JA WORD DATA

Selected distribution license: CC0-1.0. Where the upstream offers alternative licenses, this distribution selects the option stated here.

--- Original upstream notice: license ---
Creative Commons Legal Code

CC0 1.0 Universal

    CREATIVE COMMONS CORPORATION IS NOT A LAW FIRM AND DOES NOT PROVIDE
    LEGAL SERVICES. DISTRIBUTION OF THIS DOCUMENT DOES NOT CREATE AN
    ATTORNEY-CLIENT RELATIONSHIP. CREATIVE COMMONS PROVIDES THIS
    INFORMATION ON AN "AS-IS" BASIS. CREATIVE COMMONS MAKES NO WARRANTIES
    REGARDING THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS
    PROVIDED HEREUNDER, AND DISCLAIMS LIABILITY FOR DAMAGES RESULTING FROM
    THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS PROVIDED
    HEREUNDER.

Statement of Purpose

The laws of most jurisdictions throughout the world automatically confer
exclusive Copyright and Related Rights (defined below) upon the creator
and subsequent owner(s) (each and all, an "owner") of an original work of
authorship and/or a database (each, a "Work").

Certain owners wish to permanently relinquish those rights to a Work for
the purpose of contributing to a commons of creative, cultural and
scientific works ("Commons") that the public can reliably and without fear
of later claims of infringement build upon, modify, incorporate in other
works, reuse and redistribute as freely as possible in any form whatsoever
and for any purposes, including without limitation commercial purposes.
These owners may contribute to the Commons to promote the ideal of a free
culture and the further production of creative, cultural and scientific
works, or to gain reputation or greater distribution for their Work in
part through the use and efforts of others.

For these and/or other purposes and motivations, and without any
expectation of additional consideration or compensation, the person
associating CC0 with a Work (the "Affirmer"), to the extent that he or she
is an owner of Copyright and Related Rights in the Work, voluntarily
elects to apply CC0 to the Work and publicly distribute the Work under its
terms, with knowledge of his or her Copyright and Related Rights in the
Work and the meaning and intended legal effect of CC0 on those rights.

1. Copyright and Related Rights. A Work made available under CC0 may be
protected by copyright and related or neighboring rights ("Copyright and
Related Rights"). Copyright and Related Rights include, but are not
limited to, the following:

  i. the right to reproduce, adapt, distribute, perform, display,
     communicate, and translate a Work;
 ii. moral rights retained by the original author(s) and/or performer(s);
iii. publicity and privacy rights pertaining to a person's image or
     likeness depicted in a Work;
 iv. rights protecting against unfair competition in regards to a Work,
     subject to the limitations in paragraph 4(a), below;
  v. rights protecting the extraction, dissemination, use and reuse of data
     in a Work;
 vi. database rights (such as those arising under Directive 96/9/EC of the
     European Parliament and of the Council of 11 March 1996 on the legal
     protection of databases, and under any national implementation
     thereof, including any amended or successor version of such
     directive); and
vii. other similar, equivalent or corresponding rights throughout the
     world based on applicable law or treaty, and any national
     implementations thereof.

2. Waiver. To the greatest extent permitted by, but not in contravention
of, applicable law, Affirmer hereby overtly, fully, permanently,
irrevocably and unconditionally waives, abandons, and surrenders all of
Affirmer's Copyright and Related Rights and associated claims and causes
of action, whether now known or unknown (including existing as well as
future claims and causes of action), in the Work (i) in all territories
worldwide, (ii) for the maximum duration provided by applicable law or
treaty (including future time extensions), (iii) in any current or future
medium and for any number of copies, and (iv) for any purpose whatsoever,
including without limitation commercial, advertising or promotional
purposes (the "Waiver"). Affirmer makes the Waiver for the benefit of each
member of the public at large and to the detriment of Affirmer's heirs and
successors, fully intending that such Waiver shall not be subject to
revocation, rescission, cancellation, termination, or any other legal or
equitable action to disrupt the quiet enjoyment of the Work by the public
as contemplated by Affirmer's express Statement of Purpose.

