uga-rosa / cmp-dictionary Goto Github PK

View Code? Open in Web Editor NEW

237.0 2.0 15.0 188 KB

A dictionary completion source for nvim-cmp

License: MIT License

Lua 99.58% Makefile 0.42%

nvim-cmp neovim-plugin lua

cmp-dictionary's Introduction

Hi! I'm uga-rosa

Profiles

cmp-dictionary's People

Contributors

Stargazers

Watchers

Forkers

dky uncomfyhalomacro yutkat luozhiya mbaneshi v-ita simbaninja917 berkeleytrue rghvdberg npc-z farzadmf mnjm

cmp-dictionary's Issues

Russian spelllang dictionary

Hello, thank you for the plugin. Can you please hint me on how to set up suggestions from russian dictionary? I have the following init.vim:

let s:plug_dir = expand('/tmp/plugged/vim-plug')
if !filereadable(s:plug_dir . '/autoload/plug.vim')
  execute printf('!curl -fLo %s/autoload/plug.vim --create-dirs https://raw.githubusercontent.com/junegunn/vim-plug/master/plug.vim', s:plug_dir)
end

call plug#begin(s:plug_dir)
Plug 'hrsh7th/nvim-cmp'
Plug 'uga-rosa/cmp-dictionary'
call plug#end()

if len(filter(values(g:plugs), '!isdirectory(v:val.dir)'))
	echom 'Some modules are missing, running :PlugInstall'
	PlugInstall --sync
endif

lua <<EOF

local cmp = require('cmp')

require("cmp_dictionary").setup({
	dic = {
		spelllang = {
			en = "/path/to/en.dict",
			ru = "/path/to/ru.dict",
		},
	},
})

cmp.setup{
  mapping = {
		["<C-j>"] = cmp.mapping(function(fallback)
			if cmp.visible() then
				cmp.select_next_item()
			end
		end, { "i", "s" }),
		["<C-k>"] = cmp.mapping(function(fallback)
			if cmp.visible() then
				cmp.select_prev_item()
			end
		end, { "i", "s" }),
		['<C-l>'] = cmp.mapping({
			i = cmp.mapping.confirm({ select = false }),
		}),
	},

	sources = {
		{
			name = "dictionary",
			keyword_length = 2,
		},
	}
}

EOF

set keymap=russian-jcuken
set spelllang=ru

But I do not get any completions as I type in russian:

Executing CmpDictionaryUpdate after entering buffer also does not help.

While english seems to work fine:

I use aspell-en and aspell-ru packages on archlinux and the following command to get dictionaries:

aspell -d en dump master | aspell -l en expand > en.dict
aspell -d ru dump master | aspell -l ru expand > ru.dict

No suggestion for for existing words with mistake.

Hi, I'm not sure if spellcorrection is in scope of this plugin but I noticed dict suggest words only for new typed in words, but it wont suggest anything of existing words with spell mistake:

cmp_dict.mp4

In video above TabNine suggested correction, but I got not entries from cmp-dict.

Any chances for cmp-dict to suggest correction for spelling mistakes?

[Question]: Dynamically swap or change a dictionary with a command or mapping

Hi!

Is possible to change dictionaries on-demand?

I have cmp-dictionary configured with:

require('cmp_dictionary').setup({
  dic = {
    -- ['*'] = '/usr/share/dict/words',
    ['markdown'] = '/path/to/es.dict',
  },
  -- The following are default values, so you don't need to write them if you don't want to change them
  exact = 3,
  first_case_insensitive = true,
  async = true,
  -- capacity = 5,
  -- debug = false,
})

I tried:

:lua require('cmp_dictionary').setup({dic={['markdown'] = '/path/to/en.dict'}})

And after:

:lua require("cmp_dictionary.caches").update()

After that, I try to see the current config:

:lua dump(require("cmp_dictionary.config"))
{                                                                                                       
  config = {
    async = false,
    capacity = 5,
    debug = false,
    dic = {
      ["*"] = <1>{},
      markdown = "/path/to/en.dict"
    },
    exact = 2,
    first_case_insensitive = false
  },
  default = {
    async = false,
    capacity = 5,
    debug = false,
    dic = {
      ["*"] = <table 1>
    },
    exact = 2,
    first_case_insensitive = false
  },
  get = <function 1>,
  ready = true,
  setup = <function 2>
}

I see the new dictionary in config (I notice that all my other settings get back to the defaults), but I'm still seeing only the words of the previous dict (not the new one).

Is this possible?

Thank you!

When I've updated cmp-dictionary and nvim to 0.6 I get 2 errors:
The first is when I have async enabled and the mpack can't be find - I don't know if I have to install that or the new nvim should provide it.
vim.g.cmp_dictionary_async = true
cmp_dictionary/caches.lua:32 module mpack not found

When I disable async I get this error:

Error executing luv callback:
...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:49: attempt to index a nil value
stack traceback:
        ...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:49: in function 'indexing'
        ...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:119: in function '_create_cache'
        ...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:129: in function 'create_cache_sync'
        ...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:219: in function <...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:205>

Very long delay after a buffer is opened

Hi, first of all, thank you for the plugin. It's a life saver for me, being an old CoC user, which had this feature.

