Author: Karl Yngve Lervåg <karl.yngve@gmail.com>
License: MIT license {{{
Copyright (c) 2021 Karl Yngve Lervåg
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the "Software"), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.
The software is provided "as is", without warranty of any kind, express or implied, including but not limited to the warranties of merchantability, fitness for a particular purpose and noninfringement. In no event shall the authors or copyright holders be liable for any claim, damages or other liability, whether in an action of contract, tort or otherwise, arising from, out of or in connection with the software or the use or other dealings in the software.
}}}
CONTENTS
Introduction vimtex-introduction
Comment on internal tex plugin vimtex-comment-internal
Feature overview vimtex-features
Requirements vimtex-requirements
Support for multi-file projects vimtex-multi-file
Support for TeX specifiers vimtex-tex-directives
Package detection vimtex-package-detection
Integration with other plugins vimtex-and-friends
Usage vimtex-usage
Default mappings vimtex-default-mappings
Options vimtex-options
Commands vimtex-commands
Map definitions vimtex-mappings
Insert mode mappings vimtex-imaps
Events vimtex-events
Text objects vimtex-text-objects
Completion vimtex-completion
Complete citations vimtex-complete-cites
Complete labels vimtex-complete-labels
Complete commands vimtex-complete-commands
Complete environments vimtex-complete-environments
Complete file names vimtex-complete-filenames
Complete glossary entries vimtex-complete-glossary
Complete packages vimtex-complete-packages
Complete documentclasses vimtex-complete-classes
Complete bibliographystyles vimtex-complete-bibstyle
Complete beamer themes vimtex-complete-beamer
Autocomplete vimtex-complete-auto
coc.nvim vimtex-complete-coc.nvim
deoplete vimtex-complete-deoplete
Neocomplete vimtex-complete-neocomplete
ncm2 vimtex-complete-ncm2
nvim-completion-manager vimtex-complete-ncm
YouCompleteMe vimtex-complete-youcompleteme
VimCompletesMe vimtex-complete-vcm
nvim-cmp vimtex-complete-nvim-cmp
nvim-compe vimtex-complete-nvim-compe
MUcomplete vimtex-complete-mucomplete
Folding vimtex-folding
Indentation vimtex-indent
Syntax highlighting vimtex-syntax
Syntax core specification vimtex-syntax-core
Syntax package specification vimtex-syntax-packages
Syntax conceal vimtex-syntax-conceal
Syntax group reference vimtex-syntax-reference
Navigation vimtex-navigation
Include expression (gf command) vimtex-includeexpr
Table of contents vimtex-toc
Custom mappings vimtex-toc-custom-maps
Denite/Unite source vimtex-denite / vimtex-unite
fzf.vim integration vimtex-fzf
fzf-lua integration vimtex-fzf-lua
Snacks integration vimtex-snacks
Compilation vimtex-compiler
Latexmk vimtex-compiler-latexmk
Latexrun vimtex-compiler-latexrun
Tectonic vimtex-compiler-tectonic
Arara vimtex-compiler-arara
TeXpresso vimtex-compiler-texpresso
Generic vimtex-compiler-generic
Syntax Checking (Linting) vimtex-lint
Grammar Checking vimtex-grammar
textidote vimtex-grammar-textidote
vlty vimtex-grammar-vlty
View vimtex-view
Viewer configuration vimtex-view-configuration
Evince vimtex-view-evince
Galley vimtex-view-galley
MuPDF vimtex-view-mupdf
Okular vimtex-view-okular
qpdfview vimtex-view-qpdfview
Sioyek vimtex-view-sioyek
Skim vimtex-view-skim
SumatraPDF vimtex-view-sumatrapdf
TeXShop vimtex-view-texshop
Zathura vimtex-view-zathura
Zathura (simple) vimtex-view-zathura-simple
Synctex vimtex-synctex
Forward search vimtex-synctex-forward-search
Inverse search vimtex-synctex-inverse-search
LaTeX Documentation vimtex-latexdoc
Context menu vimtex-context-menu
Citation context vimtex-context-citation
Code structure vimtex-code
API vimtex-code-api
FAQ vimtex-faq
Troubleshooting vimtex-troubleshooting
Credits vimtex-credits
Changelog vimtex-changelog
INTRODUCTION
VimTeX provides convenient functionality for editing LaTeX documents. The main goal of VimTeX is to be simple, functional, and to be easy to customize and evolve.
The documentation is understandably too long for a full read through. It is recommended that new users read or skim the entire introduction, as it should give a clear idea of what VimTeX is and is not. The remaining part of the documentation should then be considered a reference for the various parts of the plugin.
COMMENT ON INTERNAL TEX PLUGIN
Vim ships with pretty decent LaTeX support out of the box. In particular, it provides syntax highlighting (ft-tex-syntax), indentation (see the source file $VIMRUNTIME/indent/tex.vim for the documentation), and some sensible options (ft-tex-plugin).
vimtex-tex-flavor
When VimTeX is active, it will override the internal TeX plugin for the filetype
tex (ft-tex-plugin), both for syntax highlighting and for filetype specific features. To prevent the unexpected behaviour where .tex files by default will be recognized as the filetype plaintex (ft-plaintex-syntax) for e.g. empty documents, VimTeX overrides the filetype detection for .tex. The user may prevent this overriding by specifying the g:tex_flavor option something different than 'latex'.FEATURE OVERVIEW
Document compilation with
latexmk, latexrun, tectonic, arara or texpresso
LaTeX log parsing for quickfix entries using
internal method
pplatex
Compilation of selected part of document
Support for several PDF viewers with forward search
MuPDF
Zathura
Okular
qpdfview
SumatraPDF
Other viewers are supported through a general interface
Completion of
citations
labels
commands
file names for figures, input/include, includepdf, includestandalone
glossary entries
package and documentclass names based on available
.sty and .cls files Document navigation through
table of contents
allow include-search and definition-search
give enhanced gf command
Easy access to (online) documentation of packages
Word count (through
texcount)
Motions vimtex-motions
Move between section boundaries with
[[, [], ][, and ]] Move between environment boundaries with
[m, [M, ]m, and ]M
Move between math environment boundaries with
[n, [N, ]n, and ]N Move between frame environment boundaries with
[r, [R, ]r, and ]R
Move between comment boundaries with
[* and ]*
Move between matching delimiters with
%
Text objects
ic ac Commands
id ad Delimiters
ie ae LaTeX environments
i$ a$ Math environments
iP aP Sections
im am Items
Other mappings
Delete the surrounding command, environment or delimiter with
dsc/dse/ds$/dsd
Change the surrounding command, environment or delimiter with
csc/cse/cs$/csd
Toggle between complementary environments with
tse
Toggle starred command or environment with
tsc/tss
Toggle inline and displaymath with
ts$
Toggle between e.g.
() and \left(\right) with tsd/tsD
Toggle (inline) fractions with
tsf
Toggle line-break macro
\\ with tsb
Close the current environment/delimiter in insert mode with
]] Add
\left ... \right) modifiers to surrounding delimiters with <F8>
Insert new command with
<F7>
Convenient insert mode mappings for faster typing of e.g. maths
Context menu on citations (e.g.
\cite{...}) mapped to <cr>
Folding
Indentation
Syntax highlighting
A consistent core syntax specification
General syntax highlighting for several popular LaTeX packages
Nested syntax highlighting for several popular LaTeX packages
Highlight matching delimiters
Support for multi-file project packages
import
subfiles
REQUIREMENTS
The following is a list of specific requirements for running VimTeX and some of its key features. Windows users should also read vimtex-faq-windows.
Vim version
vimtex_version_check
VimTeX requires Vim version 9.2 or neovim version 0.12.4. It will not load for older versions, unless one addslet g:vimtex_version_check = 0to one's
vimrc file. This might work, but issues due to older versions than the mentioned here will be ignored.Vim configuration
VimTeX requires :filetype-plugin-on and optionally :filetype-indent-on.
There are several features in VimTeX that depend on the syntax parsing used for syntax highlighting. Examples include functions like vimtex#syntax#in_mathzone and text objects like <plug>(vimtex-i$). This is important to be aware of especially for neovim users who are interested in Tree-sitter. If you use Tree-sitter for syntax highlighting and thus disable the normal Vim syntax feature, then you will also lose the VimTeX features that depend on the built-in syntax parser. For more info, see vimtex-faq-treesitter.
Some of the VimTeX scripts contain UTF-8 characters, and as such, it is necessary to have the 'encoding' option set to utf-8. This is not necessary in neovim, only in Vim. Add the following to your vimrc file:
autocmd FileType tex set encoding=utf-8
Compiler backend
VimTeX uses
latexmk, latexrun, tectonic, arara or texpresso to compile the LaTeX document.latexmk: http://users.phys.psu.edu/~collins/software/latexmk-jcc
"a perl script for running LaTeX the correct number of times to resolve
cross references, etc; it also runs auxiliary programs (e.g. bibtex). It has a number of other useful capabilities, for example to start a previewer and then run latex whenever the source files are updated, so that the previewer gives an up-to-date view of the document. The script runs on both UNIX and MS-WINDOWS (XP, etc)." [Copied from the latexmk page.] (vimtex-compiler-latexmk)latexrun: https://github.com/aclements/latexrun
Similar to latexmk in that it runs the desired LaTeX engine an
appropriate number of times, including bibtex/biber. However, it differs in philosophy in that it only does the build part. It does not support continuous builds, nor automatic starting of the viewer. However, it does parse the output log in order to provide a more concise list of relevant warnings and error messages (this has currently not been adapted to VimTeX, as of yet). (vimtex-compiler-latexrun)tectonic: https://tectonic-typesetting.github.io/
tectonic is a complete, self-contained TeX/LaTeX engine, powered by XeTeX
and TeXLive. It doesn't support continuous build like latexmk but it presents other worth mentioning features such as automatic support file downloading along with reproducible builds and full Unicode and OpenType fonts support thanks to the power of XeTeX. (vimtex-compiler-tectonic)arara: https://github.com/cereda/arara
arara is a TeX automation tool similar to the above mentioned tools,
but where the compilation behaviour is typically defined in the preamble of the document. (vimtex-compiler-arara)texpresso: https://github.com/let-def/texpresso
texpresso provides live rendering by combining a forked XeTeX engine
with a custom PDF viewer. It communicates incremental buffer changes via stdio, allowing the preview to update incrementally as you type. It supports SyncTeX forward and inverse search and quickfix integration. Unlike the other compilers it only does a single TeX pass, so bibliography and table of contents require a separate build step. (vimtex-compiler-texpresso)Clientserver
vimtex-clientserver
Vim requires |+clientserver| in order to allow inverse search from the PDF viewer to Vim (see vimtex-synctex-inverse-search). The clientserver is used by VimTeX. Thus, if one uses Vim one must ensure that it starts a server. Neovim does not have this requirement.A server will be started automatically if Vim is running on Windows or if it is running in a GUI (gVim). If you use Vim under a terminal in Linux or MacOS, a server will not be started by default. Also, MacVim users should be aware of some differences from regular Vim and should therefore read |macvim-clientserver| carefully.
You can use |remote_startserver()| to start a server from your
vimrc file. The following vimrc configuration snippet will ensure that Vim starts with a server, if possible:if empty(v:servername) && exists('*remote_startserver')
call remote_startserver('VIM')
endifAlternatively, Vim can be started with the command line option
--servername, e.g. vim --servername VIM . The simplest way to ensure this is to add an alias to your .bashrc (or similar), that is, add:alias vim='vim --servername VIM'One can use serverlist() to check whether a server was successfully started, e.g. with
:echo serverlist().Neovim does not implement the same clientserver feature. Instead, it implements the MessagePack-RPC protocol (see RPC). VimTeX relies on this protocol in the same fashion as the clientserver. Both Vim and neovim have the v:servername variable that contains the name/location of the server with which we need to communicate.
SUPPORT FOR MULTI-FILE PROJECTS
VimTeX supports most multi-file documents and has several methods to locate the
main document. Locating this file is very important, because the main file is the one that must be compiled.The default method for locating the main file uses a directory-scan algorithm that searches for a main LaTeX file, see method 6 below. It is expected to work in the vast majority of cases.
There are several alternative methods for specifying the main file that can be more flexible and are relevant for certain work flows and use cases. These methods all require some explicit declaration of the main file and are therefore tried prior to the directory scan.
The complete list of methods in the order of priority is as follows and are then described in more detail:
Buffer variable
TeX root directive
Subfiles package
File
.latexmain specifier
Local
latexmkrc file specifier (from @default_files option) Directory scan
b:vimtex_main
Buffer variable
The main file may be specified through the buffer variable b:vimtex_main. To take effect, it has to be set prior to loading the buffer. If set after the buffer is already loaded, :VimtexReloadState (by default bound to |<localleader>lX|) can be used to make VimTeX aware of its new value. A convenient way to use this feature is to add an BufReadPre autocmd in one's vimrc. An example is warranted:augroup VimTeX
autocmd!
autocmd BufReadPre /path/to/project/*.tex
\ let b:vimtex_main = '/path/to/project/main.tex'
augroup ENDNote: When writing such rules, one should be aware that the
* is not the same as regular globbing because it also includes directory separators. Also, one should use / on every OS. See file-pattern for more info on the autocmd file pattern syntax.Note: Users may be interested in the concept of project specific vim configuration. This is supported in Vim, see 'exrc' and 'secure'. There are also several plugins to help work with project specific settings, such as:
* https://github.com/embear/vim-localvimrc
* https://github.com/tpope/vim-projectionist
* https://github.com/jenterkin/vim-autosource
* https://github.com/ii14/exrc.vim
* https://github.com/MarcWeber/vim-addon-local-vimrc/
* https://github.com/MunifTanjim/exrc.nvim (neovim only)
* See also:
https://superuser.com/questions/598947/setting-vim-options-only-for-files-in-a-certain-directory-tree/598970#598970
vimtex-tex-root
TeX root directive
It is also possible to specify the main TeX file with a comment in one of the first 20 lines of the current file. This is often referred to as a TeX directive, see vimtex-tex-directives for more info. The syntax is best described by some examples:%! TEX root = /path/to/my-main.tex
% ! TeX root = ../*.tex
%!Tex Root=**/main.tex
%! TeX root: ../main.texAs can be seen, the words "tex root" are recognized regardless of casing and the spaces are ignored. Also, both a colon and an equal sign can be used.
VimTeX parses this directive during initialization. Thus, users should be aware that they need to reload (with :VimtexReload) or restart Vim/neovim if they change the TeX root directive.
Note: It is allowed to use a globbing pattern (see wildcards). If there are multiple matches, then VimTeX will ask for input when the buffer is opened.
vimtex-subfiles
vimtex-import
Subfiles package
VimTeX also supports the import [0] and the subfiles [1] packages that can be used to make it easier to work with multi-file projects. If one uses the subfiles package, the :VimtexToggleMain command is particularly useful. Also note the option g:vimtex_subfile_start_local, which can be used to automatically start in the local mode when opening a subfile document.With
subfiles, included files will typically look like this:\documentclass[<main-path>]{subfiles}
\begin{document}
...
\end{document}Here
<main-path> is the path to the main file. It must be specified as relative to the particular subfile. So, given the structure:main.tex sub/sub.tex
The header in
sub.tex should be \documentclass[../main.tex]{subfiles}. Absolute paths like /home/user/main.tex are also allowed and should work as expected.File .latexmain specifier
In some cases, it might be preferable to specify the main file by creating an indicator file. The indicator file should be an empty file, and the name must be the name of the desired main file with .latexmain appended. An example should make this clear:path/file.tex path/file.tex.latexmain path/sections/file1.tex path/sections/file2.tex
Here
path/file.tex.latexmain indicates for file1.tex and file2.tex that path/file.tex is the main LaTeX file.Local latexmkrc file specifier
It is possible to specify to latexmk which files to compile with the @default_files option in the latexmkrc configuration file. VimTeX supports reading this option in any LOCAL latexmkrc or .latexmkrc file.Note:
@default_files is a list of files, VimTeX will use the first entry that is found.Directory scan
If the above methods don't give an appropriate candidate for a main file of the present file, then a search for a suitable main file from the current directory and upwards is started.A candidate
.tex file qualifies as a main file if the following three requirements are all satisfied:It includes the present file, either directly or indirectly.
The expanded content contains a
\documentclass line near the top.
The expanded content contains
\begin{document}.
Notice that the main file itself does not need to contain the
\documentclass line and \begin{document}, since these can stem from included .tex files. The option g:vimtex_include_indicators is used by the parser to specify commands that include .tex files for the recursive expansion.In cases where automatic detection of the main file through the directory scan fails, one may explicitly set up method 1 to 5 instead. The vimtex-tex-root is usually a good alternative.
Note: Recursive directory descents are not performed to find the main file. That is, if the current file is
./B/chapter.tex then ./A/main.tex will not be found as the main file, because the descent to subdirectory ./A/ is not performed.SUPPORT FOR TEX DIRECTIVES
VimTeX supports two of the commonly used TeX directives [0]: the TeX root and the TeX program directive. The TeX root directive was already described above, see vimtex-tex-root.
vimtex-tex-program
The TeX program directive works by specifying the TeX compiler program in a comment in one of the first lines of the main project file. It is parsed only when it is required by a compiler backend.
The syntax is best explained with an example:
%! TeX program = lualatex
%! TEX TS-program = xelatex
The left-hand side must contain the text "tex program" or "tex ts-program" and as with vimtex-tex-root, the words are recognized regardless of casing and the spaces are ignored. The right-hand side must correspond to a key in the g:vimtex_compiler_latexmk_engines or g:vimtex_compiler_latexrun_engines dictionaries. See also [0,1].PACKAGE DETECTION
VimTeX maintains a list of latex packages that are required by the current project. This list is used by VimTeX for instance to determine which commands to suggest during command completion (see vimtex-complete-commands) and which packages to look up documentation for (see <plug>(vimtex-doc-package)). The list can be viewed with :VimtexInfo.
The package list is determined in two ways:
If a
.fls file exists having the name of the main file, it is scanned. This file is created by latex (or pdflatex, xelatex, ...) if it is
run with the -recorder option (which is set by default when using
latexmk, unless overridden in an initialization file). Parsing the .fls
file is done both at VimTeX initialization and after each successful
compilation, if possible.
Note: Parsing after successful compilations requires that one uses
a) continuous compilation with callbacks (see the
callback option
for g:vimtex_compiler_latexmk), or
b) single-shot compilation.Otherwise, the preamble is parsed for
\usepackage statements. This is slower and less accurate than .fls file parsing. Therefore, it is only
done during VimTeX initialization. If desired, one may manually reload
VimTeX to parse the preamble again during an editing session. See
:VimtexReload and <plug>(vimtex-reload) (by default mapped to
<localleader>lx).
INTEGRATION WITH OTHER PLUGINS
VimTeX provides a lot of convenient and useful features for working with LaTeX files. However, there are several features that one might expect to be part of VimTeX, but that are left out because they are better served by other plugins. Let's call them "friends".
The following is an overview of some such features. We also try to give hints and suggestions for how to best integrate with VimTeX experience, if that is applicable.
* Linting and syntax checking vimtex-af-linting
* Snippets/Templates vimtex-af-snippets
* Tag navigation vimtex-af-tag-nav
* Manipulate surrounding cmds/delims/envs vimtex-af-surround
* Enhanced matching and highlighting of delimiters vimtex-af-enhanced-matchparen * Formatting vimtex-af-formatting
* Filetype plugin for bib files vimtex-af-ftplugin-bib
* Language servers (texlab & ltex) vimtex-af-lsp
Linting and syntax checking
vimtex-af-linting
VimTeX has some support for linting through the :compiler command, see vimtex-lint. There exists several more dedicated, automatic linting plugins. The following plugins have support for (La)TeX syntax checking through lacheck [0], chktex [1], and proselint [2].ale https://github.com/dense-analysis/ale
neomake https://github.com/neomake/neomake
syntastic https://github.com/vim-syntastic/syntasticneomake also supports rubberinfo [3]. One may also be interested in blacktex [4], which may be used to clean up/fix LaTeX code.[0]: https://www.ctan.org/pkg/lacheck
[1]: http://www.nongnu.org/chktex/
[2]: http://proselint.com/
[3]: https://www.systutorials.com/docs/linux/man/1-rubber-info/
[4]: https://github.com/nschloe/blacktex
Snippets/Templates
vimtex-af-snippets
Snippets and/or templates are provided by for instance neosnippet and UltiSnips. See vimtex-neosnippet and vimtex-UltiSnips for more info.Tag navigation
One may navigate by tags with the CTRL-] mapping, e.g. from \eqref{eq:example} to the corresponding \label{eq:example}. However, this requires that a tag file has been generated with ctags. I recommend that one uses the maintained version of ctags [0]. In addition, I recommend that one uses a plugin that automatically generates the tag files as necessary, e.g. |gutentags| [1].See vimtex-faq-tags and vimtex-faq-tags-bibtex for concrete examples.
Manipulate surrounding commands/delimiters/environments
vimtex-af-surround
VimTeX provides mappings that change, delete and toggle commands, delimiters and environments (see the ds, cs and ts family of mappings listed under vimtex-default-mappings). These mappings are inspired by the great surround.vim [0] (|surround.txt|) by Tim Pope, which provides mappings to manipulate surrounding delimiters such as '', "", (), [], {}, and <>. As such, the mappings from VimTeX should work well together with, and as an extension of, surround.vim. Consider also the customization described under vimtex-faq-surround. The mappings may be repeated with the dot (.) command. See also g:vimtex_delim_list if you are interested in customizing the delimiter pairs that are recognized.A different possibility is to use
vim-sandwich [1] (|sandwich.txt|) by Machakann, which may be considered a generalisation of surround.vim in that it can handle much more complex sets of delimiters. vim-sandwich is relatively easy to expand with custom surroundings and has built in support for LaTeX-specific surroundings such as quotations and math delimiters. For a list of supported delimiters, see |sandwich-filetype-recipes|. vim-sandwich supports vim-repeat [2] in addition to visualrepeat.vim [3].Note: The default mappings of
vim-sandwich differ from those of surround.vim, in that they use s as the prefix. E.g., to add surroundings, one uses sa{motion/textobject}{type-of-surrounding} instead of ys{motion/textobject}{type-of-surrounding}. If one prefers the map variants from surround.vim, these are also available as an option, see |sandwich-miscellaneous|. And it is also easy to define custom mappings, if one prefers that.Note:
vim-sandwich actually consists of three plugins that work together. One should make sure to read the docs for all of them: |sandwich.txt|, |operator-sandwich.txt|, and |textobj-sandwich.txt|.[0]: https://github.com/tpope/vim-surround
[1]: https://github.com/machakann/vim-sandwich
[2]: https://github.com/tpope/vim-repeat
[3]: http://www.vim.org/scripts/script.php?script_id=3848
Enhanced matching and highlighting of delimiters
vimtex-af-enhanced-matchparen
VimTeX highlights and allows navigation between matching pairs of delimiters including those in math mode, such as \bigl( and \bigr), and the \begin and \end tags of environments. However, the implementation may be slow (see also vimtex-faq-slow-matchparen, and so one may use g:vimtex_matchparen_enabled to disable the highlighting).Alternatively, one may use the plugin |match-up| [0], which offers enhanced matchparen highlighting and
matchit.zip style motions and text-objects for a variety of file types. For LaTeX documents, it: Extends highlighting and the
% motion to a number of middle delimiters including
\bigm and \middle marked delimiters
\items in itemize and enumerate environments
\toprule, \midrule, \bottomrule in the tabular environment. \if, \else and \endif
Note: VimTeX does not support highlighting the middle delimiters.
Adds motions,
g%, [%, and ]% and text objects, a% and i% which move between matching delimiters and operate on delimited text.
For example, with match-up enabled,
\left( \frac{a}{b} \middle| q \right)the motion
% will cycle through \left(, \middle|, and \right), whereas with VimTeX only \left( and \right) will be matched. The motion g% will do the same, except in reverse.To enable the plugin match-up after installation, add the following to your vimrc:
let g:matchup_override_vimtex = 1Matching may become computationally intensive for complex LaTeX documents. If you experience slowdowns while moving the cursor, the following option is recommended to delay highlighting slightly while navigating:
let g:matchup_matchparen_deferred = 1Note: The exact set of delimiters recognized may differ between match-up and VimTeX. For example, the mappings
da% and dad will not in general be identical, particularly if you have customized VimTeX's delimiters.Formatting
vimtex-af-formatting
VimTeX has a custom |formatexpr| that may be enabled with the option g:vimtex_format_enabled. However, there are a lot of different styles for formatting LaTeX manuscripts. These are typically much more relevant when writing in collaboration with others. A good reference on this topic is [0], and note in particular the box "Directives for using LaTeX with version control systems".The most basic style is to hard wrap lines at a given column, e.g. 80 columns, and this is exactly the type of formatting that is supported by VimTeX. However, this is usually not very friendly when collaborating with others, as it tends to mess up diffs between versions of the document. Instead, one might want to consider one of these:
a) keeping each sentence on a line (use soft wrapping)
b) add additional indentation for split sentences [1]
c) use semantic line feeds [2]
In order to make it easier to use one of these styles of formatting, one may want to use an external formatter:
latexindent.pl [3]
vim-bucky [4] (note: this is an alpha version as of October 2018)
semantic-linebreaker [5] (note: this is a web-based tool)
Further, there are a range of Vim plugins that can be used to format your document with external tools. Some of these also allow autoformatting of some kind. In no particular order:
neoformat [6]
vim-codefmt [7]
vim-autoformat [8]
ale [9]
vim-sentence-chopper [10]
[0]: https://en.wikibooks.org/wiki/LaTeX/Collaborative_Writing_of_LaTeX_Documents
[1]: http://dustycloud.org/blog/vcs-friendly-patchable-document-line-wrapping/
[2]: https://rhodesmill.org/brandon/2012/one-sentence-per-line/
[3]: https://github.com/cmhughes/latexindent.pl
[4]: https://github.com/dbmrq/vim-bucky
[5]: https://github.com/waldyrious/semantic-linebreaker
[6]: https://github.com/sbdchd/neoformat
[7]: https://github.com/google/vim-codefmt
[8]: https://github.com/Chiel92/vim-autoformat
[9]: https://github.com/dense-analysis/ale
[10]: https://github.com/Konfekt/vim-sentence-chopper
Filetype plugin for bib files
vimtex-af-ftplugin-bib
VimTeX is not a full filetype plugin for bibliography files (.bib). However, it does alter the 'comments' and 'commentstring' options and provide basic indentation and folding; see g:vimtex_indent_bib_enabled and g:vimtex_fold_bib_enabled, respectively.Here are a couple of other related Vim plugins and external tools that might be of interest:
bibtool
An external tool for formatting, sorting, filtering, merging, and more of
.bib files.
http://www.gerd-neugebauer.de/software/TeX/BibTool/
GooseBib
Some simple command-line tools to clean-up / modify BibTeX files.
https://github.com/tdegeus/GooseBib
bibtex-tidy
Another tool for formatting and cleaning .bib files.
https://flamingtempura.github.io/bibtex-tidy/
tbibtools
A set of ruby-based bibtex-related utilities for sorting, reformatting,
listing contents, and so on. Has optional Vim integration.
https://www.vim.org/scripts/script.php?script_id=1915
See also https://github.com/lervag/vimtex/issues/1293 for some related discussions.
Language servers
vimtex-af-lsp
In recent years, language servers (LSPs) [0] have become very popular. There is a language server for LaTeX and bibtex called texlab [1]. It may be interesting both as an alternative to VimTeX and/or an addition.There is currently no known conflict between texlab and VimTeX, although there is some feature overlap. E.g., both texlab and VimTeX provides advanced completion in various contexts. As texlab is written in Rust and runs in a separate thread, it is no surprise that it will have a clear performance advantage. However, VimTeX does use caches to speed up completion which should in most cases work well.
To use texlab, one must use an LSP client, e.g. |vim-lsp| [2], neovim's built-in LSP client [3], or |coc-nvim| [4, 5]. See also this VimTeX issue [6] for more information.
In addition to texlab, there is also a dedicated grammar and spell checking language server called LTeX [7]. It relies on LanguageTool and supports both LaTeX and other markup languages. It may be a useful tool to use in conjunction with VimTeX as an alternative to vimtex-grammar.
[0]: https://langserver.org/
[1]: https://github.com/latex-lsp/texlab
[2]: https://github.com/prabirshrestha/vim-lsp
[3]: https://github.com/neovim/nvim-lspconfig/blob/master/doc/server_configurations.md#texlab
[4]: https://github.com/neoclide/coc.nvim
[5]: https://github.com/fannheyward/coc-texlab
[6]: https://github.com/lervag/vimtex/issues/1371
[7]: https://valentjn.github.io/ltex/index.html
USAGE
Default mappings vimtex-default-mappings
Options vimtex-options
Commands vimtex-commands
Map definitions vimtex-mappings
Insert mode mappings vimtex-imaps
Events vimtex-events
DEFAULT MAPPINGS
VimTeX is designed to be controlled by a selection of mappings. Note, though, that most of the mappings are also available as commands, see vimtex-commands.
Many of the mappings use
<localleader>l as a common prefix, where the default |<localleader>| is \. Thus, <localleader>ll will for most people mean \ll. The prefix may be changed with g:vimtex_mappings_prefix.The right-hand sides are provided as
<plug>-mappings, see |using-<plug>|. For any given <plug> map, the default mapping will only be created if it does not already exist. This means that if a user defines a custom mapping withnmap <space>li <plug>(vimtex-info)
then the corresponding default left-hand side will not be mapped.If one prefers, one may disable all the default mappings through the option g:vimtex_mappings_enabled. Custom mappings for all desired features must then be defined through the listed RHS
<plug>-maps or by mapping the available commands.In the below list of mappings, LHS is the default mapping, RHS is the corresponding
<plug>-maps, and MODE indicates in which vim mode the mappings are valid. See map-modes for an explanation of the various modes. The indicator refers to the prefix of the corresponding map command, e.g. n refers to an |nmap|, nx refers to both |nmap| and |xmap|, and so on.In addition to the mappings listed below, VimTeX provides convenient insert mode mappings to make it easier and faster to type mathematical equations. This feature is explained in more detail later, see vimtex-imaps.
