Skip to content

af12066/nvim-lsp

 
 

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

nvim-lsp

Collection of common configurations for the Nvim LSP client.

It is hoped that these configurations serve as a "source of truth", but they are strictly best effort. If something doesn't work, these configs are useful as a starting point, which you can adjust to fit your environment.

This is work-in-progress and requires Nvim HEAD/nightly. Update Nvim and nvim-lsp before reporting an issue.

Contributions are welcome!

There are many language servers in the world, and not enough time. Help us create configs for all the LSPs!

  1. Read CONTRIBUTING.md for instructions. Ask questions in Neovim Gitter.
  2. Choose a language from the coc.nvim wiki or the emacs-lsp project.
  3. Create a new file at lua/nvim_lsp/SERVER_NAME.lua. See existing configs for examples (lua/nvim_lsp/texlab.lua is an extensive example).

Install

  • Requires Nvim HEAD/nightly (v0.5 prerelease).
  • nvim-lsp is just a plugin. Install it like any other Vim plugin.
    :Plug 'neovim/nvim-lsp'
    

Usage

Each config provides a setup() function, to initialize the server with reasonable defaults and some server-specific things like commands or different diagnostics.

require'nvim_lsp'.<config>.setup{name=…, settings = {…}, …}

Find the config for your language, then paste the example given there to your init.vim. All examples are given in Lua, see :help :lua-heredoc to use Lua from your init.vim.

Some configs may define additional server-specific functions, e.g. the texlab config provides nvim_lsp.texlab.buf_build({bufnr}).

Example: using the defaults

To use the defaults, just call setup() with an empty config parameter. For the gopls config, that would be:

require'nvim_lsp'.gopls.setup{}

Example: override some defaults

To set some config properties at setup(), specify their keys. For example to change how the "project root" is found, set the root_dir key:

local nvim_lsp = require'nvim_lsp'
nvim_lsp.gopls.setup{
  root_dir = nvim_lsp.util.root_pattern('.git');
}

The documentation for each config lists default values and additional optional properties.

local nvim_lsp = require'nvim_lsp'
nvim_lsp.texlab.setup{
  name = 'texlab_fancy';
  log_level = vim.lsp.protocol.MessageType.Log;
  settings = {
    latex = {
      build = {
        onSave = true;
      }
    }
  }
}

Example: custom config

To configure a custom/private server, just require nvim_lsp/skeleton and do the same as we do if we were adding it to the repository itself.

  1. Define the config: configs.foo_lsp = { … }
  2. Call setup(): require'nvim_lsp'.foo_lsp.setup{}
local nvim_lsp = require'nvim_lsp'
local configs = require'nvim_lsp/skeleton'
-- Check if it's already defined for when I reload this file.
if not nvim_lsp.foo_lsp then
  configs.foo_lsp = {
    default_config = {
      cmd = {'/home/ashkan/works/3rd/lua-language-server/run.sh'};
      filetypes = {'lua'};
      root_dir = function(fname)
        return nvim_lsp.util.find_git_ancestor(fname) or vim.loop.os_homedir()
      end;
      log_level = vim.lsp.protocol.MessageType.Warning;
      settings = {};
    };
  }
end
nvim_lsp.foo_lsp.setup{}

Installing a language server

Configs may provide an install() function. Then you can use :LspInstall {name} to install the required language server. For example, to install the Elm language server:

:LspInstall elmls

Use :LspInstallInfo to see install info.

:LspInstallInfo

setup() function

The setup() interface:

nvim_lsp.SERVER.setup{config}

  The `config` parameter has the same shape as that of
  |vim.lsp.start_client()|, with these additions and changes:

  {root_dir}
    Required for some servers, optional for others.
    Function of the form `function(filename, bufnr)`.
    Called on new candidate buffers being attached-to.
    Returns either a root_dir or nil.

    If a root_dir is returned, then this file will also be attached. You
    can optionally use {filetype} to help pre-filter by filetype.

    If a root_dir is returned which is unique from any previously returned
    root_dir, a new server will be spawned with that root_dir.

    If nil is returned, the buffer is skipped.

    See |nvim_lsp.util.search_ancestors()| and the functions which use it:
    - |nvim_lsp.util.root_pattern(patterns...)| finds an ancestor which
    - contains one of the files in `patterns...`. This is equivalent
      to coc.nvim's "rootPatterns"
    - Related utilities for common tools:
      - |nvim_lsp.util.find_git_root()|
      - |nvim_lsp.util.find_node_modules_root()|
      - |nvim_lsp.util.find_package_json_root()|

  {name}
    Defaults to the server's name.

  {filetypes}
    Set of filetypes to filter for consideration by {root_dir}.
    May be empty.
    Server may specify a default value.

  {log_level}
    controls the level of logs to show from build processes and other
    window/logMessage events. Defaults to
    vim.lsp.protocol.MessageType.Warning instead of
    vim.lsp.protocol.MessageType.Log.

  {settings}
    Map with case-sensitive keys corresponding to `workspace/configuration`
    event responses.
    We also notify the server *once* on `initialize` with
    `workspace/didChangeConfiguration`.
    If you change the settings later on, you must emit the notification
    with `client.workspace_did_change_configuration({settings})`
    Example: `settings = { keyName = { subKey = 1 } }`

  {on_attach}
    `function(client)` executed with the current buffer as the one the {client}
    is being attached-to. This is different from
    |vim.lsp.start_client()|'s on_attach parameter, which passes the {bufnr} as
    the second parameter instead. Useful for doing buffer-local setup.

  {on_new_config}
    `function(new_config)` will be executed after a new configuration has been
    created as a result of {root_dir} returning a unique value. You can use this
    as an opportunity to further modify the new_config or use it before it is
    sent to |vim.lsp.start_client()|.

Configurations

The following LSP configs are included. Follow a link to find documentation for that config.

bashls

https://github.com/mads-hartmann/bash-language-server

Language server for bash, written using tree sitter in typescript.

Can be installed in Nvim with :LspInstall bashls

require'nvim_lsp'.bashls.setup{}

  Default Values:
    cmd = { "bash-language-server", "start" }
    filetypes = { "sh" }
    log_level = 2
    root_dir = vim's starting directory
    settings = {}

ccls

https://github.com/MaskRay/ccls/wiki

ccls relies on a JSON compilation database specified as compile_commands.json or, for simpler projects, a compile_flags.txt.

This server accepts configuration via the settings key.

Available settings:
  • ccls.cache.directory: string

    Default: ".ccls-cache"

    Absolute or relative (from the project root) path to the directory that the cached index will be stored in. Try to have this directory on an SSD. If empty, cached indexes will not be saved on disk.

    ${workspaceFolder} will be replaced by the folder where .vscode/settings.json resides.

    Cache directories are project-wide, so this should be configured in the workspace settings so multiple indexes do not clash.

    Example value: "/work/ccls-cache/chrome/"

  • ccls.cache.hierarchicalPath: boolean

    If false, store cache files as $directory/@a@b/c.cc.blob

    If true, $directory/a/b/c.cc.blob.

