This is a plugin for user to lookup synonyms of any word under cursor or phrase covered in visual mode, and replace it with an user chosen synonym. It also accepts word/phrases from manual input for synonym checkup.
Notice: Currently this plugin Supports only English (en), Chinese (cn), French (fr), Russian (ru) and German (de) thesaurus query. If you want to use the plugin for other languages, or if you're not satisfied with the performance of current backends and know of some online synonym services that I can integrate into the plugin, please come and post an issue with your suggestion.
This plugin is written in Python. So +Python or +Python3 version of Vim is required.
Deleted thesaurus_com
backend due to the legal warning from Thesaurus.com on
the upstream package thesarus.
For existing users that still have thesaurus_com
explicitly enabled, the
backend now always return exception. For English users, currently the highest
quality source is either the OpenOffice Thesaurus source or mthesaur.txt. I am sorry for any inconvenience caused.
Added two new French backend based on synonymo.fr
and cnrtl.fr. To activate them, add fr
to variable
g:tq_language
and synonymo_fr
and/or cnrtl_fr
to g:tq_enabled_backends
.
Added new German backend based on
openthesaurus.de. This
thesaurus backend is one of the German default backends. To activate it, add
de
to variable g:tq_language
. Also, if you have manual specification for
g:tq_enabled_backends
, be sure to add openthesaurus_de
to your backend
list.
Added new Russian backend based
on yarn-synsets.csv. To use, add
yarn_synsets
to Vim variable g:tq_enabled_backends
, e.g.
let g:tq_enabled_backends=["yarn_synsets", "openoffice_en", "mthesaur_txt"]
also, add ru
to variable g:tq_language
and make sure that either
yarn-synsets.csv
in downloaded to ~/.vim/thesaurus
or where variable
g:tq_yarn_synsets_file
points at.
A new variable is designed to force use certain version of Python, e.g.
let g:tq_python_version = 2
will force Plugin to use Python2 no matter if Python3 is supported.
Use your plugin manager of choice.
- Pathogen
git clone https://github.com/ron89/thesaurus_query.vim ~/.vim/bundle/thesaurus_query.vim
- Vundle
- Add
Bundle 'ron89/thesaurus_query.vim'
to.vimrc
- Run
:BundleInstall
- Add
- NeoBundle
- Add
NeoBundle 'ron89/thesaurus_query.vim'
to.vimrc
- Run
:NeoBundleInstall
- Add
- vim-plug
- Add
Plug 'ron89/thesaurus_query.vim'
to.vimrc
- Run
:PlugInstall
- Add
By default, command :ThesaurusQueryReplaceCurrentWord
is mapped to
<Leader>cs
.
nnoremap <Leader>cs :ThesaurusQueryReplaceCurrentWord<CR>
This routine check the synonyms of the word under cursor and replace it with
the candidate chosen by user. The corresponding non-replacing routine is
defined as ThesaurusQueryLookupCurrentWord
. User may choose to use it if you
prefer the split buffer display of result over the word replacement routine.
Another might-be-useful routine is the one to query synonym for and replace
a multi-word phrase covered in visual mode, using the same key mapping
<Leader>cs
vnoremap <Leader>cs y:ThesaurusQueryReplace <C-r>"<CR>
The phrase covered in visual mode can be in a same line, or wrapped in two or more lines.
Also, this plugin support Vim's builtin completefunc
insert mode autocomplete
function. To invoke it, use keybinding ctrl-x ctrl-u
in insert mode. This
function resembles Vim's own thesaurus checking function, but using online
resources for matchings.
Finally, this plugin support thesaurus checkup for manually input through
command mode command :Thesaurus
.
:Thesaurus your phrase
To ensure stability of the plugin's functionality, under the hood, this plugin uses multiple backends sequentially to query for a synonym. Backends function independently, hence the plugin will be functional as long as one of the these backends is behaving properly.
