spelllangidentifier

Vim plugin to automatically identify and set the spell language

It detects the language based on the buffer contents through 'mguesser' (http://www.mnogosearch.org/guesser/), which in turn uses "N-Gram-Based Text Categorization".

Installation

Decompress the tarball and change to directory sli-x.y.z/. From there, assuming you have a standard Vim setup:

   mkdir -p ~/.vim/plugin
   cp -r SpellLangIdentifier.vim LanguageIdentifier.sh mguesser ~/.vim/plugin
   cp vimrc.sample ~/.vimrc.sli
   cd ~/.vim/plugin/mguesser && make

Configuration (~/.vimrc)

Please, check out the sample Vim configuration file which you have just copied. If you prefer you can introduce your changes there and source it directly from your personal vimrc (usually ~/.vimrc), by inserting the line source ~/.vimrc.sli.

Make sure you remove or comment out any previous spell checking configuration that might conflict with the new ones.

Usage

1. Automatically (just open a text, mail or TeX file, as illustrated in vimrc.sli). It even works on-the-fly while you edit the file, switching the languages on a paragraph basis when working on multi-language documents (configurable).
2. Using shortcuts in normal or insert mode (F5, F6, F7 and F8; F9 to disable spell checking and on-the-fly language identification).
3. Directly: :{range}SpellLangIdentify (identify the language based on {range} lines; default is the whole buffer).

Notes

The user can create its own reference language set ("maps") which will guide/bias the language identification. An example would be to create an n-gram frequency file (a map) for words related to the medical literature or, perhaps, to some subset of the English language. Please check out mguesser's documentation for that.

Prerequisites

• Unix-like environment (POSIX shell, coreutils, awk, sed, grep, file, ...).

• In order to improve language detection according to the file type at hand, it is recommended (but not required) to have the following utilities installed: detex, ps2ascii, html2text.