3. Public License Fallback. Should any part of the Waiver for any reason
be judged legally invalid or ineffective under applicable law, then the
Waiver shall be preserved to the maximum extent permitted taking into
account Affirmer's express Statement of Purpose. In addition, to the
extent the Waiver is so judged Affirmer hereby grants to each affected
person a royalty-free, non transferable, non sublicensable, non exclusive,
irrevocable and unconditional license to exercise Affirmer's Copyright and
Related Rights in the Work (i) in all territories worldwide, (ii) for the
maximum duration provided by applicable law or treaty (including future
time extensions), (iii) in any current or future medium and for any number
of copies, and (iv) for any purpose whatsoever, including without
limitation commercial, advertising or promotional purposes (the
"License"). The License shall be deemed effective as of the date CC0 was
applied by Affirmer to the Work. Should any part of the License for any
reason be judged legally invalid or ineffective under applicable law, such
partial invalidity or ineffectiveness shall not invalidate the remainder
of the License, and in such case Affirmer hereby affirms that he or she
will not (i) exercise any of his or her remaining Copyright and Related
Rights in the Work or (ii) assert any associated claims and causes of
action with respect to the Work, in either case contrary to Affirmer's
express Statement of Purpose.

4. Limitations and Disclaimers.

 a. No trademark or patent rights held by Affirmer are waived, abandoned,
    surrendered, licensed or otherwise affected by this document.
 b. Affirmer offers the Work as-is and makes no representations or
    warranties of any kind concerning the Work, express, implied,
    statutory or otherwise, including without limitation warranties of
    title, merchantability, fitness for a particular purpose, non
    infringement, or the absence of latent or other defects, accuracy, or
    the present or absence of errors, whether or not discoverable, all to
    the greatest extent permissible under applicable law.
 c. Affirmer disclaims responsibility for clearing rights of other persons
    that may apply to the Work or any use thereof, including without
    limitation any person's Copyright and Related Rights in the Work.
    Further, Affirmer disclaims responsibility for obtaining any necessary
    consents, permissions or other rights required for any use of the
    Work.
 d. Affirmer understands and acknowledges that Creative Commons is not a
    party to this document and has no duty or obligation with respect to
    this CC0 or use of the Work.


--- Original upstream notice: readme.md ---
# TKG Japanese-English Learner's Dictionary

**Live site: https://www.tkgje.jp/**

An explanatory Japanese-English dictionary for intermediate learners of Japanese: people who can
read hiragana and katakana, know some kanji, and want to understand words well enough to use
them, not just to look up a translation. This repository holds all of the dictionary's data and
all of the code that builds the site.

The project is supervised by [Tom Gally](https://www.gally.net/about.html). Almost all of the
entry writing, checking and coding has been done by Claude (Anthropic) working in Claude Code,
with independent checks by models from other companies. The story of the project, in Tom's
words, is on the site's [About page](https://www.tkgje.jp/about.html).

Everything here is dedicated to the public domain under [CC0 1.0 Universal](LICENSE). You may copy,
adapt and redistribute the data and code for any purpose, including commercial use, without
asking permission. Credit is welcome but not required.

## What the dictionary contains

As of late September 2026:

| | |
|---|---|
| Entries | about 30,900 |
| Example sentences | about 120,000, each with an English translation |
| Inline links from words in examples and notes to their entries | about 1,000,000 |
| Cross-references between entries (synonyms, antonyms, verb pairs, keigo, …) | about 78,000 |
| Kanji in the kanji index | about 2,800 |
| Verbs and i-adjectives with full conjugation tables | about 7,800 |
| Articles on topics that single entries cannot cover (keigo, counters, giving and receiving, …) | 10 |

Existing entries are reviewed, corrected and connected to one another, and new entries are
added for words the dictionary already uses without defining them, for words found missing by
brainstorming from its headwords and subject areas, and for words that learners using
[ezFurigana](https://www.ezfurigana.com/) looked up and did not find (see
[Who uses the data](#who-uses-the-data)).

