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.

map/libhangul2hanja

@@ -1,21 +1,18 @@
 #!/usr/bin/env python3
-"""Extract single-character Hanja readings from libhangul data."""
+"""Extract Hanja readings from libhangul data."""
 
 import sys
 from pathlib import Path
 
 
 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 extract(src, name):
@@ -44,7 +41,7 @@ def extract(src, name):
         seen.add(pair)
         entries.append(pair)
     if not entries:
-        raise ValueError(f"{name}: no single-character Hanja readings")
+        raise ValueError(f"{name}: no Hanja readings")
     return comments, entries