61 KiB
YouCompleteMe: a code-completion engine for Vim
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-based engine that provides native semantic code completion for C/C++/Objective-C/Objective-C++ (from now on referred to as "the C-family languages"), a Jedi-based completion engine for Python, an OmniSharp-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.).
Here's an explanation of what happens in the short GIF demo above.
First, realize that no keyboard shortcuts had to be pressed to get the list of completion candidates at any point in the demo. The user just types and the suggestions pop up by themselves. If the user doesn't find the completion suggestions relevant and/or just wants to type, they can do so; the completion engine will not interfere.
When the user sees a useful completion string being offered, they press the TAB key to accept it. This inserts the completion string. Repeated presses of the TAB key cycle through the offered completions.
If the offered completions are not relevant enough, the user can continue typing to further filter out unwanted completions.
A critical thing to notice is that the completion filtering is NOT based on
the input being a string prefix of the completion (but that works too). The
input needs to be a subsequence match of a completion. This is a fancy way
of saying that any input characters need to be present in a completion string in
the order in which they appear in the input. So abc
is a subsequence of
xaybgc
, but not of xbyxaxxc
. After the filter, a complicated sorting system
ranks the completion strings so that the most relevant ones rise to the top of
the menu (so you usually need to press TAB just once).
All of the above works with any programming language because of the identifier-based completion engine. It collects all of the identifiers in the current file and other files you visit (and your tags files) and searches them when you type (identifiers are put into per-filetype groups).
The demo also shows the semantic engine in use. When the user presses .
, ->
or ::
while typing in insert mode (for C++; different triggers are used for
other languages), the semantic engine is triggered (it can also be triggered
with a keyboard shortcut; see the rest of the docs).
The last thing that you can see in the demo is YCM's integration with Syntastic (the little red X that shows up in the left gutter) if you are editing a C-family file. As Clang compiles your file and detects warnings or errors, they will be piped to Syntastic for display. You don't need to save your file or press any keyboard shortcut to trigger this, it "just happens" in the background.
In essence, YCM obsoletes the following Vim plugins because it has all of their features plus extra:
- clang_complete
- AutoComplPop
- Supertab
- neocomplcache
YCM also provides semantic go-to-definition/declaration commands for C-family languages & Python. Expect more IDE features powered by the various YCM semantic engines in the future.
You'll also find that YCM has filepath completers (try typing ./
in a file)
and a completer that integrates with UltiSnips.
Mac OS X super-quick installation
Please refer to the full Installation Guide below; the following commands are provided on a best-effort basis and may not work for you.
Install the latest version of MacVim. Yes, MacVim. And yes, the latest.
If you don't use the MacVim GUI, it is recommended to use the Vim binary that is
inside the MacVim.app package (MacVim.app/Contents/MacOS/Vim
). To ensure it
works correctly copy the mvim
script from the MacVim download to your
local binary folder (for example /usr/local/bin/mvim
) and then symlink it:
ln -s /usr/local/bin/mvim vim
Install YouCompleteMe with Vundle.
Remember: YCM is a plugin with a compiled component. If you update YCM using Vundle and the ycm_support_libs library APIs have changed (happens rarely), YCM will notify you to recompile it. You should then rerun the install process.
It's recommended that you have the latest Xcode installed along with the latest Command Line Tools (that you install from within Xcode).
Install CMake. Preferably with Homebrew, but here's the stand-alone CMake installer.
If you have installed a Homebrew Python and/or Homebrew MacVim, see the FAQ for details.
Compiling YCM with semantic support for C-family languages:
cd ~/.vim/bundle/YouCompleteMe
./install.sh --clang-completer
Compiling YCM without semantic support for C-family languages:
cd ~/.vim/bundle/YouCompleteMe
./install.sh
If you want semantic C# support, you should add --omnisharp-completer
to the
install script as well.
That's it. You're done. Refer to the User Guide section on how to use YCM. Don't forget that if you want the C-family semantic completion engine to work, you will need to provide the compilation flags for your project to YCM. It's all in the User Guide.
YCM comes with sane defaults for its options, but you still may want to take a look at what's available for configuration. There are a few interesting options that are conservatively turned off by default that you may want to turn on.
Ubuntu Linux x64 super-quick installation
Please refer to the full Installation Guide below; the following commands are provided on a best-effort basis and may not work for you.
Make sure you have Vim 7.3.584 with python2 support. At the time of writing, the version of Vim shipping with Ubuntu is too old. You may need to compile Vim from source (don't worry, it's easy).
Install YouCompleteMe with Vundle.
Remember: YCM is a plugin with a compiled component. If you update YCM using Vundle and the ycm_support_libs library APIs have changed (happens rarely), YCM will notify you to recompile it. You should then rerun the install process.
Install development tools and CMake: sudo apt-get install build-essential cmake
Make sure you have Python headers installed: sudo apt-get install python-dev
.
Compiling YCM with semantic support for C-family languages:
cd ~/.vim/bundle/YouCompleteMe
./install.sh --clang-completer
Compiling YCM without semantic support for C-family languages:
cd ~/.vim/bundle/YouCompleteMe
./install.sh
If you want semantic C# support, you should add --omnisharp-completer
to the
install script as well.
That's it. You're done. Refer to the User Guide section on how to use YCM. Don't forget that if you want the C-family semantic completion engine to work, you will need to provide the compilation flags for your project to YCM. It's all in the User Guide.
YCM comes with sane defaults for its options, but you still may want to take a look at what's available for configuration. There are a few interesting options that are conservatively turned off by default that you may want to turn on.
Windows Installation
YCM has no official support for Windows, but that doesn't mean you can't get it to work there. See the Windows Installation Guide wiki page. Feel free to add to it.
Full Installation Guide
These are the steps necessary to get YCM working on a Unix OS like Linux or Mac OS X. My apologies to Windows users, but I don't have a guide for them. The code is platform agnostic, so if everything is configured correctly, YCM should work on Windows without issues (but as of writing, it's untested on that platform).
See the FAQ if you have any issues.
Remember: YCM is a plugin with a compiled component. If you update YCM using Vundle and the ycm_support_libs library APIs have changed (happens rarely), YCM will notify you to recompile it. You should then rerun the install process.
Please follow the instructions carefully. Read EVERY WORD.