I think it's a recent issue I'm noticing, but when I open a buffer in Neovim, I saw a HUGE delay until my cursor becomes interactive (the UI is basically frozen).

I do have A LOT of plugins, however I managed to pin the issue down to this plugin: when I disable this, my cursor is instantly interactive and I can immediately start interacting with the buffer, but if it's enabled, I have that delay.

I'm not configuring anything for the plugin and just loading it as a dependency (like below). Is there a setting I can configure to avoid the delay?

I'm using lazy.nvim as my package manager:

-- nvim-cmp
return {
  'hrsh7th/nvim-cmp',
  event = 'VeryLazy',
  dependencies = {
    -- other sources ...
    'uga-rosa/cmp-dictionary'
  },
  -- other settings for nvim-cmp
}

source loaded but no completion

source is loaded but I am not getting any completion from cmp-dictionary. I am aware of the v2 changes and am not using any .setup()

This is my only reference to cmp-dictionary in my config file.

# cmp.lua
cmp.setup {
  sources = {
    { name = "dictionary" },
    ...
  },
  ...
}

:CmpStatus
# ready source names
- dictionary
- ...

Can you give bit more clearer example of installing aspell languages?

You can download dic from aspell.net or installing by package manager, xbps extract to

$ ls /usr/share/dict/
american-english british-english words

But all I can see is just, tar.bz2 file When I uncompress it contains quite many files.
Tried to install

sudo make install and it outputs

mkdir -p /usr/lib/aspell-0.60/
cp fi.rws fi.multi finnish.alias /usr/lib/aspell-0.60/
cd /usr/lib/aspell-0.60/ && chmod 644 fi.rws fi.multi finnish.alias
mkdir -p /usr/lib/aspell-0.60/
cp fi.dat fi_affix.dat /usr/lib/aspell-0.60/
cd /usr/lib/aspell-0.60/ && chmod 644 fi.dat fi_affix.dat

Any clue "/usr/share/dict/ language or if its in /usr/lib/aspell-0.60/, which file it should be?

I tried
["*"] = { "/usr/lib/aspell-0.60/fi.rws" },

but it didn't give me any completions

My `cmp-dictionary` takes 700+ms on startup as BufEnter

Question

Is there any tips for speed up (such as lazy load)?

Is this startuptime in the specifications?

Startup time

Before uninstall cmp-dictionary

1299.237  836.919: BufEnter autocommands

After uninstall cmp-dictionary

372.051  005.900: BufEnter autocommands

Version

NVIM v0.8.0-v0.8.0
Build type: RelWithDebInfo
LuaJIT 2.1.0-beta3

My Setting

Packer

use {
"uga-rosa/cmp-dictionary"
	["config"] = function() {
		local path = vim.fn.expand "en.dict"
    	local cmp_dict = require("cmp_dictionary")
	    cmp_dict.setup({
			["dict"] = {
				["*"] = {"/user/share/dict/words"}
				["spelling"] = {["en"] = path}
			}
	}) 
}

cmp.config.sources {
	["name"] = "dictionary"
    ["group_incex"] = 5
    ["keyword_length"] 2
}

Use dictionary cache.

Dictionaries are not updated that often, so why not convert them into a program-friendly form and store them under stdpath("cache")?

source name was not shown

   96         sources = cmp.config.sources({
   97             { name = "nvim_lsp" },
   98             { name = "luasnip" },
   99             { name = "path" },
  100             { name = "buffer" },
  101             { name = "nvim_lsp_signature_help" },
  102             { name = "dictionary", keyword_length = 3 },
  103         }),
  104
  105         diagnost
  106        ┌──────────────────────────────────┐
  107        │ diagnostic       Variable [LSP] │
  108        │ diagnostics      Variable [LSP] │em)
  109        │ diagnostics      Variable [LSP] │at("%s %s", kind_icons[vim_item.kind], vim_item.kind)
  110        │ diagnostics      Variable [LSP] │
  111        │ diagnostic       Variable [LSP] │
  112        │ diagnostics      Variable [LSP] │
  113        │ diagnosticls     Variable [LSP] │
  114        │ diagnostic       Text           │
  115        │ diagnostics      Text           │
  116        │ diagnosticate    Text           │
  117        │ diagnostician    Text           │
  118     }) │ diagnostically   Text           │
  119        │ diagnostication  Text           │
  120     -- └──────────────────────────────────┘s

the text was not shown with dictionary, how should I config for that purpose, here was my config:

  147     use({
  148         "uga-rosa/cmp-dictionary",
  149         config = function()
  150             require("cmp_dictionary").setup({
  151                 dic = {
  152                     ["*"] = { "/usr/share/dict/words" },
  153                 },
  154             })
  155         end,
  156     })

can't work with packer lazyload

If i not use lazyload in packer, cmp-dictionary work fine.
after I open lazyload local lazyload = 1, cmp-dictionary can't work.
I saw cmp-dictionary config is load. print('config dict')

test init.lua

local fn, cmd = vim.fn, vim.cmd
local root_dir = fn.fnamemodify(fn.resolve(fn.expand('<sfile>:p')), ':h')

cmd('set packpath=' .. root_dir .. '/plugged')
cmd('set runtimepath^=' .. root_dir .. '/plugged')
local install_path = root_dir .. '/plugged/pack/packer/opt/packer.nvim'
if fn.empty(fn.glob(install_path)) > 0 then
    Bootstrap = fn.system({'git', 'clone', '--depth', '1', 'https://github.com/wbthomason/packer.nvim', install_path})
end

cmd([[
    set mouse=n
    packadd packer.nvim
    nnoremap <space>pu :PackerSync<CR>
    nnoremap <space>pi :PackerInstall<CR>
    nnoremap <space>pr :PackerClean<CR>
    nnoremap <space>pc :PackerCompile<CR>
]])

function Cmp()
    local cmp = require('cmp')
    cmp.setup({
        sources = cmp.config.sources({
            { name = 'path' },
            { name = 'dictionary' },
        }),
    })
end

function Dict()
    print('config dict')
    require("cmp_dictionary").setup({
        dic = {
            ['*'] = {root_dir .. '/dictionary'},
        },
        exact = 2,
        async = false,
        capacity = 5,
        debug = false,
    })
end

local packer = require('packer')
local use = packer.use

local lazyload = 1

packer.init({
    package_root = root_dir .. '/plugged/pack',
    compile_path  = root_dir .. '/plugged/plugin/packer_compiled.lua',
    plugin_package = 'packer',
    auto_clean = false,
})

use({'wbthomason/packer.nvim', opt = true})

if lazyload == 1 then
    use {'hrsh7th/nvim-cmp', event='InsertEnter', config='Cmp()'}
    use {'hrsh7th/cmp-path', after='nvim-cmp'}
    use {'uga-rosa/cmp-dictionary', after='nvim-cmp', config='Dict()'}
else
    use {'hrsh7th/nvim-cmp'}
    use {'hrsh7th/cmp-path'}
    use {'uga-rosa/cmp-dictionary'}
end

if Bootstrap then
    packer.sync()
end


if lazyload == 0 then
    Cmp()
    Dict()
end

test dictionary

Release
Debug

incorrectly loaded dictionaries per filetype

my config (minimal):

local dict = require("cmp_dictionary")
dict.setup({
exact = 2,
first_case_insensitive = false,
document = false,
document_command = "wn %s -over",
async = true,
sqlite = false,
max_items = 1000,
capacity = 5,
debug = false,
})

local google_10k = '/usr/share/dict/google-10000-english.txt'
filetype = {
python = {vim.g.pydiction_location, google_10k},
lua = google_10k,
},
spelllang = {
en = "/usr/share/dict/google-10000-english.txt",
},
})

• with above config when I edited a python file dictionaries seems to be loaded fine, but when I switched to a lua file during editing, items from 'python only' dictionary are also present in completion menu, am I doing something wrong with config??? or what can be done to make cmp using dictionaries set for lua while in lua buffer?

[feature] support bidding filename and dict

Currently, dict bindding to the file type.
Can you add an option to specify the file name to use the corresponding dict? e.g.

filename = {
    ["xmake.lua" ] = { "xmake.dict", "lua.dict" }
}

Some special files, not use file type.

HOWTO: Configure per filetype

Hello,
Thank you for your time to create this plugin.
If you want configure this plugin per filetype, I found this way:

Setup per filetype

Example ftplugin/markdown.lua

--settings for markdown ft
--vim.opt.dictionary=vim.opt.dictionary + "......."

packer

To use cmp-user_dictionary per filetype, you need to set the ft option in packer so cmp-user_dictionary will only load after the ftplugin. This avoids that the plugin is loaded before the dictionary is set.

    use {"uga-rosa/cmp-user_dictionary",ft={"text","markdown"}}

Changes in v3

Changed dictionary specifications
- Only line breaks are used to separate words.
Stop referencing &dictionary.
- switcher is deprecated.
- Set paths in require("cmp_dictionary").setup().
Remove :CmpDictionaryUpdate.
- The dictionary is automatically loaded each time paths is set.
Fix asynchronous dictionary reading.
Organize option names.
- exact -> exact_length
- max_items -> max_number_items
- document and document_command -> document (single table)
sqlite option is deprecated.
- Maybe external option could help you.

TODO: use sqlite

https://github.com/kkharji/sqlite.lua

Unable to set keyword_length > 2

Hi @uga-rosa, thanks for the great work on this plugin. I have it working fine when I have keyword_length <= 2 but anything above that disables Dictionary source completion.

    sources = {
        { name = "dictionary", keyword_length = 5 }
    }

Error message in trie.lua

Any idea what i do wrong here?