- openoffice_en queries from local thesaurus database provided by
OpenOffice. It is an useful option when you don't have any internet access at
all. If you are using Linux and has installed OpenOffice from official repo,
you should have index file
th_en_US_v2.idx
and database fileth_en_US_v2.dat
in/usr/share/myspell/dicts
, and this plugin should work outright. But if not, you can download the thesaurus data from this link. You may then manually inform the plugin of the downloaded location on your machine by setting variableg:tq_openoffice_en_file
. Eg, if your indes and database (2 files) are~/Downloads/MyThes-1.0/th_en_US_new[.idx,.dat]
then you should set your variable by addinglet g:tq_openoffice_en_file="~/Downloads/MyThes-1.0/th_en_US_new"
in your.vimrc
. - mthesaur_txt queries from local
mthesaur.txt
. It is another useful option when you don't have any internet access at all. For this backend to work, be sure to download the file from gutenberg.org and place it under"~/.vim/thesaurus"
. If you place the file elsewhere, change global variableg:tq_mthesaur_file
to point to the file you downloaded, eg. put the following linelet g:tq_mthesaur_file="~/.config/nvim/thesaurus/mthesaur.txt"
into your.vimrc
file if yourmthesaur.txt
is placed in folder"~/.config/nvim/thesaurus/"
. NOTE thatmthesaur.txt
tend to return a lot more results than we can process. The global parameterg:tq_truncation_on_definition_num
andg:tq_truncation_on_syno_list_size
may be helpful in reduce the returned size. But some meaningful results may be lost. - datamuse_com queries from datamuse.com using its officially provided API. The returned synonym list is usually quite relavant with reasonable quality. But the synonyms list tend to be short, so it might leave out some less-frequently-used synonyms.
- cilin_txt queries from local
cilin.txt
. It makes use of a Chinese thesaurus source "cilin.txt". For this backend to work, be sure to download the file from this github repo and place it under"~/.vim/thesaurus"
. If you place the file elsewhere, change global variable |g:tq_cilin_txt_file| to point to the file you downloaded, eg. put the following linelet g:tq_cilin_txt_file="~/.config/nvim/thesaurus/cilin.txt"
into your.vimrc
file if yourcilin.txt
is placed in folder"~/.config/nvim/thesaurus/"
. Note that if you downloaded "cilin.txt" from elsewhere, make sure that your sourcecilin.txt
is utf-8 encoded. Or the plugin won't function correctly with the file. - yarn_synsets is a Russian thesaurus backend. It
uses yarn-synsets.csv as synonym
resource, providing a fast responding and decent thesaurus query solution.
Please place
yarn-synsets.csv
in~/.vim/thesaurus
, or use variableg:tq_yarn_synsets_file
to indicate the location ofyarn-synsets.csv
in your system. - jeck_ru is a Russian thesaurus backend. It queries jeck.ru for synonym resources. This website didn't provide standard API to use. Hence functionality of this backend depends on whether the website owner will change the webpage design.
- woxikon_de is a German thesaurus backend. It queries woxikon.de for synonym resources. This website didn't provide standard API to use. Hence functionality of this backend depends on whether the website owner will change the webpage design.
- openthesaurus_de is a German thesaurus backend. It queries
openthesaurus.de for synonym
resources. This thesaurus backend is one of the German default backends. To
activate it, add
de
to variableg:tq_language
. Also, if you have manual specification forg:tq_enabled_backends
, be sure to addopenthesaurus_de
to your backend list. - cnrtl_fr is a French thesaurus backend. It queries
cnrtl.fr for synonym resources. This
website didn't provide standard API to use. Hence functionality of this
backend depends on whether the website owner will change the webpage design.
This backend requires
bs4
BeautifulSoup dependency. - synonymo_fr is a French thesaurus backend. It queries
synonymo.fr for synonym resources. This
website didn't provide standard API to use. Hence functionality of this
backend depends on whether the website owner will change the webpage design.