-
Ensure that your version of Vim is at least 7.3.584 and that it has support for python2 scripting.
Inside Vim, type
:version
. Look at the first two to three lines of output; it should sayVi IMproved 7.3
and then below that,Included patches: 1-X
, where X will be some number. That number needs to be 584 or higher.If your version of Vim is not recent enough, you may need to compile Vim from source (don't worry, it's easy).
After you have made sure that you have Vim 7.3.584+, type the following in Vim:
:echo has('python')
. The output should be 1. If it's 0, then get a version of Vim with Python support. -
Install YCM with Vundle (or Pathogen, but Vundle is a better idea). With Vundle, this would mean adding a
Bundle 'Valloric/YouCompleteMe'
line to your vimrc.If you don't install YCM with Vundle, make sure you have run
git submodule update --init --recursive
after checking out the YCM repository (Vundle will do this for you) to fetch YCM's dependencies. -
[Complete this step ONLY if you care about semantic completion support for C-family languages. Otherwise it's not neccessary.]
Download the latest version of
libclang
. Clang is an open-source compiler that can compile C/C++/Objective-C/Objective-C++. Thelibclang
library it provides is used to power the YCM semantic completion engine for those languages. YCM is designed to work with libclang version 3.3 or higher, but can in theory work with 3.2 as well.You can use the system libclang only if you are sure it is version 3.3 or higher, otherwise don't. Even if it is, I recommend using the official binaries from llvm.org if at all possible. Make sure you download the correct archive file for your OS.
-
Compile the
ycm_support_libs
libraries that YCM needs. These libs are the C++ engines that YCM uses to get fast completions.You will need to have
cmake
installed in order to generate the required makefiles. Linux users can install cmake with their package manager (sudo apt-get install cmake
for Ubuntu) whereas other users can download and install cmake from its project site. Mac users can also get it through Homebrew withbrew install cmake
.You also need to make sure you have Python headers installed. On a Debian-like Linux distro, this would be
sudo apt-get install python-dev
. On Mac they should already be present.Here we'll assume you installed YCM with Vundle. That means that the top-level YCM directory is in
~/.vim/bundle/YouCompleteMe
.We'll create a new folder where build files will be placed. Run the following:
cd ~ mkdir ycm_build cd ycm_build
Now we need to generate the makefiles. If you DON'T care about semantic support for C-family languages, run the following command in the
ycm_build
directory:cmake -G "Unix Makefiles" . ~/.vim/bundle/YouCompleteMe/cpp
If you DO care about semantic support for C-family languages, then your
cmake
call will be a bit more complicated. We'll assume you downloaded a binary distribution of LLVM+Clang from llvm.org in step 3 and that you extracted the archive file to folder~/ycm_temp/llvm_root_dir
(withbin
,lib
,include
etc. folders right inside that folder). With that in mind, run the following command in theycm_build
directory:cmake -G "Unix Makefiles" -DPATH_TO_LLVM_ROOT=~/ycm_temp/llvm_root_dir . ~/.vim/bundle/YouCompleteMe/cpp
Now that makefiles have been generated, simply run:
make ycm_support_libs
For those who want to use the system version of libclang, you would pass
-DUSE_SYSTEM_LIBCLANG=ON
to cmake instead of the-DPATH_TO_LLVM_ROOT=...
flag.You could also force the use of a custom libclang library with
-DEXTERNAL_LIBCLANG_PATH=/path/to/libclang.so
flag (the library would end with.dylib
on a Mac). Again, this flag would be used instead of the other flags.Running the
make
command will also place thelibclang.[so|dylib]
in theYouCompleteMe/python
folder for you if you compiled with clang support (it needs to be there for YCM to work).
That's it. You're done. Refer to the User Guide section on how to use YCM. Don't forget that if you want the C-family semantic completion engine to work, you will need to provide the compilation flags for your project to YCM. It's all in the User Guide.
YCM comes with sane defaults for its options, but you still may want to take a look at what's available for configuration. There are a few interesting options that are conservatively turned off by default that you may want to turn on.
User Guide
General Usage
- If the offered completions are too broad, keep typing characters; YCM will continue refining the offered completions based on your input.
- Filtering is "smart-case" sensitive; if you are typing only lowercase letters, then it's case-insensitive. If your input contains uppercase letters, then the uppercase letters in your query must match uppercase letters in the completion strings (the lowercase letters still match both). So, "foo" matches "Foo" and "foo", "Foo" matches "Foo" and "FOO" but not "foo".
- Use the TAB key to accept a completion and continue pressing TAB to cycle through the completions. Use Shift-TAB to cycle backwards. Note that if you're using console Vim (that is, not Gvim or MacVim) then it's likely that the Shift-TAB binding will not work because the console will not pass it to Vim. You can remap the keys; see the Options section below.
Knowing a little bit about how YCM works internally will prevent confusion. YCM has several completion engines: an identifier-based completer that collects all of the identifiers in the current file and other files you visit (and your tags files) and searches them when you type (identifiers are put into per-filetype groups).
There are also several semantic engines in YCM. There's a libclang-based completer that provides semantic completion for C-family languages. There's a Jedi-based completer for semantic completion for Python. There's also an omnifunc-based completer that uses data from Vim's omnicomplete system to provide semantic completions when no native completer exists for that language in YCM.
There are also other completion engines, like the UltiSnips completer and the filepath completer.
YCM automatically detects which completion engine would be the best in any situation. On occasion, it queries several of them at once, merges the outputs and presents the results to you.
Client-server architecture
YCM has a client-server architecture; the Vim part of YCM is only a thin client
that talks to the ycmd
HTTP+JSON server that has the vast majority of YCM
logic and functionality. The server is started and stopped automatically as you
start and stop Vim.
Completion string ranking
The subsequence filter removes any completions that do not match the input, but then the sorting system kicks in. It's actually very complicated and uses lots of factors, but suffice it to say that "word boundary" (WB) subsequence character matches are "worth" more than non-WB matches. In effect, this means given an input of "gua", the completion "getUserAccount" would be ranked higher in the list than the "Fooguxa" completion (both of which are subsequence matches). A word-boundary character are all capital characters, characters preceded by an underscore and the first letter character in the completion string.