OPTIONS
g:vimtex_enabled
Set to 0 to disable VimTeX.
Default value: Undefined.
g:vimtex_cache_root
Specify the cache directory for VimTeX.
Default value:
'$XDG_CACHE_HOME/vimtex' if $XDG_CACHE_HOME is defined
'~/.cache/vimtex' otherwiseg:vimtex_cache_persistent
Specify whether to use persistent caching.
Default value: 1
g:vimtex_compiler_enabled
Use this option to disable/enable the
compiler interface, see vimtex-compiler.Default value: 1
g:vimtex_compiler_silent
Set this to 1 or v:true to silence the compiler messages during start, stop, and callbacks.
Default: 0
g:vimtex_compiler_method
This option sets the compiler method. There are two ways to configure this option:
i) Explicity as a string.
ii) Dynamically through a function.
In the first variant, you can choose from the following list of compiler methods:
Value Documentation Configuration
latexmk vimtex-compiler-latexmk g:vimtex_compiler_latexmk
latexrun vimtex-compiler-latexrun g:vimtex_compiler_latexrun
tectonic vimtex-compiler-tectonic g:vimtex_compiler_tectonic
arara vimtex-compiler-arara g:vimtex_compiler_arara
texpresso vimtex-compiler-texpresso g:vimtex_compiler_texpresso
generic vimtex-compiler-generic g:vimtex_compiler_genericIn the second variant, the option must be specified as the name of a function or a Funcref. Using a Funcref is only possible with Lua in Neovim. The function is passed the path to the main TeX file as a single string argument and must return the desired method as a string. The method must be one from the above list of available methods.
Using a function allows a lot of flexibility in the choice of compiler method. For instance, one could use
arara for files that have arara specifications at the top and fall back to latexmk for other files:function! SetCompilerMethod(mainfile)
if filereadable(a:mainfile)
for line in readfile(a:mainfile, '', 5)
if line =~# '^%\s*arara'
return 'arara'
endif
endfor
endif
return 'latexmk'
endfunction
let g:vimtex_compiler_method = 'SetCompilerMethod'Default value:
'latexmk'g:vimtex_compiler_clean_paths
A list of additional path expressions for generated files that you want to be cleaned by
:VimtexClean. Note that this is NOT relevant for the compiler backends vimtex-compiler-latexmk and vimtex-compiler-latexrun. These backends provide their own clean implementations.Each expression is a glob expression (see glob() and wildcards) and each path is assumed rooted to the project root. For instance, to clean any generated
_minted paths (including directories), you could use something like this:let g:vimtex_compiler_clean_paths = ['_minted*']Warning: Each resolved path will be deleted with delete() with the
"rf" flag!Default value:
[]g:vimtex_compiler_latexmk
This dictionary allows customization of the vimtex-compiler-latexmk compiler. The values set by the user will take precedence over the default values.
Default value:
let g:vimtex_compiler_latexmk = {
\ 'aux_dir' : '',
\ 'out_dir' : '',
\ 'callback' : 1,
\ 'continuous' : 1,
\ 'clean_ext' : '',
\ 'executable' : 'latexmk',
\ 'hooks' : [],
\ 'options' : [
\ '-verbose',
\ '-file-line-error',
\ '-synctex=1',
\ '-interaction=nonstopmode',
\ ],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
aux_dir
This option sets the directory for auxiliary output files. It corresponds
to the $aux_dir option of latexmk. If the path is a relative path,
then it is considered relative to the main project file. The value of this option should be either:
1) a string that represents a path, or
2) a Funcref with a single dictionary argument "file_info":
file_info = {
root = …
target = …
target_basename = …
target_name = …
jobname = …
} This makes it possible to specify a dynamic
aux_dir. It may be
easier to understand from an example:let g:vimtex_compiler_latexmk = {'aux_dir': {_ -> expand("%:t:r")}} With the above setting, the
aux_dir is set to the base name of the
current file. E.g., If you do vim test.tex, the value becomes
test. The specified auxiliary directory is created if it does not exist.
Note 1: This option only works with
latexmk version 4.27 and later.
Note 2: If $aux_dir is added to .latexmkrc, then the .latexmkrc setting
will have priority.
Note 3: If $VIMTEX_OUTPUT_DIRECTORY is defined, it will have the highest
priority.
Note 4: The -emulate_aux option will be automatically passed to
latexmk if this option is not empty.out_dir
This option sets the directory for the compilation output files. It
corresponds to the $out_dir option in latexmk. If the path is
a relative path, then it is considered relative to the main project file. The value is either a string or a Funcref, similar to the above
described
aux_dir key. The specified output directory is created if it does not exist.
Note 1: This option only works with
latexmk version 4.27 and later.
Note 2: If $out_dir is added to .latexmkrc, then the .latexmkrc setting
will have priority.
Note 3: If $VIMTEX_OUTPUT_DIRECTORY is defined, it will have the highest
priority.clean_ext
This option specifies additional file extensions to be removed by
:VimtexClean and :VimtexClean!. It corresponds to latexmk's
$clean_ext option. See the documentation for more info:
https://texdoc.org/serve/latexmk/0. It consists of a single string of space-separated extensions with no
preceding dot, for example:
"ex1 ex2 ex3" By default,
latexmk will remove aux, bcf, fls, idx, ind, lof,
lot, out, toc, blg, ilg, log, and xdv files. Note: If
$clean_ext is added to .latexmkrc, then the .latexmkrc
setting will have priority and this option is ignored.callback
If enabled, this option tells latexmk to run vimtex#compiler#callback
after compilation is finished.continuous
If enabled, latexmk will run in continuous mode, i.e. with the -pvc
argument. This means that the document is compiled automatically by
latexmk every time a related file has been changed, until the processes
is stopped. If disabled,
latexmk will run single shot compilations. Note: The events VimtexEventCompileStarted and VimtexEventCompileStopped
are only relevant when this option is enabled.
executable
The name/path to the latexmk executable. Can be a string or a list.
If a list is provided, it will be interpreted as a command sequence
(e.g., ['docker', 'exec', 'container', 'latexmk'] for running latexmk
through a container).hooks
A list of Funcrefs. If running in continuous mode, each hook will be
called for each line of output from latexmk, with that line as argument.
E.g., to show information about the compilation run numbers, one could do
this:function! Callback(msg)
let l:m = matchlist(a:msg, '\vRun number (\d+) of rule ''(.*)''')
if !empty(l:m)
echomsg l:m[2] . ' (' . l:m[1] . ')'
endif
endfunction
let g:vimtex_compiler_latexmk = { 'hooks': [function('Callback')] }options
This is a list of options that are passed to latexmk. The default
options should work well for most people. Note: Options like
-pdf or -lualatex should NOT be added to this list.
These are options used to specify the LaTeX processor/engine, see
instead g:vimtex_compiler_latexmk_engines. Note: Options may also be specified indirectly to
latexmk through both
a global and a project specific .latexmkrc file. One should know,
though, that options specified on the command line has priority, and
so if one wants to override one of the above default options, then
one has to set this key to a list that contains the desired options.g:vimtex_compiler_latexmk_engines
Defines a map between TeX program directive (vimtex-tex-program) and compiler engine. This is used by vimtex-compiler-latexmk to define the LaTeX program. The
_ key defines the default engine.Note: If the TeX program directive is not specified within the main project file, and if
$pdf_mode is added to a project-specific .latexmkrc file, then the compiler engine will be deduced from the value of $pdf_mode. The supported values of $pdf_mode are 1 (pdflatex), 4 (lualatex) and 5 (xelatex). See the latexmk documentation for details.Default value:
let g:vimtex_compiler_latexmk_engines = {
\ '_' : '-pdf',
\ 'pdfdvi' : '-pdfdvi',
\ 'pdfps' : '-pdfps',
\ 'pdflatex' : '-pdf',
\ 'luatex' : '-lualatex',
\ 'lualatex' : '-lualatex',
\ 'xelatex' : '-xelatex',
\ 'context (pdftex)' : '-pdf -pdflatex=texexec',
\ 'context (luatex)' : '-pdf -pdflatex=context',
\ 'context (xetex)' : '-pdf -pdflatex=''texexec --xtx''',
\}
g:vimtex_compiler_latexrun
This dictionary allows customization of the vimtex-compiler-latexrun compiler. The values set by the user will take precedence over the default values.Default value:
let g:vimtex_compiler_latexrun = {
\ 'out_dir' : '',
\ 'options' : [
\ '-verbose-cmds',
\ '--latex-args="-synctex=1"',
\ ],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
options
This is a list of options that are passed to latexrun. The default
options should work well for most people. Note: By default, the option
-pdf is also supplied to indicate the LaTeX
engine. This may be changed on a per project basis with TeX
directives, see vimtex-tex-program or the two compiler-specific
options g:vimtex_compiler_latexmk_engines and
g:vimtex_compiler_latexrun_engines. The latter two options may
also be used to change the default engine.g:vimtex_compiler_latexrun_engines
Defines a map between TeX program directive (vimtex-tex-program) and compiler engine, i.e. as should be specified to the
--latex-cmd argument to latexrun. This is used by vimtex-compiler-latexrun to define the LaTeX program. The _ key defines the default engine.Default value:
let g:vimtex_compiler_latexrun_engines = {
\ '_' : 'pdflatex',
\ 'pdflatex' : 'pdflatex',
\ 'lualatex' : 'lualatex',
\ 'xelatex' : 'xelatex',
\}
g:vimtex_compiler_tectonic
This dictionary allows customization of the vimtex-compiler-tectonic compiler. The values set by the user will take precedence over the default values.Default value:
let g:vimtex_compiler_tectonic = {
\ 'out_dir' : '',
\ 'hooks' : [],
\ 'options' : [
\ '--keep-logs',
\ '--synctex'
\ ],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
options
This is a list of options that are passed to tectonic. The default
options should work well for most people. For anyone who wishes to modify
these, please note:
Don't use
--outdir or -o here. Use the out_dir option instead. Without
--keep-logs (or --keep-intermediates or -k), you won't see
errors/warnings in the quickfix list when compilations finish.
By default,
tectonic cleans all auxiliary files (such as .aux, .toc, etc.). If you omit the --keep-logs (or similar) options that
specify to keep these files, :VimtexClean and <plug>(vimtex-clean)
won't delete anything (as there is nothing to delete).
g:vimtex_compiler_arara
This dictionary allows customization of the vimtex-compiler-arara compiler. The values set by the user will take precedence over the default values.
Default value:
let g:vimtex_compiler_arara = {
\ 'options' : ['--log'],
\ 'hooks' : [],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
options
This is a list of options that are passed to arara. The default options
should work well for most people.g:vimtex_compiler_texpresso
This dictionary allows customization of the vimtex-compiler-texpresso compiler. The values set by the user will take precedence over the default values.
Default value:
let g:vimtex_compiler_texpresso = {
\ 'options' : [],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
options
Extra options passed to texpresso, in addition to the -json and
-lines flags that VimTeX always supplies internally. Use this to pass
engine selection flags such as -texlive or -tectonic.g:vimtex_compiler_generic
This dictionary allows customization of the vimtex-compiler-generic compiler. This compiler is, as the name hints, generic. It allows to specify a custom command to run for compilation. As for the other compilers, the configuration values set by the user will take precedence over the default values.
Default value:
let g:vimtex_compiler_generic = {
\ 'command' : '',
\ 'hooks' : [],
\}The default value shows which entries may be changed. Here the different keys are explained in more detail:
command
This is the command to run to start compilation. This can be any command,
and the command is run from the project root. The command string will
substitute @tex with the path to the current project's main tex file.g:vimtex_complete_enabled
Use this option to disable/enable VimTeX completion.
Default value: 1
g:vimtex_complete_smart_case
If enabled, then VimTeX will filter case sensitive if there is a capital letter in the completion input. This is only relevant if g:vimtex_complete_ignore_case is also enabled.
Default value: Same as your 'smartcase' value
g:vimtex_complete_ignore_case
If enabled, then VimTeX will filter case insensitive.
Default value: Same as your 'ignorecase' value
g:vimtex_complete_close_braces
This option controls whether to append a closing brace after a label or a citation has been completed.
Default value: 0
g:vimtex_parser_bib_backend
This option sets the desired default backend for parsing bibliographies. This is used e.g. for gathering completion candidates. Possible values:
bibtex: The fastest, but most "hacky" solution. Still, time has proved
that this works well!vim: The slowest but perhaps most robust solution, as it does not
require any external utilities.lua: A Lua implementation of the Vim backend. About as fast as the
bibtex parser, but this only works on Neovim.bibparse: Also fast, but might be more robust. Note: bibparse is an executable provided by the Perl package
Text-BibTeX [0]. It should not be confused with the
similarly named Python project [1]. The latter is
deprecated in favor of GooseBib [2]. The Python projects
bibparse and GooseBib are both based on the Python library
bibtexparser. [0]: https://metacpan.org/dist/Text-BibTeX
[1]: https://github.com/tdegeus/bibparse
[2]: https://github.com/tdegeus/GooseBib
bibtexparser:
Also fast and possibly more robust. See the project Github
page for more details:
https://github.com/sciunto-org/python-bibtexparser Note: This requires that Python 3 is available to Vim/neovim
(see if_pyth and |py3|) and that the
bibtexparser
Python module is installed and available.Some people may want to conditionally change this option if a backend is available. For example:
if executable('bibparse')
let g:vimtex_parser_bib_backend = 'bibparse'
endifDefault value:
Vim:
bibtex
Neovim: luag:vimtex_parser_cmd_separator_check
This option specifies the policy for deciding whether successive groups of
[opt], {arg}, (arg), and <overlay> following a \command should be recognized as arguments to that \command.In fact, parsing a LaTeX command without additional knowledge is a hard problem. When we read
\foo{bar}{baz} — is {baz} going to be consumed as an argument to \foo? The only way to know this is to read the definition of the \foo command/macro.A pragmatic choice when we write a parser, therefore, is to rely on some heuristics and common practises. This will never be perfect, but it can be good enough for practical use. In VimTeX, the core heuristics are that a command will look like this:
\foo<overlay>[[opt]{arg}]...
\begin{name}<overlay>[[opt]{arg}]...The parser greedily swallows as many such groups as possible as long as the function specified via this option returns true for the text between successive groups.
The default function will allow a line break and possibly white space on the preceding line before a new group. E.g.:
% command number of args
% ------- --------------
\foo{bar}{baz} % 2
\foo{bar} {baz} % 1
\foo{bar}
{baz} % 2
\foo{bar}
{baz} % 2
\foo{bar}__
{baz} % 1 (_ indicates spaces)The option should be either the name of the function (a string) or a Funcref. The function takes a single argument, which is the string between successive groups. It should return v:true if the parser should continue and v:false if the parser should stop.
The check applies to all four group types:
[opt], {arg}, (arg), and <overlay>. Note that (arg) and <overlay> groups are only matched on a single line, so for these the separator string will never contain a newline; the default rule therefore captures them only when they directly touch the preceding group.A user may want to change this behaviour e.g. to specify that all whitespace should be allowed, including and up to a single newline:
function! MyCmdSeparatorRule(separator_string)
return a:separator_string =~# '^\_s\+$'
\ && count(a:separator_string, "\n") < 2
endfunction
let g:vimtex_parser_cmd_separator_check = 'MyCmdSeparatorRule'Note: This option is relevant for any feature that relies on the parsing of a command. This includes, but is not limited to the <plug>(vimtex-ac) text object (vimtex-text-objects).
Note: Funcrefs are only possible when it is used with neovim Lua configuration, because in Vimscript, variable names must be capitalized in order to point to Funcrefs.
Default:
'vimtex#cmd#parser_separator_check'g:vimtex_bibliography_commands
A list of command names for commands that include bibliography files. Each list entry is interpreted as a pattern (very magic, see /\v) to match a particular command name. This option may be useful if one defines custom commands that includes bibliography files.
Default value:
['%(no)?bibliography', 'add%(bibresource|globalbib|sectionbib)']g:vimtex_complete_bib This option is a dictionary for controlling the citation completion. The keys each control a different thing as explained below.
simple
Default value: 0
If zero, then the cite completion is "smart", i.e. not simple. This
behaviour is described in more detail in vimtex-complete-cites. Note: It is usually better to use the "simple" mode if you use an
autocomplete plugin (vimtex-complete-auto).
g:vimtex_complete_bib.match_str_fmt
match_str_fmt
Default value: '@key [@type] @author_all (@year), "@title"'
The format used for the match string for bib completion candidates. That
is, the string that the smart mode candidate matching is matched against.
See vimtex-complete-cites for more info. The following keys may be used
to define the string:@author_all Full author list @author_short Shortened author list @key The bibtex key @title Title @type Type of entry @year Publication year
Since the author list can be large, the
@author_all is truncated to 20
characters. This can be modified with the auth_len key (see below).menu_fmt
Default value: '[@type] @author_short (@year), "@title"'
The format used for the menu entry for bib completion candidates (see
complete-items). If the key is set to an empty string, then the menu
entry is not added to the completion candidates. See the description of
match_str_fmt for the allowed keys.info_fmt
Default value: "TITLE: @title\nAUTHOR: @author_all\nYEAR: @year"
The format used for the info entry for bib completion candidates (see
complete-items). See the description of match_str_fmt for the allowed
keys.abbr_fmt
Default value: ''
The format used for the abbr entry for bib completion candidates (see
complete-items). See the description of match_str_fmt for the allowed
keys.auth_len
Default value: 20
Truncation length for author list with the @author_all format key in the
format strings for match_str_fmt, menu_fmt, and abbr_fmt.custom_patterns
Default value: []
List of custom trigger patterns that may be used to allow completion for
e.g. custom macros.If one wants to overwrite one of the keys, e.g. the
simple entry, one can do:let g:vimtex_complete_bib = { 'simple': 1 }This does not modify the other keys and their default values.
g:vimtex_complete_ref
This option is a dictionary for controlling the label completion. The keys each control a different thing:
custom_patterns
Default value: []
List of custom trigger patterns that may be used to allow completion for
e.g. custom macros.For example, if one has defined the command
\figref, one could add following custom patternlet g:vimtex_complete_ref = {
\ 'custom_patterns': ['\\figref\*\?{[^}]*$']
\ }
g:vimtex_context_pdf_viewer
Specify PDF viewer to use to open PDF files with the vimtex-context-menu, for instance for citations with the file key (see vimtex-context-citation).The default value is based on the vimtex-view and is determined as follows:
* If g:vimtex_view_method is not
general, then the specified viewer is
used. However, the viewer will by default start without any of the
regular options. * Else fall back to the value of g:vimtex_view_general_viewer.
g:vimtex_delim_list
A dictionary that defines the pairs of delimiters that are recognized by VimTeX for various commands and functions. The dictionary contains 5 sub dictionaries:
env_tex Pairs of environment delimiters in normal TeX mode
env_math Pairs of special math environment delimiters
delim_tex Pairs of delimiters in normal TeX mode
delim_math Pairs of delimiters in math mode
mods Pairs of modifiers for math mode delimitersEach entry is a dictionary with the following format:
{
\ 'name' : [
\ ['\(', '\)'],
\ ['\[', '\]'],
\ ['$$', '$$'],
\ ['$', '$'],
\ ],
\ 're' : [
\ ['\\(', '\\)'],
\ ['\\\@<!\\\[', '\\\]'],
\ ['\$\$', '\$\$'],
\ ['\$', '\$'],
\ ],
\}Here the
name entry is a list of delimiter pairs as they are typed, and the re entry is a corresponding list of regexes that matches the delimiters.The default value should generally suffice for most people. If one wants to overwrite one of the main entries, e.g. the
mods entry, one can do something like this:let g:vimtex_delim_list = {
\ 'mods' : {
\ 'name' : [ "..." ],
\ }
\}Here the
re entry was not provided, in which case it will be automatically generated based on the name entry. The remaining four entries will remain the default value.Some people may be interested in adding support for e.g. german or french quotation marks. These may be added by extending the default
delim_tex entries, like this:let g:vimtex_delim_list = {
\ 'delim_tex' : {
\ 'name' : [
\ ['[', ']'],
\ ['{', '}'],
\ ['\glq', '\grq'],
\ ['\glqq', '\grqq'],
\ ['\flq', '\frq'],
\ ['\flqq', '\frqq'],
\ ]
\ }
\} g:vimtex#delim#lists
g:vimtex#delim#re
Note: This option is parsed on plugin initialization into a new variable, g:vimtex#delim#lists where the
re entries are added and that also contains some combinations such as tex_all, delim_all, and all. Further, the option is also used as a basis for the variable g:vimtex#delim#re, which contains full regexes for matching opening and/or closing delimiters of the desired type.Default value: See
s:init_delim_lists() in /autoload/vimtex/delim.vim.g:vimtex_delim_toggle_mod_list
Defines a list of delimiter modifiers to toggle through using the maps:
The list must be a subset of the
mods entry of g:vimtex_delim_list, otherwise the toggle will not work properly. Thus, if one wants to toggle non-standard delimiters, then one must also update the above option.Example 1: to toggle between no modifiers, the
\left/\right pair, and the \mleft/\mright pair, one may use the following options:let g:vimtex_delim_list = {'mods' : {}}
let g:vimtex_delim_list.mods.name = [
\ ['\left', '\right'],
\ ['\mleft', '\mright'],
\ ['\bigl', '\bigr'],
\ ['\Bigl', '\Bigr'],
\ ['\biggl', '\biggr'],
\ ['\Biggl', '\Biggr'],
\ ['\big', '\big'],
\ ['\Big', '\Big'],
\ ['\bigg', '\bigg'],
\ ['\Bigg', '\Bigg'],
\]
let g:vimtex_delim_toggle_mod_list = [
\ ['\left', '\right'],
\ ['\mleft', '\mright'],
\]Example 2: to step through no modifiers, and the pairs
\bigl/\bigr, \Bigl/\Bigr, \biggl/\biggr, and \Biggl/\Biggr, one may use:let g:vimtex_delim_toggle_mod_list = [
\ ['\bigl', '\bigr'],
\ ['\Bigl', '\Bigr'],
\ ['\biggl', '\biggr'],
\ ['\Biggl', '\Biggr'],
\]Default value:
[['\left', '\right']]g:vimtex_delim_timeout
g:vimtex_delim_insert_timeout
Timeout (in milliseconds) when searching for matching delimiters. It is used for the
{timeout} argument of search()-like function calls. If the option is increased it will make the matching more accurate, at the expense of potential lags. The default value should work well for most people.Default values: 300, 60 (respectively)
g:vimtex_delim_stopline
A tolerance for the number of lines to search for matching delimiters in each direction. It is used in an expression for the
{stopline} argument of search() function calls. If the option is increased it will make the matching more accurate, at the expense of potential lags. The default value should work well for most people.Default value: 500
g:vimtex_doc_enabled
Use this option to disable features related to vimtex-latexdoc.
Default value: 1
g:vimtex_doc_confirm_single
When enabled (set to 1 or v:true), then VimTeX will open the specified documentation only after a confirmation prompt such as:
Open documentation for usepackage: foobar? [y]es/[n]o
Thus, one may disable this option (set to 0 or v:false) to avoid the confirmation and open directly. Note that this is only relevant when there is only a single recognized documentation source.
Default value: v:true
g:vimtex_doc_handlers
With this option, one may specify a list of custom documentation handlers. The following pre-made handlers are available:
'vimtex#doc#handler#texdoc' Open documentation with local texdoc.
Fallback Open documentation online through
http://texdoc.org/pkg/packagename.To use the local
texdoc, set:let g:vimtex_doc_handlers = ['vimtex#doc#handlers#texdoc']A handler is a function that takes a single Dict argument with the following keys:
type
One of documentclass, usepackage, command or word.candidates
A list of detected packages (for the types command and usepackage,
this list may be larger than 1.selected
The currently selected entry. This is the package name that will
ultimately be passed to the lookup function.name
If the type is command, this is the name of the command. Else it is
not defined.Each handler in the list will be tried until a handler provides a return value of 1 or v:true. One may thus add handlers that only makes minor modifications of the context and passes it on to the next handler.
The context may have multiple candidates and the handlers are applied before any internal selection is made. Thus the
selected key may be not defined. This allows the handler to perform the selection itself. One may manually call the selection function vimtex#doc#make_selection to get a simple selection menu.The following shows a generic example of how to write a custom handler:
let g:vimtex_doc_handlers = ['MyHandler']
function! MyHandler(context)
call vimtex#doc#make_selection(a:context)
if empty(a:context.selected) | return 0 | endif
execute '!myhandler' a:context.selected '&'
return 1
endfunctionDefault value: []
g:vimtex_echo_verbose_input
For the set of operator mappings that change a surrounding type [0], VimTeX by default prints some information about what you are doing while waiting for user input. For advanced/experienced users, one will not need this info and can get a slightly cleaner UI by disabling this feature (set the option 0).
Default value: 1
[0]: This affects the following mappings:
<plug>(vimtex-env-change) (default map:
cse)
<plug>(vimtex-env-change-math) (default map: cs$)
<plug>(vimtex-cmd-change) (default map: csc)
<plug>(vimtex-delim-change-math) (default map: csd)g:vimtex_env_change_autofill
If enabled, the current environment value is used as a default input for <plug>(vimtex-env-change) and <plug>(vimtex-env-change-math). Some users may find this useful in order to quickly change from things like
align to aligned.Note: If enabled, one may erase the autofilled content with c_CTRL-U (i.e.
<c-u>).Default: 0
g:vimtex_env_toggle_map
Specify the toggle map for <plug>(vimtex-env-toggle). You can use this to change the desired toggle sequence.
Default value:
let g:vimtex_env_toggle_map = {
\ 'itemize': 'enumerate',
\ 'enumerate': 'itemize',
\}
g:vimtex_env_toggle_math_map
Specify toggle map for <plug>(vimtex-env-toggle-math). You can use this to change the desired toggle sequence.Default value:
let g:vimtex_env_toggle_math_map = {
\ '$': '\[',
\ '\[': 'equation',
\ '$$': '\[',
\ '\(': '$',
\}
g:vimtex_fold_enabled
Use this option to enable folding, which means VimTeX will enable the following options for LaTeX files:setlocal foldmethod=expr
setlocal foldexpr=vimtex#fold#level(v:lnum)
setlocal foldtext=vimtex#fold#text()More detailed info can be found in the section vimtex-folding.
Default value: 0
g:vimtex_fold_manual
With this option enabled, VimTeX uses fold-manual as the main |foldmethod|. It still uses the |foldexpr| function to compute the fold levels, but it only computes the fold levels on demand, see :VimtexRefreshFolds and vimtex-zx.
The reasoning behind this option is that the fold-expr method of folding may sometimes be slow, e.g. for long lines and large files. fold-manual is very fast.
An alternative to this method of speeding up is to use a dedicated plugin for optimizing the fold functionality, see e.g. https://github.com/Konfekt/FastFold.
Default value: 0
Use custom section symbol for folding.
Default value:
'*'g:vimtex_fold_types
g:vimtex_fold_types_defaults
This is a dictionary where each key configures the corresponding fold type. One may disable the fold types by setting the key
enabled to 0. If a type can be configured with a list of patterns or similar, the patterns assume that one uses very magic regexes (see |\v|).One may also customize the text that is displayed in a closed fold of a given type by setting the
text key to a Funcref (see fold-foldtext and Dictionary-function). The function that the Funcref refers to takes two arguments, the text of the current line and the fold level of the current line, and it returns the text that will be displayed in a closed fold. For example, the following configuration will set the text for a closed marker fold to Vim's default:let g:vimtex_fold_types = {
\ 'markers' : {},
\}
function! g:vimtex_fold_types.markers.text(line, level) abort dict
return foldtext()
endfunction(Note that the g:vimtex_fold_types
marker dictionary must be initialized, even if you are not setting custom markers!)Using a literal-Dict and a lambda, one could specify the same configuration more concisely:
let g:vimtex_fold_types = #{
\ markers : #{text : {line, level -> foldtext()}}
\}And for completeness, the following would work with a Lua based config:
vim.g.vimtex_fold_types = {
markers = {
text = function(line, level) return vim.fn.foldtext() end
}
}Each entry in g:vimtex_fold_types is combined with the corresponding entry of g:vimtex_fold_types_defaults. If there are conflicting entries, then g:vimtex_fold_types take precedence. This way, it is easy to customize various fold types without touching those that can stay with default configuration.
The available fold types (and keys) are listed below, and the default configurations are listed at the bottom.
<preamble> Fold the preamble.<sections> Fold sections and parts of documents. Can be
configured with the following extra keys:
parse_levels: Whether to use detailed parsing to
set fold text levels similar to how
they are displayed in vimtex-toc.
Disabled by default, because it uses
more resources and may be slow.
sections: List of sections that should be folded.
parts: List of parts that should be folded.
When a LaTeX document is opened, the document is
parsed in order to define the highest fold level based
on which parts (such as frontmatter, backmatter, and
appendix) and section types (parts, chapter, section,
etc.) are present. This parsing is done automatically
every time the folds are recomputed, if there are any
changes to the file.
The fold function also recognizes "fake" sections.
That is, it parses comments similar to:
% fakepart title % Fakechapter title % Fakesection title % fakesubsection title
The fake sections are folded at the same level as the
corresponding "real" sections. The fold title is the
provided title with the
Fake... part prepended.<comment_pkg> Fold \begin{comments} ... \end{comments} and disable
folding inside the environment.<comments> Fold multiline comments. This is disabled by default.<markers> Fold on vim-style markers inside comments, that is,
pairs of e.g. {{{ and }}} (the default markers).
|regex| patterns for the opening and closing markers
may be customized with the keys:
open
close
Note: Patterns are only searched inside comments!
<envs> Fold environments.
Can be further configured with a blacklist and
whitelist of environments to be folded. Note: The
document environment will never be folded.<env_options> This fold type allows to fold the \begin command if
it contains a long optional argument. Consider the
following example:\begin{axis}[ ---> \begin{axis}[...]
width=6cm,
height=8cm,
...,
] Here the
axis environment must not be otherwise
folded through the <envs> fold type.<items> \item blocks in itemize like environments. The
recognized environments are the same as specified by
g:vimtex_indent_lists.<cmd_single> Fold long commands with a single argument. E.g.:\hypersetup{ ---> \hypersetup{...}
option 1,
...,
option n
}<cmd_single_opt> Fold commands that opens with a single long optional
argument that is followed by a short "real" argument.
