Chinese spell checking

This is a question about Atom’s spell-check package. I use Atom to write in multiple languages (mostly programming) and I enjoy the convenience of the spell-check package for most of these purposes. But sometimes I write Chinese texts, and this is were it gets annoying.

As far as I know – due to the way the Chinese language works – there is no spell checking algorithm available (please correct me if I’m mistaken). The standard settings of spell-check result in every single character being underlined. So I’d like to disable spell checking for Chinese, but not for English and other languages where spell checking is available.

I’ve played around with the settings of spell-check a little bit and I’ve stumbled on a “hacky” solution: In the locales setting (which is empty by default), I simply list all the languages I have installed on my system, excluding Chinese. Now spell checking is disabled in a Chinese document, and still enabled for English!

What’s going on? I would expect my workaround to not change the behaviour at all. But somehow it works. The only drawback is that now in a (mainly) Chinese document, English words are not spell checked either.

I do not have an answer from experience (I write only in English) but this post is a good example of a “custom Atom workflow or profile”. That is, some users require a unique mix of packages and settings to meet their requirements.

The case for managing Atom profiles is not new. It was discussed in this forum more than a year ago to my knowledge. Search “profiles”.

The general idea is that it would be convenient to switch Atom into different modes of working (workflows, profiles, languages, projects). In the case of the OP (@Lililas) one such “custom profile” might be editing only in Chinese.

But it goes beyond that. A multi-dimensional table is envisaged to cover all the “DNA’s” of Atom workflows. And simply enabling/disabling packages is not the answer since there will be different configuration files (e.g. snippets) required for each scenario.

I have drawn up a target list of profiles which I intend to configure for my own purposes and I reason that Ansible is one way forward to automate provisioning of Atom desktop.

Returning to the OP’s question, one approach (in theory) might be to use an external tool (in a tool chain) to analyse the Chinese, and hybrid Chinese/English documents for spelling errors.

A command might be written in Atom (use process-palette) to run a Python or bash script to copy the contents of the document to be pasted into a separate running concordance analysis tool.

The tool I recommend is AntConc

The advantage of this tool is that it is multi-lingual (ongoing research in Japan) and other corpus metrics can be viewed.

This approach does not give inline spell checking but is a powerful tool for analysing any document or corpus of documents.

1 Like