okpyeon 0.1.0

Korean IME candidate data: hangul reading -> hanja candidates with hun-eum glosses (옥편), plus the MS-IME-style consonant -> special-symbol tables (ㅁ + 한자). Data from libhangul, single-character lookups only.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75

# okpyeon (옥편)

[![crates.io](https://img.shields.io/crates/v/okpyeon.svg)](https://crates.io/crates/okpyeon)
[![docs.rs](https://img.shields.io/docsrs/okpyeon)](https://docs.rs/okpyeon)

한국어 입력기의 한자 키가 필요로 하는 두 가지 정적 후보 데이터: 독음 → 한자
후보(훈음 포함), 그리고 MS IME 식 자음 → 특수문자 표 (`ㅁ` + 한자 → `※ ☆ ★ …`).

Static candidate data for Korean IMEs, covering both things the 한자(hanja)
key does: **hanja lookup** (a hangul syllable reading to hanja candidates with
hun-eum glosses, like a paper 옥편) and the **MS-IME-style special symbol
tables** (compose a lone consonant, press the hanja key, pick a symbol).
Single-`char` lookups only; word-level conversion is deliberately out of
scope. No allocation, `no_std`, zero dependencies.

## Usage

```rust
// 독음 → 한자 후보 (훈음 포함, 상용 한자 우선 순서)
let candidates = okpyeon::hanja('학').unwrap();
assert_eq!(candidates[0], ('學', "배울 학"));

// 자음 → 특수문자 (MS IME 원본 표)
let shapes = okpyeon::symbols('ㅁ').unwrap();
assert!(shapes.contains(&'※'));

// 알려진 결함 2건을 보정한 표 (날개셋 / 최신 Windows 동작)
assert_eq!(okpyeon::symbols_revised('ㄹ').unwrap()[3], '°');
assert_eq!(okpyeon::symbols_revised('ㅁ').unwrap().last(), Some(&'㉾'));
```

The symbol tables define 18 keys, matching Windows MS IME: `ㄱ` punctuation,
`ㄴ` brackets, `ㄷ` math, `ㄹ` units, `ㅁ` general symbols, `ㅂ` box drawing,
`ㅅ`/`ㅇ` enclosed hangul/latin, `ㅈ` fullwidth digits, `ㅊ` fractions,
`ㅋ`/`ㅌ` modern/archaic jamo, `ㅍ` fullwidth latin, `ㅎ` greek, `ㄲ` extended
latin, `ㄸ` hiragana, `ㅃ` katakana, `ㅆ` cyrillic. `ㅉ` has no assignment,
faithful to MS IME.

## Data provenance

Both tables are vendored in `data/` from [libhangul] (commit `a34aef7`) and
compiled into sorted static slices by `build.rs`:

- `data/hanja.txt` — libhangul's `data/hanja/hanja.txt`, filtered to entries
  whose key and value are each a single character (multi-syllable word
  entries plus two anomalous multi-character-value entries removed; run
  `scripts/update-data.sh` to regenerate). Candidate order preserves the file
  order, which libhangul curates common-first, so it is directly usable as an
  IME candidate list. libhangul's separate frequency file (`freq-hanja.txt`,
  unclear licensing) is **not** used.
- `data/mssymbol.txt` — libhangul's `data/hanja/mssymbol.txt`, verbatim.
  [`symbols`] is faithful to it, including two known quirks inherited from
  old MS IME: `ㄹ`'s 4th slot is a fullwidth `Ｆ` (a duplicate of the one in
  the `ㅍ` fullwidth-latin table) where `°` is expected, and `㉾` is missing.
  [`symbols_revised`] corrects both, matching the 날개셋 input method and
  current Windows respectively.

[libhangul]: https://github.com/libhangul/libhangul
[`symbols`]: https://docs.rs/okpyeon/latest/okpyeon/fn.symbols.html
[`symbols_revised`]: https://docs.rs/okpyeon/latest/okpyeon/fn.symbols_revised.html

## License

The crate's own code is dual-licensed under either of

- Apache License, Version 2.0 ([LICENSE-APACHE](LICENSE-APACHE))
- MIT license ([LICENSE-MIT](LICENSE-MIT))

at your option.

The vendored data files (`data/hanja.txt`, `data/mssymbol.txt`) are from
libhangul, copyright (c) Choe Hwanjin, licensed under the BSD 3-Clause
license; each file retains its original copyright and license header. The
compiled crate therefore contains BSD-3-Clause-licensed data
(SPDX: `(MIT OR Apache-2.0) AND BSD-3-Clause`).