Error on startup:

Error executing luv callback:
...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:45: bad argument #1 to 'decode' (string expected, got nil)
stack traceback:
	[C]: in function 'decode'
	...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:45: in function <...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:44>
Error in luv thread:
...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:42: too deep to serialize

Config (lazy):

  {
    "uga-rosa/cmp-dictionary", -- dictionary manager
    enabled = true,
    build =
    'brew install aspell && brew install wordnet && aspell -d en dump master | aspell -l en expand >$HOME/en.dict && aspell -d de dump master | aspell -l de expand >$HOME/de.dict',
    config = function()
      require("cmp_dictionary").setup({
        paths = {
          "/usr/share/dict/words", vim.fn.expand('$HOME/de.dict'), vim.fn.expand('$HOME/en.dict'),
        },
        exact_length = 2,
        first_case_insensitive = true,
        document = {
          enable = true,
          command = { "wn", "${label}", "-over" },
        },
      })
    end
  },

neovim:

 vim --version
NVIM v0.9.5
Build type: Release
LuaJIT 2.1.1703358377

   system vimrc file: "$VIM/sysinit.vim"
  fall-back for $VIM: "/opt/homebrew/Cellar/neovim/0.9.5/share/nvim"

Run :checkhealth for more info

Rename from cmp-user_dictionary to cmp-dictionary

Sorry. Please check README.

mtime NOT NULL constrained failed

cmp-dictionary/lua/cmp_dictionary/db.lua

Line 109 in 3a6035b

db:update("dictionary", { set = { valid = 0 } })

Possible fix is remove required?? Not quite sure tbh

Feat: Add dynamic preview using Wordnet

I just recently built out an fzf and Rofi commands that will pull from the same dictionary I'm using with cmp-dictionary that will give me a preview of the word overview or return the overview in the case of Rofi. Here is the guide I used. https://ddrscott.github.io/blog/2017/fzf-dictionary/

It would be great if we could get that same sort of preview with cmp-dictionary. Maybe just the option to configure the preview in the setup step.

Error at Nvim startup

After updating cmp-dictionary, I'm getting the following Nvim startup error:

Error detected while processing /home/smsjr/.config/nvim/init.lua:
E5113: Error while calling lua chunk: ...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:25: attempt
to index field 'uv' (a nil value)
stack traceback:
        ...vim/lazy/cmp-dictionary/lua/cmp_dictionary/dict/trie.lua:25: in function 'update'
        ...e/nvim/lazy/cmp-dictionary/lua/cmp_dictionary/source.lua:49: in function '_update'
        ...nvim/lazy/cmp-dictionary/after/plugin/cmp_dictionary.lua:10: in function 'update'
        ...e/nvim/lazy/cmp-dictionary/lua/cmp_dictionary/config.lua:94: in function 'setup'
        /home/smsjr/.config/nvim/lua/plugins/config/dictionary.lua:20: in main chunk
        [C]: in function 'require'
        /home/smsjr/.config/nvim/lua/plugins/init.lua:20: in main chunk
        [C]: in function 'require'
        /home/smsjr/.config/nvim/init.lua:30: in main chunk
Press ENTER or type command to continue

Here's my cmp-dictionary config file:

require('cmp_dictionary').setup({
  paths = { '/usr/share/dict/words' },
  exact_length = 2,
  max_number_items = 50,
  first_case_insensitive = true,
})

The file 'words' exists in the path (/usr/share/dict/words) and the cause of the error is the 'paths' config (without it, the error does not occur, but plugin does not work...).

Any help will be fully appreciated! Thx!!!

dictionary path is only set on entering the first buffer

Env

My config looks like this, and nvim-cmp is loaded on InsertEnter

{
  "hrsh7th/nvim-cmp",
  dependencies = {
    {
      "uga-rosa/cmp-dictionary",
      config = function(_, opts)
        local dict = require("cmp_dictionary")
        dict.setup(opts)
        dict.switcher({ spelllang = {
          en = "/usr/share/dict/words",
        } })
      end,
    },
  },
  opts = function(_, opts)
    local cmp = require("cmp")
    opts.sources = cmp.config.sources(vim.list_extend(opts.sources, {
      {
        name = "dictionary",
        keyword_length = 2,
      },
    }))
  end,
}

Since I set the dictionary path for spelllang=en only, the path will be appended to vim.opt_local.dictionary (notice that it's buffer-local!) on OptionSet

cmp-dictionary/lua/cmp_dictionary/init.lua

Lines 76 to 82 in f076c6b

 if opt.spelllang then 

 vim.api.nvim_create_autocmd("OptionSet", { 

 group = id, 

 pattern = "spelllang", 

 callback = callback, 

 }) 

 end

Issue

vim.opt_local.dictionary is only populated when entering the first buffer.
If the first is a "normal" buffer, completion keeps working for later buffers despite vim.opt_local.dictionary being empty, for unknown reasons.
However, if the first buffer is telescope or potentially other special buffers, completion won't work for the entire session.