E.g.:\usepackage[ ---> \usepackage[...]{name}
option 1,
...,
option n
]{name}<cmd_multi> Fold commands that start with a short regular argument
and continue with long optional and/or regular
arguments. E.g.:\newcommand{\xx}[3]{ ---> \newcommand{\xx} ...
Hello #1, #2, and #3.
}<cmd_addplot> Folding of the \addplot series of commands from the
pgfplots package. E.g.:\addplot+[] table[] { ---> \addplot+[] table[] {...};
table data
};As an example, the following configuration will disable folding of the preamble, as well as the
figure and table environments.let g:vimtex_fold_types = {
\ 'preamble' : {'enabled' : 0},
\ 'envs' : {
\ 'blacklist' : ['figure', 'table'],
\ },
\}Default value:
let g:vimtex_fold_types = {}
let g:vimtex_fold_types_defaults = {
\ 'preamble' : {},
\ 'items' : {},
\ 'comment_pkg' : {},
\ 'comments' : {'enabled' : 0},
\ 'envs' : {
\ 'blacklist' : [],
\ 'whitelist' : [],
\ },
\ 'env_options' : {},
\ 'markers' : {},
\ 'sections' : {
\ 'parse_levels' : 0,
\ 'sections' : [
\ '%(add)?part',
\ '%(chapter|addchap)',
\ '%(section|addsec)',
\ 'subsection',
\ 'subsubsection',
\ ],
\ 'parts' : [
\ 'appendix',
\ 'frontmatter',
\ 'mainmatter',
\ 'backmatter',
\ ],
\ },
\ 'cmd_single' : {
\ 'cmds' : [
\ 'hypersetup',
\ 'tikzset',
\ 'pgfplotstableread',
\ 'lstset',
\ ],
\ },
\ 'cmd_single_opt' : {
\ 'cmds' : [
\ 'usepackage',
\ 'includepdf',
\ ],
\ },
\ 'cmd_multi' : {
\ 'cmds' : [
\ '%(re)?new%(command|environment)',
\ 'providecommand',
\ 'presetkeys',
\ 'Declare%(Multi|Auto)?CiteCommand',
\ 'Declare%(Index)?%(Field|List|Name)%(Format|Alias)',
\ ],
\ },
\ 'cmd_addplot' : {
\ 'cmds' : [
\ 'addplot[+3]?',
\ ],
\ },
\}
g:vimtex_fold_bib_enabled
Use this option to enable/disable folding in .bib files. When enabled, VimTeX will set the following options for .bib files:setlocal foldmethod=expr
setlocal foldexpr=vimtex#fold#bib#level(v:lnum)
setlocal foldtext=vimtex#fold#bib#text()Note: The default value is the same as g:vimtex_fold_enabled. Thus, it suffices to enabled folds for tex files to also enable for bib files. But if you want to enable in tex files but keep bib folding disabled, then you must set this option to 0 or v:false.
Default value: g:vimtex_fold_enabled
g:vimtex_fold_bib_max_key_width
This option is used to specify a length to truncate identifiers (e.g.
@article{Key}) to, in the foldtext for bib files. The default of 0 indicates no truncation, i.e. VimTeX will parse the bib file to determine the longest such identifier and align all foldtext titles such that they come after the identifiers.If you have a handful of cite keys that are exceptionally long, setting this manually may be useful as it makes sure that there is some space for the titles in the foldtext.
Default value: 0
g:vimtex_format_enabled
If enabled, VimTeX uses a custom |formatexpr| that should handle inline comments and environments. That is, if it is enabled, comments at end of lines will not be joined with the gq command, and environments like
equation will not be joined/changed.Default value: 0
g:vimtex_format_border_begin
g:vimtex_format_border_end
Regular expressions that define the "borders" of a region that should be formatted. The defaults should be more or less OK for most people, but some people may be interested in adjusting to handle more complex LaTeX code.
Default value: See source in
/autoload/vimtex/options.vimg:vimtex_grammar_textidote
This option is used to configure the
textidote grammar and document checker, see vimtex-grammar-textidote. It is a dictionary with the following keysjar
The path to textidote.jar. This key must be defined if you want to use
the TeXtidote wrapper! Please note that if one installs textidote with
a package manager e.g. in some common Linux distributions, the .jar
file might be missing. If so, it should be possible to download it
manually.args
Specify arguments to be passed to the TeXtidote grammar checker.Default:
let g:vimtex_grammar_textidote = {
\ 'jar': '',
\ 'args': '',
\}
g:vimtex_grammar_vlty
This option is used to configure the vlty grammar checker. This checker relies on the Python package YaLafi in combination with the proofreading software LanguageTool (see vimtex-grammar-vlty for more details). The option is a dictionary with the following keys :lt_directory
Path to the LanguageTool software, if installed manually.lt_command
Name of LanguageTool executable, if installed via package manager. Note
that this has precedence over lt_directory!lt_disable
lt_enable
lt_disablecategories
lt_enablecategories
Options for LanguageTool that control application of rules and rule
categories. For more info, see:
http://wiki.languagetool.org/command-line-optionsserver
Specify whether an HTTP server should be used. This may be faster for
short texts. Possible values are:no Do not use a server.
my Use a local LanguageTool server. If not yet running, it is
started.
lt Contact the Web server provided by LanguageTool. In this case,
no local installation is necessary. Please see the following page
for conditions and restrictions:
https://dev.languagetool.org/public-http-apishell_options
Pass additional options to YaLafi, e.g., --equation-punctuation displ;
for more info, see:
https://github.com/torik42/YaLafishow_suggestions
If set to 1, then LanguageTool's replacement suggestions are included
in the quickfix or location-list messages.encoding
Encoding of the (La)TeX source file. For default value auto, the
encoding is taken from |fileencoding| or |encoding|.Default:
let g:vimtex_grammar_vlty = {
\ 'lt_directory': '~/lib/LanguageTool',
\ 'lt_command': '',
\ 'lt_disable': 'WHITESPACE_RULE',
\ 'lt_enable': '',
\ 'lt_disablecategories': '',
\ 'lt_enablecategories': '',
\ 'server': 'no',
\ 'shell_options': '',
\ 'show_suggestions': 0,
\ 'encoding': 'auto',
\}
g:vimtex_imaps_enabled
Use this option to disable/enable the insert mode mappings.Default value: 1
g:vimtex_imaps_leader
The default leader key for insert mode mappings.
Default value: "`"
g:vimtex_imaps_disabled
A list of mappings to disable. That is, any left-hand side that matches a string in this list will not be mapped to its corresponding right-hand side. This may be used to selectively disable one or more from the default list of mappings.
Default value: []
g:vimtex_imaps_list
The list of mappings to generate on start up. The list of activated mappings can be viewed with :VimtexImapsList.
Default value: See
/autoload/vimtex/options.vim (it's a long list)g:vimtex_include_indicators
VimTeX will recognize included files for a lot of different purposes. Most of these come from e.g.
\input{file} or \include{file}. This option allows to add more commands that are used to include files, e.g. custom macros.Note: This option is read during initialization of VimTeX, and so it must be set early. I.e., it can not be set in
after/ftplugin/tex.vim.Default value:
['input', 'include']g:vimtex_include_search_enabled
VimTeX sets 'includeexpr' to recognize included files. If a file isn't found in the current directory, VimTeX uses
kpsewhich to search for it in the system TeX distribution. If the 'complete' option includes "i", invoking keyword completion with i_CTRL-N will search included files for completion possibilities. In this case, there may be a lot of calls to kpsewhich while scanning for included files during the first invocation of keyword completion, and this may introduce a significant delay. Subsequent keyword completions should be faster, as the calls to kpsewhich are cached.This option allows to disable searching for included files with
kpsewhich, and with that prevent the above explained delay.Default value: 1
g:vimtex_indent_enabled
Use this option to disable/enable VimTeX indentation.
Default value: 1
g:vimtex_indent_bib_enabled
Use this option to disable/enable VimTeX indentation of bibliography files.
Default value: 1
g:vimtex_indent_conditionals
This is a dictionary that defines regexes for indenting conditionals. Set it to an empty dictionary to disable this type of indentation.
Default value:
let g:vimtex_indent_conditionals = {
\ 'open': '\v%(\\newif)@<!\\if%(f>|field|name|numequal|thenelse|toggle)@!',
\ 'else': '\\else\>',
\ 'close': '\\fi\>',
\}
g:vimtex_indent_delims
A dictionary that specifies how to indent delimiters. The dictionary has four keys:open
List of regexes for opening delimiters that should add indents.close
List of regexes for closing delimiters that should reduce indents.close_indented
Set this to 1 if you want the line with the closing delimiter to stay
indented.include_modified_math
Set this to 0 if you do not want modified math delimiters such as
\left( and \right) to add/reduce indents.Note: VimTeX does not allow indents for parentheses only in math mode or any similar kind of context aware delimiter indents.
Note: If one of the keys of the dictionary is not specified, the default value is assumed.
Default value:
let g:vimtex_indent_delims = {
\ 'open' : ['{'],
\ 'close' : ['}'],
\ 'close_indented' : 0,
\ 'include_modified_math' : 1,
\}
g:vimtex_indent_ignored_envs
List of environments that should not add/reduce indentation.Note: Each item is interpreted as a regular expression that is combined into very magic regexes like
\v<%(document|...)>, see /\v.Default value: ['document']
g:vimtex_indent_lists
List of environments that act like lists with
\item entries.Note: Each item is interpreted as a regular expression that is combined into regexes like
\\begin{\%(itemize|description|...\).Default value: [
\ 'itemize',
\ 'description',
\ 'enumerate',
\ 'thebibliography',
\]
g:vimtex_indent_on_ampersands
By default, VimTeX will align on
leading ampersands e.g. in math aligned environments or in tabular environments. If this feature is not wanted it may be disabled through this option.Note: To get a more advanced tabular like alignment feature, you may be interested in something like |vim-easy-align|: https://github.com/junegunn/vim-easy-align
Default value: 1
g:vimtex_indent_tikz_commands
Use this option to disable/enable VimTeX indentation of multi-line commands in TikZ pictures.
Default value: 1
g:vimtex_main_choose_first
When VimTeX cannot unambiguously determine the main project file, it will by default prompt the user to select the appropriate main file from a list of candidates (see vimtex-multi-file and |g:vimtex_main_filename|). This prompt blocks until the user makes a selection.
If this option is enabled, VimTeX will instead silently pick the first candidate without prompting. This may be useful e.g. in automated or non-interactive contexts where the blocking prompt is undesired.
Default value: 0
g:vimtex_mappings_enabled
Control whether or not to load the default mappings.
Default value: 1
g:vimtex_mappings_disable
A dictionary that can be used to disable specific mappings. The dictionary keys are the mapping modes, and the values are lists of default mappings that should be disabled. The following example will ensure that the default
tse and tsd mappings are disabled:let g:vimtex_mappings_disable = {
\ 'n': ['tse', 'tsd'],
\ 'x': ['tsd'],
\}Default value: {}
g:vimtex_mappings_override_existing
Control behaviour on mapping conflicts, in particular whether or not to override pre-existing mappings. By default, VimTeX does not override existing mappings. If this option is enabled, then VimTeX will override existing mappings on conflict.
Default value: 0
g:vimtex_mappings_prefix
The default prefix for
<localleader> based mappings.Default value:
'<localleader>l'g:vimtex_matchparen_enabled
Enable highlighting of matching delimiters.
Note: This is an improved version of matchparen. It should be possible to keep matchparen activated, which matches delimiters listed in 'matchpairs'. The VimTeX specific version will also match LaTeX specific delimiters, which is not possible with matchparen.
Note: If you think this feature is slow, see vimtex-faq-slow-matchparen.
Default value: 1
g:vimtex_motion_enabled
This option enables the motion mappings, see vimtex-motions. It also enables the highlighting of matching delimiters.
Default value: 1
g:vimtex_lint_chktex_ignore_warnings
A string variable of options to pass for
chktex to specify to ignore certain warning messages.Default value:
'-n1 -n3 -n8 -n25 -n36'g:vimtex_lint_chktex_parameters
A string variable of parameters to pass to
chktex.VimTeX will look for a configuration file at
$XDG_CONFIG_HOME/chktexrc, or $HOME/.config/chktexrc if $XDG_CONFIG_HOME is undefined. If this file exists, it is specified by default with --localrc=....Default value:
'--localrc=PATH/chktexrc' or '' (see above)g:vimtex_log_ignore
A list of regexes to filter info, warning, and error messages. If a logged message matches any of the regexes in this list, the message will not be printed to screen.
Note: All messages may still be viewed with :VimtexLog.
Default: []
g:vimtex_log_verbose
Whether or not to print messages to screen. Should generally be on, but may be turned off e.g. for debugging or testing purposes.
Default: 1
g:vimtex_quickfix_enabled
Use this option to disable/enable the quickfix integration.
Default value: 1
g:vimtex_quickfix_method
This option sets the quickfix method. The following methods are available:
latexlog
This is the standard method which parses the normal LaTeX output.pplatex
Uses pplatex (https://github.com/stefanhepp/pplatex) to parse the LaTeX
output file. pplatex is a command line utility used to prettify the
output of the LaTeX compiler.Note:
pplatex and pulp require that -file-line-error is NOT passed to the LaTeX compiler. g:vimtex_compiler_latexmk will be updated automatically if one uses latexmk through VimTeX. However, if one uses other compiler methods, either through VimTeX (see g:vimtex_compiler_method) or externally, this requirement must be ensured by the user.Default value:
'latexlog'g:vimtex_quickfix_blgparser
This option controls the parsing of
blg log files (created by bibtex or biber) for warnings and errors. The option is a dictionary with the following keys:disable
Disable the parsing of blg entries.Default value: {}
g:vimtex_quickfix_autojump
This option controls if vim should automatically jump to the first error whenever the quickfix window is opened.
Note: This option does not go well with continuous compilation and callbacks, since the callbacks will open the quickfix window if there are errors. Thus I recommend to keep it disabled for continuous compilation, and rather enable it if one prefers single shot compilations.
Default value: 0
g:vimtex_quickfix_ignore_filters
This option allows to provide a list of regular-expressions for filtering out undesired errors and warnings. This works regardless of which quickfix method is enabled.
The following example will ignore any messages that match "Marginpar on page":
" Disable custom warnings based on regexp
let g:vimtex_quickfix_ignore_filters = [
\ 'Marginpar on page',
\]Default: []
g:vimtex_quickfix_mode
This option controls the behaviour of the quickfix window in case errors and/or warnings are found. The recognized options are:
Value Effect
0 The quickfix window is never opened/closed automatically.
1 The quickfix window is opened automatically when there are errors,
and it becomes the active window.
2 The quickfix window is opened automatically when there are errors,
but it does not become the active window.Note: The quickfix window will only be opened automatically if the compiler is set to
continuous mode and has callbacks enabled, or if continuous mode is disabled.Default value: 2
g:vimtex_quickfix_autoclose_after_keystrokes
If set to value greater than zero, then the quickfix window will close after this number of motions (i.e. CursorMoved and CursorMovedI events). This is most useful if one sets g:vimtex_quickfix_mode to 2, in which case this option allows one to continue editing and removing the distraction of the quickfix window automatically.
Note: The count is reset when the quickfix window is entered.
Default value: 0
g:vimtex_quickfix_open_on_warning
Control whether or not to automatically open the quickfix window in case there are warning messages and no error messages.
Default value: 1
g:vimtex_subfile_start_local
This option allows to specify that one should start with the local file for subfile'd documents instead of the main project file. See vimtex-subfiles for further info.
Default value: 0.
g:vimtex_syntax_enabled
Use this option to disable/enable syntax highlighting as provided by VimTeX.
Default value: 1.
g:vimtex_syntax_conceal
A dictionary for specifying which core conceal features to activate. This mostly implies concealing particular elements with a replacement unicode character. For more info, see vimtex-syntax-conceal. To disable all conceal features in one go, use g:vimtex_syntax_conceal_disable.
The following keys are available:
accents
Conceal accented characters, e.g. \^a --> â.ligatures
Conceal ligatures such as \aa --> å and '' --> “.cites
Conceal LaTeX cite commands such as \citet[...]{ref00}. The conceal
style is specified by g:vimtex_syntax_conceal_cites.fancy
Some extra fancy replacements, e.g. \item --> ○.texTabularChar
Conceal tabular line breaks, i.e. \\ --> ⏎.spacing
Conceal spacing commands such as \quad and \hspace{1em} in both
normal mode and math mode.greek
Replace TeX greek letter commands into the equivalent unicode greek
letter.math_bounds
Conceal the TeX math bounds characters: pairs of $ and $$, \( ...
\), and \[ ... \].math_delimiters
Replace possibly modified math delimiters with a single unicode
letter. Modified means delimiters prepended with e.g. \left or
\bigl. As an example, this will perform the replacement\Biggl\langle ... \Biggr\rangle --> 〈 ... 〉math_fracs
Replace some simple fractions like \frac 1 2 --> ½.math_super_sub
Replace simple math super and sub operators, e.g. x^2 --> x².math_symbols
Replace various math symbol commands to an equivalent unicode character.
This includes quite a lot of replacements, so be warned!sections
Conceal \(sub)*section commands. The titles are replaced with Markdown
style ATX headers, e.g.:\section{Test} --> # Test
\subsection{Test} --> ## Teststyles
Conceal the LaTeX command "boundaries" for italicized and bolded style
commands, i.e. \emph, \textit, and \textbf. This means that one
will see something like:\emph{text here} --> text hereDefault value:
let g:vimtex_syntax_conceal = {
\ 'accents': 1,
\ 'ligatures': 1,
\ 'cites': 1,
\ 'fancy': 1,
\ 'texTabularChar': 1,
\ 'spacing': 1,
\ 'greek': 1,
\ 'math_bounds': 1,
\ 'math_delimiters': 1,
\ 'math_fracs': 1,
\ 'math_super_sub': 1,
\ 'math_symbols': 1,
\ 'sections': 0,
\ 'styles': 1,
\}
g:vimtex_syntax_conceal_disable
This option allows to disable all conceal features at once. For more fine tuned control, use g:vimtex_syntax_conceal and g:vimtex_syntax_packages.Default value: 0
g:vimtex_syntax_conceal_cites
A simple dictionary to control how citation conceal should work. It has three keys:
type
Specify the type of concealment. There are two options, and the
difference is best explained by example: Value LaTeX Concealed
----- ----- ---------
'icon' \cite{Knuth1981} 📖
'brackets' \cite{Knuth1981} [Knuth1981]icon
Specify an icon for icon conceal. This must be a single (possibly
multibyte) character.verbose
Specify how much to conceal in bracket mode (type set to 'bracket').
The following table shows how the concealed result depending on the
'verbose' value for \cite[Figure 1]{Knuth1981}:Default value:
let g:vimtex_syntax_conceal_cites = {
\ 'type': 'brackets',
\ 'icon': '📖',
\ 'verbose': v:true,
\}
g:vimtex_syntax_custom_cmds
A list of "simple" commands for which to apply custom styling. This includes bolded or italicized text arguments, conceals, or similar - see the below keys. Each command is expected to be of the following type:\cmdname[optional]{argument}It is important to be aware that these customizations will be applied on top of the existing syntax rules. These may therefore override both the core syntax rules and extensions from syntax packages.
Each element in the list must be a dictionary with the following keys:
name
Default: Undefined (REQUIRED)
The command to highlight (cmdname). This is also for defining the
syntax group names.cmdre
Default: Undefined
If this is defined, then it is used instead of name for matching the
cmdname part. It is interpreted as a regular expression with "very
magic" mode activated (see /\v). For example, you need to use >
instead of \> for end-of-word atom (|\>|).mathmode
Default: v:false
If true, then the command is a math mode command.concealchar
Default: Undefined
Specify a conceal character for the \cmdname part. With this, one can
easily create simple rules to display e.g. \R as ℝ (see config
example below).optconceal
Default: Same as conceal key
If true, the option group [optional] is concealed.argstyle
Default: Undefined.
Can be set to apply styling to the command argument by linking the
argument syntax group to one of the texStyle highlight groups (see
Table 5 in vimtex-syntax-reference). The following options are
available:
* bold
* ital
* under
* boldital
* boldunder
* italunder
* bolditalunder Note: Fine grained control is of course also possible. Each defined
command gets one or more match groups, e.g. if you create a custom
command named
foo, then it will usually have these groups:
texCmdCFoo, texCFooOpt, texCFooArg. One may then customize
the highlights as explained in vimtex-syntax-core.argspell
Default: v:true
Specify this as v:false or 0 to disable spell checking of the command
argument.arggreedy
Default: v:false
If v:true or 1, the syntax rule will "eat" as many arguments as
possible: \cmdname[opt]{arg1}{arg2}...{argn}nextgroup
Default: Undefined
This is a string that, if defined and not empty, specifies
a comma-separated list of possible next syntax groups.hlgroup
Default: Undefined
A string that can be used to indicate the target highlight group of the
command (\cmdname).A couple of examples may be helpful: The first in the following list shows how to use bolded style on a custom vector macro such as
\vct{v}. The second example shows how to conceal \R with ℝ; notice the use of cmdre and the end-of-word atom > to ensure it does not also match e.g. \Re. The third example shows how one may use the nextgroup key, and the fourth shows how to define a command whose argument should not be spell checked.let g:vimtex_syntax_custom_cmds = [
\ {'name': 'vct', 'mathmode': 1, 'argstyle': 'bold'},
\ {'name': 'R', 'cmdre': 'R>', 'mathmode': 1, 'concealchar': 'ℝ'},
\ {'name': 'mathnote', 'mathmode': 1, 'nextgroup': 'texMathTextArg'},
\ {'name': 'nospell', 'argspell': 0},
\]Default value: []
g:vimtex_syntax_custom_cmds_with_concealed_delims
This option works exactly as g:vimtex_syntax_custom_cmds, except it is used specifically to add conceals with custom replacement characters for single- and multi commands, e.g.
\cmd1{argument}
\cmd2 {argument1} {argument2}Each element in the list must be a dictionary with the following keys. Most of these keys are documented here: g:vimtex_syntax_custom_cmds. Only the unique keys are described in full here.
name
Default: Undefined (REQUIRED)nargs
Default: 1
Specify whether the command has 1 or more arguments.cchar_open
Default: Undefined
Specify single letter replacement for the head (\cmdname{). If left
undefined, the head is fully concealed.cchar_mid
Default: Undefined
Note: Only relevant if nargs > 1
Specify single letter replacement for the mid (}{). If left
undefined, the mid is fully concealed.cchar_close
Default: Undefined
Specify single letter replacement for the tail (}). If left
undefined, the tail is fully concealed.cmdre
Default: Undefinedmathmode
Default: v:falseargstyle
Default: Undefined.argspell
Default: v:truehlgroup
Default: UndefinedAn example may be elucidating. Given the following configuration:
let g:vimtex_syntax_custom_cmds_with_concealed_delims = [
\ {'name': 'ket',
\ 'mathmode': 1,
\ 'cchar_open': '|',
\ 'cchar_close': '>'},
\ {'name': 'binom',
\ 'nargs': 2,
\ 'mathmode': 1,
\ 'cchar_open': '(',
\ 'cchar_mid': '|',
\ 'cchar_close': ')'},
\]We should now see the following effect in a document:
$\ket{x}$ and $\binom{n}{k}$
% will now look like this:
|x> and (n|k)Default value: []
g:vimtex_syntax_custom_envs
A list of environments for which to apply custom styling. This allows to define custom math environments or to specify custom environments for which to load nested syntaxes. The latter is relevant e.g. if you use the
\lstnewenvironment from the listings package.Each environment is expected to look like this:
\begin{env_name}[optional argument]
…
\end{env_name}It is important to be aware that these customizations will be applied on top of the existing syntax rules. They may therefore override both the core syntax rules and extensions from syntax packages.
Each element in the list must be a dictionary with the following keys:
name
Default: Undefined (REQUIRED)
The name of the environment to highlight (env_name). The string is
used as a regular expression.region
Default: tex{Name}Zone
The syntax group used to match the defined region.math
Default: v:false
If true, then the environment is a math region.starred
Default: v:false
Whether the corresponding starred environment should also be matched.transparent
Default: v:false
If the matched syntax region should be transparent: |syn-transparent|.opts
Default: Undefined
A string with additional options that will be passed to the syntax
region` command (see |syn-region|).contains
Default: Undefined
A comma-separated string of syntax groups that should be contained in
within the matched region (see |syn-contains|).nested
Default: Undefined
This can be either a string or a dictionary: String: specify nested syntax to load inside the environment
Dictionary: specify "predicated" nested syntaxes (more flexible)
The dictionary uses the target syntax as the key and the "predicate" as
the value. This predicate is a string that must be contained within the
optional argument. See below for an example.
Notice that one should also be aware of g:vimtex_syntax_nested.
The following example creates three rules. The first creates the environment
MyMathEnv that opens a new math environment. The second creates a python_code environment that applies nested Python syntax rules in the environment region. The third rule creates a code environment that will open nested syntax regions if the optional group contains the specified predicate strings.let g:vimtex_syntax_custom_envs = [
\ {
\ 'name': 'MyMathEnv',
\ 'math': v:true
\ },
\ {
\ 'name': 'python_code',
\ 'region': 'texPythonCodeZone',
\ 'nested': 'python',
\ },
\ {
\ 'name': 'code',
\ 'region': 'texCodeZone',
\ 'nested': {
\ 'python': 'language=python',
\ 'c': 'language=C',
\ 'rust': 'language=rust',
\ },
\ },
\]Default value: []
g:vimtex_syntax_match_unicode
Whether to highlight unicode characters. If enabled, it will match unicode greek letters as
texCmdGreek and a lot of other unicode symbols as texMathSymbol.Default value: v:true
g:vimtex_syntax_nested
A dictionary for configuring nested syntaxes. The following keys are available for configuration:
aliases
Holds a dictionary of aliases, such as mapping C to c. This is
useful e.g. because the Vim syntax files are case sensitive.ignored
Holds a dictionary of ignore lists for each language. This is useful to
ignore some groups that may conflict in e.g. the \begin{...} or
\end{...} part of the nested syntax regions.Default value:
let g:vimtex_syntax_nested = {
\ 'aliases' : {
\ 'C' : 'c',
\ 'csharp' : 'cs',
\ },
\ 'ignored' : {
\ 'sh' : ['shSpecial'],
\ 'bash' : ['shSpecial'],
\ 'cs' : [
\ 'csBraces',
\ ],
\ 'python' : [
\ 'pythonEscape',
\ 'pythonBEscape',
\ 'pythonBytesEscape',
\ ],
\ 'java' : [
\ 'javaError',
\ ],
\ 'haskell' : [
\ 'hsVarSym',
\ ],
\ }
\}
g:vimtex_syntax_nospell_comments
Set to 1 to disable spell checking in comments.Default value: 0
g:vimtex_syntax_packages
A dictionary for package specific syntax configuration. Each key represents a single package and the values are themselves configuration dictionaries. All packages share the following options:
load Specify when to load the package syntax addon.
0 = disable this syntax package
1 = enable this syntax package if it is detected (DEFAULT)
2 = always enable this syntax packageThe following is a list of packages with additional options or packages that deviate from the above specified defaults. Notice that conceal options are affected by g:vimtex_syntax_conceal_disable.
amsmath
load is 2 by default
conceal whether to enable conceal; enabled by defaultbabel
conceal whether to enable conceal; enabled by defaultfontawesome5
conceal whether to enable conceal; enabled by default. Notice that
the only point of this package is to apply conceals to
fontawesome commands. Thus, disabling conceal here is
equivalent to disabling the package.hyperref
conceal whether to enable conceal; enabled by defaultrobust_externalize
presets list of presets and target syntaxes for things like
\begin{CacheMeCode}{bash} ... \end{CacheMeCode}.Default:
let g:vimtex_syntax_packages = {
\ 'amsmath': {'conceal': 1, 'load': 2},
\ 'babel': {'conceal': 1},
\ 'hyperref': {'conceal': 1},
\ 'fontawesome5': {'conceal': 1},
\ 'robust_externalize': {
\ 'presets': [
\ ['bash', 'bash'],
\ ['python', 'python'],
\ ['gnuplot', 'gnuplot'],
\ ['tikz', '@texClusterTikz'],
\ ['latex', 'TOP'],
\ ],
\ },
\}
g:vimtex_texcount_custom_arg
Option that makes it possible to add custom arguments to texcount for :VimtexCountWords and :VimtexCountLetters.Default value:
''g:vimtex_text_obj_enabled
Use this option to disable the text object mappings.
Default value: 1
g:vimtex_text_obj_linewise_operators
List of operators that will act linewise on the delimiter text objects (i.e.
ie/ae, i$/a$, and id/ad). Note, for inline regions the operators will not act linewise, since that would lead to side effects.Default value:
['d', 'y']g:vimtex_text_obj_variant
Select text object variants for command and environment text objects. The choice is either VimTeX or |targets.vim|. Possible configuration options are:
'auto' (select 'targets' if |targets.vim| is installed)
'vimtex'
'targets'
When using
'targets', the following additional text object kinds are available:Prefix
I and A instead of i and a for excluding inner whitespace or including outer whitespace, respectively.
Modifier
n and l for next or previous (mnemonic: last). For more details, see
doc/targets-textobj-cheatsheet.md.Default value:
'auto'g:vimtex_toggle_fractions
Specify rules for toggling fractions with <plug>(vimtex-cmd-toggle-frac), which is mapped to
tsf by default.Default value:
let g:vimtex_toggle_fractions = {
\ 'INLINE': 'frac',
\ 'frac': 'INLINE',
\ 'dfrac': 'INLINE',
\}
g:vimtex_toggle_star_cmds
This is a whitelist of the commands whose star may be toggled with <plug>(vimtex-cmd-toggle-star) (mapped to tsc by default) and, for commands, <plug>(vimtex-cmd-toggle-star-agn). The toggle is a no-op for any command whose name is not matched by the whitelist. This prevents toggling the star of commands and math literals that have no meaningful starred form, e.g. \sin, \frac, or \left.Each list entry is a /\v "very magic" pattern that is matched against the full command name (without the leading backslash or a trailing star) and case-insensitively. To extend the whitelist, copy the default and append your own entries.
