• Home
  • Line#
  • Scopes#
  • Navigate#
  • Raw
  • Download
1<?xml version="1.0"?>
2<!DOCTYPE kanjidic2 [
3	<!-- Version 1.3
4	This is the DTD of the XML-format kanji file combining information from
5	the KANJIDIC and KANJD212 files. It is intended to be largely self-
6	documenting, with each field being accompanied by an explanatory
7	comment.
8
9	The file covers the following kanji:
10	(a) the 6,355 kanji from JIS X 0208;
11	(b) the 5,801 kanji from JIS X 0212;
12	(c) the 3,625 kanji from JIS X 0213 as follows:
13		(i) the 2,741 kanji which are also in JIS X 0212 have
14		JIS X 0213 code-points (kuten) added to the existing entry;
15		(ii) the 884 "new" kanji have new entries.
16
17	At the end of the explanation for a number of fields there is a tag
18	with the format [N]. This indicates the leading letter(s) of the
19	equivalent field in the KANJIDIC and KANJD212 files.
20
21	The KANJIDIC documentation should also be read for additional
22	information about the information in the file.
23	-->
24<!ELEMENT kanjidic2 (header,character*)>
25<!ELEMENT header (file_version,database_version,date_of_creation)>
26<!--
27	The single header element will contain identification information
28	about the version of the file
29	-->
30<!ELEMENT file_version (#PCDATA)>
31<!--
32	This field denotes the version of kanjidic2 structure, as more
33	than one version may exist.
34	-->
35<!ELEMENT database_version (#PCDATA)>
36<!--
37	The version of the file, in the format YYYY-NN, where NN will be
38	a number starting with 01 for the first version released in a
39	calendar year, then increasing for each version in that year.
40	-->
41<!ELEMENT date_of_creation (#PCDATA)>
42<!--
43	The date the file was created in international format (YYYY-MM-DD).
44	-->
45<!ELEMENT character (literal,codepoint, radical, misc, dic_number?, query_code?, reading_meaning?,nanori?)*>
46<!ELEMENT literal (#PCDATA)>
47<!--
48	The character itself in UTF8 coding.
49	-->
50<!ELEMENT codepoint (cp_value+)>
51	<!--
52	The codepoint element states the code of the character in the various
53	character set standards.
54	-->
55<!ELEMENT cp_value (#PCDATA)>
56	<!--
57	The cp_value contains the codepoint of the character in a particular
58	standard. The standard will be identified in the cp_type attribute.
59	-->
60<!ATTLIST cp_value cp_type CDATA #REQUIRED>
61	<!--
62	The cp_type attribute states the coding standard applying to the
63	element. The values assigned so far are:
64		jis208 - JIS X 0208-1997 - kuten coding (nn-nn)
65		jis212 - JIS X 0212-1990 - kuten coding (nn-nn)
66		jis213 - JIS X 0213-2000 - kuten coding (p-nn-nn)
67		ucs - Unicode 4.0 - hex coding (4 or 5 hexadecimal digits)
68	-->
69<!ELEMENT radical (rad_value+)>
70<!ELEMENT rad_value (#PCDATA)>
71	<!--
72	The radical number, in the range 1 to 214. The particular
73	classification type is stated in the rad_type attribute.
74	-->
75<!ATTLIST rad_value rad_type CDATA #REQUIRED>
76	<!--
77	The rad_type attribute states the type of radical classification.
78		classical - as recorded in the KangXi Zidian.
79		nelson - as used in the Nelson "Modern Japanese-English
80		Character Dictionary" (i.e. the Classic, not the New Nelson).
81		This will only be used where Nelson reclassified the kanji.
82	-->
83<!ELEMENT misc (grade?, stroke_count+, variant*, freq*, rad_name*)>
84<!ELEMENT grade (#PCDATA)>
85	<!--
86	The Jouyou Kanji grade level. 1 through 6 indicate the grade in which
87	the kanji is taught in Japanese schools. 8 indicates it is one of the
88	remaining Jouyou Kanji to be learned in junior high school, and 9
89	indicates it is a Jinmeiyou (for use in names) kanji. [G]
90	-->
91<!ELEMENT stroke_count (#PCDATA)>
92	<!--
93	The stroke count of the kanji, including the radical. If more than
94	one, the first is considered the accepted count, while subsequent ones
95	are common miscounts. (See Appendix E. of the KANJIDIC documentation
96	for some of the rules applied when counting strokes in some of the
97	radicals.) [S]
