Status · 2026-06-22

Tsumugu Encoding Dictionary

9,663 entries across TOCFL A1–C2: 2,662 characters and 7,001 words, plus 515 grammar patterns. The 繁/简 script toggle and the Hán-Việt bridge render now ship across the whole site. Two authoring lanes stay open: word example sentences, and Vietnamese for the Simplified stories.

100%
Traditional encoding (chars + words)
100%
Vietnamese on every entry
982
Simplified stories authored
95%
Character example sentences

What it is

The public encoding dictionary for Tsumugu: 9,663 character and word entries plus 515 grammar patterns, for students of Traditional Chinese. Each entry opens with form and functional structure, then stories, mnemonics, cross-referenced grammar, and in-language explanations and examples. The site renders from a private content store, one JSON per entry, via scripts/render_site.py. Deployed to its custom domain on 2026-06-15; validation passes, with unique IDs across entries and patterns and a clean schema. The reader, wiki, and engine stay separate surfaces.

Lanes

17 tracked
Done9
✓
Traditional character encoding
Form, story, composition on every character
2,662 / 2,662
✓
Word entries authored
All TOCFL A1–C1 words
7,001 / 7,001
✓
Vietnamese (characters)
definition.vi on every character
2,662 / 2,662
✓
Vietnamese (words)
The paused VI campaign, finished
7,001 / 7,001
✓
Simplified character encoding
Fresh stories from Simplified components, every differing char
982 / 982
✓
EN / VI gloss toggle
VI-lock CSS bug fixed; shipped and deployed
shipped
✓
繁 / 简 script toggle
Entries, grammar, and UI chrome all convert; authored Simplified stories render; 53,117-string offline render map; search resolves either script
shipped
✓
Hán-Việt bridge render
Staged entry card — VI-only bridge rail + inverted sound-family folded into the FORM card; collapses cleanly in EN
shipped
✓
Pinyin / 注音 reading toggle
拼/ㄅ toggle, default pinyin; search rows honor it
shipped
In progress4
◐
Example sentences (characters)
133 left
95%
◐
Collocations (characters)
The menu driving example variety
95%
◐
Vietnamese on example sentences
Rides with example authoring; every char with examples carries vi
95%
◐
Audio / TTS
Manifest exists; per-word and per-sentence wiring pending
partial
Not started · considering8
○
Sound-series pages
A dedicated page per phonetic series (惑/或/域 → hoặc family), built from the same inversion that feeds the FORM-card sound-family — turn the generative bridge feature into browsable, cross-linked pages. Logged 2026-06-22; design later.
idea
○
In-band Chinese definitions (B2 · C1 · C2)
Considering: rewrite definition.zh so each entry's Chinese definition stays within the learner's own band vocabulary — most acute at B2/C1/C2 where the gloss can out-level the headword. Relates to the "band-verified graded definitions" concept in PRD-Trilingual-Dictionary-PUBLIC-v2; no standalone brainstorm doc yet.
considering
○
Simplified → Vietnamese
The 982 Simplified stories carry EN and 簡明中文 only; no Vietnamese yet
0 / 982
○
Example sentences (words)
The largest remaining lane
26 / 7,001
○
Collocations (words)
Pairs with the word example pass
~0 / 7,001
○
scholarNote renderer
7 notes authored (e.g. 苛's 疴 loan); no UI surfaces them
7 authored
○
Idioms (成語)
Queued
0
○
C2+ / higher-band expansion
Corpus solid A1–C1; C2 thin (35 chars), C2+ tail noted
C2 thin

Coverage matrix

TraditionalSimplified · 982 differing
Encoding (EN + 中文)100%100%
Vietnamese100%not started
Example sentences95% chars · ~0% wordsconverts by glyph

Next moves

  1. Simplified → Vietnamese (982). Bounded translation lane, the same shape as the Traditional VI campaign that just finished. Cleanest to queue while that pipeline stays warm.
  2. Word example sentences (7,001). The largest lane. Point the recipe-v3 run at words.
  3. Sound-series pages (design). Reuse the FORM-card sound-family inversion to build browsable per-series pages — the generative bridge feature as its own surface.
  4. In-band Chinese definitions (B2 · C1 · C2), under consideration. Grade definition.zh to each band's own vocabulary; sharpest at the upper bands. Needs a brainstorm doc before a work order.

Why it's free

The dictionary is free and runs no ads. The internet I want to build is the one where people make things for free, for the fun of it. That choice may cost me later, and for now I will take it.

The current goal is one product students can use and enjoy. Their attention is already pulled in every direction, so the design works to protect it.

Money can come later, when there is something here worth charging for. Until then, charging students does not fit.