  • ccls.callHierarchy.qualified: boolean

    Default: true

    If true, use qualified names in the call hiearchy

  • ccls.clang.excludeArgs: array

    Default: {}

    An set of command line arguments to strip before passing to clang when indexing. Each list entry is a separate argument.

  • ccls.clang.extraArgs: array

    Default: {}

    An extra set of command line arguments to give clang when indexing. Each list entry is a separate argument.

  • ccls.clang.pathMappings: array

    Default: {}

    Translate paths in compile_commands.json entries, .ccls options and cache files. This allows to reuse cache files built otherwhere if the source paths are different.

  • ccls.clang.resourceDir: string

    Default: ""

    Default value to use for clang -resource-dir argument. This will be automatically supplied by ccls if not provided.

  • ccls.codeLens.enabled: boolean

    Default: true

    Specifies whether the references CodeLens should be shown.

  • ccls.codeLens.localVariables: boolean

    Set to false to hide code lens on parameters and function local variables.

  • ccls.codeLens.renderInline: boolean

    Enables a custom code lens renderer so code lens are displayed inline with code. This removes any vertical-space footprint at the cost of horizontal space.

  • ccls.completion.caseSensitivity: integer

    Default: 2

    Case sensitivity when code completion is filtered. 0: case-insensitive, 1: case-folded, i.e. insensitive if no input character is uppercase, 2: case-sensitive

  • ccls.completion.detailedLabel: boolean

    When this option is enabled, the completion item label is very detailed, it shows the full signature of the candidate.

  • ccls.completion.duplicateOptional: boolean

    For functions with default arguments, generate one more item per default argument.

  • ccls.completion.enableSnippetInsertion: boolean

    If true, parameter declarations are inserted as snippets in function/method call arguments when completing a function/method call

  • ccls.completion.include.blacklist: array

    Default: {}

    ECMAScript regex that checks absolute path. If it partially matches, the file is not added to include path auto-complete. An example is "/CACHE/"

  • ccls.completion.include.maxPathSize: integer

    Default: 37

    Maximum length for path in #include proposals. If the path length goes beyond this number it will be elided. Set to 0 to always display the full path.

  • ccls.completion.include.suffixWhitelist: array

    Default: { ".h", ".hpp", ".hh" }

    Only files ending in one of these values will be shown in include auto-complete. Set to the empty-list to disable include auto-complete.

  • ccls.completion.include.whitelist: array

    Default: {}

    ECMAScript regex that checks absolute file path. If it does not partially match, the file is not added to include path auto-complete. An example is "/src/"

  • ccls.diagnostics.blacklist: array

    Default: {}

    Files that match these patterns won't be displayed in diagnostics view.

  • ccls.diagnostics.onChange: integer

    Default: 1000

    Time in milliseconds to wait before computing diagnostics on type. -1: disable diagnostics on type.

  • ccls.diagnostics.onOpen: integer

    Default: 0

    Time in milliseconds to wait before computing diagnostics when a file is opened.

  • ccls.diagnostics.onSave: integer

    Default: 0

    Time in milliseconds to wait before computing diagnostics when a file is saved.

  • ccls.diagnostics.spellChecking: boolean

    Default: true

    Whether to do spell checking on undefined symbol names when computing diagnostics.

  • ccls.diagnostics.whitelist: array

    Default: {}

    Files that match these patterns will be displayed in diagnostics view.

  • ccls.highlight.blacklist: array|null

    Default: vim.NIL

    Files that match these patterns won't have semantic highlight.

  • ccls.highlight.enum.face: array

    Default: { "variable", "member" }

  • ccls.highlight.function.colors: array

    Default: { "#e5b124", "#927754", "#eb992c", "#e2bf8f", "#d67c17", "#88651e", "#e4b953", "#a36526", "#b28927", "#d69855" }

    Colors to use for semantic highlight. A good generator is http://tools.medialab.sciences-po.fr/iwanthue/. If multiple colors are specified, semantic highlight will cycle through them for successive symbols.

  • ccls.highlight.function.face: array

    Default: {}

  • ccls.highlight.global.face: array

    Default: { "fontWeight: bolder" }

  • ccls.highlight.globalVariable.face: array

    Default: { "variable", "global" }

  • ccls.highlight.largeFileSize: integer|null

    Default: vim.NIL

    Disable semantic highlight for files larger than the size.

  • ccls.highlight.macro.colors: array

    Default: { "#e79528", "#c5373d", "#e8a272", "#d84f2b", "#a67245", "#e27a33", "#9b4a31", "#b66a1e", "#e27a71", "#cf6d49" }

    Colors to use for semantic highlight. A good generator is http://tools.medialab.sciences-po.fr/iwanthue/. If multiple colors are specified, semantic highlight will cycle through them for successive symbols.

  • ccls.highlight.macro.face: array

    Default: { "variable" }

  • ccls.highlight.member.face: array

    Default: { "fontStyle: italic" }

  • ccls.highlight.memberFunction.face: array

    Default: { "function", "member" }

  • ccls.highlight.memberVariable.face: array

    Default: { "variable", "member" }

  • ccls.highlight.namespace.colors: array

    Default: { "#429921", "#58c1a4", "#5ec648", "#36815b", "#83c65d", "#417b2f", "#43cc71", "#7eb769", "#58bf89", "#3e9f4a" }

    Colors to use for semantic highlight. A good generator is http://tools.medialab.sciences-po.fr/iwanthue/. If multiple colors are specified, semantic highlight will cycle through them for successive symbols.

  • ccls.highlight.namespace.face: array

    Default: { "type" }

  • ccls.highlight.parameter.face: array

    Default: { "variable" }

  • ccls.highlight.static.face: array

    Default: { "fontWeight: bold" }

  • ccls.highlight.staticMemberFunction.face: array

    Default: { "function", "static" }

  • ccls.highlight.staticMemberVariable.face: array

    Default: { "variable", "static" }

  • ccls.highlight.staticVariable.face: array

    Default: { "variable", "static" }

  • ccls.highlight.type.colors: array

    Default: { "#e1afc3", "#d533bb", "#9b677f", "#e350b6", "#a04360", "#dd82bc", "#de3864", "#ad3f87", "#dd7a90", "#e0438a" }

    Colors to use for semantic highlight. A good generator is http://tools.medialab.sciences-po.fr/iwanthue/. If multiple colors are specified, semantic highlight will cycle through them for successive symbols.

  • ccls.highlight.type.face: array

    Default: {}

  • ccls.highlight.typeAlias.face: array

    Default: { "type" }

  • ccls.highlight.variable.colors: array

    Default: { "#587d87", "#26cdca", "#397797", "#57c2cc", "#306b72", "#6cbcdf", "#368896", "#3ea0d2", "#48a5af", "#7ca6b7" }

    Colors to use for semantic highlight. A good generator is http://tools.medialab.sciences-po.fr/iwanthue/. If multiple colors are specified, semantic highlight will cycle through them for successive symbols.

  • ccls.highlight.variable.face: array

    Default: {}

  • ccls.highlight.whitelist: array|null

    Default: vim.NIL

    Files that match these patterns will have semantic highlight.