General Semantic Completion Engine Usage
- You can use Ctrl+Space to trigger the completion suggestions anywhere, even without a string prefix. This is useful to see which top-level functions are available for use.
C-family Semantic Completion Engine Usage
- You really also want to install the latest version of the Syntastic Vim plugin. It has support for YCM and together they will provide you with compile errors/warnings practically instantly and without saving the file.
YCM looks for a .ycm_extra_conf.py
file in the directory of the opened file or
in any directory above it in the hierarchy (recursively); when the file is
found, it is loaded (only once!) as a Python module. YCM calls a FlagsForFile
method in that module which should provide it with the information necessary to
compile the current file. You can also provide a path to a global
.ycm_extra_conf.py
file, which will be used as a fallback. To prevent the
execution of malicious code from a file you didn't write YCM will ask you once
per .ycm_extra_conf.py
if it is safe to load. This can be disabled and you can
white-/blacklist files. See the Options section for more details.
This system was designed this way so that the user can perform any arbitrary sequence of operations to produce a list of compilation flags YCM should hand to Clang.
See YCM's own .ycm_extra_conf.py
for details on how this
works. You should be able to use it as a starting point. Don't just
copy/paste that file somewhere and expect things to magically work; your project
needs different flags. Hint: just replace the strings in the flags
variable
with compilation flags necessary for your project. That should be enough for 99%
of projects.
Yes, Clang's CompilationDatabase
system is also supported. Again, see the
above linked example file.
If Clang encounters errors when compiling the header files that your file includes, then it's probably going to take a long time to get completions. When the completion menu finally appears, it's going to have a large number of unrelated completion strings (type/function names that are not actually members). This is because Clang fails to build a precompiled preamble for your file if there are any errors in the included headers and that preamble is key to getting fast completions.
Call the :YcmDiags
command to see if any errors or warnings were detected in
your file. Even better, use Syntastic.
Python semantic completion
YCM uses Jedi to power its semantic completion for Python. This should "just
work" without any configuration from the user. You do NOT need to install Jedi
yourself; YCM uses it as a git subrepo. If you're installing YCM with Vundle
(which is the recommended way) then Vundle will make sure that the subrepo is
checked out when you do :BundleInstall
. If you're installing YCM by hand, then
you need to run git submodule update --init --recursive
when you're checking
out the YCM repository. That's it.
But again, installing YCM with Vundle takes care of all of this for you.
In the future expect to see features like go-to-definition for Python as well.
C# semantic completion
YCM uses OmniSharp to provide semantic completion for C#. It's used as a git
subrepo. If you're installing YCM with Vundle (which is the recommended way)
then Vundle will make sure that the subrepo is checked out when you do
:BundleInstall
. If you're installing YCM by hand, then you need to run git submodule update --init --recursive
when you're checking out the YCM
repository.
OmniSharp is written in C# and has to be compiled. The install.sh
script takes
care of this if you pass --omnisharp-completer
as an argument.
Semantic completion for other languages
YCM will use your omnifunc
(see :h omnifunc
in Vim) as a source for semantic
completions if it does not have a native semantic completion engine for your
file's filetype. Vim comes with okayish omnifuncs for various languages like
Ruby, PHP etc. It depends on the language.
You can get stellar omnifuncs for Java and Ruby with Eclim. Just make sure
you have the latest Eclim installed and configured (this means Eclim >= 2.2.*
and Eclipse >= 4.2.*
).
After installing Eclim remember to create a new Eclipse project within your
application by typing :ProjectCreate <path-to-your-project> -n ruby
(or -n java
)
inside vim and don't forget to have let g:EclimCompletionMethod = 'omnifunc'
in your vimrc. This will make YCM and Eclim play nice; YCM will use Eclim's omnifuncs
as the data source for semantic completions and provide the auto-triggering
and subsequence-based matching (and other YCM features) on top of it.
Writing New Semantic Completers
You have two options here: writing an omnifunc
for Vim's omnicomplete system
that YCM will then use through its omni-completer, or a custom completer for YCM
using the Completer API.
Here are the differences between the two approaches:
- You have to use VimScript to write the omnifunc, but get to use Python to write for the Completer API; this by itself should make you want to use the API.
- The Completer API is a much more powerful way to integrate with YCM and it provides a wider set of features. For instance, you can make your Completer query your semantic back-end in an asynchronous fashion, thus not blocking Vim's GUI thread while your completion system is processing stuff. This is impossible with VimScript. All of YCM's completers use the Completer API.
- Performance with the Completer API is better since Python executes faster than VimScript.
If you want to use the omnifunc
system, see the relevant Vim docs with :h complete-functions
. For the Completer API, see the API docs.
If you want to upstream your completer into YCM's source, you should use the Completer API.
Syntastic integration
YCM has explicit support for Syntastic (and vice-versa) if you compiled YCM with Clang support; this means that any diagnostics (errors or warnings) that Clang encounters while compiling your file will be fed back to Syntastic for display.
YCM will recompile your file in the background updatetime
(see :h updatetime
in Vim) milliseconds after you stop typing (to be specific, on CursorHold
and
CursorHoldI
Vim events). YCM will change your updatetime
value to be 2000
milliseconds (there's an option to tell it not to do this if you wish).
The new diagnostics (if any) will be fed back to Syntastic the next time you press any key on the keyboard. So if you stop typing and just wait for the new diagnostics to come in, that will not work. You need to press some key for the GUI to update.
Having to press a key to get the updates is unfortunate, but cannot be changed due to the way Vim internals operate; there is no way that a background task can update Vim's GUI after it has finished running. You have to press a key. This will make YCM check for any pending diagnostics updates.
You can force a full, blocking compilation cycle with the
:YcmForceCompileAndDiagnostics
command (you may want to map that command to a
key; try putting nnoremap <F5> :YcmForceCompileAndDiagnostics<CR>
in your
vimrc). Calling this command will force YCM to immediately recompile your file
and display any new diagnostics it encounters. Do note that recompilation with
this command may take a while and during this time the Vim GUI will be
blocked.
After the errors are displayed by Syntastic, it will display a short diagnostic
message when you move your cursor to the line with the error. You can get a
detailed diagnostic message with the <leader>d
key mapping (can be changed in
the options) YCM provides when your cursor is on the line with the diagnostic.