98	-->
99<!ELEMENT variant (#PCDATA)>
100	<!--
101	A cross-reference code to another kanji, usually regarded as a variant.
102	The type of cross-reference is given in the var_type attribute.
103	-->
104<!ATTLIST variant var_type CDATA #REQUIRED>
105	<!--
106	The var_type attribute indicates the type of variant code. The current
107	values are:
108		jis208 - in JIS X 0208 - kuten coding
109		jis212 - in JIS X 0212 - kuten coding
110		jis213 - in JIS X 0213 - kuten coding
111		deroo - De Roo number - numeric
112		njecd - Halpern NJECD index number - numeric
113		s_h - The Kanji Dictionary (Spahn & Hadamitzky) - descriptor
114		nelson - "Classic" Nelson - numeric
115		oneill - Japanese Names (O'Neill) - numeric
116	-->
117<!ELEMENT freq (#PCDATA)>
118	<!--
119	A frequency-of-use ranking. The 2,500 most-used characters have a
120	ranking; those characters that lack this field are not ranked. The
121	frequency is a number from 1 to 2,500 that expresses the relative
122	frequency of occurrence of a character in modern Japanese. This is
123	based on a survey in newspapers, so it is biassed towards kanji
124	used in newspaper articles. The discrimination between the less
125	frequently used kanji is not strong.
126	-->
127<!ELEMENT rad_name (#PCDATA)>
128	<!--
129	When the kanji is itself a radical and has a name, this element
130	contains the name (in hiragana.) [T2]
131	-->
132<!ELEMENT dic_number (dic_ref+)>
133	<!--
134	This element contains the index numbers and similar unstructured
135	information such as page numbers in a number of published dictionaries,
136	and instructional books on kanji.
137	-->
138<!ELEMENT dic_ref (#PCDATA)>
139	<!--
140	Each dic_ref contains an index number. The particular dictionary,
141	etc. is defined by the dr_type attribute.
142	-->
143<!ATTLIST dic_ref dr_type CDATA #REQUIRED>
144	<!--
145	The dr_type defines the dictionary or reference book, etc. to which
146	dic_ref element applies. The initial allocation is:
147	  nelson_c - "Modern Reader's Japanese-English Character Dictionary",
148	  	edited by Andrew Nelson (now published as the "Classic"
149	  	Nelson).
150	  nelson_n - "The New Nelson Japanese-English Character Dictionary",
151	  	edited by John Haig.
152	  halpern_njecd - "New Japanese-English Character Dictionary",
153	  	edited by Jack Halpern.
154	  halpern_kkld - "Kanji Learners Dictionary" (Kodansha) edited by
155	  	Jack Halpern.
156	  heisig - "Remembering The  Kanji"  by  James Heisig.
157	  gakken - "A  New Dictionary of Kanji Usage" (Gakken)
158	  oneill_names - "Japanese Names", by P.G. O'Neill.
159	  oneill_kk - "Essential Kanji" by P.G. O'Neill.
160	  moro - "Daikanwajiten" compiled by Morohashi. For some kanji two
161	  	additional attributes are used: m_vol:  the volume of the
162	  	dictionary in which the kanji is found, and m_page: the page
163	  	number in the volume.
164	  henshall - "A Guide To Remembering Japanese Characters" by
165	  	Kenneth G.  Henshall.
166	  sh_kk - "Kanji and Kana" by Spahn and Hadamitzky.
167	  sakade - "A Guide To Reading and Writing Japanese" edited by
168	  	Florence Sakade.
169	  henshall3 - "A Guide To Reading and Writing Japanese" 3rd
170		edition, edited by Henshall, Seeley and De Groot.
171	  tutt_cards - Tuttle Kanji Cards, compiled by Alexander Kask.