### Vocabulary tiers

Instead of JLPT levels, every entry belongs to one of three tiers:

- **Basic** (801 entries): the words needed for simple everyday communication.
- **Core** (1,982 entries): the words an adult needs for ordinary communication.
- **General** (everything else): words a learner will meet in reading, conversation and media.

The basic and core tiers are closed. Example sentences in basic and core entries are written
with simpler vocabulary.

### What an entry offers

- Glosses and explanatory definitions, sense by sense
- Example sentences that progress from short and simple to longer and more natural
- Usage notes on grammar, register, collocations, commonly confused words and culture
- Furigana on every kanji, which readers can switch on or off
- For verbs: transitivity, the intransitive or transitive partner, how the ている form behaves, and a
  conjugation table
- Tags for part of speech, formality, politeness and meaning
- Links from every word in the examples and notes to that word's entry
- Recorded readings of the example sentences (being added; see [Example audio](#example-audio))

## Using the site

The site is fully static: HTML, CSS and JavaScript, with no server-side code. It counts page
views with GoatCounter, a privacy-friendly counter that sets no cookies.

- **Search** (the home page): look up a word by Japanese, romaji or English.
- **Browse**: entries by the first kana of their reading.
- **Kanji**: every kanji used in headwords, each with a page listing the entries that contain it.
- **Lists**: the basic and core tiers, and entries grouped by subject tag.
- **Articles**: longer explanations of topics such as keigo, counters and onomatopoeia.
- **Recent**: the most recently changed entries, marked **NEW** (a new entry), **REVISED** (its
  text was changed) or **REVISED (audio)** (recordings of its examples were added).
- **Random**: a cloud of randomly chosen headwords for browsing.

Three buttons in the header control what entry pages show: **Examples** (show or hide the example
sentences), **Furigana** (readings above kanji) and **Links** (underline the linked words in
examples and notes). The site remembers these settings in the browser.

Each example sentence has a play button. When the example has a recorded reading, the button
plays it; otherwise it uses the browser's own text-to-speech voice.

## Using the data

Each entry is a JSON file in `entries/`, named `<id>_<romaji>.json` and grouped by ID in
directories of 500 (`entries/00000/`, `entries/00500/`, …). An entry's ID is permanent: it is
also its page's URL (`https://www.tkgje.jp/entries/00000/00426_yomu.html`), so IDs are never
changed or reused. `build/schema.json` defines the format.

An abridged entry:

```json
{
  "id": "00426_yomu",
  "headword": "{読|よ}む",
  "reading": "よむ",
  "part_of_speech": "verb (godan)",
  "gloss": "to read",
  "definitions": [
    {"sense_number": 1, "gloss": "to read",
     "explanation": "To look at and comprehend written text. Used for books, newspapers, signs, emails, and any written material."}
  ],
  "examples": [
    {"id": "00426_yomu_ex1",
     "japanese": "⟦{本|ほん}→本：00111_hon⟧⟦を→を：00422_wo⟧{読|よ}む。",
     "english": "I read a book.",
     "notes": null,
     "has_audio": true,
     "sense_numbers": [1]}
  ],
  "notes": "TRANSITIVITY: ⟦{他動詞|たどうし}→他動詞：10418_tadoushi⟧ (transitive). …",
  "cross_references": [
    {"type": "contrast", "target_id": "26844_hatsuonsuru", "reading": "はつおんする",
     "headword": "{発音|はつおん}する", "label": "to pronounce"}
  ],
  "conjugation": {"…": "…"},
  "metadata": {
    "created": "2026-01-05T12:03:41Z",
    "modified": "2026-09-02T04:00:57Z",
    "vocabulary_tier": "basic",
    "tags": {"pos": ["verb-godan"], "transitivity": "transitive", "formality": "neutral",
             "politeness": "plain", "semantic": ["communication"], "verb_class": "godan-mu"}
  }
}
```

Two kinds of markup appear inside Japanese text:

- **Furigana**: `{漢字|かんじ}` puts the reading かんじ over 漢字.
- **Inline links**: `⟦surface→dictionary form：entry id⟧` marks a word and the entry it links to.
  In `⟦{読|よ}んで→読む：00426_yomu⟧`, the text says 読んで, and the link goes to the entry for 読む.