You can also see the full diagnostic message for all the diagnostics in the
current file in Vim's locationlist
, which can be opened with the :lopen
and
:lclose
commands (make sure you have set let g:syntastic_always_populate_loc_list = 1
in your vimrc). A good way to toggle
the display of the locationlist
with a single key mapping is provided by
another (very small) Vim plugin called ListToggle (which also makes it
possible to change the height of the locationlist
window), also written by
yours truly.
Commands
The :YcmRestartServer
command
If the ycmd
completion server suddenly stops for some reason, you can restart
it with this command.
The :YcmForceCompileAndDiagnostics
command
Calling this command will force YCM to immediately recompile your file and display any new diagnostics it encounters. Do note that recompilation with this command may take a while and during this time the Vim GUI will be blocked.
You may want to map this command to a key; try putting nnoremap <F5> :YcmForceCompileAndDiagnostics<CR>
in your vimrc.
The :YcmDiags
command
Calling this command will fill Vim's locationlist
with errors or warnings if
any were detected in your file and then open it.
A better option would be to use Syntastic which will keep your locationlist
up to date automatically and will also show error/warning notifications in Vim's
gutter.
The :YcmShowDetailedDiagnostic
command
This command shows the full diagnostic text when the user's cursor is on the line with the diagnostic.
The :YcmDebugInfo
command
This will print out various debug information for the current file. Useful to see what compile commands will be used for the file if you're using the semantic completion engine.
The :YcmCompleter
command
This command can be used to invoke completer-specific commands. If the first
argument is of the form ft=...
the completer for that file type will be used
(for example ft=cpp
), else the native completer of the current buffer will be
used.
Call YcmCompleter
without further arguments for information about the
commands you can call for the selected completer.
See the YcmCompleter subcommands section for more information on the available subcommands.
YcmCompleter subcommands
[See the docs for the YcmCompleter
command before tackling this section.]
The invoked subcommand is automatically routed to the currently active semantic
completer, so :YcmCompleter GoToDefinition
will invoke the GoToDefinition
subcommand on the Python semantic completer if the currently active file is a
Python one and on the Clang completer if the currently active file is a
C/C++/Objective-C one.
You may also want to map the subcommands to something less verbose; for
instance, nnoremap <leader>jd :YcmCompleter GoToDefinitionElseDeclaration<CR>
maps the <leader>jd
sequence to the longer subcommand invocation.
The various GoTo*
subcommands add entries to Vim's jumplist
so you can use
CTRL-O
to jump back to where you where before invoking the command (and
CTRL-I
to jump forward; see :h jumplist
for details).
The GoToDeclaration
subcommand
Looks up the symbol under the cursor and jumps to its declaration.
Supported in filetypes: c, cpp, objc, objcpp, python, cs
The GoToDefinition
subcommand
Looks up the symbol under the cursor and jumps to its definition.
NOTE: For C-family languages this only works in certain situations, namely when
the definition of the symbol is in the current translation unit. A translation
unit consists of the file you are editing and all the files you are including
with #include
directives (directly or indirectly) in that file.
Supported in filetypes: c, cpp, objc, objcpp, python, cs
The GoToDefinitionElseDeclaration
subcommand
Looks up the symbol under the cursor and jumps to its definition if possible; if the definition is not accessible from the current translation unit, jumps to the symbol's declaration.
Supported in filetypes: c, cpp, objc, objcpp, python, cs
The ClearCompilationFlagCache
subcommand
YCM caches the flags it gets from the FlagsForFile
function in your
ycm_extra_conf.py
file if you return them with the do_cache
parameter set to
True
. The cache is in memory and is never invalidated (unless you restart Vim
of course).
This command clears that cache entirely. YCM will then re-query your
FlagsForFile
function as needed in the future.
Supported in filetypes: c, cpp, objc, objcpp
The StartServer
subcommand
Starts the semantic-engine-as-localhost-server for those semantic engines that work as separate servers that YCM talks to.
Supported in filetypes: cs
The StopServer
subcommand
Stops the semantic-engine-as-localhost-server for those semantic engines that work as separate servers that YCM talks to.
Supported in filetypes: cs
The RestartServer
subcommand
Restarts the semantic-engine-as-localhost-server for those semantic engines that work as separate servers that YCM talks to.
Supported in filetypes: cs
Options
All options have reasonable defaults so if the plug-in works after installation you don't need to change any options. These options can be configured in your vimrc script by including a line like this:
let g:ycm_min_num_of_chars_for_completion = 1
Note that after changing an option in your [vimrc script] vimrc you have to restart Vim for the changes to take effect.
The g:ycm_min_num_of_chars_for_completion
option
This option controls the number of characters the user needs to type before
identifier-based completion suggestions are triggered. For example, if the
option is set to 2
, then when the user types a second alphanumeric character
after a whitespace character, completion suggestions will be triggered. This
option is NOT used for semantic completion.
Setting this option to a high number like 99
effectively turns off the
identifier completion engine and just leaves the semantic engine.
Default: 2
let g:ycm_min_num_of_chars_for_completion = 2
The g:ycm_min_num_identifier_candidate_chars
option
This option controls the minimum number of characters that a completion candidate coming from the identifier completer must have to be shown in the popup menu.
A special value of 0
means there is no limit.
NOTE: This option only applies to the identifier completer; it has no effect on the various semantic completers.
Default: 0
let g:ycm_min_num_identifier_candidate_chars = 0
The g:ycm_filetype_whitelist
option
This option controls for which Vim filetypes (see :h filetype
) should YCM be
turned on. The option value should be a Vim dictionary with keys being filetype
strings (like python
, cpp
etc) and values being unimportant (the dictionary
is used like a hash set, meaning that only the keys matter).
The *
key is special and matches all filetypes. By default, the whitelist
contains only this *
key.
YCM also has a g:ycm_filetype_blacklist
option that lists filetypes for which
YCM shouldn't be turned on. YCM will work only in filetypes that both the
whitelist and the blacklist allow (the blacklist "allows" a filetype by not
having it as a key).
For example, let's assume you want YCM to work in files with the cpp
filetype.
The filetype should then be present in the whitelist either directly (cpp
key
in the whitelist) or indirectly through the special *
key. It should not be
present in the blacklist.
Filetypes that are blocked by the either of the lists will be completely ignored by YCM, meaning that neither the identifier-based completion engine nor the semantic engine will operate in them.