If the list is empty, no filtering is done and the star of any command may be toggled.
Default value:
let g:vimtex_toggle_star_cmds = [
\ 'part',
\ '%(sub)*section',
\ '%(sub)*paragraph',
\ '[vh]space',
\ '\w*cite\w*',
\ '\w*ref',
\ '%(re)?newcommand',
\ 'providecommand',
\ 'DeclareRobustCommand',
\ 'DeclareMathOperator',
\ '%(re)?newenvironment',
\ 'includegraphics',
\ 'verb',
\]g:vimtex_toc_enabled
Use this option to disable/enable table of contents (ToC).
Default value: 1
g:vimtex_toc_config
This is a dictionary that can be used to configure the ToC. Each key specifies a configuration option that can be changed. For configuration of specific matchers, see g:vimtex_toc_config_matchers.
In the following, the possible configuration keys are explained briefly and the default values are indicated.
name : Table of contents (VimTeX)
The name of the ToC buffer.mode : 1
The ToC display mode, one of:
1: Separate window.
2: Separate window and location list.
3: Location list (and don't open it).
4: Location list (and open it).fold_enable : 0
Whether to enable folding in the ToC window.fold_level_start : -1
The starting fold level. The value -1 indicates that the start level is
the same as the tocdepth value.hide_line_numbers : 1
If enabled, then line numbers will be hidden in the ToC window by
setting |nonumber| and |norelativenumber| locally.hotkeys_enabled : 0
Set to 1 to enable individual hotkeys for ToC entries.hotkeys : abcdegijklmnopuvxyz
A string of keys that are used to create individual hotkeys.hotkeys_leader : ;
The hotkey leader. Set to empty string to disable the leader.indent_levels : 0
Set to 1 to indent the section levels in the ToC window.layers : Undefined
layer_status : Dictionary{ 'content': 1,
'label': 1,
'todo': 1,
'include': 1 } The initial state of the layers (1 for active, 0 for inactive). The
layers key may be used as a shorthand: it accepts a list of layers
that should be active.layer_keys : Dictionary{ 'content': 'C',
'label': 'L',
'todo': 'T',
'include': 'I'} Specify hotkeys for enabling/disabling the different layers.
resize : 0
Whether or not to automatically resize vim when index windows are
opened. Note: This option makes sense if the index window is vertically split.
refresh_always : 1
Set to 0 to manually refresh ToC entries. This may be useful for very
large projects where generating the ToC entries becomes slow. It may be useful to combine manually refreshing with a BufWritePost
autocommand, e.g.:
augroup VimTeX
autocmd!
autocmd BufWritePost *.tex call vimtex#toc#refresh()
augroup END Or, if preferred, one may use a mapping such as:
nnoremap <silent> <localleader>lf :call vimtex#toc#refresh()show_help : 1
Whether to display help text on top when the ToC is opened. If this is
disabled, we only show "Press h to toggle help text.".show_numbers : 1
Set whether or not to show section numbers in ToC.split_pos : vert leftabove
Define where index windows should be opened. This is a string that
contains either the word "full" to open in the current window, or
a position command. Use :vert if a vertical split is desired, and one
of :leftabove, :rightbelow, :topleft, and :botright to specify
the desired split position.split_width : 30
For vertically split windows: Set width of index window.tocdepth : 3
Define the depth of section levels to display. This attempts to mimic
the corresponding latex variable tocdepth. For more info, see:
https://en.wikibooks.org/w/index.php?title=LaTeX/Document_Structure Note: This will also change the width of the number column according to
the space needed to show the section numbers.
todo_sorted : 1
Whether or not to sort the TODOs at the top of the ToC window.g:vimtex_toc_config_matchers
This is a dictionary that can be used to configure the built-in ToC matchers. See below for a specification of the ToC matcher "objects" and the various keys that can be defined/changed (toc_matcher_specification).
To configure/alter a built-in matcher, one can do this:
let g:vimtex_toc_config_matchers = {
\ 'MATCHER1': {OPTIONS},
\ 'MATCHER2': {OPTIONS},
\}The available options are described in toc_matcher_specification. Please note that the built-in matchers should generally just work well for most people. However, this option allows at least two useful things: to disable a built-in matcher and to change the priority of a built-in matcher. The following is a full example that shows how this could be used:
let g:vimtex_toc_config_matchers = {
\ 'beamer_frame': {'disable': 1},
\ 'todo_fixme': {'priority': -1},
\ 'index': {'title': 'MyFancy Index Title'},
\}Note: The available built-in matchers are defined in separate files under
/autoload/vimtex/parser/toc/*.vim.Default value: {}
toc_matcher_specification
A ToC matcher is defined as a Dictionary where the possible keys are specified below. In order to write a matcher, one should also be aware of the
context argument that is passed to the matcher functions, as well as the specification of the toc entry return value. However, since this kind of customization is advanced I refer users to the source file for further specification of these objects. In particular, see the function s:toc.parse(...) in /autoload/vimtex/toc.vim.re
Type: String
Required: yes
This specifies a regular expression that should match the current line
for the desired ToC entry.prefilter_re
Type: String
Required: maybe (this or prefilter_cmds should be specified)
This specifies a regular expression that must match the current line for
the desired ToC entry. This is used as a prefilter to make things
faster, and it does not need to be a perfect match. The re key may
often be a complex and therefore slow regular expression. This key
should represent a simple and fast regular expression that may match
more than the desired entry.prefilter_cmds
Type: List of String
Required: maybe (this or prefilter_re should be specified)
This is similar to prefilter_re, except it specifies a list of command
names (regular expressions). For instance, it should contain todo for
a ToC matcher for \todo commands.priority
Type: expr-number (default: 0)
Required: no
Priority is used for sorting the ToC matchers. High priority matchers
will be tried first, and only one matcher will match a given line. Note
that the built-in matchers have priority values between 0 and 2.in_preamble
Type: 0 or 1 (default: 0)
Required: no
If the entry may appear in the preamble.in_content
Type: 0 or 1 (default: 1)
Required: no
If the entry may appear in the main content.title
Type: String
Required: no
If the matcher does not have a get_entry key, then it will use
a simple, general matcher function to generate the entry. In this case,
the title key should be specified to give the title of the ToC entry.get_entry
Type: Dictionary-function
Arguments: context
Returns: toc entry
Required: no
This is the general way to define ToC entries. It allows to define the
ToC entry based on the context. See /autoload/vimtex/parser/toc.vim
for examples on how to use this.continue
Type: Dictionary-function
Arguments: context
Returns: toc entry
Required: no
Some entries may be specified over several lines, in which case this key
becomes necessary in combination with the get_entry key. See the built
in s:matcher_section matcher for an example on how to use this.name
Type: String
Required: no
Mostly for making it easier to debug a specific matcher. Without a name,
the matcher will be registered with a semi random numbered name like
custom1.g:vimtex_toc_custom_matchers
This option is a list of custom ToC matchers, see toc_matcher_specification.
As an example, one can use this option to add ToC entries for a custom environment. Say you have defined an environment
mycustomenv, then instances of this environment could be added to the ToC with the following configuration:let g:vimtex_toc_custom_matchers = [
\ { 'title' : 'My Custom Environment',
\ 're' : '\v^\s*\\begin\{mycustomenv\}' }
\]Default value: []
g:vimtex_toc_todo_labels
Dictionary of keywords that should be recognized in comments for the todo layer. The values represent the labels used in the ToC.
Default value:
{'TODO': 'TODO: ', 'FIXME': 'FIXME: '}g:vimtex_toc_show_preamble
Whether to include the preamble in the ToC.
Default value: 1
g:vimtex_ui_method
A dictionary that specifies the backend for various input methods. The method names are the keys of the dictionary and the backend choices are the values.
The available methods:
confirm: Confirm dialogues (e.g. before opening documentation)
input: Input dialogues (e.g. for <plug>(vimtex-cmd-change))
select: Selection dialogues (e.g. to select when there are multiple
choices for documentation)
The available backends:
nvim: Popup menu created with neovim APIs.
vim: Currently there is no Vim-specific implementation. Setting the
backend to "vim" will currently fallback to "legacy".
legacy: Legacy backends that are created by :echoing the menus and
using input() and similar for getting input.
Default:
" On neovim
let g:vimtex_ui_method = {
\ 'confirm': 'nvim',
\ 'input': 'nvim',
\ 'select': 'nvim',
\}
" Otherwise
let g:vimtex_ui_method = {
\ 'confirm': 'legacy',
\ 'input': 'legacy',
\ 'select': 'legacy',
\}
g:vimtex_view_enabled
Use this option to disable/enable the VimTeX viewer interface.Default value: 1
g:vimtex_view_automatic
If enabled, the viewer should open automatically when compilation has started in
continuous mode and if callback is enabled, or if continuous mode is disabled. This should work for the following compilers: * vimtex-compiler-latexmk * vimtex-compiler-latexrun * vimtex-compiler-arara * vimtex-compiler-tectonicDefault value: 1
g:vimtex_view_use_temp_files
When enabled, this option specifies to copy the
.pdf and .synctex.gz files after successful compilation. The viewer will use the copies, which helps to avoid issues such that as the pdf becoming unavailable during compilation.The copies are named similar to the original files with a
_ prefix.Note: This option is only relevant for the
latexmk compiler backend. latexrun already ensures that the output file is updated only after the compilation is completed.Default value: v:false
g:vimtex_view_forward_search_on_start
If disabled, the first invocation of the viewer will not perform a forward search to the current cursor position.
Note: This option is only relevant when g:vimtex_view_method is set to either
mupdf, zathura. See also the specific viewer sections for more info: vimtex-view-mupdf, vimtex-view-zathura.Default value: 1
g:vimtex_view_reverse_search_edit_cmd
When working in a multi-file project, initiating inverse search (see vimtex-synctex-inverse-search) may require opening a file that is not currently open in a window. This option controls the command that is used to open files as a result of an inverse search.
Examples:
*
edit open buffer in current window
* tabedit open buffer in new tab page
* split split current window to open bufferDefault value:
editg:vimtex_view_method
Set the viewer method. By default, a generic viewer is used through the general view method (e.g.
xdg-open on Linux).Possible values:
*
'general'
* 'galley' vimtex-view-galley
* 'mupdf' vimtex-view-mupdf
* 'skim' vimtex-view-skim
* 'zathura' vimtex-view-zathura
* 'zathura_simple' vimtex-view-zathura-simpleSee vimtex-view-configuration for more information on various popular viewers and on how to configure them.
Default:
generalg:vimtex_view_general_options
Set options for the specified general viewer. The options are parsed to substitute the following keywords:
@pdf Path to pdf file
@tex Path to tex file
@line Current line number
@col Current column numberFor some examples on how to use this, see
* vimtex-view-evince
* vimtex-view-okular
* vimtex-view-sumatrapdf
Default value:
'@pdf'g:vimtex_view_mupdf_options
g:vimtex_view_zathura_options
Set options for mupdf and Zathura, respectively. See also: * vimtex-view-mupdf * vimtex-view-zathura * vimtex-view-zathura-simple
Default value:
''g:vimtex_view_general_viewer
Use generic viewer application.
Default value:
Linux:
xdg-open
macOS: open
Windows: SumatraPDF or mupdf if available, else start ""g:vimtex_view_mupdf_send_keys
A string of keys that will be sent to MuPDF just after the PDF file has been opened.
Default value:
''g:vimtex_view_sioyek_exe
The name or path of the Sioyek executable. The default should usually work, but in some cases it can be useful or necessary to specify the executable directly. E.g., if one downloads a release version, it may be named something like
'Sioyek-x86_64.AppImage'. Unless the executable location is available in PATH one must use an absolute path here.Default value:
'sioyek'g:vimtex_view_sioyek_options
Set additional command-line options for Sioyek (vimtex-view-sioyek). This can e.g. be used to add the
--reuse-instance or --reuse-window option, which some users prefer.Default value:
''g:vimtex_view_galley_activate
Set this option to 1 to make Galley have focus and be moved to the foreground after command :VimtexView or compiler callback. When 0, Galley is opened and updated in the background.
Default value: 0
g:vimtex_view_galley_sync
Set this option to 1 to make Galley perform a forward search after successful compilation.
Default value: 0
g:vimtex_view_skim_activate
Set this option to 1 to make Skim have focus after command :VimtexView in addition to being moved to the foreground.
Default value: 0
g:vimtex_view_skim_sync
Set this option to 1 to make Skim perform a forward search after successful compilation.
Default value: 0
g:vimtex_view_skim_reading_bar
Set this option to 1 to highlight current line in PDF after command :VimtexView or compiler callback.
Default value: 0
g:vimtex_view_skim_no_select
Set this option to 1 to prevent Skim from selecting the text after command :VimtexView or compiler callback.
Default value: 0
g:vimtex_view_texshop_activate
Set this option to 1 to make TeXShop have focus after command :VimtexView in addition to being moved to the foreground.
Default value: 0
g:vimtex_view_texshop_sync
Set this option to 1 to make TeXShop perform a forward search after successful compilation.
Default value: 0
g:vimtex_view_zathura_check_libsynctex
Check on startup if Zathura is compiled with libsynctex. This is done by default because Zathura on some systems is compiled without libsynctex support, in which case forward and inverse search will not work. When this is the case, the startup check will provide a notification to the user.
If this option is set to 0 or v:false, then the check is skipped.
Default value: 1
g:vimtex_view_zathura_use_synctex
Set to 0 or v:false to disable synctex for the Zathura viewer. This can be useful e.g. for MacOS users who struggle with getting synctex to work properly (see vimtex-faq-zathura-macos).
Default value: 1
g:vimtex_callback_progpath
The path to the Vim/neovim executable. This is currently passed to Zathura and Sioyek for use with synctex callbacks; see vimtex-view-zathura and vimtex-view-sioyek.
You usually don't have to touch this variable; VimTeX will v:progpath if the option is not defined. But some people may use wrappers and similar to load Vim/neovim, in which case one may want to specify the executable directly.
Default value: Undefined
$VIMTEX_OUTPUT_DIRECTORY
This environment variable allows to specify the output directory of generated LaTeX files. If it exists and is a valid path, this path will be used as the output directory. This has two main use cases:
It allows to use a custom output directory for different projects.
It allows to specify an output directory for projects where one uses
compiler backends such as vimtex-compiler-arara. This makes it possible
to make e.g. the vimtex-view feature to work as expected if output
directories are used with arara.
Note: This will override
out_dir (and aux_dir) of options like g:vimtex_compiler_latexmk and g:vimtex_compiler_latexrun.COMMANDS
:VimtexContextMenu
:VimtexContextMenu Show a context menu on the item below cursor. See
vimtex-context-menu for more information.
:VimtexInfo
<plug>(vimtex-info)
:VimtexInfo Show information that is stored by VimTeX about the
current LaTeX project (available mostly for debug
purposes).
:VimtexInfo!
<plug>(vimtex-info-full)
:VimtexInfo! Show information that is stored by VimTeX about all
open LaTeX projects (available mostly for debug
purposes).
:VimtexDocPackage
<plug>(vimtex-doc-package)
:VimtexDocPackage Show documentation for packages. The command takes
one optional argument, which is the name of the
package to show docs for. If no argument is
supplied, it parses the command under the cursor and
opens the most relevant documentation.
:VimtexRefreshFolds
:VimtexRefreshFolds Refresh folds, see vimtex-zx.
:VimtexTocOpen
<plug>(vimtex-toc-open)
:VimtexTocOpen Open table of contents.
:VimtexTocToggle
<plug>(vimtex-toc-toggle)
:VimtexTocToggle Toggle table of contents.
:VimtexLog
<plug>(vimtex-log)
:VimtexLog Open a scratch buffer to show message log with
timestamps and traces from where the messages were
raised. To close the log buffer, one may press
q
or <esc>. :VimtexCompile
<plug>(vimtex-compile)
:VimtexCompile[!] [opts] If the compiler supports and is set to run in continuous mode, then this command works as a compiler toggle. If not, this command will run a single shot compilation.
If [!] is given, the command will only start
compilation, never stop it.
Arguments to the command will be passed on as options
when starting the compiler. This allows the user to
start the compiler with different options without
changing any configuration. That is, if the user uses
the latexmk backend, then adding any option argument
is equivalent to adding them to the
'options' key
of g:vimtex_compiler_latexmk. Note: Special items in the arguments will be
expanded as explained in |expandcmd|.
:VimtexCompileSS
<plug>(vimtex-compile-ss)
:VimtexCompileSS [opts] Start single shot compilation.
:VimtexCompileSelected
<plug>(vimtex-compile-selected)
:VimtexCompileSelected Compile the selected part of the current LaTeX file.
When used as a command, it takes a range, e.g.:
:start,end VimtexCompileSelected
When used as a normal mode mapping, the mapping
will act as an operator on the following motion or
text object. Finally, when used as a visual mode
mapping, it will act on the selected lines.
Note: This always works linewise!
The command compiles the selected text by copying it
to a temporary file with the same preamble as the
current file. It will be compiled similarly to
a single shot compile (see :VimtexCompileSS. If
there are errors, they will be shown in the quickfix
list.
One may specify a custom template with a template
file in which any (single!) line with the exact
content
%%% VIMTEX PLACEHOLDER will be
interchanged with the selected lines. This allows to
customize the preamble and surrounding content. The
template file should be named vimtex-template.tex
or <head>-vimtex-template.tex, where <head>
implies the head of the current file name with the
extension removed. E.g., for a file foo.tex, one
may specify a custom template
foo-vimtex-template.tex. This will have a higher
priority than vimtex-template.tex. :VimtexCompileOutput
<plug>(vimtex-compile-output)
:VimtexCompileOutput Open file where compiler output is redirected.
:VimtexStop
<plug>(vimtex-stop)
:VimtexStop Stop compilation for the current project.
:VimtexStopAll
<plug>(vimtex-stop-all)
:VimtexStopAll Stop compilation for all open projects in the
current vim instance.
:VimtexStatus
<plug>(vimtex-status)
:VimtexStatus Show compilation status for current project.
:VimtexStatus!
<plug>(vimtex-status-all)
:VimtexStatus! Show compilation status for all open projects in the
current vim instance.
:VimtexClean
<plug>(vimtex-clean)
:VimtexClean Clean auxiliary files.
Note: If compilation is running continuously in the
background (which is the default behaviour),
then this command will first temporarily stop
compilation, then execute the clean command,
and finally restart the compilation.
:VimtexClean!
<plug>(vimtex-clean-full)
:VimtexClean! As :VimtexClean, but also remove output files.
:VimtexErrors
<plug>(vimtex-errors)
:VimtexErrors Open quickfix window if there are errors or
warnings.
:VimtexView
<plug>(vimtex-view)
:VimtexView View
pdf for current project, perform forward
search if available. :VimtexReload
<plug>(vimtex-reload)
:VimtexReload Reload VimTeX scripts. This is primarily useful
when developing and debugging VimTeX itself.
:VimtexReloadState
<plug>(vimtex-reload-state)
:VimtexReloadState Reload the state for the current buffer.
:VimtexCountLetters
:VimtexCountWords
vimtex#misc#wordcount(opts)
:VimtexCountLetters Shows the number of letters/characters or words in
:VimtexCountWords the current project or in the selected region. The
count is created with
texcount through a call on
the main project file similar to:texcount -nosub -sum [-letter] -merge -q -1 FILE
Note: Default arguments may be controlled with
g:vimtex_texcount_custom_arg.
Note: One may access the information through the
function
vimtex#misc#wordcount(opts), where
opts is a dictionary with the following
keys (defaults indicated):'range' : [1, line('$')]
'count_letters' : 0/1
'detailed' : 0 If
detailed is 0, then it only returns the
total count. This makes it possible to use for
e.g. statusline functions. If the opts dict
is not passed, then the defaults are assumed. :VimtexCountLetters!
:VimtexCountWords!
:VimtexCountLetters! Similar to the un-banged versions
:VimtexCountWords! :VimtexCountLetters and :VimtexCountWords,
except these show separate reports for included
files. I.e. presents the result of:
texcount -nosub -sum [-letter] -inc FILE :VimtexImapsList
<plug>(vimtex-imaps-list)
:VimtexImapsList Show the list of insert mode mappings created by the
vimtex-imaps feature. The mappings are displayed
in a scratch buffer. Press
q or <esc> to close
the buffer. :VimtexToggleMain
<plug>(vimtex-toggle-main)
:VimtexToggleMain In general, VimTeX detects the main file for the
current LaTeX project and uses it for compilation
and many other features. However, in some cases it
may be useful to instead focus on the current file,
for instance in large projects. In such cases, one
can use :VimtexToggleMain to change which file to
use as the "current project". It is easy to toggle
back and forth, and both the "main project" and the
"local project" can be used simultaneously if
desired (e.g. for compilation).
Note: To compile the current file when it is part of
a larger project, one must of course include
a preamble and the
\begin/\end{document}! It is
possible to have a working preamble in every
file in a multi-file project with subfiles,
see vimtex-subfiles. See also
g:vimtex_subfile_start_local. :VimtexClearCache
:VimtexClearCache
{name} Clear cache files that matches name. The cache files are located at g:vimtex_cache_root and can also be deleted manually.:VimtexClearCache ALL clears all cache files.MAP DEFINITIONS
vimtex-zx
When VimTeX folding is enabled and when the manual mode is turned on (g:vimtex_fold_manual), then VimTeX remaps zx and zX in such that the folds are refreshed appropriately.
<plug>(vimtex-env-delete)
<plug>(vimtex-env-delete-math)
<plug>(vimtex-env-change)
<plug>(vimtex-env-change-math)
Delete/Change surrounding environment. When changing, there will be sensible completion candidates, see cmdline-completion. See also g:vimtex_env_change_autofill and g:vimtex_echo_verbose_input.
<plug>(vimtex-cmd-delete)
<plug>(vimtex-cmd-delete-math)
<plug>(vimtex-cmd-change)
Delete/Change surrounding command. See also g:vimtex_echo_verbose_input.
<plug>(vimtex-delim-delete)
<plug>(vimtex-delim-change-math)
Delete/Change surrounding (math) delimiter. See also
g:vimtex_echo_verbose_input.
<plug>(vimtex-cmd-toggle-frac)
Toggle fractions between inline mode (
num/den) and command mode (\frac{num}{den}). Fractions are toggled according to the map specified by g:vimtex_toggle_fractions.In visual mode, the selected text is toggled if it matches either a
\frac{}{} command or a numerator / denominator string. In normal mode, we try to detect the surrounding fraction command or inline fraction expression. If successful, the detected fraction is toggled.<plug>(vimtex-env-toggle)
Toggle environment, e.g. between
enumerate and itemize. The toggle sequence/map can be customized with g:vimtex_env_toggle_map.<plug>(vimtex-cmd-toggle-star)
<plug>(vimtex-env-toggle-star)
Toggle starred command or environment. For commands, the toggle only applies to commands matched by g:vimtex_toggle_star_cmds.
<plug>(vimtex-cmd-toggle-star-agn)
Toggle starred command or environment, whichever is closest to the cursor. This is not mapped by default. Some users may prefer the combined behaviour and may want to map it, e.g. like this in their personal
ftplugin/tex.vim:nmap tss <plug>(vimtex-cmd-toggle-star-agn)<plug>(vimtex-env-toggle-math)
Toggle between inline math and displayed math, e.g.:
ts$ \[
$f(x) = 1$ ↔ f(x) = 1
\]One may change the toggle sequence with g:vimtex_env_toggle_math_map.
<plug>(vimtex-cmd-toggle-break)
Toggle the line-break macro
\\ at the end of current line. This may be convenient when working with array and math environments.<plug>(vimtex-env-surround-line)
<plug>(vimtex-env-surround-operator)
<plug>(vimtex-env-surround-visual)
Surround the current line, operated text, or visually selected text, with an environment specified during execution. More specifically, this adds
\begin{ENV} on the line above and \end{ENV} on the line below the specified region. The resulting region is filtered to apply proper indentation (see ==).Note: This only works linewise!
Note: There is no default for the operator version.
<plug>(vimtex-delim-toggle-modifier)
<plug>(vimtex-delim-toggle-modifier-reverse)
Toggle delimiter modifiers, by default alternating between
(...) and \left(...\right). The normal mode mapping toggles the closest surrounding delimiter, whereas the visual mode mapping toggles all delimiters that are fully contained in the visual selection. The visual selection is preserved.When g:vimtex_delim_toggle_mod_list is set to contain more than one set of modifiers, these mappings iterate through the list instead of just toggling. For example, one may alternate between
(...), \bigl(...\bigr), \Bigl(...\Bigr), and so on. These mappings accept a [count], which allows the modifier to be incremented multiple steps at a time. The -reverse mapping goes backwards through the modifier list instead of forwards.See also g:vimtex_delim_toggle_mod_list and g:vimtex_delim_list.
<plug>(vimtex-cmd-create)
This mapping works in both insert mode, normal mode and visual mode. It is mapped by default to
<f7>. See below for the behaviour in the different modes.Insert mode:
Convert the preceding text into a LaTeX command. That is, it prepends
a backslash and adds an opening brace. It also moves the cursor to the end
of the word. If you also want the closing brace (e.g. to emulate
delimitMate [0] or any of its like), you can add the following to your
~/.vim/after/ftplugin/tex.vim:imap <buffer> <f7> <plug>(vimtex-cmd-create)}<left>Normal/Visual mode:
Surrounds the word under the cursor/visual selection by the command
provided in an input prompt.
<plug>(vimtex-delim-close)
Close the current environment or delimiter (insert mode), except the top-level
document environment.<plug>(vimtex-delim-add-modifiers)
Add
\left and \right) modifiers to all surrounding "unmodified" delimiters in the current math scope.<plug>(vimtex-reverse-search)
Do reverse search for the MuPDF viewer, see vimtex-view-mupdf.
<plug>(vimtex-ac) Commands
<plug>(vimtex-ic)
<plug>(vimtex-ad) Delimiters
<plug>(vimtex-id)
<plug>(vimtex-ae) Environments (except top-level
document)
<plug>(vimtex-ie)
<plug>(vimtex-a$) Math environments
<plug>(vimtex-i$)
<plug>(vimtex-aP) Sections
<plug>(vimtex-iP)
<plug>(vimtex-am) Items
<plug>(vimtex-im)
These are all text object mappings for the indicated types of objects , see vimtex-text-objects for more info.<plug>(vimtex-%)
Find matching pair.
<plug>(vimtex-]])
go to [count] next beginning of a section.
exclusive motion.
<plug>(vimtex-][)
go to [count] next end of a section.
exclusive motion.
<plug>(vimtex-[[)
go to [count] previous beginning of a section.
exclusive motion.
<plug>(vimtex-[])
go to [count] previous end of a section.
exclusive motion.
<plug>(vimtex-]n)
go to [count] next start of a math zone.
exclusive motion.
<plug>(vimtex-]N)
go to [count] next end of a math zone.
exclusive motion.
<plug>(vimtex-[n)
go to [count] previous start of a math zone.
exclusive motion.
<plug>(vimtex-[N)
go to [count] previous end of a math zone.
exclusive motion.
<plug>(vimtex-]r)
go to [count] next start of a frame environment.
exclusive motion.
<plug>(vimtex-]R)
go to [count] next end of a frame environment.
exclusive motion.
<plug>(vimtex-[r)
go to [count] previous start of a frame environment.
exclusive motion.
<plug>(vimtex-[R)
go to [count] previous end of a frame environment.
exclusive motion.
<plug>(vimtex-]/)
go to [count] next start of a LaTeX comment "%".
exclusive motion.
<plug>(vimtex-]star)
go to [count] next end of a LaTeX comment "%".
exclusive motion.
<plug>(vimtex-[/)
go to [count] previous start of a LaTeX comment "%".
exclusive motion.
<plug>(vimtex-[star)
go to [count] previous end of a LaTeX comment "%".
exclusive motion.
INSERT MODE MAPPINGS
Some LaTeX commands are very common, and so it is both natural and convenient to have insert mode mappings/abbreviations for them. VimTeX therefore provides a list of such mappings that are enabled by default, see g:vimtex_imaps_list. The mappings utilize a map leader defined by g:vimtex_imaps_leader. The default list of maps are all math mode mappings, but one may also add mappings that are available and useful outside of math mode. To see the list of mappings that are created, one can use the command :VimtexImapsList, which is by default mapped to
<localleader>lm.It is of course possible to customize the list of mappings. First, one may specifically disable the entire imaps feature with g:vimtex_imaps_enabled or specific default mappings through g:vimtex_imaps_disabled. Second, one may specify g:vimtex_imaps_list, which will overwrite the default list. Finally, one may add new maps through calls to the function vimtex#imaps#add_map. The following are some examples of how to customize the mappings:
" Disable \alpha and \beta mappings
let g:vimtex_imaps_disabled = ['a', 'b']
" Add custom mapping through vimtex#imaps#add_map
call vimtex#imaps#add_map({
\ 'lhs' : 'test',
\ 'rhs' : '\tested',
\ 'wrapper' : 'vimtex#imaps#wrap_trivial'
\})
" Add custom mapping: #rX -> \mathrm{X}
call vimtex#imaps#add_map({
\ 'lhs' : 'r',
\ 'rhs' : 'vimtex#imaps#style_math("mathrm")',
\ 'expr' : 1,
\ 'leader' : '#',
\ 'wrapper' : 'vimtex#imaps#wrap_math'
\})This function is used to add new insert mode mappings. It takes a single dictionary argument:
let add_map_arg = {
\ 'lhs' : lhs,
\ 'rhs' : rhs,
\ 'expr' : bool,
\ 'leader' : leader_key,
\ 'wrapper' : function_name,
\ 'context' : value,
\ }
Explanation of the keys:lhs
Mandatory argument. The left-hand side part of the map.rhs
Mandatory argument. The right-hand side part of the map. There is one
utility function that can be useful: vimtex#imaps#style_math
Wraps the RHS inside a specified command, e.g.