To reproduce:

# Step	Operation	`:setlocal dictionary`	Has word completion?
1	start nvim		N/A
2	open a file and enter insert mode	/usr/share/dict/words	Y
3	open another file and enter insert mode		Y

# Step	Operation	`:setlocal dictionary`	Has word completion?
1	start nvim		N/A
2	`:Telescope`, then press <esc> to switch to normal mode but not quit telescope	/usr/share/dict/words	N/A
3	quit telescope, open a file and enter insert mode		N

If the number of elements in the dictionary is small, the candidate will not be obtained correctly.

Error executing vim.schedule lua callback

Hi!
Firstly, I'd like to thank you fort this amazing plugin!

I just have an issue when I open neovim I got this error:

Error executing vim.schedule lua callback: .../packer/start/cmp-dictionary/lua/cmp_dictionary/util.lua:48: attempt to index upvalue 'timer' (a nil value)
stack traceback:
.../packer/start/cmp-dictionary/lua/cmp_dictionary/util.lua:48: in function ''
vim/_editor.lua: in function <vim/_editor.lua:0>

I will appreciate a lot if you could have a look at my settings to figure it out where I made a mistake.

local dict = require("cmp_dictionary")

dict.setup({
	exact = 3,
	first_case_insensitive = false,
	document = false,
	document_command = "wn %s -over",
	async = false,
	max_items = -1,
	capacity = 5,
	debug = false,
})

dict.switcher({
	spelllang = {
		en = "~/dotfiles/en.dict",
	},
	filepath = {
		["*"] = "~/dotfiles/en.dict",
	},
})
cmp.setup({
	snippet = {
		expand = function(args)
			if not ls then
				return
			end
			ls.lsp_expand(args.body)
		end,
	},
	sources = {
		{ name = "luasnip" },
		{ name = "path" },
		{ name = "buffer", keyword_length = 3 },
		{ name = "cmdline", keyword_length = 3 },
		{ name = "dictionary", keyword_length = 3 },
	},
	window = {
		completion = cmp.config.window.bordered(),
		documentation = cmp.config.window.bordered(),
	},
	formatting = {
		expandable_indicator = true,
		fields = { "kind", "abbr", "menu" },
		format = lspkind.cmp_format({
			mode = "symbol_text",
			maxwidth = 50,
			ellipsis_char = "...",
			menu = {
				nvim_lsp = "   ",
				luasnip = "   ",
				buffer = "  ",
				path = "  ",
				cmdline = " ",
				dictionary = "  ",
			},
			before = function(_, vim_item)
				return vim_item
			end,
		}),
	},

plugin not loaded in lunarvim

Hi, as the title said. my config is:

table.insert(lvim.builtin.cmp.sources, { name = 'dictionary', keyword_length = 2 })
lvim.builtin.cmp.formatting.source_names['dictionary'] = ''

lvim.plugins = {
  { 'uga-rosa/cmp-dictionary', before = 'nvim-cmp',
    config = function()
      require 'cmp_dictionary'.setup {
        async = true,
        capacity = 5,
        dic = { ['*'] = '/path/to/words' }
      }
    end }, -- vim dictionary source for cmp
}

when i run: PackerLoad cmp-dictionary, it reports:

Error: attempted to load cmp-dictionary which is not present in plugins table!
Error in packer_compiled: ...m/site/pack/packer/start/packer.nvim/lua/packer/load.lua:135: Error: attempted to load cmp-dictionary which is not present in plugins table!
Please check your config for correctness

Could you please take a look or correct my error, mant thanks

screenshot

default

set to -1

Accept function for dic and/or file type or provide other means to set buffer-specific dictionaries

I have the following configuration:

require('cmp_dictionary').setup {
  dic = {
    ['*'] = '/usr/share/dict/words',
    ledger = '<secret location>/accounting/accounts.txt'
  }
}

While it mostly works there are a number of issues with this configuration.

This is a static configuration that can't be customized in the runtime. Think set dictionary+=asd.txt so one can set dictionaries specific to the particular buffer.
The configuration is "global", i.e. you can't customize the configuration per buffer.
All the configuration made using 'dictionary and 'thesaurus is ignored :(

I've been using a lot of dictionaries and have quite a few configurations for file types, frameworks, and so forth. It would be great to see the support of 'dictionary and/or 'thesaurus in cmp-dictionary. If that's too much work then at least accept a function instead of the string for specifying the dictionary would be very appreciated.

Multi-word entries & a dictionary separated by newlines.