  • ccls.index.blacklist: array

    Default: {}

    A translation unit (cc/cpp file) is not indexed if any of the ECMAScript regexes in this list partially matches translation unit's the absolute path.

  • ccls.index.initialBlacklist: array

    Default: {}

    Files matched by the regexes should not be indexed on initialization. Indexing is deferred to when they are opened.

  • ccls.index.initialWhitelist: array

    Default: {}

    Files matched by the regexes should be indexed on initialization.

  • ccls.index.maxInitializerLines: integer

    Default: 15

    Number of lines of the initializer / macro definition showed in hover.

  • ccls.index.multiVersion: integer

    Default: 0

    If not 0, a file will be indexed in each tranlation unit that includes it.

  • ccls.index.onChange: boolean

    Allow indexing on textDocument/didChange. May be too slow for big projects, so it is off by default.

  • ccls.index.threads: number

    Default: 0

    Number of indexer threads. If 0, 80% of cores are used.

  • ccls.index.trackDependency: integer

    Default: 2

    Whether to reparse a file if write times of its dependencies have changed. The file will always be reparsed if its own write time changes.

    0: no, 1: only during initial load of project, 2: yes

  • ccls.index.whitelist: array

    Default: {}

    If a translation unit's absolute path partially matches any ECMAScript regex in this list, it will be indexed. The whitelist takes priority over the blacklist. To only index files in the whitelist, make "ccls.index.blacklist" match everything, ie, set it to ".*".

  • ccls.launch.args: array

    Default: {}

    Array containing extra arguments to pass to the ccls binary

  • ccls.launch.command: string

    Default: "ccls"

    Path to the ccls binary (default assumes the binary is in the PATH)

  • ccls.misc.compilationDatabaseCommand: string

    Default: ""

    If not empty, the compilation database command to use

  • ccls.misc.compilationDatabaseDirectory: string

    Default: ""

    If not empty, the compilation database directory to use instead of the project root

  • ccls.misc.showInactiveRegions: boolean

    Default: true

    If true, ccls will highlight skipped ranges.

  • ccls.statusUpdateInterval: integer

    Default: 2000

    Interval between updating ccls status in milliseconds. Set to 0 to disable.

  • ccls.theme.dark.skippedRange.backgroundColor: string

    Default: "rgba(18, 18, 18, 0.3)"

    CSS color to apply to the background when the code region has been disabled by the preprocessor in a dark theme.

  • ccls.theme.dark.skippedRange.textColor: string

    Default: "rgb(100, 100, 100)"

    CSS color to apply to text when the code region has been disabled by the preprocessor in a dark theme.

  • ccls.theme.light.skippedRange.backgroundColor: string

    Default: "rgba(220, 220, 220, 0.3)"

    CSS color to apply to the background when the code region has been disabled by the preprocessor in a light theme.

  • ccls.theme.light.skippedRange.textColor: string

    Default: "rgb(100, 100, 100)"

    CSS color to apply to text when the code region has been disabled by the preprocessor in a light theme.

  • ccls.trace.websocketEndpointUrl: string

    Default: ""

    When set, logs all LSP messages to specified WebSocket endpoint.

  • ccls.treeViews.doubleClickTimeoutMs: number

    Default: 500

    If a tree view entry is double-clicked within this timeout value, vscode will navigate to the entry.

  • ccls.workspaceSymbol.caseSensitivity: integer

    Default: 1

    Case sensitivity when searching workspace symbols. 0: case-insensitive, 1: case-folded, i.e. insensitive if no input character is uppercase, 2: case-sensitive

  • ccls.workspaceSymbol.maxNum: number|null

    Default: vim.NIL

    The maximum number of global search (ie, Ctrl+P + #foo) search results to report. For small search strings on large projects there can be a massive number of results (ie, over 1,000,000) so this limit is important to avoid extremely long delays. null means use the default value provided by the ccls language server.

require'nvim_lsp'.ccls.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "ccls" }
    filetypes = { "c", "cpp", "objc", "objcpp" }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("compile_commands.json", "compile_flags.txt", ".git")
    settings = {}

clangd

https://clang.llvm.org/extra/clangd/Installation.html

NOTE: Clang >= 9 is recommended! See this issue for more.

clangd relies on a JSON compilation database specified as compile_commands.json or, for simpler projects, a compile_flags.txt.

require'nvim_lsp'.clangd.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "clangd", "--background-index" }
    filetypes = { "c", "cpp", "objc", "objcpp" }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("compile_commands.json", "compile_flags.txt", ".git")
    settings = {}

cssls

https://github.com/vscode-langservers/vscode-css-languageserver-bin

css-languageserver can be installed via :LspInstall cssls or by yourself with npm:

npm install -g vscode-css-languageserver-bin

Can be installed in Nvim with :LspInstall cssls

require'nvim_lsp'.cssls.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "css-languageserver", "--stdio" }
    filetypes = { "css", "scss", "less" }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("package.json")
    settings = {
      css = {
        validate = true
      },
      less = {
        validate = true
      },
      scss = {
        validate = true
      }
    }

dockerls

https://github.com/rcjsuen/dockerfile-language-server-nodejs

docker-langserver can be installed via :LspInstall dockerls or by yourself with npm:

npm install -g dockerfile-language-server-nodejs

Can be installed in Nvim with :LspInstall dockerls

require'nvim_lsp'.dockerls.setup{}

  Default Values:
    cmd = { "docker-langserver", "--stdio" }
    filetypes = { "Dockerfile", "dockerfile" }
    log_level = 2
    root_dir = root_pattern("Dockerfile")
    settings = {}

elmls

https://github.com/elm-tooling/elm-language-server#installation

If you don't want to use Nvim to install it, then you can use:

npm install -g elm elm-test elm-format @elm-tooling/elm-language-server

Can be installed in Nvim with :LspInstall elmls This server accepts configuration via the settings key.

Available settings:
  • elmLS.elmAnalyseTrigger: enum { "change", "save", "never" }

    Default: "change"

    When do you want the extension to run elm-analyse? Might need a restart to take effect.

  • elmLS.elmFormatPath: string

    Default: ""

    The path to your elm-format executable. Should be empty by default, in that case it will assume the name and try to first get it from a local npm installation or a global one. If you set it manually it will not try to load from the npm folder.

  • elmLS.elmPath: string

    Default: ""

    The path to your elm executable. Should be empty by default, in that case it will assume the name and try to first get it from a local npm installation or a global one. If you set it manually it will not try to load from the npm folder.

  • elmLS.elmTestPath: string

    Default: ""

    The path to your elm-test executable. Should be empty by default, in that case it will assume the name and try to first get it from a local npm installation or a global one. If you set it manually it will not try to load from the npm folder.

  • elmLS.trace.server: enum { "off", "messages", "verbose" }

    Default: "off"

    Traces the communication between VS Code and the language server.

require'nvim_lsp'.elmls.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "elm-language-server" }
    filetypes = { "elm" }
    init_options = {
      elmAnalyseTrigger = "change",
      elmFormatPath = "elm-format",
      elmPath = "elm",
      elmTestPath = "elm-test"
    }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("elm.json")
    settings = {}

flow

https://flow.org/ https://github.com/facebook/flow

See below for how to setup Flow itself. https://flow.org/en/docs/install/

See below for lsp command options.

npm run flow lsp -- --help

This server accepts configuration via the settings key.