You can get the filetype of the current file in Vim with :set ft?
.
Default: {'*' : 1}
let g:ycm_filetype_whitelist = { '*': 1 }
The g:ycm_filetype_blacklist
option
This option controls for which Vim filetypes (see :h filetype
) should YCM be
turned off. The option value should be a Vim dictionary with keys being filetype
strings (like python
, cpp
etc) and values being unimportant (the dictionary
is used like a hash set, meaning that only the keys matter).
See the g:ycm_filetype_whitelist
option for more details on how this works.
Default: {'notes': 1, 'markdown': 1, 'text': 1}
let g:ycm_filetype_blacklist = {
\ 'notes' : 1,
\ 'markdown' : 1,
\ 'text' : 1,
\ 'unite' : 1,
\}
The g:ycm_filetype_specific_completion_to_disable
option
This option controls for which Vim filetypes (see :h filetype
) should the YCM
semantic completion engine be turned off. The option value should be a Vim
dictionary with keys being filetype strings (like python
, cpp
etc) and
values being unimportant (the dictionary is used like a hash set, meaning that
only the keys matter). The listed filetypes will be ignored by the YCM semantic
completion engine, but the identifier-based completion engine will still trigger
in files of those filetypes.
Note that even if semantic completion is not turned off for a specific filetype, you will not get semantic completion if the semantic engine does not support that filetype.
You can get the filetype of the current file in Vim with :set ft?
.
Default: {}
let g:ycm_filetype_specific_completion_to_disable = {}
The g:ycm_register_as_syntastic_checker
option
When set, this option makes YCM register itself as the Syntastic checker for the
c
, cpp
, objc
and objcpp
filetypes. This enables the YCM-Syntastic
integration.
If you're using YCM's identifier completer in C-family languages but cannot use the clang-based semantic completer for those languages and want to use the GCC Syntastic checkers, unset this option.
Don't unset this option unless you're sure you know what you're doing.
Default: 1
let g:ycm_register_as_syntastic_checker = 1
The g:ycm_allow_changing_updatetime
option
When this option is set to 1
, YCM will change the updatetime
Vim option to
2000
(see :h updatetime
). This may conflict with some other plugins you have
(but it's unlikely). The updatetime
option is the number of milliseconds that
have to pass before Vim's CursorHold
(see :h CursorHold
) event fires. YCM
runs the completion engines' "file comprehension" systems in the background on
every such event; the identifier-based engine collects the identifiers whereas
the semantic engine compiles the file to build an AST.
The Vim default of 4000
for updatetime
is a bit long, so YCM reduces
this. Set this option to 0
to force YCM to leave your updatetime
setting
alone.
Default: 1
let g:ycm_allow_changing_updatetime = 1
The g:ycm_complete_in_comments
option
When this option is set to 1
, YCM will show the completion menu even when
typing inside comments.
Default: 0
let g:ycm_complete_in_comments = 0
The g:ycm_complete_in_strings
option
When this option is set to 1
, YCM will show the completion menu even when
typing inside strings.
Note that this is turned on by default so that you can use the filename
completion inside strings. This is very useful for instance in C-family files
where typing #include "
will trigger the start of filename completion. If you
turn off this option, you will turn off filename completion in such situations
as well.
Default: 1
let g:ycm_complete_in_strings = 1
The g:ycm_collect_identifiers_from_comments_and_strings
option
When this option is set to 1
, YCM's identifier completer will also collect
identifiers from strings and comments. Otherwise, the text in comments and
strings will be ignored.
Default: 0
let g:ycm_collect_identifiers_from_comments_and_strings = 0
The g:ycm_collect_identifiers_from_tags_files
option
When this option is set to 1
, YCM's identifier completer will also collect
identifiers from tags files. The list of tags files to examine is retrieved from
the tagfiles()
Vim function which examines the tags
Vim option. See :h 'tags'
for details.
YCM will re-index your tags files if it detects that they have been modified.
The only supported tag format is the Exuberant Ctags format. The
format from "plain" ctags is NOT supported. Ctags needs to be called with the
--fields=+l
option (that's a lowercase L
, not a one) because YCM needs the
language:<lang>
field in the tags output.
See the FAQ for pointers if YCM does not appear to read your tag files.
This option is off by default because it makes Vim slower if your tags are on a network directory.
Default: 0
let g:ycm_collect_identifiers_from_tags_files = 0
The g:ycm_seed_identifiers_with_syntax
option
When this option is set to 1
, YCM's identifier completer will seed its
identifier database with the keywords of the programming language you're
writing.
Since the keywords are extracted from the Vim syntax file for the filetype, all keywords may not be collected, depending on how the syntax file was written. Usually at least 95% of the keywords are successfully extracted.
Default: 0
let g:ycm_seed_identifiers_with_syntax = 0
The g:ycm_server_use_vim_stdout
option
By default, the ycmd
completion server writes logs to logfiles. When this
option is set to 1
, the server writes logs to Vim's stdout (so you'll see them
in the console).
Default: 0
let g:ycm_server_use_vim_stdout = 0
The g:ycm_server_keep_logfiles
option
When this option is set to 1
, the ycmd
completion server will keep the
logfiles around after shutting down (they are deleted on shutdown by default).
To see where the logfiles are, call :YcmDebugInfo
.
Default: 0
let g:ycm_server_keep_logfiles = 0
The g:ycm_server_log_level
option
The logging level that the ycmd
completion server uses. Valid values are the
following, from most verbose to least verbose:
debug
info
warning
error
critical
Note that debug
is very verbose.
Default: info
let g:ycm_server_log_level = 'info'
The g:ycm_server_idle_suicide_seconds
option
This option sets the number of seconds of ycmd
server idleness (no requests
received) after which the server stops itself. NOTE: the YCM Vim client sends a
shutdown request to the server when Vim is shutting down.
If your Vim crashes for instance, ycmd
never gets the shutdown command and
becomes a zombie process. This option prevents such zombies from sticking around
forever.
The default option is 43200
seconds which is 12 hours. The reason for the
interval being this long is to prevent the server from shutting down if you
leave your computer (and Vim) turned on during the night.
The server "heartbeat" that checks whether this interval has passed occurs every 10 minutes.