This backend requires
bs4
BeautifulSoup dependency.
The thesaurus query plugin will go through the list g:tq_enabled_backends
in
sequence until a match is found. Unless user explicitly instruct, Next query
will be conducted only when the previous query return empty synonym list or
failed to query. You may remove unwanted backend or lower their priority by
removing them/putting them on latter position in variable
g:tq_enabled_backends
as following example:
let g:tq_enabled_backends=["woxikon_de","jeck_ru","openoffice_en","mthesaur_txt"]
let g:tq_enabled_backends=["cilin_txt",
\"openthesaurus_de",
\"yarn_synsets",
\"openoffice_en",
\"mthesaur_txt",
\"datamuse_com",]
Non-Engligh backends are currently not activated by
default, due to the default setting g:tq_language='en'
. To enable Russian, German,
or Chinese backend, add 'ru', 'de' or 'cn' to the tq_language
list:
let g:tq_language=['en', 'ru', 'de', 'cn']
Or if you want to use only German thesaurus engine in specific/current buffer
let b:tq_language=['de']
To ensure the best user experience, the backend that reports error during query will have its priority automatically lowered. If user want to restore originally defined priority, simply invoke command
:ThesaurusQueryReset
Timeout mechanism (configurable with g:tq_online_backends_timeout
) is added
to all online query backends to reduce query time(in seconds). Default value is
1.0
(second), so that user using slower internet or having long response time
with certain backend servers could still has reliable performance. For users
with faster internet, you may want to lower the value (mine is 0.4
).
let g:tq_online_backends_timeout = 0.4
Q: Why would this help?
A: Usually when thesaurus is found, the server respond quickly. However, when the word is not found, it will take a while before server return 404 error. And our plugin will freeze Vim before the error is returned. By setting timeout, we may cut the waiting time down and start query from next backend sooner. So that waiting-time for user can be drastically reduced if set properly.
Q: What does it mean by long response time to certain servers
A: Depending on where you are, some thesaurus query service might not have
server built up close to you. In this case, the response time between you and
the server might varied greatly. For example, woxikon_de
might only have
server in Germany. And I am in Singapore. Based on my experience, it might take
400ms-1.0s
for the server to respond to my request. If I want to use its
service reliably, I might want to set the timeout to 1.0
(second), so that
most of the valid query can yield response before timeout.
Synonym replacing interface(shown in first screenshot) is the key feature of this plugin. I will make my best effort to make sure it work sensibly. If user has any complain about the current layout or otherwise, please draft an issue to let me know. Currently, I have drafted two variables to help reducing the candidate list when the number of synonym is too overwhelming.
Synonyms are grouped by definitions. If there are too many groups to your
liking, you may reduce the number of groups shown to 3
by setting
let g:tq_truncation_on_definition_num = 3
Sometimes synonyms in a single group is also too long to choose from, in this
case, to reduce the number of synonym shown in each group to no more than
200
, you can set
let g:tq_truncation_on_syno_list_size = 200
Know that if query result is truncated by your rule, and you want to browse
through the result being truncated, you can still access the complete synonym
list by typing A<Enter>
in your input prompt.
Notice Due to current lack of user feedback, and that I do not want to arbitrarily make up truncation threshold that may damage user experience, both truncation methods are inactive unless variables stated above are set explicitly by user.
This plugin is functionally based on and extended from the idea behind vim-online-thesaurus by Anton Beloglazov:
Several key improvements were made comparing to his plugin:
- added synonym replacing mechanism
- added fallback backend to ensure functionality even when server down or lack of internet connection.
- added alternative thesaurus sources and switching mechanism
- added multilanguage support. With proper backends, it is possible to extend the functionality of this plugin to any western letter based languages.
- Add more thesaurus source and try to parallelize the query process with a timeout limit.
- Implement algorithm to make synonym candidates in a same form(tense, plurality, etc.). This could take a while... :-|
- Update documentation