\myarg{RHS}, if the
cursor is inside math mode.expr
Either 0/v:false or 1/v:true (default: 0). If true, then the
right-hand side is evaluated before it is passed to the wrapper.
This is necessary e.g. for use with vimtex#imaps#style_math.leader
Custom leader key. If the key is not present, then g:vimtex_imaps_leader
is used as leader key.wrapper
The name of a wrapper function that is used to generate the rhs. Two
functions are available from VimTeX: vimtex#imaps#wrap_trivial
Trivial wrapper: Simply returns
rhs. vimtex#imaps#wrap_math
Only define
rhs if inside a math environment. This is the default
wrapper function and will be used if no other wrapper is supplied. vimtex#imaps#wrap_environment
Only define
rhs if inside a specified environment. The wrapper works
by utilizing the context key, which is a list that contains strings
and/or dictionaries: i. If the entry is a string, then the
lhs is mapped to rhs
inside the specified environment.
ii. If the entry is a dictionary, then we assume it has two entries,
envs and rhs, where envs is a list of environment names.
If inside any environment in this list, then we expand to the
corresponding rhs. This allows one to create a mapping that
expands to different rhss in different environments. Of course, one may use custom wrapper functions. To write a custom wrapper
function, please see the source for examples on how the VimTeX wrappers
are written.
context
A value that can be used by the chosen wrapper function. vimtex-neosnippet
vimtex-UltiSnips
Note: that this feature is not the same as the snippet feature of |UltiSnips| or |neosnippet|. The imaps feature of VimTeX previously supported
automatic snippets, but these have been removed after careful considerations and input from VimTeX users, please see VimTeX issue #295: https://github.com/lervag/vimtex/issues/295#issuecomment-164262446It has been decided that the best approach is to only provide basic mappings, and to let users manually create automatic snippets through the anonymous snippet functions in |UltiSnips| and |neosnippet|, please see |UltiSnips#Anon| and |neosnippet#anonymous|, respectively (these will work if the respective plugins are installed). Here are a couple of examples that show how to create such mappings:
" Using neosnippet#anonymous
inoremap <silent><expr> __ neosnippet#anonymous('_${1}${0}')
inoremap <silent><expr> ^^ neosnippet#anonymous('^${1}${0}')
" Using UltiSnips#Anon
inoremap <silent> __ __<c-r>=UltiSnips#Anon('_{$1}$0', '__', '', 'i')<cr>
inoremap <silent> ^^ ^^<c-r>=UltiSnips#Anon('^{$1}$0', '^^', '', 'i')<cr>
A drawback with the anonymous UltiSnips snippets is that they do not nest. That is, if you did __ twice in a row, only the second one could be escaped. In recent versions of |UltiSnips|, one may set normal snippets to trigger automatically, see |UltiSnips-autotrigger|. This allows nesting, and is therefore a better approach than using the anonymous snippet function.EVENTS
VimTeX defines some events using the User autocmd that may be used for further customization.
VimtexEventQuit
This event is triggered when the last buffer for a particular LaTeX project
is wiped (for example, using
:bwipeout) and when Vim is quit. The event
may be used, for instance, to cleanup up auxiliary build files or close
open viewers (see Examples below). With Vim defaults, this event is not
triggered when using :quit or :bdelete since these commands merely hide
the buffer. In multi-file projects, the event may be triggered multiple
times. The b:vimtex variable contains context data for the quitting
file or project. For example, b:vimtex.tex identifies the tex file being
wiped, or the main tex file of a multi-file project. Note: Commands such as :VimtexClean does not always work as expected
with this event. This is because, when quitting vim, the current
buffer does not necessarily have filetype "tex".
VimtexEventInitPre
VimtexEventInitPost
These events are triggered at the start/end of VimTeX initialization. The
post event may e.g. be used to automatically start compiling a document.
VimtexEventCompileStarted
This event is triggered after compilation is started.
VimtexEventCompileStopped
This event is triggered after compilation is stopped.
VimtexEventCompiling
This event is triggered when the compiler backend triggers a new
compilation. This is only supported by vimtex-compiler-latexmk.
VimtexEventCompileSuccess
VimtexEventCompileFailed
These events are triggered after successful/failed compilation and
allows users to add custom callback functionality.
VimtexEventTocCreated
This event is triggered after a ToC window is created.
VimtexEventTocActivated
This event is triggered when a ToC entry has been activated. This allows
to add custom behaviour after opening an entry, e.g. positioning the
buffer window with the zt or zz mappings.
VimtexEventView
This event is triggered after the viewer has opened/forward search has
been performed by the command :VimtexView or the related mapping.
VimtexEventViewReverse
This event is triggered at the end of the vimtex#view#inverse_search
function, which can be used as the callback function for reverse goto from
a PDF viewer.
Examples (Vimscript) - see below for a Lua example:
" Compile on initialization, cleanup on quit
augroup vimtex_event_1
autocmd!
autocmd User VimtexEventQuit VimtexClean
autocmd User VimtexEventInitPost VimtexCompile!
augroup END
" Close viewers when VimTeX buffers are closed
function! CloseViewers()
if executable('xdotool')
\ && exists('b:vimtex.viewer.xwin_id')
\ && b:vimtex.viewer.xwin_id > 0
call system('xdotool windowclose '. b:vimtex.viewer.xwin_id)
endif
endfunction
augroup vimtex_event_2
autocmd!
autocmd User VimtexEventQuit call CloseViewers()
augroup END
" Add custom mappings in ToC buffer
function! TocMappings()
" You probably don't want to do this, though...
nnoremap <silent><buffer><nowait> q :quitall!
endfunction
augroup vimtex_event_3
autocmd!
autocmd User VimtexEventTocCreated call TocMappings()
augroup END
" Specify window position when opening ToC entries
augroup vimtex_event_4
autocmd!
autocmd User VimtexEventTocActivated normal! zt
augroup END
function! CenterAndFlash() abort
" Close all folds, then open only the folds at cursor position, then
" center the cursor in window.
normal! zMzvzz
let save_cursorline_state = &cursorline
" Add simple flashing effect, see
" * https://vi.stackexchange.com/a/3481/29697
" * https://stackoverflow.com/a/33775128/38281
for i in range(1, 3)
set cursorline
redraw
sleep 200m
set nocursorline
redraw
sleep 200m
endfor
let &cursorline = save_cursorline_state
endfunction
" Specify additional behaviour after inverse search
augroup vimtex_event_5
autocmd!
" Use a custom center and flash function
autocmd User VimtexEventViewReverse call CenterAndFlash()
" Or perhaps just use a vimtex blink function
autocmd User VimtexEventViewReverse call vimtex#ui#blink()
augroup END
" Focus the terminal after inverse search
augroup vimtex_event_6
autocmd!
autocmd User VimtexEventViewReverse call b:vimtex.viewer.xdo_focus_vim()
augroup END
Examples - Lua:local au_group = vim.api.nvim_create_augroup("vimtex_events", {}),
-- Cleanup on quit
vim.api.nvim_create_autocmd("User", {
pattern = "VimtexEventQuit",
group = au_group,
command = "VimtexClean"
})
-- Focus the terminal after inverse search
vim.api.nvim_create_autocmd('User', {
pattern = 'VimtexEventViewReverse',
group = au_group,
command = "call b:vimtex.viewer.xdo_focus_vim()"
})
TEXT OBJECTS
Text objects (and motions) are a fundamental feature in Vim. Operations can be combined with motions or text objects in endless ways and can be repeated with the dot operator (repeat.txt). If you are reading this and do not know about these things, then it is strongly advised to read the help section about text-objects and the famous Stack Overflow post "Your problem with Vim is that you don't grok vi": http://stackoverflow.com/questions/1218390/what-is-your-most-productive-shortcut-with-vim/1220118#1220118
VimTeX defines LaTeX specific text objects (and motions). These are all mappings, and as such, they are also described in the sections vimtex-mappings and vimtex-default-mappings.
The usual convention for text object mappings is to prepend "a" to select "a"n object, including the whitespace/delimiters/etc, and to prepend "i" to select the corresponding "inner" object. This is the case for VimTeX text objects, e.g. by default,
vie will visually select the inner part of an environment, whereas vae will select the entire environment including the boundaries.VimTeX supports the well known |targets.vim| as a "backend" for the command and environment text objects (
ie/ae and ic/ac). This should work automatically, see g:vimtex_text_obj_variant for more info.Some examples of how to use the text objects can be useful. The following is a simple table that shows the original text on the left, the keys that are typed in the middle, and the result on the right. The bar "█" indicates the cursor position before the operation.
BEFORE KEYS AFTER
\comm█and{arg} dic \{arg}
\command{a█rg} gUac \COMMAND{ARG}
\lef█t( asd \right) cid \left(█ \right)
\begin{x} die \begin{x}
hello world█ \end{x}
\end{x}
$math █ here$ da$
\begin{itemize} \begin{itemize}
\item hello moon█ cim \item █
\end{itemize} \end{itemize}
\begin{itemize} \begin{itemize}
\item hello moon█ dam \end{itemize}
\end{itemize}
Note: The "greediness" of the command text objects (ic and ac) can be controlled with g:vimtex_parser_cmd_separator_check.Note: Some of the text objects rely on syntax highlighting (vimtex-syntax) to work. That is, some text objects check the syntax groups to determine the proper regions. Examples include the math text objects (e.g. <plug>(vimtex-a$) and <plug>(vimtex-i$)).
Associated settings:
* g:vimtex_text_obj_enabled
* g:vimtex_text_obj_linewise_operators
* g:vimtex_text_obj_variant
COMPLETION
If g:vimtex_complete_enabled is 1 (default), then VimTeX sets the 'omnifunc' to provide omni completion, see compl-omni. Omni completion is then accessible with i_CTRL-X_CTRL-O. If desired, one may set g:vimtex_complete_close_braces, which makes the completion include closing braces.
The omni completion completes citations, labels, glossary entries and filenames. The following sections document the various kinds of completions provided by VimTeX's completion function.
A lot of people expect VimTeX and Vim/neovim to provide
automatic completion, aka autocomplete. However, autocompletion is not a built-in feature of Vim/neovim. The last section of this chapter presents some alternatives to plugins that provide an autocomplete engine and how to configure it with VimTeX, see vimtex-complete-auto.Associated settings:
* g:vimtex_complete_bib
* g:vimtex_complete_close_braces
* g:vimtex_complete_enabled
* g:vimtex_complete_ignore_case
* g:vimtex_complete_ref
* g:vimtex_complete_smart_case
COMPLETE CITATIONS
Citation completion is triggered by
'\cite{' commands. The completion parses included bibliography files (*.bib) and thebibliography environments to gather the completion candidates.By default, cite completion is "smart" in that it allows to complete on author names, title, and similar by matching against a match string defined by g:vimtex_complete_bib.match_str_fmt. If one prefers, one may set the
simple key of g:vimtex_complete_bib to only allow completion on the bibkeys directly. This should typically work better with autocomplete plugins.As an example of the smart completion, assume that a bibliography file is included with the following entry:
@book{knuth1981,
author = "Donald E. Knuth",
title = "Seminumerical Algorithms",
publisher = "Addison-Wesley",
year = "1981"
}
Then, with the default configuration, the bibliography key knuth1981 will be completed with e.g.:\cite{Knuth 1981<CTRL-X><CTRL-O>
\cite{algo<CTRL-X><CTRL-O>
\cite{Don.*Knuth<CTRL-X><CTRL-O>
As is shown in the last example, the search string (e.g. Don.*Knuth) is applied as a regular expression.COMPLETE LABELS
Label completion is triggered by
\ref{ commands. The completion parses every relevant aux file to gather the completion candidates. This is important, because it means that the completion only works when the LaTeX document has been compiled.As an example:
\ref{sec:<CTRL-X><CTRL-O>
offers a list of all matching labels with a menu that contains the associated value and page number.The completion base is matched as a regex in the following order:
\ref{<base><CTRL-X><CTRL-O>The menu, which contains the reference value and page number.
The actual labels.
The menu and label, separated by whitespace. An example:
\ref{eq 2<CTRL-X><CTRL-O> This matches "eq" in the label and "2" in the menu.
Finally, it should also be mentioned that for
\eqref, the candidates will automatically be filtered to only show equation references.COMPLETE COMMANDS AND ENVIRONMENTS
vimtex-complete-environmentsCommand completion is available after
\ and should provide completion candidates for relevant LaTeX commands. The document's preamble is analysed, and commands will be completed for the loaded packages as well as those defined within the preamble using \newcommand, \let and \def. Environment completion is also available after \begin{ or \end{. As with commands, the suggested environment names come from the loaded packages and \newenvironment definitions in the preamble.A lot of packages are supported, see the path
/autoload/vimtex/complete for a relevant file listing.COMPLETE FILE NAMES
File name completion is available for the following macros:
\includegraphics{
Completes image file names.\input{
\include{
\includeonly{
Complete .tex files.\includepdf{
Complete .pdf files. This macro is provided by the pdfpages package.\includestandalone{
Complete .tex files. This macro is provided by the standalone package.COMPLETE INCLUDE GLOSSARY ENTRIES
Glossary entry completion from the
glossaries package are triggered by the commands \gls{, \glspl{ and their variations.Package completion is available for the
\usepackage, \RequirePackage, and \PassOptionsToPackage commands. Similarly, documentclass completion is available for \documentclass and \PassOptionsToClass, and bibliography style completion is available for \bibliographystyle.These completion types all rely on the contents of
ls-R files that are found with:kpsewhich --all ls-RPackages and documentclasses installed at
TEXMFHOME will also be searched. The default value can be found with:kpsewhich --var-value TEXMFHOMENote: If you want to change the default value of
TEXMFHOME in your shell startup file and use gvim started from the desktop environment, please read vimtex-faq-texmfhome.COMPLETE BEAMER THEMES
Theme completion is available for the following
beamer commands:\usetheme{
Complete presentation themes (beamertheme*).\usecolortheme{
Complete color themes (beamercolortheme*).\useinnertheme{
Complete inner themes (beamerinnertheme*).\useoutertheme{
Complete outer themes (beameroutertheme*).\usefonttheme{
Complete font themes (beamerfonttheme*).The candidates are gathered from the same
ls-R and TEXMFHOME sources as the package completion, see vimtex-complete-packages for more details.AUTOCOMPLETE
Vim does not provide automatic completion by itself, but there exist at least several good plugins that provide this: |coc-nvim|, |deoplete|, |neocomplete|, |ncm2|, |nvim-completion-manager|, |youcompleteme|, and |nvim-compe|. Moreover, there is |VimCompletesMe| that overrides
<tab> to trigger different built-in completions, such as the omni-completion by VimTeX, depending on the context. See below for descriptions on how to setup these with VimTeX.coc.nvim
vimtex-complete-coc.nvim
|coc-nvim| is an intellisense engine for Vim8 & Neovim. It's a completion framework and language server client which supports extension features of Visual Studio Code. The project is here: https://github.com/neoclide/coc.nvim.|coc-nvim| can be installed using vim-plug:
Plug 'neoclide/coc.nvim'
However, it does require some more steps, and users are recommended to read the installation instructions in the coc.nvim wiki: https://github.com/neoclide/coc.nvim/wiki/Install-coc.nvimTo configure for VimTeX, one should use the extension plugin |coc-vimtex|, which may be found here: https://github.com/neoclide/coc-vimtex. To use it, first make sure you have |coc-nvim| installed, then just run:
:CocInstall coc-vimtex
The coc-vimtex extension has a few options that can be configured in the coc-settings.json file. See the documentation for |coc-nvim| to learn how to apply the configurations. The following is a list of the options with a short description as it is specified on the project web page:coc.source.vimtex.disableSyntaxes disabled syntax names coc.source.vimtex.enable set to false to disable this source coc.source.vimtex.priority priority of source, default 99 coc.source.vimtex.shortcut shortcut used in menu of completion itemNote: The README of
coc.nvim suggests using noremap K to show documentation. K is also used by VimTeX as one of the default maps (see vimtex-default-mappings) for the same purpose. To enable VimTeX's mapping for .tex files (since coc.nvim does not have a doc source), do one of the following:* Manually remap for
.tex files: Put the following in your $HOME/.vim/after/ftplugin/tex.vim:map <buffer> K <Plug>(vimtex-doc-package)
* Use a custom function in your vimrc file, something like this:nnoremap <silent> K :call <sid>show_documentation()<cr>
function! s:show_documentation()
if index(['vim', 'help'], &filetype) >= 0
execute 'help ' . expand('<cword>')
elseif &filetype ==# 'tex'
VimtexDocPackage
else
call CocAction('doHover')
endif
endfunctiondeoplete
vimtex-complete-deoplete
|deoplete| is a modern remake of |neocomplete|, and was originally written specifically for Neovim, see here: https://github.com/Shougo/deoplete.nvim. It is a highly customizable and flexible completion manager.To configure for VimTeX, one may use:
" This is new style
call deoplete#custom#var('omni', 'input_patterns', {
\ 'tex': g:vimtex#re#deoplete
\})
" This is old style (deprecated)
if !exists('g:deoplete#omni#input_patterns')
let g:deoplete#omni#input_patterns = {}
endif
let g:deoplete#omni#input_patterns.tex = g:vimtex#re#deoplete
neocomplete
vimtex-complete-neocomplete
|neocomplete| is also a flexible automatic completion engine for vim, although active development has been stopped. Users are recommended to change to |deoplete|, see also vimtex-complete-deoplete. The plugin is available here: https://github.com/Shougo/neocomplete.vim.The following options may be used to enable automatic completion for LaTeX documents with |neocomplete| and VimTeX's omni completion function:
if !exists('g:neocomplete#sources#omni#input_patterns')
let g:neocomplete#sources#omni#input_patterns = {}
endif
let g:neocomplete#sources#omni#input_patterns.tex =
\ g:vimtex#re#neocomplete
ncm2
vimtex-complete-ncm2
|ncm2| is a modern remake and replacement of |nvim-completion-manager| and is supposed to be a "Slim, Fast and Hackable Completion Framework for Neovim": https://github.com/ncm2/ncm2The following simple configuration should work well with VimTeX:
" include the following plugins (here using junnegun/vim-plug)
Plug 'roxma/nvim-yarp'
Plug 'ncm2/ncm2'
set completeopt=noinsert,menuone,noselect
augroup my_cm_setup
autocmd!
autocmd BufEnter * call ncm2#enable_for_buffer()
autocmd Filetype tex call ncm2#register_source({
\ 'name': 'vimtex',
\ 'priority': 8,
\ 'scope': ['tex'],
\ 'mark': 'tex',
\ 'word_pattern': '\w+',
\ 'complete_pattern': g:vimtex#re#ncm2,
\ 'on_complete': ['ncm2#on_complete#omni', 'vimtex#complete#omnifunc'],
\ })
augroup ENDFor more lenient, omni-complete-like, filtering of completion candidates, use the following setup (in your
init.vim or a personal ftplugin) instead:augroup my_cm_setup
autocmd!
autocmd BufEnter * call ncm2#enable_for_buffer()
autocmd Filetype tex call ncm2#register_source({
\ 'name' : 'vimtex-cmds',
\ 'priority': 8,
\ 'complete_length': -1,
\ 'scope': ['tex'],
\ 'matcher': {'name': 'prefix', 'key': 'word'},
\ 'word_pattern': '\w+',
\ 'complete_pattern': g:vimtex#re#ncm2#cmds,
\ 'on_complete': ['ncm2#on_complete#omni', 'vimtex#complete#omnifunc'],
\ })
autocmd Filetype tex call ncm2#register_source({
\ 'name' : 'vimtex-labels',
\ 'priority': 8,
\ 'complete_length': -1,
\ 'scope': ['tex'],
\ 'matcher': {'name': 'combine',
\ 'matchers': [
\ {'name': 'substr', 'key': 'word'},
\ {'name': 'substr', 'key': 'menu'},
\ ]},
\ 'word_pattern': '\w+',
\ 'complete_pattern': g:vimtex#re#ncm2#labels,
\ 'on_complete': ['ncm2#on_complete#omni', 'vimtex#complete#omnifunc'],
\ })
autocmd Filetype tex call ncm2#register_source({
\ 'name' : 'vimtex-files',
\ 'priority': 8,
\ 'complete_length': -1,
\ 'scope': ['tex'],
\ 'matcher': {'name': 'combine',
\ 'matchers': [
\ {'name': 'abbrfuzzy', 'key': 'word'},
\ {'name': 'abbrfuzzy', 'key': 'abbr'},
\ ]},
\ 'word_pattern': '\w+',
\ 'complete_pattern': g:vimtex#re#ncm2#files,
\ 'on_complete': ['ncm2#on_complete#omni', 'vimtex#complete#omnifunc'],
\ })
autocmd Filetype tex call ncm2#register_source({
\ 'name' : 'bibtex',
\ 'priority': 8,
\ 'complete_length': -1,
\ 'scope': ['tex'],
\ 'matcher': {'name': 'combine',
\ 'matchers': [
\ {'name': 'prefix', 'key': 'word'},
\ {'name': 'abbrfuzzy', 'key': 'abbr'},
\ {'name': 'abbrfuzzy', 'key': 'menu'},
\ ]},
\ 'word_pattern': '\w+',
\ 'complete_pattern': g:vimtex#re#ncm2#bibtex,
\ 'on_complete': ['ncm2#on_complete#omni', 'vimtex#complete#omnifunc'],
\ })
augroup ENDnvim-completion-manager
vimtex-complete-ncm
Note: |nvim-completion-manager| has been replaced by |ncm2|, and users are recommended to change. See vimtex-complete-ncm2 for hints on how to setup |ncm2| for VimTeX.|nvim-completion-manager| is a fast, extensible, async completion framework for neovim (and Vim version 8.0 and above). The project is available here: https://github.com/roxma/nvim-completion-manager
To configure for VimTeX, one can use the following code:
augroup my_cm_setup
autocmd!
autocmd User CmSetup call cm#register_source({
\ 'name': 'vimtex',
\ 'priority': 8,
\ 'scoping': 1,
\ 'scopes': ['tex'],
\ 'abbreviation': 'tex',
\ 'cm_refresh_patterns': g:vimtex#re#ncm,
\ 'cm_refresh': {'omnifunc': 'vimtex#complete#omnifunc'},
\ })
augroup END
YouCompleteMe
vimtex-complete-youcompleteme
|youcompleteme| is probably the most popular code-completion engine for Vim. The github repository is here: https://github.com/ycm-core/YouCompleteMe. It is described as:> YouCompleteMe is a fast, as-you-type, fuzzy-search code completion engine > for Vim. It has several completion engines: an identifier-based engine that > works with every programming language, a semantic, Clang [3]-based engine > that provides native semantic code completion for the C-family languages, > a Jedi [4]-based completion engine for Python, an OmniSharp [5]-based > completion engine for C# and an omnifunc-based completer that uses data from > Vim's omnicomplete system to provide semantic completions for many other > languages (Ruby, PHP etc.).
To enable automatic completion with |youcompleteme|, use the following options:
if !exists('g:ycm_semantic_triggers')
let g:ycm_semantic_triggers = {}
endif
au VimEnter * let g:ycm_semantic_triggers.tex=g:vimtex#re#youcompleteme
VimCompletesMe
vimtex-complete-vcm
A plugin that maps <tab> to trigger the built-in completion that is most suitable to the current context. The plugin is available here: https://git.sr.ht/~ackyshake/VimCompletesMe.vimThe following options may be used to enable completion with the
<tab> trigger for LaTeX documents with |VimCompletesMe| and VimTeX's omni completion function:augroup VimCompletesMeTex
autocmd!
autocmd FileType tex
\ let b:vcm_omni_pattern = g:vimtex#re#neocomplete
augroup END
nvim-cmp
vimtex-complete-nvim-cmp
|nvim-cmp| [0] is the successor of |nvim-compe|. It provides completion for a lot of different sources. VimTeX support is provided either by using the dedicated cmp-vimtex extension [1] or through the general omni-completion source, cmp-omni [2]. Both approaches are explained briefly below in the form of a Lua snippet for configuration. It is assumed that the reader already knows how to install and configure plugins, though![0]: https://github.com/hrsh7th/nvim-cmp
[1]: https://github.com/micangl/cmp-vimtex
[2]: https://github.com/hrsh7th/cmp-omni
nvim-cmp with cmp-vimtex
cmp-vimtex is a dedicated nvim-cmp extension that gives some extra benefits compared to the more general cmp-omni:* Allows fuzzy matching against all info provided by Vimtex (including bibliographic details, useful for citations). * Parses and display all details contained in bibtex files for cite completions. * Trims long strings in the completion menu (to add space for the documentation window). * Triggers the completion menu automatically after typing
\cite{. * Allows more granular configuration of the menus, and more.To use
cmp-vimtex, you should add something like the following to your configuration. For more info, check the README of the github repo ([1] above). local cmp = require('cmp')
cmp.setup {
-- global configuration goes here
sources = {
{ name = 'buffer' },
-- other sources (GLOBAL)
},
}
cmp.setup.filetype("tex", {
sources = {
{ name = 'vimtex' },
{ name = 'buffer' },
-- other sources
},
}
nvim-cmp with cmp-omni
local cmp = require('cmp')
cmp.setup {
-- global configuration goes here
sources = {
{ name = 'buffer' },
-- other sources (GLOBAL)
},
}
cmp.setup.filetype("tex", {
formatting = {
-- nvim-cmp overrides the standard completion-menu formatting. We use
-- a custom format function to preserve the format as provided by
-- VimTeX's omni completion function:
format = function(entry, vim_item)
vim_item.menu = ({
omni = (vim.inspect(vim_item.menu):gsub('%"', "")),
buffer = "[Buffer]",
-- formatting for other sources
})[entry.source.name]
return vim_item
end,
},
sources = {
{ name = "omni", trigger_characters = { "{", "\\" } },
{ name = 'buffer' },
-- other sources
},
}
nvim-compe
vimtex-complete-nvim-compe
|nvim-compe| is an automatic completion plugin for Neovim. It has support for many different completion sources, including omni-completion. However, |nvim-compe| has been deprecated in favor of the more recent |nvim-cmp|, see vimtex-complete-nvim-cmp. The following configuration examples should still be valid for anyone who for some reason would still wish to use |nvim-compe|. https://github.com/hrsh7th/nvim-compe.The omni-completion source can be enabled for TeX/LaTeX files by adding
omni to the completion sources and specifying the tex filetype. For example, in Lua, it should look something like this:require('compe').setup({
source = {
omni = {
filetypes = {'tex'},
},
},
-- the rest of your compe config...
})
And in Vimscript, it should look something like this:let g:compe.source = {
\ 'omni': {
\ 'filetypes': ['tex'],
\ }
\}
MUcomplete
vimtex-complete-mucomplete
MUcomplete is an implementation of chained (fallback) completion, whereby several completion methods are attempted one after another until a result is returned. The plugin is available here: https://github.com/lifepillar/vim-mucomplete.To enable automatic completion with MUcomplete, use the following options:
let g:mucomplete#can_complete = {}
let g:mucomplete#can_complete.tex =
\ { 'omni': { t -> t =~# g:vimtex#re#neocomplete . '$' } }
FOLDING
vimtex-bib-foldingVimTeX can fold documents according to the LaTeX structure (part, chapter, section and subsection). Folding in tex files is turned off by default, but can be enabled if desired, either through the option g:vimtex_fold_enabled, or manually with
set foldmethod=expr
set foldexpr=vimtex#fold#level(v:lnum)
set foldtext=vimtex#fold#text()
The folding is mainly configured through the dictionary option g:vimtex_fold_types.Note: The fold-expr method of folding is well known to be slow, e.g. for long lines and large files. To speed things up, the user may want to enable the g:vimtex_fold_manual option. An alternative is to add a dedicated plugin that improves folding speed for the slow fold methods, e.g. https://github.com/Konfekt/FastFold.
In order to get slightly cleaner fold text, I recommend setting the global 'fillchars' option to a single space for folds:
set fillchars=fold:\
Note: Remember to include the whitespace after backslash!In addition, VimTeX also provides basic folding in bibtex files. This is disabled by default, but will be enabled if g:vimtex_fold_enabled is set to 1 or v:true. It can also be individually configured with g:vimtex_fold_bib_enabled.
Associated settings:
* g:vimtex_fold_enabled
* g:vimtex_fold_manual
* g:vimtex_fold_levelmarker
* g:vimtex_fold_types
* g:vimtex_fold_types_defaults
* g:vimtex_fold_bib_enabled
* g:vimtex_fold_bib_max_key_width
INDENTATION
vimtex-bib-indentVimTeX provides custom indentation functions both for LaTeX documents and for bibliography files (
.bib files).Associated settings:
* g:vimtex_indent_enabled
* g:vimtex_indent_bib_enabled
* g:vimtex_indent_delims
* g:vimtex_indent_ignored_envs
* g:vimtex_indent_lists
* g:vimtex_indent_on_ampersands
* g:vimtex_indent_tikz_commands
SYNTAX HIGHLIGHTING
VimTeX provides a core syntax plugin combined with package specific addons. The syntax plugin aims to be both consistent, structured, and efficient. The package specific addons are generally only loaded when applicable.
In these modern times, a lot of people have started to rely on Tree-sitter for syntax highlighting (and more). Thus, it is pertinent to mention that a couple of the text objects rely on the syntax highlighting rules to work. See also vimtex-text-objects and vimtex-faq-treesitter.