Available settings:
  • flow.coverageSeverity: enum { "error", "warn", "info" }

    Default: "info"

    Type coverage diagnostic severity

  • flow.enabled: boolean

    Default: true

    Is flow enabled

  • flow.fileExtensions: array

    Default: { ".js", ".mjs", ".jsx", ".flow", ".json" }

    Array items: {type = "string"}

    (Supported only when useLSP: false). File extensions to consider for flow processing

  • flow.lazyMode: string

    Default: vim.NIL

    Set value to enable flow lazy mode

  • flow.logLevel: enum { "error", "warn", "info", "trace" }

    Default: "info"

    Log level for output panel logs

  • flow.pathToFlow: string

    Default: "flow"

    Absolute path to flow binary. Special var ${workspaceFolder} or ${flowconfigDir} can be used in path (NOTE: in windows you can use '/' and can omit '.cmd' in path)

  • flow.runOnAllFiles: boolean

    (Supported only when useLSP: false) Run Flow on all files, No need to put //@flow comment on top of files.

  • flow.runOnEdit: boolean

    Default: true

    If true will run flow on every edit, otherwise will run only when changes are saved (Note: 'useLSP: true' only supports syntax errors)

  • flow.showStatus: boolean

    Default: true

    (Supported only when useLSP: false) If true will display flow status is the statusbar

  • flow.showUncovered: boolean

    If true will show uncovered code by default

  • flow.stopFlowOnExit: boolean

    Default: true

    Stop Flow on Exit

  • flow.trace.server

    Default: "off"

    Traces the communication between VSCode and the flow lsp service.

  • flow.useBundledFlow: boolean

    Default: true

    If true will use flow bundled with this plugin if nothing works

  • flow.useCodeSnippetOnFunctionSuggest: boolean

    Default: true

    Complete functions with their parameter signature.

  • flow.useLSP: boolean

    Default: true

    Turn off to switch from the official Flow Language Server implementation to talking directly to flow.

  • flow.useNPMPackagedFlow: boolean

    Default: true

    Support using flow through your node_modules folder, WARNING: Checking this box is a security risk. When you open a project we will immediately run code contained within it.

require'nvim_lsp'.flow.setup{}

  Default Values:
    cmd = { "npm", "run", "flow", "lsp" }
    filetypes = { "javascript", "javascriptreact", "javascript.jsx" }
    log_level = 2
    root_dir = root_pattern(".flowconfig")
    settings = {}

fortls

https://github.com/hansec/fortran-language-server

Fortran Language Server for the Language Server Protocol

This server accepts configuration via the settings key.

Available settings:
  • fortran-ls.autocompletePrefix: boolean

    Filter autocomplete suggestions with variable prefix

  • fortran-ls.displayVerWarning: boolean

    Default: true

    Provides notifications when the underlying language server is out of date.

  • fortran-ls.enableCodeActions: boolean

    Enable experimental code actions (requires v1.7.0+).

  • fortran-ls.executablePath: string

    Default: "fortls"

    Path to the Fortran language server (fortls).

  • fortran-ls.hoverSignature: boolean

    Show signature information in hover for argument (also enables 'variableHover').

  • fortran-ls.includeSymbolMem: boolean

    Default: true

    Include type members in document outline (also used for 'Go to Symbol in File')

  • fortran-ls.incrementalSync: boolean

    Default: true

    Use incremental synchronization for file changes.

  • fortran-ls.lowercaseIntrinsics: boolean

    Use lowercase for intrinsics and keywords in autocomplete requests.

  • fortran-ls.maxCommentLineLength: number

    Default: -1

    Maximum comment line length (requires v1.8.0+).

  • fortran-ls.maxLineLength: number

    Default: -1

    Maximum line length (requires v1.8.0+).

  • fortran-ls.notifyInit: boolean

    Notify when workspace initialization is complete (requires v1.7.0+).

  • fortran-ls.useSignatureHelp: boolean

    Default: true

    Use signature help instead of snippets when available.

  • fortran-ls.variableHover: boolean

    Show hover information for variables.

require'nvim_lsp'.fortls.setup{}

  Default Values:
    cmd = { "fortls" }
    filetypes = { "fortran" }
    log_level = 2
    root_dir = root_pattern(".fortls")
    settings = {
      nthreads = 1
    }

ghcide

https://github.com/digital-asset/ghcide

A library for building Haskell IDE tooling.

This server accepts configuration via the settings key.

Available settings:
  • hic.arguments: string

    Default: "--lsp"

    The arguments you would like to pass to the executable

  • hic.executablePath: string

    Default: "ghcide"

    The location of your ghcide executable

require'nvim_lsp'.ghcide.setup{}

  Default Values:
    cmd = { "ghcide", "--lsp" }
    filetypes = { "haskell", "lhaskell" }
    log_level = 2
    root_dir = root_pattern("stack.yaml", "hie-bios", "BUILD.bazel", "cabal.config", "package.yaml")
    settings = {}

gopls

https://github.com/golang/tools/tree/master/gopls

Google's lsp server for golang.

require'nvim_lsp'.gopls.setup{}

  Default Values:
    cmd = { "gopls" }
    filetypes = { "go" }
    log_level = 2
    root_dir = root_pattern("go.mod", ".git")
    settings = {}

hie

https://github.com/haskell/haskell-ide-engine

the following init_options are supported (see https://github.com/haskell/haskell-ide-engine#configuration):

init_options = {
  languageServerHaskell = {
    hlintOn = bool;
    maxNumberOfProblems = number;
    diagnosticsDebounceDuration = number;
    liquidOn = bool (default false);
    completionSnippetsOn = bool (default true);
    formatOnImportOn = bool (default true);
    formattingProvider = string (default "brittany", alternate "floskell");
  }
}

This server accepts configuration via the settings key.

Available settings:
  • languageServerHaskell.completionSnippetsOn: boolean

    Default: true

    Show snippets with type information when using code completion

  • languageServerHaskell.diagnosticsOnChange: boolean

    Default: true

    Compute diagnostics continuously as you type. Turn off to only generate diagnostics on file save.

  • languageServerHaskell.enableHIE: boolean

    Default: true

    Enable/disable HIE (useful for multi-root workspaces).

  • languageServerHaskell.formatOnImportOn: boolean

    Default: true

    When adding an import, use the formatter on the result

  • languageServerHaskell.formattingProvider: enum { "brittany", "floskell", "none" }

    Default: "brittany"

    The tool to use for formatting requests.

  • languageServerHaskell.hieExecutablePath: string

    Default: ""

    Set the path to your hie executable, if it's not already on your $PATH. Works with ~, ${HOME} and ${workspaceFolder}.

