Offline command-line dictionary based on Wiktionary, thus supporting basically every word in every language ever.
Usage
- Download the Wiktionary
pages-articles.xml.bz2
dump. - Extract the bz2.
- Run
build_definitions_db
on the resulting xml file. - Run
define
to define words.
TODO
- Detect languages and parts of speech automatically (is currently hardcoded)
- Render Wiki templates
- This is super hard because many templates eventually resolve to Scribunto scripts, so we have to call into Lua
- Resolve inflections
- Probably has to be language-specific
- Additional data sets
- The ones Tangorin uses seem quite nice for Japanese