LaTeX is a macro expansion language and it is impossible to write a fully correct syntax parser without running the
tex compiler itself. VimTeX aims to be pragmatic and provide a best-effort syntax highlighting - a decent trade off between simplicity and completeness. vimtex-synignore
There will probably always be situations where the parser will fail, and in some cases it may be hard to "recover". It is therefore possible to manually activate a severely reduced syntax zone to handle such situations. The zone is activated with the directive
% VimTeX: SynIgnore on and disabled with the similar directive % VimTeX: SynIgnore off, e.g.:% VimTeX: SynIgnore on
\catcode`\$=11
$
\catcode`\$=3
% VimTeX: SynIgnore off
The above LaTeX code will look plain, but OK with VimTeX. Note: The directive is matched case insensitive, and the synonyms enable and disable may be used instead of on and off.Overleaf's magic comment [0] is also supported similarly as the
SynIgnore method described above, i.e.:%%begin novalidate
\catcode`\$=11
$
\catcode`\$=3
%%end novalidate
The VimTeX syntax plugin is loosely based on Dr Chip's syntax plugin for LaTeX which is shipped by default with Vim and neovim (ft-tex-syntax) [1]. There are several major differences that users may want to be aware of: * VimTeX syntax use different names for almost all syntax groups. * VimTeX syntax does not support syntax based folding. * VimTeX syntax does not lint @ in commands, e.g. \@cmd (you should know what you are doing).[0]: https://www.overleaf.com/learn/how-to/Code_Check
[1]: http://www.drchip.org/astronaut/vim/index.html#SYNTAX_TEX
Associated settings:
* g:vimtex_syntax_enabled
* g:vimtex_syntax_custom_cmds
* g:vimtex_syntax_custom_cmds_with_concealed_delims
* g:vimtex_syntax_conceal
* g:vimtex_syntax_conceal_cites
* g:vimtex_syntax_conceal_disable
* g:vimtex_syntax_nested
* g:vimtex_syntax_packages
SYNTAX CONCEAL
VimTeX utilizes the |syn-conceal| feature of Vim to allow displaying commands like
\alpha as α. That is, various elements/commands can be concealed or substituted with a unicode symbol.This feature is mostly enabled by default. Various types of concealments can be enabled/disabled with g:vimtex_syntax_conceal. The entire feature can be fully disabled with g:vimtex_syntax_conceal_disable.
For conceals to work properly, one must set the option 'conceallevel' to 2. It is also good to be aware of the 'concealcursor' option.
It is very important to note that not all fonts are suitable for this feature. That is, for this feature to work well, you should install and use a font that includes unicode characters. For Vim or Neovim in a terminal, this means you must configure your terminal to use such a font. This is, of course, an exercise for the reader, but here is a list of some possibly useful links:
* https://www.programmingfonts.org/
* A convenient site to test different "programming" fonts. Not always easy to see if the unicode support is good, but at least you can see examples of how they look. * https://wiki.archlinux.org/index.php/Fonts * The Arch Wiki is famous for being useful, and it does not fail. But it is rather technical and of course refers to Arch Linux packages. It may still be a good source of knowledge and inspiration. * https://www.binarytides.com/gorgeous-looking-fonts-ubuntu-linux/ * This is a blog post on how to install some modern/good looking/better fonts on a Ubuntu system. * https://github.com/cormullion/juliamono * This is the font that I personally use (2021-03-24, @lervag).
SYNTAX CORE SPECIFICATION
As it is relatively common for some users to customize their colorschemes, it is useful to describe some of the underlying "philosophy" of the syntax rules. Note that, for the following information to be useful, the reader should have at least some basic understanding of how to customize their colorschemes and syntax highlighting. The best resources to learn about this are:
i) |usr_06| "Using syntax highlighting" (READ THIS FIRST)
ii) https://gist.github.com/romainl/379904f91fa40533175dfaec4c833f2f "The right way to override any highlighting if you don't want to edit the colorscheme file directly" This is a good resource that describes how to properly customize the highlighting of syntax groups on top of a basic colorscheme. iii) https://github.com/lervag/vimtex/wiki/Syntax This wiki page gives an example of how to customize and fine-tune syntax highlighting of TeX and BibTeX files. iv) |usr_44| "Your own syntax highlighted" (ADVANCED)
The main philosophy of the VimTeX syntax plugin is to keep things simple, structured, and consistent. There is a small set of primitive syntax elements whose highlighting rules are linked to conventional highlight groups (see group-name). More specialized syntax elements are then linked to a corresponding primitive syntax element. This allows a user to change the highlighting of primitives with the effect that corresponding elements are automatically also updated. It is also possible to override specialized groups to link them to other conventional groups or set colors directly. This gives a high degree of flexibility with regards to customizing colorschemes for LaTeX files. See vimtex-syntax-reference for tables of the most important syntax groups with examples and descriptions.
Most of LaTeX syntax is based around the macro expansion where forms are of the type
\name + [optional group(s)] + {argument group(s)}, where there can often (not always) be white spaces and newlines between the elements. An argument group can often consist of other top level elements, but not always. Further, since LaTeX is designed to have very strong support for typing mathematical equations, there are several ways to start math mode, e.g.$ ... $`, $$ ... $$, \( ... \), \[ ... \], and \begin{equation} matched with \end{equation}. Within math mode, there's a different subset of commands available, and it is common to want a slightly different highlighting of the math mode regions.VimTeX's syntax script is implemented to support these basic structures as well as a large set of more specific commands and elements. The more specific rules define groups whose names are more specific, and it is usually possible to define custom highlighting of specific commands and argument groups.
Finally, it is useful to explain the naming scheme of specialized groups. The general idea can be described as follows.
texCmd{type}
tex{type}Opt
tex{type}Arg
A lot of LaTeX macros and commands are specified specifically with
a given number of optional and real arguments. They may also specify what
those arguments are. In most cases, the highlighting of texCmd{type} is
linked to texCmd and the highlighting of tex{type}Opt and
tex{type}Arg are respectively linked to texOpt and texArg. An
example of this scheme is texCmdAuthor, texAuthorOpt, and
texAuthorArg for \author[...]{...}.
Often, but not always, texCmd{name} is coupled with tex{name}*
groups. For example, \include{...} wants a file argument. The command is
matched as texCmdInput, but it is followed by a texFileArg argument
group.tex{type}Zone
Some commands open specific syntax regions that have different rules. Math
mode is a good example. Math mode is highlighted differently, and the main
syntax regions are named texMathZone*. The tex{type}Zones may
typically contain their own (sub)sets of syntax groups that are only
matched within the specific region. Another example is the inclusion of
nested syntax highlighting with e.g. the minted or listings packages.tex{type}{element}
Some regions or commands include other types of elements, e.g. parameters
like in \def\name #1 where #1 is matched as texDefParm. For
completeness: \def is matched as texCmdDef and \name is matched as
texDefArgName.SYNTAX PACKAGE SPECIFICATION
VimTeX provides several package specific syntax addons that provide richer syntax highlighting. These are built around the same principles as explained in vimtex-syntax-core.
The syntax improvements for a specific package are by default loaded only if that package is detected in the current document (as explained in vimtex-package-detection). This generally works well when a document is compiled, but VimTeX may fail to detect packages for new documents or documents that are not compiled. It is therefore possible to configure that individual syntax packages should always load. One may also disable individual syntax packages. See g:vimtex_syntax_packages for a full list of which syntax addons exist and how to configure them.
SYNTAX GROUP REFERENCE
The following is a reference of the main syntax groups and its default highlighting, as well as one or more examples of what it matches. Most of the primitive groups are linked to conventional syntax groups as listed in group-name. In the examples, capital letters are used to indicate which parts are matched by the current group. For even more details, please refer to the code itself: * Core elements: The
vimtex#syntax#core#init_highlights() function in the file /autoload/vimtex/syntax/core.vim specifies the default highlighting of the core groups. * Package specific groups and elements are defined in the package specific scripts: /autoload/vimtex/syntax/p/*.vim.Note:
* This is only a reference of the main groups. There are also other groups available. See the source files for the full lists. * The following lists might not be always completely up to date. If you find inconsistencies or errors, please open an issue.
Table 1: A list of groups that are only primitive link targets.
GROUP DEFAULT ---------------------------------------------------------------------------- VimtexMsg ModeMsg VimtexInfo Question VimtexTodo Todo VimtexWarning WarningMsg VimtexError ErrorMsg VimtexFatal Error VimtexSuccess Statement texCmdType Type texParm Special texZone PreCondit texSymbol SpecialChar texError VimtexError
Table 2: A list of the most common normal LaTeX groups.
GROUP DEFAULT EXAMPLE
----------------------------------------------------------------------------
texComment Comment % COMMENT
texCommentTodo VimtexTodo % TODO
texDelim Delimiter {, }, [, and ]
texCmd Statement \CMD
texOpt Identifier \cmd[OPT]
texOptSep NormalNC [a, b] (commas)
texOptEqual texSymbol [a=b]
texArg Include \cmd[...]{ARG}
texSpecialChar SpecialChar \S, \P, \$, \;, ...
texCmdInput texCmd \INPUT
\INCLUDE
\INCLUDEONLY
\INCLUDEGRAPHICS
texCmdBib texCmd \BIBLIOGRAPHY
\BIBLIOGRAPHYSTYLE
texCmdClass texCmd \DOCUMENTCLASS
texCmdPackage texCmd \USEPACKAGE
\REQUIREPACKAGE
texFileOpt texOpt \includegraphics[PACKAGE OPTIONS]
\documentclass[CLASS OPTIONS]
texFileArg texArg \input{FILE}
\include{FILE}
\includegraphics[...]{FILE}
\bibliographystyle{FILE}
\documentclass[...]{CLASS}
texFilesOpt texFileOpt \usepackage[PACKAGE OPTIONS]
\RequirePackage[PACKAGE OPTIONS]
texFilesArg texFileArg \includeonly{FILE1, FILE2}
\bibliography{FILE1, FILE2}
\usepackage[...]{PACKAGE1, PACKAGE2}
\RequirePackage[...]{PACKAGE1, PACKAGE2}
texCmdTitle texCmd \TITLE
texTitleArg Underlined \title{MAIN TITLE}
texCmdAuthor texCmd \AUTHOR
texAuthorOpt texOpt \author[OPT]
texAuthorArg NONE \author[...]{AUTHOR LIST}
texCmdPart texCmd \(SUB*)SECTION
texPartArgTitle String \(sub*)section{TITLE}
texCmdEnv texCmd \BEGIN; \END
texEnvArgName PreCondit \begin{ENVNAME}
texCmdRef texCmd \CITE; \LABEL
texRefArg Special \cite{REFERENCE}; \label{REF}
texE3Variable texCmd \G_MYFILE_NAME_STR
texE3Constant texE3Variable
\C_MYFILE_NAME_STR
texE3Function texCmdType \STR_NEW:n
texE3Type texParm \str_new:N
texCmdParbox texCmd \PARBOX[p][h][i]{w}{c}
texBoxOptPosVal texSymbol \parbox[P][h][i]{w}{c}
\begin{minipage}[P][h][i]{w}
texBoxOptIPosVal texBoxOptPosVal
\parbox[p][h][I]{w}{c}
\begin{minipage}[p][h][I]{w}
Table 3: A list of math mode groups.
GROUP DEFAULT EXAMPLE
----------------------------------------------------------------------------
texMathZone Special
texMathZoneLI texMathZone \( HERE \)
texMathZoneLD texMathZone \[ HERE \]
texMathZoneTI texMathZone $ HERE $
texMathZoneTD texMathZone $$ HERE $$
texMathZoneEnv texMathZone \begin{menv} HERE \end{menv}
texMathZoneEnvStarred texMathZone \begin{menv*} HERE \end{menv*}
texMathZoneEnsured texMathZone \ensuremath{HERE}
texMathDelimZone texDelim
texMathDelimZoneLI texMathDelimZone \(; \)
texMathDelimZoneLD texMathDelimZone \[; \]
texMathDelimZoneTI texMathDelimZone $
texMathDelimZoneTD texMathDelimZone $$
texCmdMathEnv texCmdEnv \BEGIN; \END
(Only for math environments.)
texMathEnvArgName Delimiter \begin{EQUATION}
texCmdMath texCmd \ENSUREMATH
texMathDelim Type \LVERT
texMathDelimMod texMathDelim \LEFT\lvert \RIGHT\rvert
texMathOper Operator Basic operators: +-=/
texMathSuperSub texMathOper Sub and super operators (^, _)
texMathError texError Unmatched region endings
Table 4: A list of other important groups.
GROUP DEFAULT EXAMPLE
----------------------------------------------------------------------------
texLength Number Length units, e.g. "4 cm". Only when
contained e.g. in option groups.
texLigature texSymbol --; ---; ``; ''; ,,
texCmdAccent texCmd \"{a}
texCmdLigature texSpecialChar \ss; \ae
texCmdSpaceCodeChar Special Catcodes. For more info, see:
https://en.wikibooks.org/wiki/TeX/catcode
texCmdTodo VimtexTodo \TODOSOMETHING
texCmdVerb texCmd \VERB
texVerbZoneInline texZone \verb+VERB TEXT+
texVerbZone texZone \begin{verbatim} VERB TEXT \end{verbatim}
texCmdDef texCmdNew \DEF
texDefArgName texArgNew \def\NAME
texDefParm texParm \def\name #1
texCmdItem texCmd \item
Table 5: Bold, italic and underline groups.
These groups are used as targets for various (nested) commands, e.g.
\emph{\textbf{...}}. GROUP EFFECT ---------------------------------------------------------------------------- texStyleBold Bold texStyleItal Italic texStyleUnder Underlined texStyleBoth Bold + italic texStyleBoldUnder Bold + underlined texStyleItalUnder Italic + underlined texStyleBoldItalUnder Bold + italic + underlined texMathStyleBold Bold [ONLY in math mode] texMathStyleItal Italic [ONLY in math mode]
NAVIGATION
Vim already has a lot of useful navigation related features, such as tags-and-searches and include-search. VimTeX improves the latter feature by setting the 'include' and 'includeexpr' options, see vimtex-includeexpr.
VimTeX also provides a separate table-of-content feature. This works by parsing the LaTeX project and displaying a table of contents in a separate window. For more info, see vimtex-toc.
The "engine" for collecting the table-of-content entries may also be used as a backend for external plugins. There are sources for |denite.nvim|, |unite.vim| and |fzf.vim| that should work well. The source code may be used as inspiration to write custom sources or sources for other, similar plugins.
INCLUDE EXPRESSION
VimTeX provides an advanced |includeexpr| that makes it possible to open source files for e.g. packages and documentclasses with the gf command. The implementation relies on
kpsewhich to find the source files. Consider the following example:\documentclass{article}
\usepackage{MyLocalPackage}
\usepackage{SomeOtherPackage,YetAnotherPackage}
...
With the cursor on the documentclass name article or one of the package names, gf will take you to the TeX source files (typically .cls file for documentclass and .sty files for packages).TABLE OF CONTENTS
vimtex-toc displays a table of contents (ToC) for the current LaTeX document. The ToC entries may be activated/jumped to with
<cr> or <space>. There are currently four different "layers" of entries:* content This is the main part and the "real" ToC
* todo This shows TODOs from comments,
\todo{...} commands, and commands
from the fixme package
* label This shows \label{...} commands
* include This shows included filesThe ToC is configured with g:vimtex_toc_config. One may change things from where the ToC window is positioned to which layers to show and more. Please read the option help for details.
The ToC parser uses a list of matchers to parse the LaTeX project for the ToC entries. One may add custom matchers through the g:vimtex_toc_custom_matchers option. The syntax of a custom matcher is specified here: toc_matcher_specification.
Note: By setting the
mode configuration key to > 2, the separate ToC window is not opened and most of the features mentioned here will be irrelevant.One may force file input entries of the "include" type into the ToC through comments with the following syntax:
% vimtex-include: /path/to/fileThe path may be absolute or relative. In the latter case, it will be relative to the current root (as printed by :VimtexInfo). This will add an entry in the ToC which makes it easy to open any file. Any file opened through the ToC that was included in this manner will be linked to the current VimTeX project, and thus the ToC and similar commands will be available, even if the file is not a LaTeX file.
vimtex-toc-custom-maps
Some people may want to have separate mappings for different ToC contents, e.g. one mapping to open a table of labels and todos and a different mapping to open a table of include files. This may be easily added with custom mappings:
augroup vimtex_customization
autocmd!
autocmd FileType tex call CreateTocs()
augroup END
function CreateTocs()
let g:custom_toc1 = vimtex#toc#new({
\ 'layers' : ['label', 'todo'],
\ 'todo_sorted' : 0,
\ 'show_help' : 0,
\ 'show_numbers' : 0,
\ 'mode' : 4,
\})
nnoremap <silent> \ly :call g:custom_toc1.open()<cr>
let g:custom_toc2 = vimtex#toc#new({
\ 'layers' : ['include'],
\ 'show_help' : 0,
\})
nnoremap <silent> \lY :call g:custom_toc2.open()<cr>
endfunction
The vimtex#toc#new function takes a dictionary argument that may be used to override the one main configuration (i.e. the combination of the default values and g:vimtex_toc_config).Associated settings:
* g:vimtex_toc_enabled
* g:vimtex_toc_custom_matchers
* g:vimtex_toc_todo_labels
* g:vimtex_toc_show_preamble
* g:vimtex_toc_config
DENITE AND UNITE SOURCES
vimtex-unite https://github.com/Shougo/denite.nvim
https://github.com/Shougo/unite.vim
|denite.nvim| is a popular interface for many things, including outlines. Although VimTeX includes a simple interface for a tables of contents, it also makes sense to provide these as a source to |denite.nvim|. The source name is simply
vimtex.|unite.vim| is the predecessor to |denite.nvim|. As for denite, there is a source called
vimtex.If one prefers the |denite.nvim| or |unite.vim| source to the VimTeX interface, one may override the default mapping, e.g.:
nnoremap <localleader>lt :<c-u>Denite vimtex<cr>
nnoremap <localleader>lt :<c-u>Unite vimtex<cr>
FZF INTEGRATION
https://github.com/junegunn/fzf.vim
https://github.com/junegunn/fzf
|fzf.vim| integrates the general-purpose command-line fuzzy finder |fzf| into vim and neovim. Similar to the |denite.vim| and |unite.vim| source it may be used to quickly navigate VimTeX's built-in ToC feature. To use it, just define a mapping to
vimtex#fzf#run() in your .vimrc, e.g.:nnoremap <localleader>lt :call vimtex#fzf#run()<cr>
You can also choose to only show certain entry "layers", according to this table (see vimtex-toc for detailed explanation of the "layers"):c: content
t: todo
l: label
i: includeThe default behavior is to show all layers, i.e.
'ctli'. To only show content and labels use::call vimtex#fzf#run('cl')
On Windows the python package Colorama is required for colored output. For Linux and MacOS colors should work out-of-the-box, even without Colorama.A second argument can be passed to this function to customize the FZF options. It should be an object containing the parameters passed to
fzf#run(). For example, if you've defined g:fzf_layout, then those options can be passed to vimtex#fzf#run::call vimtex#fzf#run('ctli', g:fzf_layout)
FZF-LUA INTEGRATION
https://github.com/ibhagwan/fzf-lua
|fzf-lua| integrates the general-purpose command-line fuzzy finder |fzf| into neovim through Lua. It may be used to quickly navigate VimTeX's built-in ToC feature. To use it, define a mapping to
require("vimtex.fzf-lua").run(options), e.g.vim.keymap.set("n", "<localleader>lt", function()
return require("vimtex.fzf-lua").run()
end)
Currently two options are supported: "layers" and "fzf_opts". You can pass custom options to fzf-lua via "fzf_opts". You can also choose to only show certain entry "layers" by passing a layer string. By default, all layers are displayed. To only show content and labels use require("vimtex.fzf-lua").run({layers ="cl"}).Note: This requires fzf version 0.60.2 or newer!
SNACKS INTEGRATION
https://github.com/folke/snacks.nvim
|Snacks| is a collection of quality of life plugins for Neovim. It includes a picker, which can be used with the table of contents in VimTeX.
The VimTeX table of contents is registered as a snacks picker source called
vimtex_toc. This allows you to customize the picker with any snacks.picker.Config options. You can open the picker with:Snacks.picker "vimtex_toc"
You can pass options to customize the picker. For example, to filter by content, todos, and labels, and use a dropdown layout:Snacks.picker("vimtex_toc", {
layers = "ctl",
layout = { preset = "dropdown" },
})
The available options are a combination of |snacks.picker.Config| and the vimtex-specific layers option:* layers: Filter by TOC layers. Can be a substring of "ctli" corresponding to content, todos, labels, and includes (similar to vimtex-fzf-lua). Default is "ctli" (all layers).
For more options, see |snacks-picker-config| in the snacks documentation.
For backward compatibility, you can also use:
vim.keymap.set("n", "<localleader>lt", function()
return require("vimtex.snacks").toc()
end)
This accepts the same layers option (and more options are passed to the snacks picker).COMPILER
VimTeX provides an interface to the following LaTeX compilers/compiler backends:
* vimtex-compiler-latexmk http://users.phys.psu.edu/~collins/software/latexmk-jcc
* vimtex-compiler-latexrun https://github.com/aclements/latexrun
* vimtex-compiler-tectonic https://tectonic-typesetting.github.io/
* vimtex-compiler-arara https://github.com/cereda/arara
* vimtex-compiler-texpresso https://aur.archlinux.org/packages/texpresso-git * vimtex-compiler-generic
The interface is implemented in a general way, which makes it relatively easy to add new compilers.
Compilation is started and stopped with :VimtexCompile and :VimtexStop. Although, :VimtexStop stopping is only relevant for continuous compilations, and in this case, :VimtexCompile itself works as a toggle. Single shot compilation is always available through :VimtexCompileSS. The default mappings for these commands are listed here: vimtex-default-mappings.
It is also possible to compile a selection of the file. To do this, one may either use the mapping, <plug>(vimtex-compile-selected), or the command :VimtexCompileSelected.
The compilers should respect the TeX program directive as described here: vimtex-tex-program, except for vimtex-compiler-arara, which uses its own set of directives and rules.
Associated commands:
* :VimtexCompile
* :VimtexCompileSS
* |:VimtexCompileSS!|
* :VimtexCompileSelected
* :VimtexCompileOutput
* :VimtexStatus
* :VimtexStatus!
* :VimtexStop
* :VimtexStopAll
* :VimtexErrors
* :VimtexErrors
* :VimtexClean
* :VimtexClean!
* :VimtexErrors
Associated settings:
* g:vimtex_compiler_enabled
* g:vimtex_compiler_method
* g:vimtex_compiler_latexmk
* g:vimtex_compiler_latexmk_engines
* g:vimtex_compiler_latexrun
* g:vimtex_compiler_latexrun_engines
* g:vimtex_compiler_tectonic
* g:vimtex_compiler_arara
* g:vimtex_compiler_texpresso
* g:vimtex_compiler_generic
* $VIMTEX_OUTPUT_DIRECTORY
Associated events:
* VimtexEventCompileStarted
* VimtexEventCompileStopped
* VimtexEventCompileSuccess
* VimtexEventCompileFailed
LATEXMK
http://users.phys.psu.edu/~collins/software/latexmk-jcc
> latexmk is a perl script for running LaTeX the correct number of times to > resolve cross references, etc; it also runs auxiliary programs (e.g. > bibtex). It has a number of other useful capabilities, for example to start > a previewer and then run latex whenever the source files are updated, so > that the previewer gives an up-to-date view of the document. The script runs > on both UNIX and MS-WINDOWS (XP, etc).
latexmk is a compiler backend that handles recompilation of LaTeX documents when source files have been changed. VimTeX uses the continuous mode by default, but latexmk also allows single shot compilations. The compiler may be configured through the g:vimtex_compiler_latexmk option.If the
callback key is enabled (it is by default and there is really no reason to disable it!), then compilation errors will be parsed automatically. This is done by utilizing the tricks explained below. Although latexmk can control viewers directly, VimTeX disables this feature with -view=none to get full control of the viewers.As stated, one may customize the
latexmk options through g:vimtex_compiler_latexmk. However, one may also configure latexmk explicitly through a global ~/.latexmkrc file, or a project specific .latexmkrc file. It is important to know that command line arguments have priority, so one may want to use custom options if one wants to specify particular things in a configuration file.A particular set of options are very convenient for a good coupling between
latexmk and Vim: $compiling_cmd, $success_cmd, and $failure_cmd. These options can be used to specify commands that are run by latexmk before and after compilation. They are used by VimTeX to achieve callbacks after compilation has finished through vimtex#compiler#callback.Another neat way to use these options is to use
xdotool to change the window title of the viewer to indicate the compilation status:$compiling_cmd = "xdotool search --name \"%D\" " .
"set_window --name \"%D compiling...\"";
$success_cmd = "xdotool search --name \"%D\" " .
"set_window --name \"%D OK\"";
$failure_cmd = "xdotool search --name \"%D\" " .
"set_window --name \"%D FAILURE\"";
Note: If you define these options similar to the above xdotool trick and still want to enable the VimTeX callbacks, then one must include a semicolon at the end of the cmd strings so that VimTeX may append safely to the options.Note: More info on
xdotool here: https://www.semicomplete.com/projects/xdotool.LATEXRUN
https://github.com/aclements/latexrun
> See LaTeX run. Run latexrun.
> latexrun fits LaTeX into a modern build environment. It hides LaTeX's > circular dependencies, surfaces errors in a standard and user-friendly > format, and generally enables other tools to do what they do best.
latexrun is a compiler backend that handles recompilation of LaTeX documents when source files have been changed. However, it is a much simpler backend, and does not support e.g. continuous mode.The compiler may be configured through the g:vimtex_compiler_latexrun option.
TECTONIC
https://tectonic-typesetting.github.io/
> Tectonic is a modernized, complete, self-contained TeX/LaTeX engine, powered > by XeTeX and TeXLive.
tectonic is a compiler backend that features automatic support file downloading along with reproducible builds and full Unicode and OpenType fonts support thanks to the power of XeTeX. It does not support continuous compilation like vimtex-compiler-latexmk, so the only relevant commands are :VimtexCompile to start (single shot) compilation, and :VimtexCompileOutput to see the compilation output.tectonic cleans up intermediate files like .aux and log files by default. However, VimTeX's backend invoke it with the flags --keep-logs and --keep-synctex which enables us to see the errors on the quickfix and it gives us synctex support. Therefore, by default, <plug>(vimtex-clean) and :VimtexClean clean these files.The compiler may be configured through the g:vimtex_compiler_tectonic option.
Some users may be interested in using a custom command for running
tectonic in a forced continuous mode by use of external tools like entr [0]. This could be achieved with the vimtex-compiler-generic interface, e.g. like this:let g:vimtex_compiler_method = 'generic'
let g:vimtex_compiler_generic = {
\ 'command': 'ls *.tex | entr -c tectonic /_ --synctex --keep-logs',
\}
[0]: http://eradman.com/entrproject/ARARA
https://github.com/cereda/arara
> arara is a TeX automation tool based on rules and directives. It gives you > subsidies to enhance your TeX experience.
arara is a TeX automation tool that uses rules and directives that are defined in the preamble of a LaTeX project. The user manual can be found here: https://ctan.uib.no/support/arara/doc/arara-manual.pdfarara does not do continuous compilation, so the only relevant commands are :VimtexCompile to start (single shot) compilation, and :VimtexCompileOutput to see the compilation output.The compiler may be configured through the g:vimtex_compiler_arara option.
Note: It is not possible to directly specify an output directory from VimTeX. This is a restriction caused by the design of arara. However, since one may still want to specify custom output directories, VimTeX allows to customize the output directory through the environment variable $VIMTEX_OUTPUT_DIRECTORY.
TEXPRESSO
https://github.com/let-def/texpresso
> TeXpresso is a live rendering and error reporting tool for LaTeX.
texpresso is a compiler backend that provides a "live rendering" experience: the preview updates incrementally as the buffer changes, without any manual save or recompile step. It combines a forked XeTeX engine with a custom rendering pipeline backed by MuPDF.Unlike the other compiler backends,
texpresso always runs in continuous mode and communicates with VimTeX through a JSON protocol over stdio. This enables: Incremental preview updates as the buffer changes (including unsaved edits)
SyncTeX forward search (cursor position → PDF page)
SyncTeX inverse search (PDF click → cursor position)
Quickfix integration for compiler errors and warnings
texpresso works by forking the TeX process to take periodic snapshots. When the buffer changes, compilation resumes from the nearest snapshot to the edit, so on average only the current page needs to be recomputed.Requirements
texpresso must be installed and available as texpresso in $PATH. The bundled texpresso-xetex engine must also be present (it is included in the standard upstream installation). See the upstream installation instructions:Note: TeXpresso performs only a single TeX pass. Documents that require a table of contents or bibliography must have these built separately (e.g. with
latexmk or biber). TeXpresso will pick up the resulting .aux files when sent SIGUSR1:killall -SIGUSR1 texpressoNote: TeXpresso currently requires a XeTeX-compatible document.
The compiler may be configured through the g:vimtex_compiler_texpresso option.
GENERIC COMPILER
There are a lot of various compiler backends for LaTeX, and it is also possible to simply use things like a Makefile. The generic backend allows to use mostly whatever you want. However, since it is a generic implementation, it will not be as well integrated as e.g. vimtex-compiler-latexmk.
An example may be illuminating. Let's say you want to use a Makefile to compile your project. Then the following shows how to configure with the generic interface, including how to include a simple callback function that reacts to the compiler program output.
function! Callback(msg)
" Use a regex match on the compiler output to get automatic VimtexErrors
" functionality. The below conditional must likely be changed to be
" useful, of course!
if a:msg =~# 'error'
call vimtex#compiler#callback(!vimtex#qf#inquire(b:vimtex.tex))
endif
endfunction
let g:vimtex_compiler_method = 'generic'
let g:vimtex_compiler_generic = {
\ 'command' : 'make',
\ 'hooks': [function('Callback')],
\}See also vimtex-compiler-tectonic for another example that relies on the generic interface to run a "continuous" tectonic command.
Some examples of build tools that can be used with the generic backend:
* Light LaTeX Make (llmk)
https://ctan.org/pkg/light-latex-make
> This program is yet another build tool specific for LaTeX documents. Its > aim is to provide a simple way to specify a workflow of processing LaTeX > documents and encourage people to always explicitly show the right > workflow for each document.
* spix (Yet another TeX compilation tool: simple, human readable, no option, no magic) https://ctan.org/pkg/spix
> SpiX offers a way to store information about the compilation process for > a tex file inside the tex file itself. Just write the commands as comments > in the tex files, and SpiX will extract and run those commands. > Everything is stored in the tex file (so that you are not missing some
> piece of information that is located somewhere else), in a human-readable > format (no need to know SpiX to understand it).
