chars visually selected (only in Visual mode) visual_words Number of words visually selected (only in Visual mode)
writefile({object}, {fname} [, {flags}]) writefile() When {object} is a List write it to file {fname}. Each list item is separated with a NL. Each list item must be a String or Number. All NL characters are replaced with a NUL character. Inserting CR characters needs to be done before passing {list} to writefile().
When {object} is a Blob write the bytes to file {fname} unmodified, also when binary mode is not specified.
{flags} must be a String. These characters are recognized:
'b' Binary mode is used: There will not be a NL after the last list item. An empty item at the end does cause the last line in the file to end in a NL.
'a' Append mode is used, lines are appended to the file:
call writefile(["foo"], "event.log", "a")
call writefile(["bar"], "event.log", "a")
'D' Delete the file when the current function ends. This works like:
defer delete({fname})
Fails when not in a function. Also see :defer.
's' fsync() is called after writing the file. This flushes the file to disk, if possible. This takes more time but avoids losing the file if the system crashes.
'S' fsync() is not called, even when 'fsync' is set.
When {flags} does not contain "S" or "s" then fsync() is called if the 'fsync' option is set.
An existing file is overwritten, if possible.
When the write fails -1 is returned, otherwise 0. There is an error message if the file can't be created or when writing fails.
Also see readfile(). To copy a file byte for byte:
let fl = readfile("foo", "b")
call writefile(fl, "foocopy", "b")
xor({expr}, {expr}) xor() Bitwise XOR on the two arguments. The arguments are converted to a number. A List, Dict or Float argument causes an error. Also see and() and or(). Example:
let bits = xor(bits, 0x80)

2. Matching a pattern in a String string-match

This is common between several functions. A regexp pattern as explained at pattern is normally used to find a match in the buffer lines. When a pattern is used to find a match in a String, almost everything works in the same way. The difference is that a String is handled like it is one line. When it contains a "\n" character, this is not seen as a line break for the pattern. It can be matched with a "\n" in the pattern, or with ".". Example:
let a = "aaaa\nxxxx"
echo matchstr(a, "..\n..")
" aa
" xx
echo matchstr(a, "a.x")
" a
" x
Don't forget that "^" will only match at the first character of the String and "$" at the last character of the string. They don't match after or before a "\n".
vim:tw=78:ts=8:noet:ft=help:norl:
Main
Commands index
Quick reference