To get plain text, replace each link with its surface part (the text between `⟦` and `→`), then
each `{漢字|かんじ}` with 漢字 (or with かんじ for a kana reading).

Other data files:

- `entries_index.json`: every entry's ID, headword, reading, gloss and tier.
- `kanji/`: the kanji index (`kanji_list.json` and one file per kanji).
- `articles/`: the articles, as JSON with a markdown body using the same markup.
- `audio/manifest/`: the recorded readings (see below).

### Who uses the data

[ezFurigana](https://www.ezfurigana.com/) adds furigana to Japanese texts that its users paste
in and links the words in those texts to this dictionary's entries. Its owner sends Tom reports
of the words his users looked up that the dictionary did not have, and those words are
considered for new entries (`external/ezfurigana/`).

If you use the data in a project of your own, Tom would be glad to hear about it.

## Example audio

Example sentences are being given recorded readings, starting with the basic and core tiers.
Each recording is made by a text-to-speech model (Gemini TTS) and is accepted only after four
automatic checks, using models from different companies, agree that every word was pronounced
as the example's furigana indicate. The MP3 files are kept in separate repositories, each served from its own GitHub Pages
site and kept under 900 MB:
[tkgally/je-dict-audio-1](https://github.com/tkgally/je-dict-audio-1) (full),
[tkgally/je-dict-audio-2](https://github.com/tkgally/je-dict-audio-2) (in use), and
[tkgally/je-dict-audio-3](https://github.com/tkgally/je-dict-audio-3) and
[tkgally/je-dict-audio-4](https://github.com/tkgally/je-dict-audio-4) (waiting). The list and
each store's status are in `audio/config.json`.

`audio/manifest/<range>.jsonl` lists every recording: the example ID, a hash of the text that was
recorded, the file's path and the voice used. A recording counts only while the example's text
and furigana are unchanged; when an example is edited, the site shows no play button for it
until the example is recorded again. An example's `has_audio` field says whether it currently
has a valid recording. The whole workflow, the evidence for it and its changelog are in
[AUDIO_WORKFLOW.md](AUDIO_WORKFLOW.md).

## How the dictionary is made and maintained

The dictionary is written and maintained by Claude in scheduled sessions (Claude Code
"Routines"). Each session follows `prompts/routine2.md`: a selector (`pipeline/routine_next.py`)
chooses one kind of work, such as polishing a batch of entries, a cross-model accuracy review,
a fix for a problem found across many entries, new entries, the vetting of candidate words for
new entries, or example audio. The session then:

1. makes its changes following the guidelines in `.claude/skills/`;
2. runs the deterministic passes (`make mechanical`): note formatting
   (`build/normalize_notes.py`), inline links (`build/auto_link.py`, which uses the SudachiPy
   morphological analyzer), and cross-references named in notes (`build/harvest_crossrefs.py`);
3. has a model from another company review the changed entries (`build/review_accuracy.py` and
   `build/review_links.py`, through OpenRouter) and decides on each point it raises, logging the
   decision in `reviews/decisions.jsonl`;
4. runs the same checks as continuous integration (`make gate`), updates the indexes
   (`make index`) and opens a pull request, which it merges once the checks pass.

When a pull request is merged, GitHub Actions builds the site from the JSON
(`.github/workflows/pages.yml`) and deploys it to GitHub Pages. The generated site is not stored
in the repository.

Tom reads the pull requests and the questions sessions leave for him in
`reviews/needs_curator.txt`.