Default: 43200
let g:ycm_server_idle_suicide_seconds = 43200
The g:ycm_csharp_server_port
option
The port number (on localhost
) on which the OmniSharp server should be
started.
Default: 2000
let g:ycm_csharp_server_port = 2000
The g:ycm_auto_start_csharp_server
option
When set to 1
, the OmniSharp server will be automatically started (once per
Vim session) when you open a C# file.
Default: 1
let g:ycm_auto_start_csharp_server = 1
The g:ycm_auto_stop_csharp_server
option
When set to 1
, the OmniSharp server will be automatically stopped upon
closing Vim.
Default: 1
let g:ycm_auto_stop_csharp_server = 1
The g:ycm_add_preview_to_completeopt
option
When this option is set to 1
, YCM will add the preview
string to Vim's
completeopt
option (see :h completeopt
). If your completeopt
option
already has preview
set, there will be no effect. You can see the current
state of your completeopt
setting with :set completeopt?
(yes, the question
mark is important).
When preview
is present in completeopt
, YCM will use the preview
window at
the top of the file to store detailed information about the current completion
candidate (but only if the candidate came from the semantic engine). For
instance, it would show the full function prototype and all the function
overloads in the window if the current completion is a function name.
Default: 0
let g:ycm_add_preview_to_completeopt = 0
The g:ycm_autoclose_preview_window_after_completion
option
When this option is set to 1
, YCM will auto-close the preview
window after
the user accepts the offered completion string. If there is no preview
window
triggered because there is no preview
string in completeopt
, this option is
irrelevant. See the g:ycm_add_preview_to_completeopt
option for more details.
Default: 0
let g:ycm_autoclose_preview_window_after_completion = 0
The g:ycm_autoclose_preview_window_after_insertion
option
When this option is set to 1
, YCM will auto-close the preview
window after
the user leaves insert mode. This option is irrelevant if
g:ycm_autoclose_preview_window_after_completion
is set or if no preview
window is triggered. See the g:ycm_add_preview_to_completeopt
option for more
details.
Default: 0
let g:ycm_autoclose_preview_window_after_insertion = 0
The g:ycm_max_diagnostics_to_display
option
This option controls the maximum number of diagnostics shown to the user when errors or warnings are detected in the file. This option is only relevant if you are using the semantic completion engine and have installed the version of the Syntastic plugin that supports YCM.
Default: 30
let g:ycm_max_diagnostics_to_display = 30
The g:ycm_key_list_select_completion
option
This option controls the key mappings used to select the first completion string. Invoking any of them repeatedly cycles forward through the completion list.
Some users like adding <Enter>
to this list.
Default: ['<TAB>', '<Down>']
let g:ycm_key_list_select_completion = ['<TAB>', '<Down>']
The g:ycm_key_list_previous_completion
option
This option controls the key mappings used to select the previous completion string. Invoking any of them repeatedly cycles backwards through the completion list.
Note that one of the defaults is <S-TAB>
which means Shift-TAB. That mapping
will probably only work in GUI Vim (Gvim or MacVim) and not in plain console Vim
because the terminal usually does not forward modifier key combinations to Vim.
Default: ['<S-TAB>', '<Up>']
let g:ycm_key_list_previous_completion = ['<S-TAB>', '<Up>']
The g:ycm_key_invoke_completion
option
This option controls the key mapping used to invoke the completion menu for
semantic completion. By default, semantic completion is trigged automatically
after typing .
, ->
and ::
in insert mode (if semantic completion support
has been compiled in). This key mapping can be used to trigger semantic
completion anywhere. Useful for searching for top-level functions and classes.
Console Vim (not Gvim or MacVim) passes <Nul>
to Vim when the user types
<C-Space>
so YCM will make sure that <Nul>
is used in the map command when
you're editing in console Vim, and <C-Space>
in GUI Vim. This means that you
can just press <C-Space>
in both console and GUI Vim and YCM will do the right
thing.
Setting this option to an empty string will make sure no mapping is created.
Default: <C-Space>
let g:ycm_key_invoke_completion = '<C-Space>'
The g:ycm_key_detailed_diagnostics
option
This option controls the key mapping used to show the full diagnostic text when
the user's cursor is on the line with the diagnostic. It basically calls
:YcmShowDetailedDiagnostic
.
Setting this option to an empty string will make sure no mapping is created.
Default: <leader>d
let g:ycm_key_detailed_diagnostics = '<leader>d'
The g:ycm_global_ycm_extra_conf
option
Normally, YCM searches for a .ycm_extra_conf.py
file for compilation flags
(see the User Guide for more details on how this works). This option specifies
a fallback path to a config file which is used if no .ycm_extra_conf.py
is
found.
You can place such a global file anywhere in your filesystem.
Default: ''
let g:ycm_global_ycm_extra_conf = ''
The g:ycm_confirm_extra_conf
option
When this option is set to 1
YCM will ask once per .ycm_extra_conf.py
file
if it is safe to be loaded. This is to prevent execution of malicious code
from a .ycm_extra_conf.py
file you didn't write.
To selectively get YCM to ask/not ask about loading certain .ycm_extra_conf.py
files, see the g:ycm_extra_conf_globlist
option.
Default: 1
let g:ycm_confirm_extra_conf = 1
The g:ycm_extra_conf_globlist
option
This option is a list that may contain several globbing patterns. If a pattern
starts with a !
all .ycm_extra_conf.py
files matching that pattern will be
blacklisted, that is they won't be loaded and no confirmation dialog will be
shown. If a pattern does not start with a !
all files matching that pattern
will be whitelisted. Note that this option is not used when confirmation is
disabled using g:ycm_confirm_extra_conf
and that items earlier in the list
will take precedence over the later ones.
Rules:
*
matches everything?
matches any single character[seq]
matches any character in seq[!seq]
matches any char not in seq
Example:
let g:ycm_extra_conf_globlist = ['~/dev/*','!~/*']
- The first rule will match everything contained in the
~/dev
directory so.ycm_extra_conf.py
files from there will be loaded. - The second rule will match everything in the home directory so a
.ycm_extra_conf.py
file from there won't be loaded. - As the first rule takes precedence everything in the home directory excluding
the
~/dev
directory will be blacklisted.