SYNTAX CHECKING (LINTING)
VimTeX provides syntax checking (linting) for TeX and BibTeX files through three compilers:
lacheck [1], chktex [2], and biber [3]. These may be activated with the :compiler command, see compiler-select. A selected compiler may then be used e.g. with :make or :lmake. See the following text for some tips on how one may use this feature.It is possible to use more automatic linting through dedicated plugins. For more information, see vimtex-af-linting.
Associated settings:
* g:vimtex_lint_chktex_parameters
* g:vimtex_lint_chktex_ignore_warnings
A common workflow is to utilize the location-list with :lmake:
To lint the currently open TeX file with
lacheck, run
:compiler lacheck|lmake
To lint the currently open TeX file with
chktex, run
:compiler chktex|lmake
To lint the currently open BibTeX file with
biber, run
:compiler bibertool|lmake
After linting, the compiler or linter messages are added to the location list. This list may be displayed in the location-list window with :lwindow, and one may jump between the entries with :lne and :lp. To automatically open the location-list window after linting is finished, one may add the following to one's vimrc:
augroup VimTeX
autocmd!
autocmd QuickFixCmdPost lmake lwindow
augroup END
For convenience, one may also define a command for linting for each file type and add an autocmd to automatically lint on save. The following gives an example for bibertool and BibTeX, but one may of course do the same with lacheck and/or chktex for TeX files as well. First, add the following to ~/.vim/after/ftplugin/bib.vim:command! -buffer -bang Lint compiler bibertool | lmake<bang>
Then, add to ~/.vim/after/ftplugin/bib.vim:augroup VimTeX
autocmd!
autocmd BufWrite <buffer=abuf> compiler bibertool | lmake!
augroup END
If one minds that Vim becomes unresponsive while linting, then one may utilize plugins like |vim-dispatch| [4], |AsyncRun| [5] or |tasks.vim| [6]. With vim-dispatch, one may replace the :lmake call with :Make. Note that this may conflict with the listing of compilation errors, since :Make from vim-dispatch uses the quickfix window. tasks.vim provide :LMake which allows one to use the location list. For AsyncRun, one may define a custom :Make command with:command! -bang -nargs=* -complete=file Make
\ AsyncRun<bang> -auto=make -program=make
The quickfix window that lists the linter errors and warnings can then be opened by :cwindow and they can be jumped to by :cn and :cp respectively.Often, a syntax error in a BibTeX file is due to a missing comma after an entry. One may define a command to automatically add such missing commas, e.g. by adding the following lines in
~/.vim/after/ftplugin/bib.vim:command! -buffer -range=% -bar AddMissingCommas keeppatterns
\ <line1>,<line2>substitute:\v([}"])(\s*\n)+(\s*\a+\s*\=):\1,\2\3:giep
To call this automatically after saving a BibTeX file, add the following autocommand inside a proper autocommand group (e.g. augroup VimTeX as suggested above) in ~/.vim/after/ftplugin/bib.vim:autocmd BufWrite <buffer> exe
\ 'normal! m`' | silent AddMissingCommas | silent! exe 'normal! g``'
Finally, for more full-fledged linting in Vim, see the plug-ins mentioned in vimtex-and-friends.[1] https://ctan.org/pkg/lacheck
[2] https://www.nongnu.org/chktex/
[3] https://github.com/plk/biber
[4] https://github.com/tpope/vim-dispatch
[5] https://github.com/skywind3000/asyncrun.vim
[6] https://github.com/mg979/tasks.vim
GRAMMAR CHECKING
VimTeX provides several compilers for grammar checking TeX files through the compiler-select feature in Vim. A compiler may be activated with the :compiler command (see vimtex-lint above for some more tips on how to use this feature). The selected compiler may then be used e.g. with :make or :lmake. As an example, one may do the following to use the location-list with a given checker:
:compiler {checker}|lmake
The following is a list of the available checkers:textidote
See more details here: vimtex-grammar-textidotevlty
See more details here: vimtex-grammar-vltystyle-check
https://github.com/nspring/style-check.gitThe language of the Tex file is determined by the option 'spelllang'. This option can be specified in one's vimrc file, but it can also be specified in a modeline (see also the user manual section 21.6 for a gentle introduction to the use of modelines).
Other possibilities for grammar and language checking are: * The LTeX project: a grammar and spell checking tool available as a language server. See vimtex-af-lsp for more info. * Angry Reviewer: An off-line vim plugin for the AngryReviewer service that provides style suggestions for academic and scientific text in the quickfix list. https://github.com/anufrievroman/vim-angry-reviewer
TEXTIDOTE
The
textidote compiler is a VimTeX wrapper over TeXtidote [1]. TeXtidote is a correction tool for LaTeX documents to check grammar, style, and perform spell checking.Configuration of the wrapper is controlled by the Vim dictionary g:vimtex_grammar_textidote. In particular, it is important to specify the
jar key to the path of the executable jar file textidote.jar. Please note that if one installs textidote with a package manager e.g. in some common Linux distributions, the .jar file might be missing. If so, it should be possible to download it manually from [1]. However, before one does that, it can be smart to check the top lines of the installed executable, as it may be a simple Bash script wrapper.VLTY
The
vlty compiler uses the Python package YaLafi [1] for extracting the plain text and combines this with the proofreading software LanguageTool [2]. The name vlty comes from VimTeX + LanguageTool + YaLafi.In order to use
vlty, you need local installations of both components. An archive of LanguageTool can be downloaded from [3]. After uncompressing at a suitable place, the path to it is specified as shown below. On a system like Arch Linux, LanguageTool may also be installed with:sudo pacman -S languagetool
YaLafi itself can be installed with:pip install --user yalafiConfiguration is controlled by the Vim dictionary g:vimtex_grammar_vlty. As a minimal example, one could write in vimrc:
let g:vimtex_grammar_vlty = {'lt_directory': 'path/to/LanguageTool'}
set spelllang=en_gb
The given directory has to contain the LanguageTool software, including for instance the file languagetool-server.jar. If instead LanguageTool is installed through a package manager as mentioned above, one could write:let g:vimtex_grammar_vlty = {'lt_command': 'languagetool'}
set spelllang=en_gb
Calling :compiler vlty will raise an error message if some component cannot be found.Note: Spell checking with
LanguageTool is only enabled if a country code is specified in 'spelllang'.[1] https://github.com/torik42/YaLafi
[2] https://www.languagetool.org
[3] https://www.languagetool.org/download/
VIEW
VimTeX provides the command :VimtexView to open the output PDF in a desired viewer specified by g:vimtex_view_method. The command is mapped to
<localleader>lv by default. The supported viewers are described in vimtex-view-configuration, which also explains how to configure them.Many viewers support synctex for navigating between the PDF and the source tex file. If possible, :VimtexView will perform forward search when the viewer is opened. See vimtex-synctex for more details.
Associated settings:
* g:vimtex_view_enabled
* g:vimtex_view_automatic
* g:vimtex_view_forward_search_on_start
* g:vimtex_view_use_temp_files
* g:vimtex_view_method
* g:vimtex_view_general_options
* g:vimtex_view_general_viewer
Associated events:
* VimtexEventView
* VimtexEventViewReverse
VIEWER CONFIGURATION
g:vimtex_view_method is the main configuration variable. It allows to choose between a set of predefined viewers, including a generic customizable interface. For the predefined viewers, forward search with synctex should usually work without any further configuration. With the general viewer, one may often specify options to enable forward search. Inverse search requires configuration on the viewer side in most cases.
The generic interface is flexible. It relies on three options: * g:vimtex_view_general_viewer Specify the viewer executable. * g:vimtex_view_general_options Specify the viewer options (e.g. to specify forward search configuration).
The following is a list of popular PDF viewers, in alphabetic order, and how they can be configured to work with VimTeX.
vimtex-view-evince
Evince
https://wiki.gnome.org/Apps/Evince
Evince is a document viewer for viewing multiple document formats, including PDFs. It comes by default with Gnome.Configuration:
let g:vimtex_view_general_viewer = 'evince'
Note: Evince only supports synctex through DBus, which is not supported by VimTeX. Thus VimTeX does not support forward and inverse search with Evince. However, the Vim plugin SVED by Peter Jorgensen is reported to work well in combination with VimTeX. See the plugin page [0] for more information. vimtex-view-galley
Galley
https://github.com/munepi/Galley
Galley is a lightweight PDF previewer for macOS, designed for TeX/LaTeX authors with SyncTeX support as a core feature. It enforces a single-window policy (one source, one PDF, one window) and reloads the PDF automatically after recompilation. The VimTeX implementation supports forward search and uses a callback to update Galley after successful compilations.Configuration:
let g:vimtex_view_method = 'galley'By default Galley is opened and updated in the background. Set g:vimtex_view_galley_activate to 1 to bring it to the foreground and give it focus after a forward search.
To configure inverse search: Select
Custom... from the SyncTeX menu in the menu bar, then set the custom editor command from a terminal. With Vim or Neovim, respectively:defaults write com.github.munepi.galley customEditorCommand \
"/path/to/vim -v --not-a-term -T dumb -c \"VimtexInverseSearch %line '%file'\""
defaults write com.github.munepi.galley customEditorCommand \
"/path/to/nvim --headless -c \"VimtexInverseSearch %line '%file'\""Replace
/path/to/vim (or nvim) with the absolute path to your binary, e.g. /opt/homebrew/bin/nvim for a homebrew installation.Inverse search is then activated by
Command-clicking in the PDF. The first forward search will trigger a macOS prompt asking for Automation permissions, which must be allowed. For more info on inverse search, see vimtex-synctex-inverse-search.Associated settings:
* g:vimtex_view_galley_activate
* g:vimtex_view_galley_sync
vimtex-view-mupdf
MuPDF
https://www.mupdf.com/
MuPDF is a very minimalistic and quick PDF viewer. It does not support synctex itself, but VimTeX provides both forward and inverse search by abusing xdotool. Inverse search must be used from within VimTeX with the mapping <plug>(vimtex-reverse-search) (default mapping: '<localleader>lr').One can also use g:vimtex_view_mupdf_send_keys to specify a set of keys that is sent to MuPDF on startup.
Configuration:
let g:vimtex_view_method = 'mupdf'
Associated settings:
* g:vimtex_view_mupdf_options
* g:vimtex_view_mupdf_send_keysNote: Both forward and inverse search requires
xdotool to work. Forward search will only take you to the correct page. Inverse search will take you to the line in Vim that corresponds to the first line of the current page in MuPDF.Note: Viewer handling uses window title matching. If there exists another pdf viewer with the same name as the current project pdf file, then there might be conflicts, and so MuPDF might not work as expected.
vimtex-view-okular
Okular
https://okular.kde.org/
Okular is a very feature rich PDF viewer that supports both forward and inverse search.Configuration:
let g:vimtex_view_general_viewer = 'okular'
let g:vimtex_view_general_options = '--unique file:@pdf\#src:@line@tex'
Inverse search can be set up within Okular in the settings pane under "Settings > Editor > Custom Text Editor" [0]. The following is the recommended settings for Vim and neovim, respectively:vim -v --not-a-term -T dumb -c "VimtexInverseSearch %l '%f'"
nvim --headless -c "VimtexInverseSearch %l '%f'"
To perform an inverse search in Okular, do shift + click while browse mode is enabled. For more info, see vimtex-synctex-inverse-search. vimtex-view-qpdfview
qpdfview
https://launchpad.net/qpdfview
qpdfview is a tabbed document viewer. It supports both forward and inverse search.Configuration:
let g:vimtex_view_general_viewer = 'qpdfview'
let g:vimtex_view_general_options
\ = '--unique @pdf\#src:@tex:@line:@col'
Inverse search must be set up from within qpdfview under "Edit -> Settings -> Source Editor". The following is the recommended settings for Vim and neovim, respectively:vim -v --not-a-term -T dumb -c "VimtexInverseSearch %2 '%1'"
nvim --headless -c "VimtexInverseSearch %2 '%1'"
Use right click to perform an inverse search. For more info, see vimtex-synctex-inverse-search. vimtex-view-sioyek
Sioyek
https://sioyek.info/
Sioyek is a PDF viewer designed for reading research papers and technical books.Configuration:
let g:vimtex_view_method = 'sioyek'
Inverse search should be automatically configured and work out of the box. VimTeX will try to pass options to Sioyek to automatically configure inverse search. This means that, in most cases, inverse search should work as expected without any further configuration. One may still be interested in learning how inverse-search configuration works, in which case one should read vimtex-synctex-inverse-search.Note: The interpolation variables for Sioyek inverse search configuration are
%2 and %1, not %l and %f.Associated settings:
* g:vimtex_callback_progpath
* g:vimtex_view_sioyek_exe
vimtex-view-skim
Skim
https://skim-app.sourceforge.net/
https://sourceforge.net/p/skim-app/wiki/TeX_and_PDF_Synchronization Skim is a PDF reader and note-taker for OS X. It is designed to help you read and annotate scientific papers in PDF, but is also great for viewing any PDF file. The VimTeX implementation supports forward search and uses a callback to update Skim after successful compilations.Configuration:
let g:vimtex_view_method = 'skim'
To configure inverse search: Open the Sync tab in the settings panel in Skim and set the options according to your desired version of Vim. With MacVim, one may use the MacVim preset. However, it may be more convenient to use a Custom setting and configure the inverse search option tovim -v --not-a-term -T dumb -c "VimtexInverseSearch %line '%file'"
nvim --headless -c "VimtexInverseSearch %line '%file'"
Depending on your system configuration, it may be necessary to specify the path to the Vim/Neovim binary. For example, if Neovim was installed via homebrew, one can specify the following in the Skim settings panel.Command = /opt/homebrew/bin/nvim
Arguments = --headless -c "VimtexInverseSearch %line '%file'"
Inverse search is activated by pressing Shift and Command, then clicking the text you want to search. For more info on inverse search, see vimtex-synctex-inverse-search.Associated settings:
* g:vimtex_view_skim_activate
* g:vimtex_view_skim_reading_bar
* g:vimtex_view_skim_no_select
vimtex-view-sumatrapdf
SumatraPDF
https://www.sumatrapdfreader.org/free-pdf-reader.html SumatraPDF is a PDF viewer for windows that is powerful, small, portable and starts up very fast. It supports both forward and inverse search.Configuration:
let g:vimtex_view_general_viewer = 'SumatraPDF'
let g:vimtex_view_general_options
\ = '-reuse-instance -forward-search @tex @line @pdf'
For convenience, the above configuration is used by default on Windows if SumatraPDF is detected as executable.To configure inverse search:
Install SumatraPDF for all users and add to
PATH.
Open "Settings -> Advanced Options" in SumatraPDF and edit the following variables:
InverseSearchCmdLine = cmd /c start /min "" nvim --headless -c "VimtexInverseSearch %l '%f'"
EnableTeXEnhancements = true
For Vim instead of Neovim, use:
InverseSearchCmdLine = cmd /c start /min "" vim -v --not-a-term -T dumb -c "VimtexInverseSearch %l '%f'"
If you are using an old version of SumatraPDF where Advanced Options is not available: open "Settings -> Options", Find the section "Set inverse search command-line" in the bottom, and use the appropriate viewer configuration, i.e. one of the InverseSearchCmdLine from step 2 above.Inverse search is activated with a double click in the PDF file. See vimtex-synctex-inverse-search for more info on inverse search.
Note: If you want to use SumatraPDF with VimTeX from within WSL, please read vimtex-faq-sumatrapdf-wsl.
Note: There is a known issue with VimTeX + SumatraPDF when you use
xelatex, where the pdf file in SumatraPDF is not refreshed after compilation. A workaround was found and posted by @Whitebeard0 here: https://github.com/lervag/vimtex/issues/1410#issuecomment-506143020 vimtex-view-texshop
TeXShop
https://pages.uoregon.edu/koch/texshop/index.html TeXShop is a TeX front-end program for macOS. It provides an editor (which can be replaced by other external editors, such as vim), a front-end to call TeX-related programs that process source files, and a PDF viewer with support for syncing between source and PDF. VimTeX supports both forward and inverse searches with TeXShop.Configuration:
let g:vimtex_view_method = 'texshop'
From TeXShop, inverse search is activated by Command-clicking in the viewer. Furthermore, if TeXShop's compiler front-end is used, clicking Goto Error in the console (keyboard shortcut Command-Control-E) will also activate inverse search and jump to the line containing the error in the source.To configure inverse search:
Adjust TeXShop's preference from a terminal:
defaults write TeXShop OtherEditorSync YES
defaults write TeXShop UseExternalEditor -bool true These commands only need to be run once as they modify TeXShop's preference
file in
~/Library/Preferences/TeXShop.plist, which persists through
sessions. While the first preference is a hidden preference and must be set
via a terminal command, the second can be toggled from within TeXShop's
preferences pane.Create a shell script
/usr/local/bin/othereditor that contains a call to either Vim or neovim, i.e. one of the following lines, respectively:vim -v --not-a-term -T dumb -c "VimtexInverseSearch $1 '$2'"
nvim --headless -c "VimtexInverseSearch $1 '$2'"
The script must be executable:
chmod +x /usr/local/bin/othereditor MacVim users should read vimtex-faq-texshopviewer for additional setup
instructions and limitations.
For more info on inverse search, please see vimtex-synctex-inverse-search.
Associated settings:
* g:vimtex_view_texshop_activate
* g:vimtex_view_texshop_sync
vimtex-view-zathura
vimtex-view-zathura-simple
Zathura
https://pwmt.org/projects/zathura/
Zathura is, like MuPDF, a very fast and minimalistic viewer. Compared to MuPDF, it allows more user configuration. Zathura has full support for both forward and inverse search. Zathura should be straightforward to install and use on Linux with Xorg. It should also work on macOS and WSL, but users may want to consider to use either the native Skim viewer (vimtex-view-skim) or Sioyek (vimtex-view-sioyek). The macOS users who still want to use Zathura should read vimtex-faq-zathura-macos or vimtex-faq-zathura-windows-wsl.Configuration:
" Main variant with xdotool
let g:vimtex_view_method = 'zathura'
" For simple variant without xdotool
let g:vimtex_view_method = 'zathura_simple'
The main variant uses xdotool to help avoid duplicate Zathura instances. However, in some environments, xdotool is not available. Here the simple variant should work well.VimTeX will start Zathura with the
-x argument to specify the inverse search options automatically. This means that, in most cases, inverse search should work as expected without any further configuration. One may still be interested in learning how inverse-search configuration works, in which case one should read vimtex-synctex-inverse-search.Zathura also supports
%{column}, so if your synctex implementation supports getting the column, then you can put the following in your zathurarc:set synctex-editor-command "vim -v --not-a-term -T dumb -c \"VimtexInverseSearch %{line}:%{column} '%{input}'\""
Associated settings:
* g:vimtex_callback_progpath
* g:vimtex_view_zathura_check_libsynctex
* g:vimtex_view_zathura_use_synctex
* g:vimtex_view_zathura_optionsNote: The interpolation variables for Zathura configuration of inverse search are
%{line} and %{input}, not %l and %f.Note: Recent versions of Zathura no longer ensures synctex support. This has resulted in synctex support being dropped on some platforms, e.g. on OpenSUSE, cf. https://github.com/lervag/vimtex/issues/384. A workaround is to build Zathura from source manually.
Note: Viewer handling uses window title matching. If there exists another pdf viewer with the same name as the current project pdf file, then there might be conflicts. In particular, this might affect forward/inverse searching for Zathura.
SYNCTEX SUPPORT
Synctex is a tool that enables synchronization of the text editor position and the pdf viewer position. The tool may be used to add mappings in vim to go to the current position in the compiled pdf document (forward search), and also to go from a specific position in the pdf file to the corresponding position in vim (inverse search).
To make synctex work, it must be enabled. VimTeX enables this by default by passing
-synctex=1 on the command line, unless the user overrides the option (see the options key for g:vimtex_compiler_latexmk or g:vimtex_compiler_latexrun).Alternatively, for vimtex-compiler-latexmk, one can put this in one's
~/.latexmkrc file:$pdflatex = 'pdflatex -synctex=1 %O %S';
Forward search
vimtex-synctex-forward-search
For supported viewers, :VimtexView (<localleader>lv) will issue a forward search if the viewer is already opened. The forward search will take you to the page or position in the viewer that corresponds to the current line in your vim session. See g:vimtex_view_method for a list of supported viewers.In supported viewers, one may set up inverse search, which allows one to go directly from a selected line in the viewer (typically by double clicking with the mouse or something similar) to the corresponding line inside the Vim instance. This is sometimes also called backward search or reverse search.
Inverse search relies on communicating with Vim/neovim from the viewer by use of shell commands executed by the viewer. It is usually configured within the specific viewer through an option named something like "inverse search command-line". The option specifies the necessary shell command to perform the inverse search. The target line and file are provided as interpolation variables. A typical shell command looks like this:
vim --remote-silent +%l %f
Luckily, VimTeX provides a convenience function to simplify the viewer configuration. The command VimtexInverseSearch will execute vimtex#view#inverse_search with the target line and file as arguments inside the desired Vim or neovim instance. The latter function is the one that really performs the inverse search. The combined effect is a more robust experience that will seamlessly handle multiple Vim or neovim instances and multiple VimTeX instances. The user doesn't need to worry about passing the correct servernames.To configure with
VimtexInverseSearch, use:vim -v --not-a-term -T dumb -c "VimtexInverseSearch %l '%f'"
nvim --headless -c "VimtexInverseSearch %l '%f'"
# Or, if you also have the column number in %c
nvim --headless -c "VimtexInverseSearch %l:%c '%f'"
On Windows, the above commands may lead to an annoying command window "popup". This may be avoided, or at least reduced, with the following variants:cmd /c start /min "" vim -v --not-a-term -T dumb -c "VimtexInverseSearch %l '%f'"
cmd /c start /min "" nvim --headless -c "VimtexInverseSearch %l '%f'"
Note: In the above, we used %l and %f. However, the interpolation variables may be named different in some viewers. The correct names are given for each supported viewer (see e.g. vimtex-view-skim).Note: Vim users should be aware that one may need to ensure that the server is really running, see vimtex-clientserver.
Note: Many plugin managers provide mechanisms to lazy load plugins. There is no need to use such a mechanism for VimTeX, and in fact, doing it will prevent Vim/neovim from loading the
:VimtexInverseSearch command.Note: You may need to add the installation prefix for Vim/Neovim, for example
/opt/homebrew/bin, in your PDF viewer if inverse search does not work.LATEX DOCUMENTATION
VimTeX provides the command :VimtexDocPackage to open documentation for packages and documentclasses. The command is mapped to
K by default.For simplicity, the standard method provided by VimTeX is to look up documentation online through http://texdoc.org/. However, this can be customized with the option g:vimtex_doc_handlers. The option allows much flexibility for advanced users. For users that want to use a local
texdoc installation as the main method, they may use the following config:let g:vimtex_doc_handlers = ['vimtex#doc#handlers#texdoc']See https://www.tug.org/texdoc/doc/texdoc.pdf for more info about
texdoc.In the following, I list some relevant online and offline alternatives for accessing LaTeX documentation. Please note that these methods are not integrated into VimTeX and are listed purely for the readers convenience.
ONLINE
I recommend the LaTeX Wikibook [0] as a good source of documentation for LaTeX. One should also know about the Comprehensive TeX Archive Network, or CTAN [1], which is the central place for all kinds of material around TeX. The long-existing unofficial LaTeX(2e) reference manual (latexref) can be found online at [2].
OFFLINE
One may use offline documentation by using a different handler with the g:vimtex_doc_handlers option. An example:
let g:vimtex_doc_handlers = ['vimtex#doc#handlers#texdoc']
One may also use a more dedicated offline documentation system. On macOS, Dash [0] is a non-free but high-quality system. On Linux, one may use Zeal [1] or dasht [2], both of which access the Dash documentation sets. Zeal should also work well on Windows.The above systems may be accessed from vim through dash.vim [3], zeavim.vim [4] or vim-dasht [5], respectively. Other alternative vim plugins include investigate.vim [6].
The unofficial LaTeX(2e) reference manual (latexref) should also be mentioned, since it may be easily downloaded in various formats from [7].
[0]: https://kapeli.com/dash
[1]: https://zealdocs.org/
[2]: https://github.com/sunaku/dasht
[3]: https://github.com/rizzatti/dash.vim
[4]: https://github.com/sunaku/vim-dasht
[5]: https://github.com/KabbAmine/zeavim.vim
[6]: https://github.com/keith/investigate.vim
[7]: https://latexref.xyz/dev/
CONTEXT MENU
VimTeX provides the command :VimtexContextMenu to open a context menu for the item below the cursor. The menu allows various actions relevant to the current context. It is mapped by default to
<localleader>la.The available contexts are listed below.
Associated settings:
* g:vimtex_context_pdf_viewer
CITATION CONTEXT
When the cursor is over a citations, e.g.
\textcite{myRef}, then the context menu will show choices relevant to the current citation entry. This works by parsing the relevant bib file for metadata and providing menu actions depending on the available metadata. The actions are only displayed when they are relevant.Possible actions:
Edit entry
Go to the entry location in the relevant bib file.Show entry
Show the registered data for the current entry.Open PDF
Open associated PDF file from the file key of the bib entry.Open DOI
Open associated DOI url from the doi key of the bib entry.Open URL
Open associated URL from the url key of the bib entry.CODE STRUCTURE
The VimTeX code is based on the autoload feature of vim. For each new latex buffer, the function vimtex#init initializes a state variable as well as buffer local mappings and commands, all based on the desired options (see vimtex-options).
The main init function calls
vimtex#mymodule#init_buffer for each submodule, if it exists. This function should take care of defining buffer local mappings, commands, and autocommands.The state variable is a Dictionary that contains data that is specific to a single LaTeX project. Such a project may consist of several buffers for different files if the project is a multi-file project (see vimtex-multi-file). A submodule may add to the state during initialization with
vimtex#mymodule#init_state, which takes the state object as a single argument.The command :VimtexInfo (mapped to
<localleader>li by default) will show the (relevant) contents of the local state, as well as some auxiliary information that may be useful for debugging purposes.See also the supplementary high-level code documentation [0] for more detailed information about the VimTeX code.
API REFERENCE
This is an API reference of the most useful VimTeX functions available to users for customization.
Note: This reference is currently a work in progress!
vimtex#cite#get_entry
Returns a citation entry. It takes an option
key argument. If a key is supplied, then the corresponding bib entry for that key is returned (a Dict). Else it returns the entry for the key under the cursor. If no key is found, an empty Dict is returned.vimtex#cite#get_key
Returns the citation key under the cursor. Can be useful e.g. to create a function to open a citation in another program such as BibDesk or Zotero. For example:
function! OpenInBibDesk() abort
let l:key = vimtex#cite#get_key()
if empty(l:key) | return | endif
call vimtex#util#www('x-bdsk://' .. vimtex#util#url_encode(l:key))
endfunction
vimtex#compiler#callback
Utility function to be used as a compiler callback function. Takes a single argument, which is the compiler status:1: Compilation cycle has started
2: Compilation complete - Success
3: Compilation complete - Failed
The function does several useful things based on the status, such as running the VimtexEventCompiling, VimtexEventCompileFailed and VimtexEventCompileSuccess events.
vimtex#env#get_inner
vimtex#env#get_outer
vimtex#env#get_all
Functions that return the surrounding inner or outer environment, or all surrounding environments. The return value is a dictionary with the following keys:
name: The environment name
open: The environment opening delimiter object
close: The environment closing delimiter objectThe delimiter objects contain information about the locations and arguments of the corresponding
\begin{...}[...] and \end{...} commands.vimtex#env#is_inside
A function that returns the start position of the
\begin{environment} of the environment name that was passed as the only mandatory argument. That is, the return value is a list of two numbers: The line number and the column number. These are both 0 if no surrounding environment was found.vimtex#syntax#in
vimtex#syntax#in(name) -> Boolean
vimtex#syntax#in(name, line, column) -> Boolean
Returns v:true if the cursor position or the specified position is inside the named group. name is a regex that is used to matched against the syntax group stack.vimtex#syntax#in_mathzone
vimtex#syntax#in_mathzone() -> Boolean
vimtex#syntax#in_mathzone(line, column) -> Boolean
Returns v:true if the cursor position or the specified position is inside a math zone.vimtex#view#inverse_search
Utility function for reverse search from pdf viewer. Takes two arguments: the line number and a filename. The function runs the event VimtexEventViewReverse at the end, which allows more user customization. E.g., the user could call
vimtex#ui#blink() to have a blinking effect after doing the reverse search. See the docs for VimtexEventViewReverse for an example.FAQ
This is a section of some frequently asked questions whose answers may be of help to users.
Contents:
* vimtex-faq-windows
* vimtex-faq-neovim
* vimtex-faq-slow-matchparen
* vimtex-faq-surround
* vimtex-faq-isfname
* vimtex-faq-tags
* vimtex-faq-tags-bibtex
* vimtex-faq-texmfhome
* vimtex-faq-wsl
* vimtex-faq-sumatrapdf-wsl
* vimtex-faq-zathura-macos
* vimtex-faq-zathura-windows-wsl
* vimtex-faq-texshopviewer
* vimtex-faq-treesitter
Q: Does VimTeX support Windows?
A: Yes. But there are some "gotchas":
* It is highly recommended to install a dedicated Perl distribution (e.g.
Strawberry Perl [0]).
* Several features rely on having executables like
latexmk and
SumatraPDF.exe "readily available" by adding the parent directories of
the executables to your PATH environment variable. See [1] for more info
on PATH.
* VimTeX does not work well with the 'shell' setting set to Windows
PowerShell. It is therefore recommended to use the default 'shell'
settings. See [2] for more information.
* :VimtexInfo might fail on the first attempt to run because Windows
doesn't natively support UTF-8. This support can be enabled by [3]:
navigating to
Control Panel/Clock and Region/Region,
going to the
Administrative tab and clicking on the
Change system locale button,
checking the box for
Use Unicode UTF-8 for worldwide language support. Note that this Windows feature is still in Beta and could cause side
effects on other programs. Use it with care and revert the process if you
encounter more trouble.
