// the find
FreeLanguageTools/vocabsieve
Simple sentence mining tool for language learning
VocabSieve is a desktop GUI tool for sentence mining — capturing vocabulary from reading or watching content and turning it into Anki flashcards. It sits between your reading material and Anki, automating the tedious parts: looking up definitions, pulling audio, and formatting cards. Aimed at intermediate language learners doing immersion-style study.
Genuinely local-first: dictionaries, frequency lists, and audio all run offline once downloaded, which means no subscription and no service going away mid-study. The ereader integration (KOReader and Kindle highlights → Anki batch import) is the feature that actually saves time for heavy readers — most tools ignore this workflow entirely. Lemmatization via simplemma/pymorphy3 is handled correctly rather than left as an exercise for the user. Dictionary format support is wide: StarDict, MDX, DSL, Migaku, plain JSON — you're not locked into one source.
520 stars for a 4+ year-old tool that fills a genuine niche suggests adoption has plateaued; the community is small enough that edge-language support will have real gaps even if the architecture supports it. The built-in web reader bundles jQuery, Bulma, epub.js, and jszip as vendored statics — no lockfile or integrity check, so those are just bit-rotting in-tree. Test coverage is thin: one test file for local dictionaries and a `run_tests.sh` that looks nearly empty, which means refactoring the dictionary parsing layer is risky. Configuration is a nine-tab dialog; the 'sane defaults' claim is undermined by the volume of knobs a new user has to understand before getting a working card template.