Default: []
let g:ycm_extra_conf_globlist = []
The g:ycm_filepath_completion_use_working_dir
option
By default, YCM's filepath completion will interpret relative paths like ../
as being relative to the folder of the file of the currently active buffer.
Setting this option will force YCM to always interpret relative paths as being
relative to Vim's current working directory.
Default: 0
let g:ycm_filepath_completion_use_working_dir = 0
The g:ycm_semantic_triggers
option
This option controls the character-based triggers for the various semantic completion engines. The option holds a dictionary of key-values, where the keys are Vim's filetype strings delimited by commas and values are lists of strings, where the strings are the triggers.
Setting key-value pairs on the dictionary adds semantic triggers to the internal default set (listed below). You cannot remove the default triggers, only add new ones.
A "trigger" is a sequence of one or more characters that trigger semantic
completion when typed. For instance, C++ (cpp
filetype) has .
listed as a
trigger. So when the user types foo.
, the semantic engine will trigger and
serve foo
's list of member functions and variables. Since C++ also has ->
listed as a trigger, the same thing would happen when the user typed foo->
.
Default: [see next line]
let g:ycm_semantic_triggers = {
\ 'c' : ['->', '.'],
\ 'objc' : ['->', '.'],
\ 'ocaml' : ['.', '#'],
\ 'cpp,objcpp' : ['->', '.', '::'],
\ 'perl' : ['->'],
\ 'php' : ['->', '::'],
\ 'cs,java,javascript,d,vim,python,perl6,scala,vb,elixir,go' : ['.'],
\ 'ruby' : ['.', '::'],
\ 'lua' : ['.', ':'],
\ 'erlang' : [':'],
\ }
The g:ycm_cache_omnifunc
option
Some omnicompletion engines do not work well with the YCM cache—in particular, they might not produce all possible results for a given prefix. By unsetting this option you can ensure that the omnicompletion engine is requeried on every keypress. That will ensure all completions will be presented, but might cause stuttering and lagginess if the omnifunc is slow.
Default: 1
let g:ycm_cache_omnifunc = 1
FAQ
I get a linker warning regarding libpython
on Mac when compiling YCM
If the warning is ld: warning: path '/usr/lib/libpython2.7.dylib' following -L not a directory
, then feel free to ignore it; it's caused by a limitation of
CMake and is not an issue. Everything should still work fine.
I get a weird window at the top of my file when I use the semantic engine
This is Vim's preview
window. Vim uses it to show you extra information about
something if such information is available. YCM provides Vim with such extra
information. For instance, when you select a function in the completion list,
the preview
window will hold that function's prototype and the prototypes of
any overloads of the function. It will stay there after you select the
completion so that you can use the information about the parameters and their
types to write the function call.
If you would like this window to auto-close after you select a completion
string, set the g:ycm_autoclose_preview_window_after_completion
option to 1
in your vimrc
file. Similarly, the g:ycm_autoclose_preview_window_after_insertion
option can be set to close the preview
window after leaving insert mode.
If you don't want this window to ever show up, add set completeopt-=preview
to
your vimrc
. Also make sure that the g:ycm_add_preview_to_completeopt
option
is set to 0
.
It appears that YCM is not working
In Vim, run :messages
and carefully read the output. YCM will echo messages to
the message log if it encounters problems. It's likely you misconfigured
something and YCM is complaining about it.
Also, you may want to run the :YcmDebugInfo
command; it will make YCM spew out
various debugging information, including the compile flags for the file if the
file is a C-family language file and you have compiled in Clang support.
I cannot get the Syntastic integration to work
Try to update your version of Syntastic. At the time of writing (Jan 2013), the YCM integration is very recent and it's likely that your version of Syntastic does not have it.
Sometimes it takes much longer to get semantic completions than normal
This means that libclang (which YCM uses for C-family semantic completion)
failed to pre-compile your file's preamble. In other words, there was an error
compiling some of the source code you pulled in through your header files. I
suggest calling the :YcmDiags
command to see what they were (even better, have
Syntastic installed and call :lopen
).
Bottom line, if libclang can't pre-compile your file's preamble because there were errors in it, you're going to get slow completions because there's no AST cache.
Vim flickers every time I move the cursor or moving the cursor is slow
You probably have an old version of Syntastic installed. If you are using
Vundle, make sure that your bundle command is Bundle 'scrooloose/syntastic'
and not Bundle 'Syntastic'
. The first command pulls in the latest version of
Syntastic from GitHub while the second one pulls in an old version from vim.org.
Because of a Vundle bug, make sure you have completely removed
everything in your Vundle bundle directory (~/.vim/bundle
by default) before
switching from one Syntastic bundle command to the other.
YCM auto-inserts completion strings I don't want!
This means you probably have some mappings that interfere with YCM's internal
ones. Make sure you don't have something mapped to <C-p>
, <C-x>
or <C-u>
(in insert mode).
YCM never selects something for you; it just shows you a menu and the user has to explicitly select something. If something is being selected automatically, this means there's a bug or a misconfiguration somewhere.
I get a E227: mapping already exists for <blah>
error when I start Vim
This means that YCM tried to set up a key mapping but failed because you already
had something mapped to that key combination. The <blah>
part of the message
will tell you what was the key combination that failed.
Look in the Options section and see if any of the default mappings conflict with your own. Then change that option value to something else so that the conflict goes away.
I get 'GLIBC_2.XX' not found (required by libclang.so)
when starting Vim
Your system is too old for the precompiled binaries from llvm.org. Compile
Clang on your machine and then link against the libclang.so
you just produced.
See the full installation guide for help.
I'm trying to use a Homebrew Vim with YCM and I'm getting segfaults
Something (I don't know what) is wrong with the way that Homebrew configures and
builds Vim. I recommend using MacVim. Even if you don't like the MacVim GUI,
you can use the Vim binary that is inside the MacVim.app package (it's
MacVim.app/Contents/MacOS/Vim
) and get the Vim console experience.
I have a Homebrew Python and/or MacVim; can't compile/SIGABRT when starting
You should probably run brew rm python; brew install python
to get the latest
fixes that should make YCM work with such a configuration. Also rebuild Macvim
then. If you still get problems with this, see issue #18 for
suggestions.
Vim segfaults when I use the semantic completer in Ruby files
This was caused by a Vim bug. Update your version of Vim (Vim 7.3.874 is known to work, earlier versions may also fix this issue).