[0]: https://strawberryperl.com/
[1]: https://www.rapidee.com/en/path-variable
[2]: https://github.com/lervag/vimtex/issues/1507
[3]: https://github.com/lervag/vimtex/issues/2671
Q: Does VimTeX support neovim?
A: Yes, but some people may complain that VimTeX is not written in Lua!
Q: Why is matching parens so slow?
A: Because it is complicated and requires some expensive searches for matching parentheses. It uses the syntax information to skip commented delimiters, which is expensive. You can tune the timeout and stopline parameters for the searches with g:vimtex_delim_timeout and g:vimtex_delim_stopline, which may help. If it is still too slow, you can also try to use vim-matchup [0], see also vimtex-af-enhanced-matchparen.
Q: VimTeX provides
dse, dsc, cse, and csc. These seem to be inspired by |surround.vim|. Does VimTeX also provide the corresponding yse and ysc? A: The mentioned mappings are indeed inspired by |surround.vim|. However, VimTeX does not provide ys<text-object>e and ys<text-object>c. If you use |surround.vim|, then the asked for mappings may be easily added if one adds the following lines to ~/.vim/after/ftplugin/tex.vim or any other ftplugin/tex.vim in your |runtimepath|:let b:surround_{char2nr('e')}
\ = "\\begin{\1environment: \1}\n\t\r\n\\end{\1\1}"
let b:surround_{char2nr('c')} = "\\\1command: \1{\r}" Remark also that, by default, |surround.vim| already provides the mapping
ys<text-object>l for encapsulating a text object in a LaTeX environment. Note: Please also read the section vimtex-af-surround!
Note: An alternative is to use
vim-sandwich (see |sandwich.txt| or
https://github.com/machakann/vim-sandwich), which has built-in
support for LaTeX-specific surroundings.Q: Vim throws error when jumping to file with gf.
A: This might be due to the |isfname| setting, which by default contains
{,} on windows. |isfname| is a global option, and can therefore not be set by VimTeX. Suggested solution is to remove {,} from |isfname| by:set isfname-={,}Q: How can I jump from a
\ref{label} to the corresponding label? A: This is not a feature provided by VimTeX itself, but vim has very good support for tag navigation, see tags-and-searches. It is worth mentioning that the ctags support for LaTeX is somewhat lacking. This can be amended by adding some lines to your ~/.ctags configuration file (or .ctags.d/default.ctags if you use Universal ctags), e.g.:--langdef=tex2
--langmap=tex2:.tex
--regex-tex2=/\\label[ \t]*\*?\{[ \t]*([^}]*)\}/\1/l,label/ See [0,1] for references. I also find |gutentags| [2] to be very convenient
for automatically generating and updating tag files.
[0]: http://stackoverflow.com/q/8119405/51634
[1]: https://github.com/lervag/vimtex/issues/348
[2]: https://github.com/ludovicchabant/vim-gutentags
Q: How can I jump from a
\cite{key} to the corresponding bibtex entry? A: This is not a feature provided by VimTeX itself. Similar to vimtex-faq-tags, the feature is available through tags-and-searches. The following ~/.ctags configuration will be useful (or .ctags.d/default.ctags if you use Universal ctags):--langdef=bib
--langmap=bib:.bib
--regex-bib=/^@[A-Za-z]+\{([^,]+),/\1/e,entry/i
--regex-bib=/^@article\{([^,]*)/\1/a,article/i
--regex-bib=/^@book\{([^,]*)/\1/b,book/i
--regex-bib=/^@booklet\{([^,]*)/\1/L,booklet/i
--regex-bib=/^@conference\{([^,]*)/\1/c,conference/i
--regex-bib=/^@inbook\{([^,]*)/\1/B,inbook/i
--regex-bib=/^@incollection\{([^,]*)/\1/C,incollection/i
--regex-bib=/^@inproceedings\{([^,]*)/\1/P,inproceedings/i
--regex-bib=/^@manual\{([^,]*)/\1/m,manual/i
--regex-bib=/^@mastersthesis\{([^,]*)/\1/T,mastersthesis/i
--regex-bib=/^@misc\{([^,]*)/\1/M,misc/i
--regex-bib=/^@phdthesis\{([^,]*)/\1/t,phdthesis/i
--regex-bib=/^@proceedings\{([^,]*)/\1/p,proceedings/i
--regex-bib=/^@string\{([^ "#%')(,=}{]+)/\1/s,string/i
--regex-bib=/^@techreport\{([^,]*)/\1/r,techreport/i
--regex-bib=/^@unpublished\{([^,]*)/\1/u,unpublished/i
Q: How can I change TEXMFHOME?
A: If you change TEXMFHOME in your .bashrc or .zshrc or similar and use gvim invoked from the desktop environment (from menus, hotkeys, etc.), gvim does not know about the new value of TEXMFHOME. The reason for this is that vim invokes shells (e.g. with ! or system()) as non-interactive and non-login shell, which means .bashrc or .zshrc are not read. If you start gvim from an interactive shell which has read .bashrc or .zshrc, gvim inherits these values and therefore they are consistent. One can make the invoked shells interactive by setting |shellcmdflag| to
"-ic". If you want to keep them non-interactive, you can create an
additional shell startup file where you keep your environment variables:
If bash is your default shell, create e.g. the file
.bashenv containing your customized TEXMFHOME variable and add $BASH_ENV=$HOME/.bashenv to
$MYVIMRC and source $HOME/.bashenv to .bashrc [0].
If zsh is your default shell, use
.zshenv for customizing TEXMFHOME. This file is always read by zsh. Nothing has to be added to $MYVIMRC [1].
For more information on how to correctly set environment variables, see e.g.
the SO answer by @Rmano [2].
[0]: https://www.gnu.org/software/bash/manual/html_node/Bash-Startup-Files.html
[1]: http://zsh.sourceforge.net/Intro/intro_3.html
[2]: http://askubuntu.com/a/356973/16395
Q: Does VimTeX support WSL (the Windows Subsystem for Linux)? A: For the moment, rudimentarily, as follows: To set up the viewer, install Sioyek [0], MuPDF [1], or SumatraPDF [2], add the executable to
%PATH%, say by Rapidee [3]. In your vimrc, the linesif has('win32') || (has('unix') && exists('$WSLENV'))
if executable('sioyek.exe')
let g:vimtex_view_method = 'sioyek'
let g:vimtex_view_sioyek_exe = 'sioyek.exe'
let g:vimtex_callback_progpath = 'wsl nvim'
elseif executable('mupdf.exe')
let g:vimtex_view_general_viewer = 'mupdf.exe'
elseif executable('SumatraPDF.exe')
let g:vimtex_view_general_viewer = 'SumatraPDF.exe'
endif
endif make :VimtexView work under Windows, and also under WSL, provided that at
least Windows 10 version 1903 [4] of WSL is installed, the current work dir
and the compiled file is contained in the Linux home directory
$HOME (as
opposed to %USERPROFILE% in Windows). With this configuration, both forward and inverse search should work with
Sioyek and neovim. However, forward search does not work for SumatraPDF. It
seems that, while SumatraPDF is able to find the PDF file in
\\wsl$\<DistroName>\..., the corresponding source file in
\\wsl$\<DistroName>\... is not available. To set up a LaTeX distribution, while reusing that of Windows as proposed
at [5] seems efficient, in practice accessing files on mounted NTFS drives
from WSL is slow, even more so under WSL2 [6]. Therefore a full TeXLive
installation is recommended.
If only basic functionality is required, then a minimal TeXLive
installation, such as TinyTeX [7] or a minimal set of packages to compile
LaTeX as provided by your distribution, is an option, as discussed at [8].
For example, under openSUSE, it suffices to install the packages
texlive-scheme-basic, texlive-latexmk, texlive-collection-fontsrecommended.
[0]: https://sioyek.info/
[1]: https://chocolatey.org/packages/mupdf
[2]: https://chocolatey.org/packages/sumatrapdf
[3]: https://www.rapidee.com/en/about
[4]: https://devblogs.microsoft.com/commandline/whats-new-for-wsl-in-windows-10-version-1903/
[5]: https://github.com/lervag/vimtex/issues/1380
[6]: https://vxlabs.com/2019/12/06/wsl2-io-measurements/
[7]: https://yihui.org/tinytex/
[8]: https://tex.stackexchange.com/questions/397174/minimal-texlive-installation
Q: Does VimTeX work with SumatraPDF from within WSL (the Windows Subsystem for Linux)? A: SumatraPDF expects all path arguments to be in the regular Windows format, for instance
C:\Path\To\file.pdf. When you work in WSL you will typically work in a Bash shell where the corresponding path would be /mnt/c/Path/To/file.pdf. This means SumatraPDF will not work with VimTeX inside WSL without some adjustments. Luckily, there is a CLI tool
wslpath that can be used to translate these
paths. This means we can make VimTeX work with SumatraPDF from within WSL
by writing a simple wrapper script for SumatraPDF and using it instead.
Essentially, we will achieve most of the functionality we expected, except
backward search from SumatraPDF to VimTeX. That may also be possible, but
it is currently uncharted territory. First, create a script called
sumatrapdf.sh and put it under
~/.local/bin inside your WSL environment. You may copy the script
provided by @Liampor on GitHub [0] or write something similar on your own.
The main idea is to let the wrapper of the script do two things:Convert all paths to Windows style paths with
wslpath.
Update the Synctex file (
*.synctex.gz ) correspondingly to allow forward search.
Now, configure the viewer like this:
let g:vimtex_view_general_viewer = '~/.local/bin/sumatrapdf.sh'
let g:vimtex_view_general_options
\ = '-reuse-instance -forward-search @tex @line @pdf'Q: Does Zathura + VimTeX work on macOS?
A: Yes, it works. It is recommended to use the
'zathura_simple' variant. Running Zathura without synctex support should "just work". To do so, set the g:vimtex_view_zathura_use_synctex to 0. To have synctex support, one needs to have
dbus working properly. This
seems to be quite hard, but the following recipe has been reported to work
for some people [0]. However, users with Apple Silicon CPUs have reported
difficulties starting dbus. These users may want to try the more involved
recipe given in [3] if the below procedure fails. The steps assume the user has installed and knows how to use Homebrew [1].
Zathura needs
dbus to work properly. Install it with the following: brew install dbus, or, if it is already installed, reinstall (this
seems necessary for some unknown reason): brew reinstall dbus
The
DBUS_SESSION_BUS_ADDRESS environment variable must be set for Zathura to work with VimTeX; see [2] for details. This can be done by
adding the following to your .bashrc or .zshrc file (or similar):export DBUS_SESSION_BUS_ADDRESS="unix:path=$DBUS_LAUNCHD_SESSION_BUS_SOCKET"
Change the value of
<auth><\auth> in
/usr/local/opt/dbus/share/dbus-1/session.conf from EXTERNAL to
DBUS_COOKIE_SHA1.
Run
brew services start dbus, and use brew services info dbus to double-check that dbus is running. Apple Silicon users that encounter
problems at this step should try the recipe in [3]. Also, if the user
encounter problems with running dbus with brew services start, they
could try to start it manually as explained in [4].
If
brew services start dbus does not work, which seems to be
a relatively common problem, then the user may be more lucky if they try
to start dbus with launchctl:launchctl start org.freedesktop.dbus-session NB: This may need to be done after each restart of your system.
Now install Zathura (most recent version, aka HEAD):
brew tap zegervdv/zathura
brew install girara
brew install zathura --with-synctex
brew install zathura-pdf-poppler
mkdir -p $(brew --prefix zathura)/lib/zathura
ln -s $(brew --prefix zathura-pdf-poppler)/libpdf-poppler.dylib $(brew --prefix zathura)/lib/zathura/libpdf-poppler.dylibReboot and enjoy.
Note: If you already had Zathura and girara installed and things don't
work, then first uninstall and unlink them and try to follow the
above steps from step 1.
[0]: https://github.com/lervag/vimtex/issues/1737#issuecomment-759953886
[1]: https://brew.sh
[2]: https://github.com/lervag/vimtex/issues/2391#issuecomment-1127678531
[3]: https://github.com/zegervdv/homebrew-zathura/issues/99
[4]: https://github.com/lervag/vimtex/issues/2889#issuecomment-1974827512
Q: Does Zathura + VimTeX work on WSL2?
A: Yes, but
systemd or D-Bus must be enabled to make inverse search work properly.systemd is the default for Ubuntu 23.04 running on WSL 2. In this case, there is nothing to do: Zathura and VimTeX should work fine.
Otherwise, if WSL version is 0.67.6 or newer, enable
systemd by adding the following content to the config file /etc/wsl.conf (create the
file if it doesn't exist).[boot]
systemd=true
Then close WSL by using the command
wsl.exe --shutdown in PowerShell to
restart all WSL instances [0].Another approach which also works with older version of WSL consists in setting up
D-Bus daemons that can be shared in all your WSL consoles.
These steps are described in [1].
[0]: https://learn.microsoft.com/en-us/windows/wsl/systemd
[1]: https://x410.dev/cookbook/wsl/sharing-dbus-among-wsl2-consoles/
Q: How does VimTeX compare to Tree-sitter?
A: VimTeX implements a traditional syntax script for syntax highlighting of LaTeX documents (see syntax.txt). The implementation is quite complete and has support for a lot of packages. It should work well for most people. See vimtex-syntax for detailed information.
Tree-sitter [0] is a modern library for incremental parsing of code. Neovim
has built-in support for Tree-sitter (see treesitter). With the
additional Tree-sitter plugin [1], one can have syntax highlighting based
on the Tree-sitter parser. The general benefit of this is that it should be
very fast, and that, for a lot of languages, it can provide high quality
results.
However, it is hard to write a general parser for LaTeX. This is because
LaTeX is a semantic language with a large amount of different commands and
macros from thousands of available packages, many of which make sense to be
highlighted in a different manner. That is, we need to handle a whole lot of
special cases and edge cases!
Furthermore, Tree-sitter highlighting does not currently support
concealing, which many people find useful (see vimtex-syntax-conceal).
Finally, some features of VimTeX relies on the VimTeX syntax highlighting
to work. Examples include the math text objects (e.g. <plug>(vimtex-a$)
and <plug>(vimtex-i$)) as these text objects check the syntax groups to
determine a math region.
Thus, for people who use Tree-sitter, it is strongly advised to disable
Tree-sitter highlighting for LaTeX buffers. This can be done with the
ignore_install option for the setup part of nvim-treesitter, e.g.:require 'nvim-treesitter.configs'.setup {
ignore_install = { "latex" },
-- more stuff here
} Or, alternatively, to only disable the highlighting:
require 'nvim-treesitter.configs'.setup {
highlight = {
enable = true,
disable = { "latex" },
},
-- more stuff here
} Of course, some people may still want to use Tree-sitter for highlighting
regardless of the arguments raised above. In this case, it is advised to
use these options for VimTeX to avoid the startup warning:
let g:vimtex_syntax_enabled = 0
let g:vimtex_syntax_conceal_disable = 1Q: Can I use VimTeX with Markdown plugins that need Tree-sitter enabled? A: Yes, it is possible by using the "additional_vim_regex_highlighting" option in the
nvim-treesitter setup (|nvim-treesitter-highlight-mod|), e.g.:require 'nvim-treesitter.configs'.setup {
ensure_installed = { "markdown" },
highlight = {
enable = true,
disable = { "latex" },
additional_vim_regex_highlighting = { "latex", "markdown" },
},
--other treesitter settings
} Be warned, though, that this will run both Tree-sitter and regex
highlighting in parallel. This negates any performance benefit, and it can
possibly lead to weird results since both types of highlighting are
applied, one over the other.
[0]: https://tree-sitter.github.io/tree-sitter/
[1]: https://github.com/nvim-treesitter/nvim-treesitter
[2]: https://github.com/nvim-treesitter/nvim-treesitter#available-modules
Q: How do I set up VimTeX to work with TeXShop and MacVim (macOS)? A: Start by reading the section on vimtex-view-texshop. The examples below apply to MacVim [0].
Here is an example
/usr/local/bin/othereditor script that uses the
VimTeX's convenience function VimtexInverseSearch:#!/bin/bash
/usr/local/bin/mvim -v --not-a-term -T dumb -c "VimtexInverseSearch $1 '$2'" The call with the convenience function
VimtexInverseSearch offers the
advantage that the buffer will be found if it is open, regardless of how it
was opened. If the buffer is not active, in a hidden tab or if the file is
not open at all, inverse search will fail. As an alternative to
VimtexInverseSearch, if the .tex document was opened
with the --remote-silent option, i.e.:/usr/local/bin/mvim --remote-silent foo.tex the following script
/usr/local/bin/othereditor can be used instead:#!/bin/bash
/usr/local/bin/mvim --remote-silent +$1 "$2" It will activate the buffer even if it is hidden, or open the file if it is
not yet loaded in any buffer.
The TeXShop release notes for versions 4.24 and 4.25 contain information
about backward searches with external editors in general [1].
[0]: https://macvim-dev.github.io/macvim/
[1]: https://pages.uoregon.edu/koch/texshop/changes_3.html
TROUBLESHOOTING
Here are some pitfalls that one may experience if one of these assumptions are broken:
Completion may not work properly for exotic file encodings, such as for UTF-16LE (see https://github.com/lervag/vimtex/issues/615)
With different operating systems and different plugin configurations, there are a few things that must be considered for system interoperability. A number of common problems and suggested solutions are included in the following troubleshooting section.
Problem: Continuous compilation seems to hang
Upon starting continuous compilation the status bar indicates "VimTeX: Compiler started in continuous mode", but the compilation never terminates and the quickfix window does not load.
Tips:
Ensure that a latexmk process and a Perl process have started. If they have not been started, then these two programs may not be accessible given your
operating system's PATH environment variable.
Ensure that the option
-interaction=nonstopmode is provided to latexmk. This is done by default by VimTeX, unless the user provides custom options
through g:vimtex_compiler_latexmk (see the options key). In the latter
case, the user must ensure that the said option is also provided.
Problem: Text objects on Windows
In Windows, environment text object commands, like
vae and vie, do not select the entire body of the environment. More specifically, given:\begin{someenv}
some content
\end{someenv}
The command dae results in:}and
die results in:\begin{someenv}
t
\end{someenv}
Solution: It seems that vim for Windows comes with some options set by default in the vimrc file. One of these has been reported to be :behave mswin (see |:behave|) which, among other things, sets the 'selection' option to "exclusive". This can be ameliorated by pursuing one of two options:Add
:behave xterm to your vimrc file.
Add
:set selection=inclusive to your vimrc file.
Problem: Typing
<Tab> or <C-n> causes Vim to hang before making a completionVimTeX may be scanning included files with
kpsewhich while collecting completion candidates for keyword completion. Try disabling this feature by setting g:vimtex_include_search_enabled to 0 in your vimrc:let g:vimtex_include_search_enabled = 0
Note: Plugins like |supertab| [0], which often maps the i_<Tab> key, will typically use keyword completion "behind the scenes" to gather completion candidates.CREDITS
VimTeX is developed by Karl Yngve Lervåg <karl.yngve@gmail.com>, and is distributed under the MIT license. The project is available as a Git repository: https://github.com/lervag/vimtex.
VimTeX was developed from scratch, but much of the code has been based on LaTeX-Box: https://github.com/LaTeX-Box-Team/LaTeX-Box. LaTeX-suite was also an inspiration: http://vim-latex.sourceforge.net/.
I do accept donations through PayPal (see link below [0]). As there are no expenses related to VimTeX (except time), any money I receive would be spent on coffee, beer or chocolate. These things make me happy. However, I will also be happy if one should choose to donate to a charity instead, as there are a lot of people in more need of money than me! Examples of charities may be ICCF (the organisation that Vim specifically supports) or Medicins sans Frontieres [1]. Feel free to let me know if you should donate to a charity due to VimTeX, as I would be happy to hear of it.
[0]: https://www.paypal.com/cgi-bin/webscr?cmd=_s-xclick&hosted_button_id=5N4MFVXN7U8NW
[1]: https://www.msf.org/
CHANGELOG
The following changelog only logs particularly important changes, such as changes that break backwards compatibility. See the git log for the detailed changelog.
2022-11-06: Better main file detection algorithm
The detection whether a file is a main file has seen gradual improvements. In course of this, g:vimtex_disable_recursive_main_file_detection has been removed.2021-10-25: Better inverse search
Deprecate g:vimtex_compiler_progname as it is no longer necessary.2021-10-09: Better options for syntax conceal
Deprecate g:vimtex_syntax_conceal_default in favor of g:vimtex_syntax_conceal_disable. The new option makes things more explicit and makes better sense (at least to me).2020-11-16: More flexible package syntax options
Deprecate g:vimtex_syntax_autoload_packages in favor of g:vimtex_syntax_packages, which allows more fine grained control over each package.2020-09-24: More concise grammar options
Deprecate g:vimtex_textidote_jar in favor of g:vimtex_grammar_textidote.2020-08-11: Remove g:vimtex_quickfix_latexlog
The option g:vimtex_quickfix_latexlog was deprecated in favor of the more general mechanism provided by g:vimtex_quickfix_ignore_filters.2020-07-31: Use events for callback hooks
The events VimtexEventCompileSuccess, VimtexEventCompileFailed, and VimtexEventView have been added to make it easier to hook personal customizations. This deprecates the following options: * g:vimtex_compiler_callback_hooks * g:vimtex_view_general_callback * g:vimtex_view_general_hook_callback * g:vimtex_view_general_hook_view * g:vimtex_view_mupdf_hook_callback * g:vimtex_view_mupdf_hook_view * g:vimtex_view_skim_hook_callback * g:vimtex_view_skim_hook_view * g:vimtex_view_zathura_hook_callback * g:vimtex_view_zathura_hook_view2020-07-19: Released version 1.0
Version 1.0 (and earlier) works on Vim 7.4 and with neovim 0.1.7. Later versions require Vim 8.0 or neovim 0.4.3.2018-08-15: Refactored the ToC interface
I've made a large update to the code for the ToC window in order to simplify and unify the interface. In the new version, g:vimtex_toc_config replaces all of the following options: * g:vimtex_index_split_width * g:vimtex_index_split_pos * g:vimtex_index_show_help * g:vimtex_index_resize * g:vimtex_index_hide_line_numbers * g:vimtex_index_mode * g:vimtex_toc_layers * g:vimtex_toc_fold * g:vimtex_toc_fold_level_start * g:vimtex_toc_hotkeys * g:vimtex_toc_refresh_always * g:vimtex_toc_show_numbers * g:vimtex_toc_tocdepth2017-07-27: Major refactoring of the folding feature
I've made a large update to the code for folding. The configuration of the various folded elements is now done through a single option: g:vimtex_fold_types.Deprecated options:
* g:vimtex_fold_comments
* g:vimtex_fold_preamble
* g:vimtex_fold_envs
* g:vimtex_fold_env_blacklist
* g:vimtex_fold_env_whitelist
* g:vimtex_fold_markers
* g:vimtex_fold_parts
* g:vimtex_fold_sections
* g:vimtex_fold_commands
* g:vimtex_fold_commands_default
vimtex-lacheck
2017-06-05: Removed Lacheck support
Removed support for using lacheck for checking LaTeX syntax. The reason is that there exist several (good) external plugins for syntax checking files. These are general purpose plugins that work for multiple file types. For more info, see vimtex-and-friends.2017-05-20: Updated TOC options
There's been a few updates to the TOC. During this work, I removed some unnecessary options.Deprecated options:
* g:vimtex_toc_fold_levels (was not necessary)
* g:vimtex_toc_number_width (see g:vimtex_toc_tocdepth)
2017-03-31: Refactored quickfix related features
I've added a more general layer for handling different error parsers. Currently there are few or now changes from the user point of view, but this should make it possible to add other methods for showing errors in a LaTeX project than the current one that parses the .log file directly.Deprecated options:
* g:vimtex_quickfix_warnings (see g:vimtex_quickfix_latexlog)
2017-03-28: Major refactoring of initialization
Added a general compiler interface, see vimtex-compiler. To configure the latexmk compiler, see g:vimtex_compiler_latexmk.Deprecated options:
* g:vimtex_latexmk_enabled (use g:vimtex_compiler_enabled) * g:vimtex_latexmk_progname (use g:vimtex_compiler_progname) * g:vimtex_latexmk_callback_hooks (use g:vimtex_compiler_callback_hooks) * g:vimtex_latexmk_callback * g:vimtex_latexmk_autojump * g:vimtex_latexmk_continuous * g:vimtex_latexmk_background * g:vimtex_latexmk_options
Deprecated commands:
* VimtexCompileToggle (use :VimtexCompile)
2017-03-28: Major refactoring of initialization
The initialization has been refactored in order to provide a more consistent separation of buffer initialization and state initialization. This has no major consequence for users, but it makes maintenance and further development easier.2017-03-02: Changed how to set ignored warnings
I'm updating the changelog to notify of a change to the quickfix settings.Deprecated options:
* g:vimtex_quickfix_ignore_all_warnings
* g:vimtex_quickfix_ignored_warnings
See instead:
g:vimtex_quickfix_warnings
2016-05-31: A lot of things have updated
I know that people would like to see a simple list of changes. Unfortunately, I am bad at keeping this changelog updated. All details are available in the git log, though. The reason I added this entry is to note that I have removed an option:* g:vimtex_env_complete_list --- It is no longer necessary. Completion candidates are instead parsed from the project.
2016-02-06: Large refactoring of delimiter parsing
I've refactored a lot of the code in order to make the parsing of delimiters and features that rely on delimiter detection and similar more consistent. This results in some changes in option names and similar, but it should make it easier to provide improved and more robust features.There is one feature change: The delimiter toggle now consistently toggles the modifier, not the delimiter itself, and it toggles between a range of modifiers by default. For customization, see g:vimtex_delim_toggle_mod_list.
The following options have changed names:
* g:vimtex_change_set_formatexpr ---> g:vimtex_format_enabled * g:vimtex_change_complete_envs ---> g:vimtex_env_complete_list * g:vimtex_change_toggled_delims ---> g:vimtex_delim_toggle_mod_list
The following options have been removed:
* g:vimtex_change_ignored_delims_pattern --- It was no longer necessary
The following mappings have been renamed:
* <plug>(vimtex-delete-env) ---> <plug>(vimtex-env-delete)
* <plug>(vimtex-delete-cmd) ---> <plug>(vimtex-cmd-delete)
* <plug>(vimtex-change-env) ---> <plug>(vimtex-env-change)
* <plug>(vimtex-change-cmd) ---> <plug>(vimtex-cmd-change)
* <plug>(vimtex-toggle-star) ---> <plug>(vimtex-env-toggle-star) * <plug>(vimtex-toggle-delim) ---> <plug>(vimtex-delim-toggle-modifier) * <plug>(vimtex-create-cmd) ---> <plug>(vimtex-cmd-create)
* <plug>(vimtex-close-env) ---> <plug>(vimtex-delim-close)
2015-10-19: Added convenient insert mode mappings
I've merged the math_mappings branch (see #172 and #251). It adds the feature that is explained in vimtex-imaps.2015-06-06: Minor but convenient restructuring (++)
I've changed a lot of the code structure in relatively small ways. For instance, instead of referring to the particular data blobs through the global array, I instead linked a buffer variable to the correct global array element.One particular change is that all modules are now initialized in three steps:
Initialize module options
Initialize script variables and single execution functionalities
Initialize buffer options
Finally, I've cleaned up a lot of the code by removing some deprecation warnings and similar.
2015-03-21: Implemented index buffers, deprecated vimtex_toc filetype
The system for displaying the table of content relied on a dedicated filetype plugin. This was inherited from LaTeX-Box, and worked quite well. However, I intend to implement more functionality that uses the same kind of buffer to display similar things, such as a list of labels. I realized I wanted the ToC window to be more adaptable, so I implemented the index interface for such buffers. The index in itself may be used to create ToC-like buffers with simple actions. The vimtex-toc uses and expands the index in such a way that the changes should barely be noticeable from the user perspective. Note however the following variable name changes:* g:vimtex_toc_numbers_width ---> g:vimtex_toc_number_width
* g:vimtex_toc_hide_preamble ---> g:vimtex_toc_show_preamble
* g:vimtex_toc_numbers ---> g:vimtex_toc_show_numbers
* g:vimtex_toc_hide_line_numbers ---> g:vimtex_index_hide_line_numbers * g:vimtex_toc_resize ---> g:vimtex_index_resize
* g:vimtex_toc_hide_help ---> g:vimtex_index_show_help
* g:vimtex_toc_split_pos ---> |g:vimtex_index_split|
* g:vimtex_toc_width -/
vim-latex-namechange
2015-03-08: Changed the name to VimTeX
The old name vim-latex was already used by LaTeX-Suite. I was not aware of the name clash in the beginning. Due to the rising popularity of this plugin, it has become clear that such a name clash is very inconvenient. The present change is therefore very much needed.The name change is reflected throughout the plugin in the names of commands, mappings, functions, and options. People should update their
vimrc settings accordingly. For instance, every option name should be changed fromg:latex_... = ...to
g:vimtex_... = ...
2014-12-07: Added more general view functionality
Added new module for view functionality. This allows more complex view functions (and commands), for instance to do forward (and possibly inverse) searching through synctex. In the first version, I added forward search for mupdf by use of the synctex command and xdotool.The
g:latex_viewer option has now been deprecated. Instead one should use g:vimtex_view_method and g:vimtex_view_general_viewer.Deprecated option:
* g:latex_viewer
2014-06-13: Changed some option names
Some VimTeX option names were changed in an attempt to make the names more consistent. These options are listed here for reference: * g:latex_errorformat_ignore_warnings * g:latex_errorformat_show_warnings * g:latex_latexmk_autojump * g:latex_latexmk_quickfixThe new names are, respectively:
* g:vimtex_quickfix_ignored_warnings
* g:vimtex_quickfix_ignore_all_warnings
* g:vimtex_quickfix_autojump
* g:vimtex_quickfix_mode
2013-10-05: First public release
VimTeX was first released on github on this date. The initial version was named vim-latex, which conflicted with Vim LaTeX-Suite which is also known as vim-latex.