  • languageServerHaskell.hlintOn: boolean

    Default: true

    Get suggestions from hlint

  • languageServerHaskell.liquidOn: boolean

    Get diagnostics from liquid haskell

  • languageServerHaskell.logFile: string

    Default: ""

    If set, redirects the logs to a file.

  • languageServerHaskell.maxNumberOfProblems: number

    Default: 100

    Controls the maximum number of problems produced by the server.

  • languageServerHaskell.showTypeForSelection.command.location: enum { "dropdown", "channel" }

    Default: "dropdown"

    Determines where the type information for selected text will be shown when the showType command is triggered (distinct from automatically showing this information when hover is triggered). dropdown: in a dropdown channel: will be revealed in an output channel

  • languageServerHaskell.showTypeForSelection.onHover: boolean

    Default: true

    If true, when an expression is selected, the hover tooltip will attempt to display the type of the entire expression - rather than just the term under the cursor.

  • languageServerHaskell.trace.server: enum { "off", "messages", "verbose" }

    Default: "off"

    Traces the communication between VSCode and the languageServerHaskell service.

  • languageServerHaskell.useCustomHieWrapper: boolean

    Use your own custom wrapper for hie (remember to specify the path!). This will take precedence over useHieWrapper and hieExecutablePath.

  • languageServerHaskell.useCustomHieWrapperPath: string

    Default: ""

    Specify the full path to your own custom hie wrapper (e.g. ${HOME}/.hie-wrapper.sh). Works with ~, ${HOME} and ${workspaceFolder}.

require'nvim_lsp'.hie.setup{}

  Default Values:
    cmd = { "hie-wrapper" }
    filetypes = { "haskell" }
    log_level = 2
    root_dir = root_pattern("stack.yaml", "package.yaml", ".git")
    settings = {}

intelephense

https://intelephense.com/

intelephense can be installed via :LspInstall intelephense or by yourself with npm:

npm install -g intelephense

Can be installed in Nvim with :LspInstall intelephense

require'nvim_lsp'.intelephense.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "intelephense", "--stdio" }
    filetypes = { "php" }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("composer.json", ".git")

leanls

https://github.com/leanprover/lean-client-js/tree/master/lean-language-server

Lean language server.

This server accepts configuration via the settings key.

Available settings:
  • lean.executablePath: string

    Default: "lean"

    Path to the Lean executable to use.

  • lean.extraOptions: array

    Default: {}

    Array items: {description = "a single command-line argument",type = "string"}

    Extra command-line options for the Lean server.

  • lean.infoViewAllErrorsOnLine: boolean

    Info view: show all errors on the current line, instead of just the ones on the right of the cursor.

  • lean.infoViewAutoOpen: boolean

    Default: true

    Info view: open info view when Lean extension is activated.

  • lean.infoViewAutoOpenShowGoal: boolean

    Default: true

    Info view: auto open shows goal and messages for the current line (instead of all messages for the whole file)

  • lean.infoViewFilterIndex: number

    Default: -1

    Index of the filter applied to the tactic state (in the array infoViewTacticStateFilters). An index of -1 means no filter is applied.

  • lean.infoViewStyle: string

    Default: ""

    Add an additional CSS snippet to the info view.

  • lean.infoViewTacticStateFilters: array

    Default: { {flags = "",match = false,regex = "^_"}, {flags = "",match = true,name = "goals only",regex = "^(⊢|\\d+ goals|case|$)"} }

    Array items: {description = "an object with required properties 'regex': string, 'match': boolean, and 'flags': string, and optional property 'name': string",properties = {flags = {description = "additional flags passed to the RegExp constructor, e.g. 'i' for ignore case",type = "string"},match = {description = "whether tactic state lines matching the value of 'regex' should be included (true) or excluded (false)",type = "boolean"},name = {description = "name displayed in the dropdown",type = "string"},regex = {description = "a properly-escaped regex string, e.g. '^_' matches any string beginning with an underscore",type = "string"}},required = { "regex", "match", "flags" },type = "object"}

    An array of objects containing regular expression strings that can be used to filter (positively or negatively) the tactic state in the info view. Set to an empty array '[]' to hide the filter select dropdown.

    Each object must contain the following keys: 'regex': string, 'match': boolean, 'flags': string. 'regex' is a properly-escaped regex string, 'match' = true (false) means blocks in the tactic state matching 'regex' will be included (excluded) in the info view, 'flags' are additional flags passed to the JavaScript RegExp constructor. The 'name' key is optional and may contain a string that is displayed in the dropdown instead of the full regex details.

  • lean.input.customTranslations: object

    Default: {}

    Array items: {description = "Unicode character to translate to",type = "string"}

    Add additional input Unicode translations. Example: {"foo": "☺"} will correct \foo to .

  • lean.input.enabled: boolean

    Default: true

    Enable Lean input mode.

  • lean.input.languages: array

    Default: { "lean" }

    Array items: {description = "the name of a language, e.g. 'lean', 'markdown'",type = "string"}

    Enable Lean Unicode input in other file types.

  • lean.input.leader: string

    Default: "\\"

    Leader key to trigger input mode.

  • lean.leanpkgPath: string

    Default: "leanpkg"

    Path to the leanpkg executable to use.

  • lean.memoryLimit: number

    Default: 4096

    Set a memory limit (in megabytes) for the Lean server.

  • lean.progressMessages: boolean

    Show error messages where Lean is still checking.

  • lean.roiModeDefault: string

    Default: "visible"

    Set the default region of interest mode (nothing, visible, lines, linesAndAbove, open, or project) for the Lean extension.

  • lean.timeLimit: number

    Default: 100000

    Set a deterministic timeout (it is approximately the maximum number of memory allocations in thousands) for the Lean server.

  • lean.typeInStatusBar: boolean

    Default: true

    Show the type of term under the cursor in the status bar.

  • lean.typesInCompletionList: boolean

    Display types of all items in the list of completions. By default, only the type of the highlighted item is shown.

require'nvim_lsp'.leanls.setup{}

  Default Values:
    cmd = { "lean-language-server", "--stdio" }
    filetypes = { "lean" }
    log_level = 2
    root_dir = util.root_pattern(".git")
    settings = {}

ocamlls

https://github.com/ocaml-lsp/ocaml-language-server

ocaml-language-server can be installed via :LspInstall ocamlls or by yourself with npm

npm install -g ocaml-langauge-server

Can be installed in Nvim with :LspInstall ocamlls

require'nvim_lsp'.ocamlls.setup{}

  Default Values:
    cmd = { "ocaml-language-server", "--stdio" }
    filetypes = { "ocaml", "reason" }
    log_level = 2
    root_dir = root_pattern(".merlin", "package.json")
    settings = {}

pyls

https://github.com/palantir/python-language-server

python-language-server, a language server for Python.

This server accepts configuration via the settings key.

Available settings:
  • pyls.configurationSources: array

    Default: { "pycodestyle" }

    Array items: {enum = { "pycodestyle", "pyflakes" },type = "string"}

    List of configuration sources to use.

  • pyls.executable: string

    Default: "pyls"

    Language server executable

  • pyls.plugins.jedi.environment: string

    Default: vim.NIL

    Define environment for jedi.Script and Jedi.names.

