data(hanja): words, not only syllables

The import kept only readings of one syllable, so the Hanja search could
convert 한 but never 한자, 학교, or 대한민국 — the conversion every other
Korean input method offers.  libhangul's table has 187k readings; both
scripts now keep them all, and the search finds a word as readily as a
syllable.  The daemon pays for it: 24 MB instead of 12, and 170 ms to
start instead of 30.
This commit is contained in:
2026-08-17 12:50:05 +09:00
parent 3527bcd489
commit 2e53627b7d
7 changed files with 424873 additions and 45 deletions

View File

@@ -1,5 +1,5 @@
#!/usr/bin/env python3
"""Write hanja.dict from a one-Hanja-per-row UTF-8 source."""
"""Write hanja.dict from a Hanja-per-row UTF-8 source."""
import sys
from pathlib import Path
@@ -10,16 +10,13 @@ MAXCANDIDATES = 128
def ishangul(s):
return len(s) == 1 and 0xAC00 <= ord(s) <= 0xD7A3
return s != "" and all(0xAC00 <= ord(c) <= 0xD7A3 for c in s)
def ishanja(s):
if len(s) != 1:
return False
c = ord(s)
return (0x3400 <= c <= 0x4DBF
or 0x4E00 <= c <= 0x9FFF
or 0xF900 <= c <= 0xFAFF)
return s != "" and all(0x3400 <= ord(c) <= 0x4DBF
or 0x4E00 <= ord(c) <= 0x9FFF
or 0xF900 <= ord(c) <= 0xFAFF for c in s)
def read(path):
@@ -42,9 +39,9 @@ def read(path):
raise ValueError(f"{path}:{lineno}: need Hanja<TAB>reading")
hanja, reading = fields
if not ishanja(hanja):
raise ValueError(f"{path}:{lineno}: need one BMP Hanja character")
raise ValueError(f"{path}:{lineno}: need BMP Hanja")
if not ishangul(reading):
raise ValueError(f"{path}:{lineno}: need one Hangul syllable")
raise ValueError(f"{path}:{lineno}: need Hangul syllables")
pair = (hanja, reading)
if pair in seen:
raise ValueError(f"{path}:{lineno}: duplicate Hanja reading")