Hi, lovely plugin. I am configuring Neovim into an editor for prose mathematics. I am creating a dictionary of math terminology, and I'd like to use multi-word phrases in said dictionary. For example, the phrase 'finite-dimensional vector space' occurs with high frequency. I have a dictionary of such entries, possibly multi-word, separated by newlines. Right now, the plugin separates all of these into their own entries ('finite-dimensional', 'vector', 'space'). Is there a way to use such a dictionary with your plugin? I poked around for a bit and didn't see a quick way to enable such an option, though I'm not familiar with the codebase.

opt:dictionary and opt:thesaurus

Howdy,

This just so happens to fill an exact need I have, but had a quick design question. I have my setup configured to add all dictionaries/wordlists to opt:dictionary as well as do something similar with opt:thesaurus (see here). I was curious if it is possible to just pull from those sources instead of having to mark them in the setup for this plugin?

Error with text filetype when multiple dictionaries defined for ["*"]

When I define multiple dictionaries for ["*"]

        dic = {
            ["*"] = { "/usr/share/dict/words", "/usr/share/dict/american-english"},
        },

then I get the following exception when triggering completion in a "text" filetype.
Dictionary completion works for other filetypes (e.g., py, md).

Error detected while processing TextChangedI Autocommands for "*":                                                                                                                                                   
E5108: Error executing lua ...packer/start/cmp-dictionary/lua/cmp_dictionary/merge.lua:160: attempt to get length of field 'array' (a nil value)                                                                     
stack traceback:                                                                                                                                                                                                     
        ...packer/start/cmp-dictionary/lua/cmp_dictionary/merge.lua:160: in function 'merge'                                                                                                                         
        .../packer/start/cmp-dictionary/lua/cmp_dictionary/init.lua:38: in function 'get_candidate'                                                                                                                  
        .../packer/start/cmp-dictionary/lua/cmp_dictionary/init.lua:53: in function 'complete'                                                                                                                       
        .../nvim/site/pack/packer/start/nvim-cmp/lua/cmp/source.lua:290: in function 'complete'                                                                                                                      
        ...re/nvim/site/pack/packer/start/nvim-cmp/lua/cmp/core.lua:250: in function 'complete'                                                                                                                      
        ...re/nvim/site/pack/packer/start/nvim-cmp/lua/cmp/core.lua:163: in function 'callback'                                                                                                                      
        ...re/nvim/site/pack/packer/start/nvim-cmp/lua/cmp/core.lua:213: in function 'autoindent'                                                                                                                    
        ...re/nvim/site/pack/packer/start/nvim-cmp/lua/cmp/core.lua:155: in function 'on_change'                                                                                                                     
        ...re/nvim/site/pack/packer/start/nvim-cmp/lua/cmp/init.lua:301: in function 'callback'                                                                                                                      
        ...ite/pack/packer/start/nvim-cmp/lua/cmp/utils/autocmd.lua:31: in function 'emit'                                                                                                                           
        [string ":lua"]:1: in main chun

Some words are not displaying

I have added several words to my general dictionary, but some of them are not displaying

dictionary

x86_64-windows
x86_64-windows-gnu
x86_64-windows-musl
x86_64-linux-gnu
x86_64-linux-musl
aarch64-linux-gnu
aarch64-linux-musl

result:

miss x86_64-windows-musl

my config:

    local dict = require("cmp_dictionary")
    local dict_path = g.config_dir .. "/dict/"
    dict.setup({
        exact = 2,
        first_case_insensitive = false,
        document = false,
        document_command = "wn %s -over",
        async = false,
        sqlite = false,
        max_items = -1,
        capacity = 5,
        debug = true,
    })
    dict.switcher({
        filetype = {
            go = { dict_path .. "go.dict" },
        },
        filepath = {
            [".*"] = { dict_path .. "dictionary" },
            [".*xmake.lua"] = { dict_path .. "xmake.dict" },
        },
    })

Doubled entries when configuring vim.opt.dictionary

Hi,

Thanks a lot for this plugin !

Since a recent update, I noticed that my entries were duplicated when setting up a dictionary with vim.opt.dictionary.

I tested with this minimal config. How can I correct this?

local lazypath = vim.fn.stdpath("data") .. "/lazy/lazy.nvim"
if not vim.loop.fs_stat(lazypath) then
  vim.fn.system({
    "git", 
    "clone",
    "--filter=blob:none",
    "https://github.com/folke/lazy.nvim.git",
    "--branch=stable",
    lazypath,
  })
end
vim.opt.rtp:prepend(lazypath)

require("lazy").setup({
  'uga-rosa/cmp-dictionary',
  'hrsh7th/nvim-cmp',
})

local cmp = require'cmp'
local gl = vim.api.nvim_buf_get_lines
local has_words_before = function()
  unpack = unpack or table.unpack
  local line, col = unpack(vim.api.nvim_win_get_cursor(0))
  return col ~= 0 and 
    gl(0, line -1, line, true)[1]:sub(col, col):match("%s") == nil
end

cmp.setup({
  mapping = cmp.mapping.preset.insert({
    ["<CR>"] = cmp.mapping.confirm {
      behavior = cmp.ConfirmBehavior.Replace,
      select = false,
    },
    ["<Tab>"] = cmp.mapping(function(fallback)
      if cmp.visible() then
        cmp.select_next_item()
      elseif has_words_before() then
        cmp.complete()
      else
        fallback()
      end
    end, { "i", "s" }),
    ["<S-Tab>"] = cmp.mapping(function(fallback)
      if cmp.visible() then
        cmp.select_prev_item()
      else
        fallback()
      end
    end, { "i", "s" }),
  }),
  sources = cmp.config.sources({
    { name = 'dictionary' },
  })
})

vim.opt.dictionary = '$HOME/test.dict'

iskeyword

Hi, is there a way to use iskeyword fonction ?

