listentosadhu.app

Listen to Sadhu — public, read-only access to the Lectorium corpus of Vedic scripture and recorded lectures. Scripture: verses (original Devanagari/Bengali + IAST transliteration + translations) and documents (commentaries/purports, prose chapters, letters). Lectures: transcribed talks and conversations with metadata and references. Addressing: a reference is a book code plus a position, e.g. "BG 2.13", "SB 5.5.3", "CC Madhya 8.128" (Cyrillic codes like "ШБ 5.5.3" also work). Tools that take a reference accept either the "ref" string or a resolved source id + tokens. In the tokens filter (search / track_list / verse_list), a BARE CHAPTER number covers the whole chapter — use "7" for all of chapter 7, "7.1" only for that one verse. How to use the tools: - search: semantic + lexical search over verses, documents, tracks and titles. - source_get/list/resolve, author_list/resolve, location_list/resolve: turn a name into an id used by the filters. - verse_get/list, document_get/list: read scripture. verse_get(lang) returns the original + IAST (line arrays) + the canonical translation inline, plus an alternatives manifest of other translation kinds. verse_translation(lang, kind) fetches a specific translation; verse_synonyms(lang, kind) gives the word-by-word ("Synonyms") array. A verse's purport is document_list(source, tokens, kind:"commentary"). - track_get/list, transcript_window: read lectures. Find the spoken moment of a verse with search(types:["track"]) then transcript_window. Inline components (render an interactive UI, not just text): - To SHOW a scripture verse as a card — call verse_render(ref/source+tokens/id, lang). It renders the original script, transliteration (in that language), word-by-word and translation inline. - To SHOW a video clip the user wants to WATCH — from a search(types:["media"]) hit — call media_render(id) with the hit's media_id. It renders an inline video player. - To let the user HEAR a lecture passage — from a search(types:["track"]) hit — call excerpt_render(track_id, start_ms, end_ms). It renders an inline audio player that generates and plays that exact passage (max 10 minutes).