https://github.com/tpope/vim-haystack

haystack.vim: fuzzy matching algorithm
https://github.com/tpope/vim-haystack

Last synced: 4 months ago
JSON representation

haystack.vim: fuzzy matching algorithm

• Host: GitHub
• URL: https://github.com/tpope/vim-haystack
• Owner: tpope
• Created: 2015-02-04T19:31:44.000Z (over 11 years ago)
• Default Branch: master
• Last Pushed: 2019-11-13T00:48:33.000Z (over 6 years ago)
• Last Synced: 2025-06-29T21:05:59.707Z (about 1 year ago)
• Language: Vim script
• Homepage: http://www.vim.org/scripts/script.php?script_id=5103
• Size: 4.88 KB
• Stars: 73
• Watchers: 3
• Forks: 3
• Open Issues: 1
• Metadata Files:
  • Readme: README.markdown
  • Funding: .github/FUNDING.yml

Awesome Lists containing this project

README

          
# haystack.vim

Haystack provides a fuzzy matching algorithm for use by other Vim plugins.

It's based on [flx][] for Emacs but adapted to my own real world experience.

The algorithm gives priority to runs of consecutive letters and letters after

punctuation marks.

[flx]: https://github.com/lewang/flx

## Usage

Haystack defines `g:completion_filter` with a reference to a function for

filtering a list of items based on a user provided query.  Plugins can check

for this variable and use it if it exists.  This level of indirection allows

for alternative matching algorithms without the need for plugins to be aware

of each one.

So far, out of the box support is included with [projectionist.vim][],

including dependent plugins like [rails.vim][].

[projectionist.vim]: https://github.com/tpope/vim-projectionist

[rails.vim]: https://github.com/tpope/vim-rails

### Use with fuzzy finders

Here's a proof of concept for [ctrlp.vim][].

    function! CtrlPMatch(items, str, limit, mmode, ispath, crfile, regex) abort

      let items = copy(a:items)

      if a:ispath

        call filter(items, 'v:val !=# a:crfile')

      endif

      return haystack#filter(items, a:str)

    endfunction

    let g:ctrlp_match_func = {'match': function('CtrlPMatch')}

Note that the algorithm is slow enough to feel sluggish with larger data sets.

You might consider a hybrid approach that uses a different algorithm above a

certain number of items.

[ctrlp.vim]: https://github.com/kien/ctrlp.vim

## License

Copyright © Tim Pope.  Distributed under the same terms as Vim itself.

See `:help license`.