I have a dictionary with words like \word and another, and it seems impossible to make the iskeyword function working.
I have already try that in my ftplugin :

vim.cmd[[set iskeyword+=\]]

and

vim.opt.iskeyword:append([[\]])

Is there a solution ?
Thanks !

Thanks !

is_setup returns false

cmp-dictionary/lua/cmp_dictionary/caches.lua

Line 167 in b60df1d

 echo("The configuration method has changed, please use setup (check README for details).", true) 

Is it on purpose you want this to return false? I only ask because it seems to me that when update is called, more often than not setup would have already been called and config wouldn't be empty?

How to configure filetype specific dictionaries in version 3?

Current config:

local dict = require("cmp_dictionary")
dict.setup({
    dic = {
        ['gap'] = '/Users/kiryph/.vim/GAPWORDS',
    },
    -- The following are default values, so you don't need to write them if you don't want to change them
    exact_length = 3, -- how many characters at the beginning are used as the exact match.
    first_case_insensitive = false,
    async = false,
    debug = false,
    max_nubmer_items = 0,
})
dict.switcher({
  filetype = {
    gap = '/Users/kiryph/.vim/GAPWORDS',
  },
})

As far as I see it stopped working. No completion for gap words.

https://github.com/gap-system/gap/blob/master/etc/vim/gap.vim#L170-L173

New version complains that switcher is deprecated and I see no help in the documentation.

How to generate these two files in `/usr/share/dict/` ?

I have installed aspell and compile a dict from aspell.net.

aspell dicts | grep en

>  aspell dicts | grep en

ca-general
ca-valencia
en
en-variant_0
en-variant_1
en-variant_2
en-w_accents
en-wo_accents
en_AU
en_AU-variant_0
en_AU-variant_1
en_AU-w_accents
en_AU-wo_accents
en_CA
en_CA-variant_0
en_CA-variant_1
en_CA-w_accents
en_CA-wo_accents
en_GB
en_GB-ise
en_GB-ise-w_accents
en_GB-ise-wo_accents
en_GB-ize
en_GB-ize-w_accents
en_GB-ize-wo_accents
en_GB-variant_0
en_GB-variant_1
en_GB-w_accents
en_GB-wo_accents
en_US
en_US-variant_0
en_US-variant_1
en_US-w_accents
en_US-wo_accents
fa-generic
fa-scientific

my dict directory:

> ls /usr/share/dict
      README       connectives       propernames       web2       web2a       words

[question] not support set dictionary in last version?

after update to last version, set dictionary+=xx.dict can't work, set dict file in setup can work.
Is this a bug? Or will it only support write in setup?

How to show a word meaning as a document?

Hi!

Thanks for the great plugin!

I use it with my anki word deck to use the vocabulary I'm learning. Extremely useful!

Please add LSP preview what can display a word meaning or translation.

Slow for large dictionaries

I use two dictionaries, one for language with inflection which means number of entries is huge (4493307 vs 127466 for British English, which means it is over 35 times larger). cmp-dictionary slows down nvim startup by little more than six seconds for me.
I took a look at source, and it seems that file is read asynchronously, yet it blocks the editor. I'm using nvim v0.5.1.

Can't load the 'dictionary' source

The Neovim version:

Other sources work normally and 'dictionary' can't work:

Here's my settings:

The .dic file is downloaded from github, I don't know if it would be the cause.

"async" option is not explained in the help file

async boolean option is not explained in the help file under Options (cmp-dictionary-option).

Would be nice to know what it exactly does, if anything.

ANNOUNCEMENT

v2.0.0

Remove dic from setup.
Instead of, refer to the value of 'dictionoary'.
Add switcher, utility to switch 'dictionary' values according to conditions.

See help in details.

Possibility to show which dictionary does a word belong to ?

For example, with the cmp-path plugin, the type (file/folder) of each choice is displayed:

Is there any way to get the source dictionary in the same way ?

For example, I have two dictionary files: path/to/dic/EN and path/to/dic/FR. I would like, instead of the Text placeholder, have the corresponding file (EN/FR in my case) :

I don't have any dev experience in lua nor vim so I really don't know how feasible that would be based on your implementation.