I get LONG_BIT definition appears wrong for platform
when compiling
Look at the output of your CMake call. There should be a line in it like the
following (with .dylib
in place of .so
on a Mac):
-- Found PythonLibs: /usr/lib/libpython2.7.so (Required is at least version "2.5")
That would be the correct output. An example of incorrect output would be the following:
-- Found PythonLibs: /usr/lib/libpython2.7.so (found suitable version "2.5.1", minimum required is "2.5")
Notice how there's an extra bit of output there, the found suitable version "<version>"
part, where <version>
is not the same as the version of the
dynamic library. In the example shown, the library is version 2.7 but the second
string is version 2.5.1
.
This means that CMake found one version of Python headers and a different version for the library. This is wrong. It can happen when you have multiple versions of Python installed on your machine.
You should probably add the following flags to your cmake call (again, dylib
instead of so
on a Mac):
-DPYTHON_INCLUDE_DIR=/usr/include/python2.7 -DPYTHON_LIBRARY=/usr/lib/libpython2.7.so
This will force the paths to the Python include directory and the Python library to use. You may need to set these flags to something else, but you need to make sure you use the same version of Python that your Vim binary is built against, which is highly likely to be the system's default Python.
I get libpython2.7.a [...] relocation R_X86_64_32
when compiling
The error is usually encountered when compiling YCM on Centos or RHEL. The full error looks something like the following:
/usr/bin/ld: /usr/local/lib/libpython2.7.a(abstract.o): relocation R_X86_64_32 against `a local symbol' can not be used when making a shared object; recompile with -fPIC
It's possible to get a slightly different error that's similar to the one above. Here's the problem and how you solve it:
Your libpython2.7.a
was not compiled with -fPIC
so it can't be linked into
ycm_core.so
. Use the -DPYTHON_LIBRARY=
CMake flag to point it to a .so
version of libpython on your machine (for instance,
-DPYTHON_LIBRARY=/usr/lib/libpython2.7.so
). Naturally, this means you'll have
to go through the full installation guide by hand.
I get Vim: Caught deadly signal SEGV
on Vim startup
This can happen on some Linux distros. If you encounter this situation, run Vim
under gdb
. You'll probably see something like this in the output when Vim
crashes:
undefined symbol: clang_CompileCommands_dispose
This means that Vim is trying to load a libclang.so
that is too old. You need
at least a 3.2 libclang. Some distros ship with a system libclang.so
that
identifies itself as 3.2 but is not; it was cut from the upstream sources before
the official 3.2 release and some API changes (like the addition of the
CompileCommands API) were added after their cut.
So just go through the installation guide and make sure you are using a correct
libclang.so
. I recommend downloading prebuilt binaries from llvm.org.
YCM does not read identifiers from my tags files
First, put let g:ycm_collect_identifiers_from_tags_files = 1
in your vimrc.
Make sure you are using Exuberant Ctags to produce your tags
files since the only supported tag format is the Exuberant Ctags
format. The format from "plain" ctags is NOT supported. The
output of ctags --version
should list "Exuberant Ctags".
Ctags needs to be called with the --fields=+l
(that's a lowercase L
, not a
one) option because YCM needs the language:<lang>
field in the tags output.
NOTE: Mac OS X comes with "plain" ctags installed by default. brew install ctags
will get you the Exuberant Ctags version.
Also make sure that your Vim tags
option is set correctly. See :h 'tags'
for
details. If you want to see which tag files YCM will read for a given buffer,
run :echo tagfiles()
with the relevant buffer active. Note that that function
will only list tag files that already exist.
CTRL-U
in insert mode does not work
YCM keeps you in a completefunc
completion mode when you're typing in insert
mode and Vim disables <C-U>
in completion mode as a "feature." Sadly there's
nothing I can do about this.
YCM conflicts with UltiSnips TAB key usage
YCM comes with support for UltiSnips (snippet suggestions in the popup menu),
but you'll have to change the UltiSnips mappings. See :h UltiSnips-triggers
in
Vim for details. You'll probably want to change some/all of the following
options:
g:UltiSnipsExpandTrigger
g:UltiSnipsJumpForwardTrigger
g:UltiSnipsJumpBackwardTrigger
Why isn't YCM just written in plain VimScript, FFS?
Because of the identifier completion engine and subsequence-based filtering. Let's say you have many dozens of files open in a single Vim instance (I often do); the identifier-based engine then needs to store thousands (if not tens of thousands) of identifiers in its internal data-structures. When the user types, YCM needs to perform subsequence-based filtering on all of those identifiers (every single one!) in less than 10 milliseconds.
I'm sorry, but that level of performance is just plain impossible to achieve with VimScript. I've tried, and the language is just too slow. No, you can't get acceptable performance even if you limit yourself to just the identifiers in the current file and simple prefix-based fitering.
Why does YCM demand such a recent version of Vim?
During YCM's development several show-stopper bugs were encountered in Vim.
Those needed to be fixed upstream (and were). A few months after those bugs were
fixed, Vim trunk landed the pyeval()
function which improved YCM performance
even more since less time was spent serializing and deserializing data between
Vim and the embedded Python interpreter. A few critical bugfixes for pyeval()
landed in Vim 7.3.584 (and a few commits before that).
I get annoying messages in Vim's status area when I type
If you're referring to the User defined completion <bla bla> back at original
and similar, then sadly there's no fix for those. Vim will emit them no matter
how hard I try to silence them.
You'll have to learn to ignore them. It's a shitty "solution", I know.
Nasty bugs happen if I have the vim-autoclose
plugin installed
Use the delimitMate plugin instead. It does the same thing without conflicting with YCM.
Is there some sort of YCM mailing list? I have questions
If you have questions about the plugin or need help, please use the ycm-users mailing list, don't create issues on the tracker. The tracker is for bug reports and feature requests.
Contact
If you have questions about the plugin or need help, please use the ycm-users mailing list.
If you have bug reports or feature suggestions, please use the issue tracker.
The latest version of the plugin is available at http://valloric.github.io/YouCompleteMe/.
The author's homepage is http://val.markovic.io.
License
This software is licensed under the GPL v3 license. © 2012 Strahinja Val Markovic <val@markovic.io>.