## Repository layout

```
entries/          The dictionary entries (JSON), 500 per directory
articles/         The articles (JSON)
audio/            Audio manifest, workflow configuration and test data (no audio files)
kanji/            Kanji index data
build/            Site build, validation, deterministic passes, review tools, unit tests
  schema.json       Entry format
  data/             Controlled vocabularies and baselines used by the checks
  templates/        Site CSS and JavaScript
  tests/            Unit tests
  COMMANDS.md       Every command, grouped by task
prompts/          Instructions for the scheduled sessions (routine2.md and its mode prompts)
pipeline/         Session selector, metrics and spending ledgers
.claude/skills/   Editorial guidelines by entry type
planning/wiki/    Research notes on Japanese lexicography and the project backlog
polishing/        Progress of the polishing passes, session logs, observations
reviews/          Review decisions, accuracy flags, questions for the editor
candidate_words.json  Vetted words waiting to become entries
external/         Reports received from users of the data (ezfurigana/: lookup reports and verdicts)
.github/workflows CI checks on pull requests; site build and deployment
archive/          Prompts, plans and logs from earlier stages of the project, kept for the record
```

`CLAUDE.md` is the working guide for Claude sessions; `PROJECT_STATUS.md` records recent history.
`planning/LLM-wiki.md` is Andrej Karpathy's "LLM wiki" pattern, on which the knowledge base in
`planning/wiki/` is modelled.

## Building the site locally

Python 3.10 or later is needed.

```bash
pip install -r build/requirements.txt
make build                 # validate, update indexes, generate the site in docs/
open docs/index.html       # or open the file in any browser; no web server is needed
make gate                  # the checks CI runs on a pull request (unit tests, validation)
```

The audio tools need extra packages (`make audio-deps`) and an OpenRouter API key; the site
build does not.

## Contact

Suggestions and corrections are welcome: write to [Tom Gally](https://www.gally.net/about.html).


Modifications by Danzu, 2026: extraction of playable word forms, Unicode normalization, filtering to the game alphabet and 3–7 tiles, deduplication, and selection of daily/campaign words. No dictionary definitions, corpus sentences, or audio are included.

Editable word data, source attribution, original notices and license texts: https://kickbid.lol/word-drop/word-data/ja.zip
These word-data files retain their stated open licenses. App terms do not restrict rights granted by those licenses. You may use, modify and redistribute the data under the applicable license; no payment or account is required for source downloads. No endorsement by upstream authors is implied.

--- Applicable license text ---
Creative Commons Legal Code

CC0 1.0 Universal

    CREATIVE COMMONS CORPORATION IS NOT A LAW FIRM AND DOES NOT PROVIDE
    LEGAL SERVICES. DISTRIBUTION OF THIS DOCUMENT DOES NOT CREATE AN
    ATTORNEY-CLIENT RELATIONSHIP. CREATIVE COMMONS PROVIDES THIS
    INFORMATION ON AN "AS-IS" BASIS. CREATIVE COMMONS MAKES NO WARRANTIES
    REGARDING THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS
    PROVIDED HEREUNDER, AND DISCLAIMS LIABILITY FOR DAMAGES RESULTING FROM
    THE USE OF THIS DOCUMENT OR THE INFORMATION OR WORKS PROVIDED
    HEREUNDER.

Statement of Purpose

The laws of most jurisdictions throughout the world automatically confer
exclusive Copyright and Related Rights (defined below) upon the creator
and subsequent owner(s) (each and all, an "owner") of an original work of
authorship and/or a database (each, a "Work").

Certain owners wish to permanently relinquish those rights to a Work for
the purpose of contributing to a commons of creative, cultural and
scientific works ("Commons") that the public can reliably and without fear
of later claims of infringement build upon, modify, incorporate in other
works, reuse and redistribute as freely as possible in any form whatsoever
and for any purposes, including without limitation commercial purposes.
These owners may contribute to the Commons to promote the ideal of a free
culture and the further production of creative, cultural and scientific
works, or to gain reputation or greater distribution for their Work in
part through the use and efforts of others.

For these and/or other purposes and motivations, and without any
expectation of additional consideration or compensation, the person
associating CC0 with a Work (the "Affirmer"), to the extent that he or she
is an owner of Copyright and Related Rights in the Work, voluntarily
elects to apply CC0 to the Work and publicly distribute the Work under its
terms, with knowledge of his or her Copyright and Related Rights in the
Work and the meaning and intended legal effect of CC0 on those rights.