Regardless, thank you for your time and efforts with this nice plugin 😃

Error in decode when opening a file

The plugin was working fine but suddenly I got the following error msg:

Error executing luv callback:                                                                                                                                    
...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:139: bad argument #1 to 'decode' (string expected, got nil)
stack traceback:
        [C]: in function 'decode'
        ...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:139: in function <...acker/start/cmp-dictionary/lua/cmp_dictionary/caches.lua:138>

I tried to use different dictionaries and reduce to few words but the error still persists.

Unclear step to generate a dictionaries

Please write a foolproof "How to"

This line of code is not clear enough to me:

aspell -d <lang> dump master | aspell -l <lang> expand > my.dict

I' m not common with aspell What is "lang"? Also, How can I list the avalible one I have in my OS?

Each time I use the command I'm getting an error that says:

Error: The file "/usr/lib/aspell-0.60/en_US" can not be opened for reading.   # What I understand as a permission error

But, in a docker container I made I get the expected long list of words from aspell

Is it possible to change loaded dictionary dynamically based `filetype` change?

Is it possible to change loaded dictionary dynamically based filetype change?

I've 2 workflows.

use tabnnew to create a scratch window/buffer to do some work, such as edit/format a message before sending, drop the buffer after paste the message to IM. In this case, the &filetype is an empty string.
Is it possible to define some dict for &filetype is empty string in the config something like:

    dic = {
      [""] =  string.format('%s/site/pack/packer/start/vim-dict/dict/text.dict',vim.fn.stdpath('data')),
      ["text"] =  string.format('%s/site/pack/packer/start/vim-dict/dict/text.dict',vim.fn.stdpath('data')),
      ["javascript,typescript"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/javascript.dict',vim.fn.stdpath('data')) },
      ["lua"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/lua.dict',vim.fn.stdpath('data')) },
      ["vim"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/vim.dict',vim.fn.stdpath('data')) },
    }

I'm working on some combined file type project, such as AWS cloudformation YAML file, it can contain shell/json code fragments.
I have defined some key binding to change the &filetype on the fly

        nnoremap <leader>1 :set filetype=html<CR>
        nnoremap <leader>2 :set filetype=javascript<CR>
        nnoremap <leader>3 :set filetype=json<CR>
        nnoremap <leader>4 :set filetype=sh<CR>
        nnoremap <leader>5 :set filetype=sql<CR>
        nnoremap <leader>6 :set filetype=vim<CR>
        nnoremap <leader>7 :set filetype=txt<CR>
        nnoremap <leader>8 :set filetype=lua<CR>
        nnoremap <leader>9 :set filetype=tdx<CR>
        nnoremap <leader>f1 :set filetype=log<CR>
        nnoremap <leader>f2 :set filetype=txt<CR>
        nnoremap <leader>f3 :set filetype=markdown<CR>
        " Set buffer filetype
        nnoremap <Leader>f= :set filetype=

For example, when I'm editing a yaml file which contains a lot of shell code, I press 4 to change the filetype to shell, take advantage of shell syntax highlighting.

After the &filetype changed, is it possible to detect the change, and load the specific dicts based on the changed filetype?

Set dictionary based on vim `spelllang`?

Sometimes I write prose in English, or in other languages within vim, and adjust vim's spelllang option to get spelling corrections.

Is it possible to configure cmp-dictionary so that the dictionary (and list of words for completion) changes, depending on what spelllang option is set?

Notice of Destructive Change

Remove dic from an option of setup.
Add an utility function for set 'dictionary'.

Currently, dictionaries are registered using the setup function, but this should be changed to refer to the 'dictionary' option in neovim itself.
Instead, I will add a function require('cmp_dictionary').register_dictionary (name to be determined) to switch 'dictionary' with the same degree of freedom as before.

I have released v1.0.0 in preparation for this disruptive change.

[Feature]Auto detect/change case based on input

Is it possible to auto detect and change based on input, just like this plugin did: https://github.com/skywind3000/vim-auto-popmenu

for example, the dictionary contains the word "example", if I type "exa", it will prompt a match in the dropdown list. However, if I type "Exam", there's none matched.

this is my config:
` use 'skywind3000/vim-dict'
use {
'uga-rosa/cmp-dictionary',
after='nvim-cmp',

config = function()
  require('cmp_dictionary').setup ({
    dic = {
      ["text"] =  string.format('%s/site/pack/packer/start/vim-dict/dict/text.dict',vim.fn.stdpath('data')),
      ["javascript,typescript"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/javascript.dict',vim.fn.stdpath('data')) },
      ["lua"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/lua.dict',vim.fn.stdpath('data')) },
      ["vim"] = { string.format('%s/site/pack/packer/start/vim-dict/dict/vim.dict',vim.fn.stdpath('data')) },
    },
    exact = 2,
    async = false,
    capacity = 5,
    debug = false,
  })
end

}
`

	if opt.spelllang then
	vim.api.nvim_create_autocmd("OptionSet", {
	group = id,
	pattern = "spelllang",
	callback = callback,
	})
	end