  • pyls.plugins.jedi.extra_paths: array

    Default: {}

    Define extra paths for jedi.Script.

  • pyls.plugins.jedi_completion.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.jedi_completion.include_params: boolean

    Default: true

    Auto-completes methods and classes with tabstops for each parameter.

  • pyls.plugins.jedi_definition.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.jedi_definition.follow_builtin_imports: boolean

    Default: true

    If follow_imports is True will decide if it follow builtin imports.

  • pyls.plugins.jedi_definition.follow_imports: boolean

    Default: true

    The goto call will follow imports.

  • pyls.plugins.jedi_hover.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.jedi_references.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.jedi_signature_help.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.jedi_symbols.all_scopes: boolean

    Default: true

    If True lists the names of all scopes instead of only the module namespace.

  • pyls.plugins.jedi_symbols.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.mccabe.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.mccabe.threshold: number

    Default: 15

    The minimum threshold that triggers warnings about cyclomatic complexity.

  • pyls.plugins.preload.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.preload.modules: array

    Default: vim.NIL

    Array items: {type = "string"}

    List of modules to import on startup

  • pyls.plugins.pycodestyle.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.pycodestyle.exclude: array

    Default: vim.NIL

    Array items: {type = "string"}

    Exclude files or directories which match these patterns.

  • pyls.plugins.pycodestyle.filename: array

    Default: vim.NIL

    Array items: {type = "string"}

    When parsing directories, only check filenames matching these patterns.

  • pyls.plugins.pycodestyle.hangClosing: boolean

    Default: vim.NIL

    Hang closing bracket instead of matching indentation of opening bracket's line.

  • pyls.plugins.pycodestyle.ignore: array

    Default: vim.NIL

    Array items: {type = "string"}

    Ignore errors and warnings

  • pyls.plugins.pycodestyle.maxLineLength: number

    Default: vim.NIL

    Set maximum allowed line length.

  • pyls.plugins.pycodestyle.select: array

    Default: vim.NIL

    Array items: {type = "string"}

    Select errors and warnings

  • pyls.plugins.pydocstyle.addIgnore: array

    Default: vim.NIL

    Array items: {type = "string"}

    Ignore errors and warnings in addition to the specified convention.

  • pyls.plugins.pydocstyle.addSelect: array

    Default: vim.NIL

    Array items: {type = "string"}

    Select errors and warnings in addition to the specified convention.

  • pyls.plugins.pydocstyle.convention: enum { "pep257", "numpy" }

    Default: vim.NIL

    Choose the basic list of checked errors by specifying an existing convention.

  • pyls.plugins.pydocstyle.enabled: boolean

    Enable or disable the plugin.

  • pyls.plugins.pydocstyle.ignore: array

    Default: vim.NIL

    Array items: {type = "string"}

    Ignore errors and warnings

  • pyls.plugins.pydocstyle.match: string

    Default: "(?!test_).*\\.py"

    Check only files that exactly match the given regular expression; default is to match files that don't start with 'test_' but end with '.py'.

  • pyls.plugins.pydocstyle.matchDir: string

    Default: "[^\\.].*"

    Search only dirs that exactly match the given regular expression; default is to match dirs which do not begin with a dot.

  • pyls.plugins.pydocstyle.select: array

    Default: vim.NIL

    Array items: {type = "string"}

    Select errors and warnings

  • pyls.plugins.pyflakes.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.pylint.args: array

    Default: vim.NIL

    Array items: {type = "string"}

    Arguments to pass to pylint.

  • pyls.plugins.pylint.enabled: boolean

    Enable or disable the plugin.

  • pyls.plugins.rope_completion.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.plugins.yapf.enabled: boolean

    Default: true

    Enable or disable the plugin.

  • pyls.rope.extensionModules: string

    Default: vim.NIL

    Builtin and c-extension modules that are allowed to be imported and inspected by rope.

  • pyls.rope.ropeFolder: array

    Default: vim.NIL

    Array items: {type = "string"}

    The name of the folder in which rope stores project configurations and data. Pass null for not using such a folder at all.

require'nvim_lsp'.pyls.setup{}

  Default Values:
    cmd = { "pyls" }
    filetypes = { "python" }
    log_level = 2
    root_dir = vim's starting directory
    settings = {}

pyls_ms

https://github.com/Microsoft/python-language-server

python-language-server, a language server for Python.

Requires .NET Core to run. On Linux or macOS:

curl -L https://dot.net/v1/dotnet-install.sh | sh

This server accepts configuration via the settings key.

Can be installed in Nvim with :LspInstall pyls_ms

require'nvim_lsp'.pyls_ms.setup{}

  Default Values:
    filetypes = { "python" }
    init_options = {
      analysisUpdates = true,
      asyncStartup = true,
      displayOptions = {},
      interpreter = {
        properties = {
          InterpreterPath = "/usr/bin/python",
          Version = "2.7"
        }
      }
    }
    log_level = 2
    on_new_config = <function 1>
    root_dir = vim's starting directory
    settings = {
      python = {
        analysis = {
          disabled = {},
          errors = {},
          info = {}
        }
      }
    }

rls

https://github.com/rust-lang/rls

rls, a language server for Rust

See https://github.com/rust-lang/rls#setup to setup rls itself. See https://github.com/rust-lang/rls#configuration for rls-specific settings.

If you want to use rls for a particular build, eg nightly, set cmd as follows:

cmd = {"rustup", "run", "nightly", "rls"}

This server accepts configuration via the settings key.

Available settings:
  • rust-client.channel: enum { "stable", "beta", "nightly" }

    Default: vim.NIL

    Rust channel to invoke rustup with. Ignored if rustup is disabled. By default, uses the same channel as your currently open project.

  • rust-client.disableRustup: boolean

    Disable usage of rustup and use rustc/rls from PATH.

  • rust-client.enableMultiProjectSetup: boolean

    Allow multiple projects in the same folder, along with remove the constraint that the cargo.toml must be located at the root. (Experimental: might not work for certain setups)

  • rust-client.logToFile: boolean

    When set to true, RLS stderr is logged to a file at workspace root level. Requires reloading extension after change.

  • rust-client.nestedMultiRootConfigInOutermost: boolean

    Default: true

    If one root workspace folder is nested in another root folder, look for the Rust config in the outermost root.

  • rust-client.revealOutputChannelOn: enum { "info", "warn", "error", "never" }

    Default: "never"

    Specifies message severity on which the output channel will be revealed. Requires reloading extension after change.

  • rust-client.rlsPath: string|null

    Default: vim.NIL

    Override RLS path. Only required for RLS developers. If you set this and use rustup, you should also set rust-client.channel to ensure your RLS sees the right libraries. If you don't use rustup, make sure to set rust-client.disableRustup.

  • rust-client.rustupPath: string

    Default: "rustup"

    Path to rustup executable. Ignored if rustup is disabled.

  • rust-client.trace.server: enum { "off", "messages", "verbose" }

    Default: "off"

    Traces the communication between VS Code and the Rust language server.