1. Copyright and Related Rights. A Work made available under CC0 may be
protected by copyright and related or neighboring rights ("Copyright and
Related Rights"). Copyright and Related Rights include, but are not
limited to, the following:

  i. the right to reproduce, adapt, distribute, perform, display,
     communicate, and translate a Work;
 ii. moral rights retained by the original author(s) and/or performer(s);
iii. publicity and privacy rights pertaining to a person's image or
     likeness depicted in a Work;
 iv. rights protecting against unfair competition in regards to a Work,
     subject to the limitations in paragraph 4(a), below;
  v. rights protecting the extraction, dissemination, use and reuse of data
     in a Work;
 vi. database rights (such as those arising under Directive 96/9/EC of the
     European Parliament and of the Council of 11 March 1996 on the legal
     protection of databases, and under any national implementation
     thereof, including any amended or successor version of such
     directive); and
vii. other similar, equivalent or corresponding rights throughout the
     world based on applicable law or treaty, and any national
     implementations thereof.

2. Waiver. To the greatest extent permitted by, but not in contravention
of, applicable law, Affirmer hereby overtly, fully, permanently,
irrevocably and unconditionally waives, abandons, and surrenders all of
Affirmer's Copyright and Related Rights and associated claims and causes
of action, whether now known or unknown (including existing as well as
future claims and causes of action), in the Work (i) in all territories
worldwide, (ii) for the maximum duration provided by applicable law or
treaty (including future time extensions), (iii) in any current or future
medium and for any number of copies, and (iv) for any purpose whatsoever,
including without limitation commercial, advertising or promotional
purposes (the "Waiver"). Affirmer makes the Waiver for the benefit of each
member of the public at large and to the detriment of Affirmer's heirs and
successors, fully intending that such Waiver shall not be subject to
revocation, rescission, cancellation, termination, or any other legal or
equitable action to disrupt the quiet enjoyment of the Work by the public
as contemplated by Affirmer's express Statement of Purpose.

3. Public License Fallback. Should any part of the Waiver for any reason
be judged legally invalid or ineffective under applicable law, then the
Waiver shall be preserved to the maximum extent permitted taking into
account Affirmer's express Statement of Purpose. In addition, to the
extent the Waiver is so judged Affirmer hereby grants to each affected
person a royalty-free, non transferable, non sublicensable, non exclusive,
irrevocable and unconditional license to exercise Affirmer's Copyright and
Related Rights in the Work (i) in all territories worldwide, (ii) for the
maximum duration provided by applicable law or treaty (including future
time extensions), (iii) in any current or future medium and for any number
of copies, and (iv) for any purpose whatsoever, including without
limitation commercial, advertising or promotional purposes (the
"License"). The License shall be deemed effective as of the date CC0 was
applied by Affirmer to the Work. Should any part of the License for any
reason be judged legally invalid or ineffective under applicable law, such
partial invalidity or ineffectiveness shall not invalidate the remainder
of the License, and in such case Affirmer hereby affirms that he or she
will not (i) exercise any of his or her remaining Copyright and Related
Rights in the Work or (ii) assert any associated claims and causes of
action with respect to the Work, in either case contrary to Affirmer's
express Statement of Purpose.

4. Limitations and Disclaimers.

 a. No trademark or patent rights held by Affirmer are waived, abandoned,
    surrendered, licensed or otherwise affected by this document.
 b. Affirmer offers the Work as-is and makes no representations or
    warranties of any kind concerning the Work, express, implied,
    statutory or otherwise, including without limitation warranties of
    title, merchantability, fitness for a particular purpose, non
    infringement, or the absence of latent or other defects, accuracy, or
    the present or absence of errors, whether or not discoverable, all to
    the greatest extent permissible under applicable law.
 c. Affirmer disclaims responsibility for clearing rights of other persons
    that may apply to the Work or any use thereof, including without
    limitation any person's Copyright and Related Rights in the Work.
    Further, Affirmer disclaims responsibility for obtaining any necessary
    consents, permissions or other rights required for any use of the
    Work.
 d. Affirmer understands and acknowledges that Creative Commons is not a
    party to this document and has no duty or obligation with respect to
    this CC0 or use of the Work.
