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.

tests/mkhanja_test.py

@@ -42,25 +42,27 @@ class MkhanjaTest(unittest.TestCase):
         self.addCleanup(tmp.cleanup)
         return path
 
-    def test_imports_single_bmp_hanja_rows(self):
+    def test_imports_bmp_hanja_rows(self):
         result = run(
             IMPORT,
             self.source(
                 "# Copyright holder\n"
                 "# BSD license\n"
-                "한:漢:first\n"
-                "가:㐀:extension A\n"
-                "김:金:compatibility\n"
-                "방학:放:word reading\n"
-                "학:學校:word value\n"
-                "한:𠀀:astral\n"
-                "ㄱ:加:jamo\n"
+                "\ud55c:\u6f22:first\n"
+                "\uac00:\u3400:extension A\n"
+                "\uae40:\u91d1:compatibility\n"
+                "\ud55c\uc790:\u6f22\u5b57:word\n"
+                "\ud55c:\U00020000:astral\n"
+                "\u3131:\u52a0:jamo\n"
+                "\ud55c\uae00:\u97d3glyph:mixed\n"
             ),
         )
         self.assertEqual(result.returncode, 0, result.stderr)
         lines = result.stdout.splitlines()
         self.assertEqual(lines[:2], ["# Copyright holder", "# BSD license"])
-        self.assertEqual(lines[2:], ["", "漢\t한", "㐀\t가", "金\t김"])
+        self.assertEqual(lines[2:], ["", "\u6f22\t\ud55c", "\u3400\t\uac00",
+                                     "\u91d1\t\uae40",
+                                     "\u6f22\u5b57\t\ud55c\uc790"])
 
     def test_import_rejects_malformed_and_duplicate_rows(self):
         bad = [
@@ -91,6 +93,7 @@ class MkhanjaTest(unittest.TestCase):
                 "㐀\t가\n"
                 "家\t가\n"
                 "金\t금\n"
+                "漢字\t한자\n"
             ),
         )
         self.assertEqual(result.returncode, 0, result.stderr)
@@ -103,13 +106,14 @@ class MkhanjaTest(unittest.TestCase):
                 "한\t漢 韓",
                 "가\t㐀 家",
                 "금\t金",
+                "한자\t漢字",
             ],
         )
 
     def test_generator_rejects_bad_rows(self):
         bad = [
-            "漢韓\t한\n",
-            "漢\t한자\n",
+            "漢a\t한\n",
+            "漢\t한a\n",
             "𠀀\t한\n",
             "漢\t한\n漢\t한\n",
             "漢 한\n",