  • rust-client.updateOnStartup: boolean

    Update the RLS whenever the extension starts up.

  • rust-client.useWSL: boolean

    When set to true, RLS is started within Windows Subsystem for Linux.

  • rust.all_features: boolean

    Enable all Cargo features.

  • rust.all_targets: boolean

    Default: true

    Checks the project as if you were running cargo check --all-targets (I.e., check all targets and integration tests too).

  • rust.build_bin: string|null

    Default: vim.NIL

    Specify to run analysis as if running cargo check --bin <name>. Use null to auto-detect. (unstable)

  • rust.build_command: string|null

    Default: vim.NIL

    EXPERIMENTAL (requires unstable_features) If set, executes a given program responsible for rebuilding save-analysis to be loaded by the RLS. The program given should output a list of resulting .json files on stdout. Implies rust.build_on_save: true.

  • rust.build_lib: boolean|null

    Default: vim.NIL

    Specify to run analysis as if running cargo check --lib. Use null to auto-detect. (unstable)

  • rust.build_on_save: boolean

    Only index the project when a file is saved and not on change.

  • rust.cfg_test: boolean

    Build cfg(test) code. (unstable)

  • rust.clear_env_rust_log: boolean

    Default: true

    Clear the RUST_LOG environment variable before running rustc or cargo.

  • rust.clippy_preference: enum { "on", "opt-in", "off" }

    Default: "opt-in"

    Controls eagerness of clippy diagnostics when available. Valid values are (case-insensitive):

    • "off": Disable clippy lints.
    • "on": Display the same diagnostics as command-line clippy invoked with no arguments (clippy::all unless overridden).
    • "opt-in": Only display the lints explicitly enabled in the code. Start by adding #![warn(clippy::all)] to the root of each crate you want linted. You need to install clippy via rustup if you haven't already.
  • rust.crate_blacklist: array|null

    Default: { "cocoa", "gleam", "glium", "idna", "libc", "openssl", "rustc_serialize", "serde", "serde_json", "typenum", "unicode_normalization", "unicode_segmentation", "winapi" }

    Overrides the default list of packages for which analysis is skipped. Available since RLS 1.38

  • rust.features: array

    Default: {}

    A list of Cargo features to enable.

  • rust.full_docs: boolean|null

    Default: vim.NIL

    Instructs cargo to enable full documentation extraction during save-analysis while building the crate.

  • rust.jobs: number|null

    Default: vim.NIL

    Number of Cargo jobs to be run in parallel.

  • rust.no_default_features: boolean

    Do not enable default Cargo features.

  • rust.racer_completion: boolean

    Default: true

    Enables code completion using racer.

  • rust.rustflags: string|null

    Default: vim.NIL

    Flags added to RUSTFLAGS.

  • rust.rustfmt_path: string|null

    Default: vim.NIL

    When specified, RLS will use the Rustfmt pointed at the path instead of the bundled one

  • rust.show_hover_context: boolean

    Default: true

    Show additional context in hover tooltips when available. This is often the type local variable declaration.

  • rust.show_warnings: boolean

    Default: true

    Show warnings.

  • rust.sysroot: string|null

    Default: vim.NIL

    --sysroot

  • rust.target: string|null

    Default: vim.NIL

    --target

  • rust.target_dir: string|null

    Default: vim.NIL

    When specified, it places the generated analysis files at the specified target directory. By default it is placed target/rls directory.

  • rust.unstable_features: boolean

    Enable unstable features.

  • rust.wait_to_build: number

    Default: 1500

    Time in milliseconds between receiving a change notification and starting build.

require'nvim_lsp'.rls.setup{}

  Default Values:
    cmd = { "rls" }
    filetypes = { "rust" }
    log_level = 2
    root_dir = root_pattern("Cargo.toml")
    settings = {}

rust_analyzer

https://github.com/rust-analyzer/rust-analyzer

rust-analyzer (aka rls 2.0), a language server for Rust

See docs for extra settings.

This server accepts configuration via the settings key.

Available settings:
  • rust-analyzer.cargo-watch.allTargets: boolean

    Default: true

    Check all targets and tests (will be passed as --all-targets)

  • rust-analyzer.cargo-watch.arguments: array

    Default: {}

    cargo-watch arguments. (e.g: --features="shumway,pdf" will run as cargo watch -x "check --features="shumway,pdf"" )

  • rust-analyzer.cargo-watch.command: string

    Default: "check"

    cargo-watch command. (e.g: clippy will run as cargo watch -x clippy )

  • rust-analyzer.cargo-watch.enable: boolean

    Default: true

    Run cargo check for diagnostics on save

  • rust-analyzer.cargoFeatures.allFeatures: boolean

    Default: true

    Activate all available features

  • rust-analyzer.cargoFeatures.features: array

    Default: {}

    List of features to activate

  • rust-analyzer.cargoFeatures.noDefaultFeatures: boolean

    Do not activate the default feature

  • rust-analyzer.displayInlayHints: boolean

    Default: true

    Display additional type information in the editor

  • rust-analyzer.enableEnhancedTyping: boolean

    Default: true

    Enables enhanced typing. NOTE: If using a VIM extension, you should set this to false

  • rust-analyzer.excludeGlobs: array

    Default: {}

    Paths to exclude from analysis

  • rust-analyzer.featureFlags: object

    Default: {}

    Fine grained feature flags to disable annoying features

  • rust-analyzer.highlightingOn: boolean

    Highlight Rust code (overrides built-in syntax highlighting)

  • rust-analyzer.lruCapacity: number

    Default: vim.NIL

    Number of syntax trees rust-analyzer keeps in memory

  • rust-analyzer.maxInlayHintLength: number

    Default: 20

    Maximum length for inlay hints

  • rust-analyzer.raLspServerPath: string

    Default: "ra_lsp_server"

    Path to ra_lsp_server executable

  • rust-analyzer.rainbowHighlightingOn: boolean

    When highlighting Rust code, use a unique color per identifier

  • rust-analyzer.trace.server: enum { "off", "messages", "verbose" }

    Default: "off"

    Trace requests to the ra_lsp_server

  • rust-analyzer.useClientWatching: boolean

    Default: true

    client provided file watching instead of notify watching.

require'nvim_lsp'.rust_analyzer.setup{}

  Default Values:
    capabilities = {
      offsetEncoding = { "utf-8", "utf-16" },
      textDocument = {
        completion = {
          completionItem = {
            commitCharactersSupport = false,
            deprecatedSupport = false,
            documentationFormat = { "markdown", "plaintext" },
            preselectSupport = false,
            snippetSupport = false
          },
          completionItemKind = {
            valueSet = { 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25 }
          },
          contextSupport = false,
          dynamicRegistration = false
        },
        documentHighlight = {
          dynamicRegistration = false
        },
        hover = {
          contentFormat = { "markdown", "plaintext" },
          dynamicRegistration = false
        },
        references = {
          dynamicRegistration = false
        },
        signatureHelp = {
          dynamicRegistration = false,
          signatureInformation = {
            documentationFormat = { "markdown", "plaintext" }
          }
        },
        synchronization = {
          didSave = true,
          dynamicRegistration = false,
          willSave = false,
          willSaveWaitUntil = false
        }
      }
    }
    cmd = { "ra_lsp_server" }
    filetypes = { "rust" }
    log_level = 2
    on_init = <function 1>
    root_dir = root_pattern("Cargo.toml")
    settings = {}

solargraph

https://solargraph.org/

solargraph, a language server for Ruby

You can install solargraph via gem install.

gem install solargraph

This server accepts configuration via the settings key.

Available settings:
  • solargraph.autoformat: enum { true, false }

    Enable automatic formatting while typing (WARNING: experimental)

  • solargraph.bundlerPath: string

    Default: "bundle"

    Path to the bundle executable, defaults to 'bundle'

  • solargraph.checkGemVersion: enum { true, false }

    Default: true

    Automatically check if a new version of the Solargraph gem is available.

  • solargraph.commandPath: string

    Default: "solargraph"

    Path to the solargraph command. Set this to an absolute path to select from multiple installed Ruby versions.

  • solargraph.completion: enum { true, false }

    Default: true

    Enable completion

  • solargraph.definitions: enum { true, false }

    Default: true

    Enable definitions (go to, etc.)

  • solargraph.diagnostics: enum { true, false }

    Enable diagnostics

  • solargraph.externalServer: object

    Default: {host = "localhost",port = 7658}

    The host and port to use for external transports. (Ignored for stdio and socket transports.)

  • solargraph.folding: boolean

    Default: true

    Enable folding ranges

  • solargraph.formatting: enum { true, false }

    Enable document formatting

  • solargraph.hover: enum { true, false }

    Default: true

    Enable hover

  • solargraph.logLevel: enum { "warn", "info", "debug" }

    Default: "warn"

    Level of debug info to log. warn is least and debug is most.

  • solargraph.references: enum { true, false }

    Default: true

    Enable finding references

  • solargraph.rename: enum { true, false }

    Default: true

    Enable symbol renaming

  • solargraph.symbols: enum { true, false }

    Default: true

    Enable symbols

  • solargraph.transport: enum { "socket", "stdio", "external" }

    Default: "socket"

    The type of transport to use.

  • solargraph.useBundler: boolean

    Use bundle exec to run solargraph. (If this is true, the solargraph.commandPath setting is ignored.)

require'nvim_lsp'.solargraph.setup{}

  Default Values:
    cmd = { "solargraph", "stdio" }
    filetypes = { "ruby" }
    log_level = 2
    root_dir = root_pattern("Gemfile", ".git")
    settings = {}

sumneko_lua

https://github.com/sumneko/lua-language-server

Lua language server. By default, this doesn't have a cmd set. This is because it doesn't provide a global binary. We provide an installer for Linux and macOS using :LspInstall. If you wish to install it yourself, here is a guide.

Can be installed in Nvim with :LspInstall sumneko_lua This server accepts configuration via the settings key.

Available settings:
  • Lua.awakened.cat: boolean

  • Lua.completion.callSnippet: enum { "Disable", "Both", "Replace" }

    Default: "Disable"

  • Lua.completion.enable: boolean

    Default: true

  • Lua.completion.keywordSnippet: enum { "Disable", "Both", "Replace" }

    Default: "Replace"

  • Lua.develop.debuggerPort: integer

    Default: 11412

  • Lua.develop.debuggerWait: boolean

  • Lua.develop.enable: boolean

  • Lua.diagnostics.disable: array

    Array items: {type = "string"}

  • Lua.diagnostics.enable: boolean

    Default: true

  • Lua.diagnostics.globals: array

    Array items: {type = "string"}

  • Lua.diagnostics.severity: object

  • Lua.runtime.path: array

    Default: { "?.lua", "?/init.lua", "?/?.lua" }

    Array items: {type = "string"}

  • Lua.runtime.version: enum { "Lua 5.1", "Lua 5.2", "Lua 5.3", "Lua 5.4", "LuaJIT" }

    Default: "Lua 5.3"

  • Lua.workspace.ignoreDir: array

    Default: { ".vscode" }

    Array items: {type = "string"}

  • Lua.workspace.ignoreSubmodules: boolean

    Default: true

  • Lua.workspace.library: object

  • Lua.workspace.maxPreload: integer

    Default: 300

  • Lua.workspace.preloadFileSize: integer

    Default: 100

  • Lua.workspace.useGitIgnore: boolean

    Default: true

require'nvim_lsp'.sumneko_lua.setup{}

  Default Values:
    filetypes = { "lua" }
    log_level = 2
    root_dir = root_pattern(".git") or os_homedir
    settings = {}

texlab

https://texlab.netlify.com/

A completion engine built from scratch for (La)TeX.

See https://texlab.netlify.com/docs/reference/configuration for configuration options.

require'nvim_lsp'.texlab.setup{}

  Commands:
  - TexlabBuild: Build the current buffer
  
  Default Values:
    cmd = { "texlab" }
    filetypes = { "tex", "bib" }
    log_level = 2
    root_dir = vim's starting directory
    settings = {
      bibtex = {
        formatting = {
          lineLength = 120
        }
      },
      latex = {
        build = {
          args = { "-pdf", "-interaction=nonstopmode", "-synctex=1" },
          executable = "latexmk",
          onSave = false
        },
        forwardSearch = {
          args = {},
          onSave = false
        },
        lint = {
          onChange = false
        }
      }
    }

tsserver

https://github.com/theia-ide/typescript-language-server

typescript-language-server can be installed via :LspInstall tsserver or by yourself with npm:

npm install -g typescript-language-server

Can be installed in Nvim with :LspInstall tsserver

require'nvim_lsp'.tsserver.setup{}

  Default Values:
    capabilities = default capabilities, with offsetEncoding utf-8
    cmd = { "typescript-language-server", "--stdio" }
    filetypes = { "javascript", "javascriptreact", "javascript.jsx", "typescript", "typescriptreact", "typescript.tsx" }
    log_level = 2
    on_init = function to handle changing offsetEncoding
    root_dir = root_pattern("package.json")
    settings = {}

vimls

require'nvim_lsp'.vimls.setup{}

  Default Values:
    cmd = { "vim-language-server", "--stdio" }
    docs = {
      description = "https://github.com/iamcco/vim-language-server\n\nIf you don't want to use Nvim to install it, then you can use:\n```sh\nnpm install -g vim-language-server\n```\n"
    }
    filetypes = { "vim" }
    init_options = {
      diagnostic = {
        enable = true
      },
      indexes = {
        count = 3,
        gap = 100,
        projectRootPatterns = { "runtime", "nvim", ".git", "autoload", "plugin" },
        runtimepath = true
      },
      iskeyword = "@,48-57,_,192-255,-#",
      runtimepath = "",
      suggest = {
        fromRuntimepath = true,
        fromVimruntime = true
      },
      vimruntime = ""
    }
    log_level = 2
    on_new_config = <function 1>
    root_dir = <function 1>
    settings = {}

About

Nvim LSP client configurations

Resources

License

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Lua 99.4%
  • Other 0.6%