172	  crowley - "The Kanji Way to Japanese Language Power" by
173	  	Dale Crowley.
174	  kanji_in_context - "Kanji in Context" by Nishiguchi and Kono.
175	  busy_people - "Japanese For Busy People" vols I-III, published
176		by the AJLT. The codes are the volume.chapter.
177	  kodansha_compact - the "Kodansha Compact Kanji Guide".
178	-->
179<!ATTLIST dic_ref m_vol CDATA #IMPLIED>
180	<!--
181	See above under "moro".
182	-->
183<!ATTLIST dic_ref m_page CDATA #IMPLIED>
184	<!--
185	See above under "moro".
186	-->
187<!ELEMENT query_code (q_code+)>
188	<!--
189	These codes contain information relating to the glyph, and can be used
190	for finding a required kanji. The type of code is defined by the
191	qc_type attribute.
192	-->
193<!ELEMENT q_code (#PCDATA)>
194	<!--
195	The q_code contains the actual query-code value, according to the
196	qc_type attribute.
197	-->
198<!ATTLIST q_code qc_type CDATA #REQUIRED>
199	<!--
200	The q_code attribute defines the type of query code. The current values
201	are:
202	  skip -  Halpern's SKIP (System  of  Kanji  Indexing  by  Patterns)
203	  	code. The  format is n-nn-nn.  See the KANJIDIC  documentation
204	  	for  a description of the code and restrictions on  the
205	  	commercial  use  of this data. [P]
206
207	  sh_desc - the descriptor codes for The Kanji Dictionary (Tuttle
208	  	1996) by Spahn and Hadamitzky. They are in the form nxnn.n,
209	  	e.g.  3k11.2, where the  kanji has 3 strokes in the
210	  	identifying radical, it is radical "k" in the SH
211	  	classification system, there are 11 other strokes, and it is
212	  	the 2nd kanji in the 3k11 sequence. (I am very grateful to
213	  	Mark Spahn for providing the list of these descriptor codes
214	  	for the kanji in this file.) [I]
215	  four_corner - the "Four Corner" code for the kanji. This is a code
216	  	invented by Wang Chen in 1928. See the KANJIDIC documentation
217	  	for  an overview of  the Four Corner System. [Q]
218
219	  deroo - the codes developed by the late Father Joseph De Roo, and
220	  	published in  his book "2001 Kanji" (Bojinsha). Fr De Roo
221	  	gave his permission for these codes to be included. [DR]
222	  misclass - a possible misclassification of the kanji according
223		to one of the code types. (See the "Z" codes in the KANJIDIC
224		documentation for more details.)
225
226	-->
227<!ELEMENT reading_meaning (rmgroup*, nanori*)>
228	<!--
229	The readings for the kanji in several languages, and the meanings, also
230	in several languages. The readings and meanings are grouped to enable
231	the handling of the situation where the meaning is differentiated by
232	reading. [T1]
233	-->
234<!ELEMENT nanori (#PCDATA)>
235	<!--
236	Japanese readings that are now only associated with names.
237	-->
238<!ELEMENT rmgroup (reading*, meaning*)>
239<!ELEMENT reading (#PCDATA)>
240	<!--
241	The reading element contains the reading or pronunciation
242	of the kanji.
243	-->
244<!ATTLIST reading r_type CDATA #REQUIRED>
245	<!--
246	The r_type attribute defines the type of reading in the reading
247	element. The current values are:
248	  pinyin - the modern PinYin romanization of the Chinese reading
249	  	of the kanji. The tones are represented by a concluding
250	  	digit. [Y]
251	  korean_r - the romanized form of the Korean reading(s) of the
252	  	kanji.  The readings are in the (Republic of Korea) Ministry
253	  	of Education style of romanization. [W]
254	  korean_h - the Korean reading(s) of the kanji in hangul.
255	  ja_on - the "on" Japanese reading of the kanji, in katakana. A
256	  	second attribute r_status, if present, will indicate with
257	  	a value of "jy" whether the reading is approved for a
258	  	"Jouyou kanji".
259	  ja_kun - the "kun" Japanese reading of the kanji, in hiragana.
260	  	Where relevant the okurigana is also included separated by a
261	  	".". Readings associated with prefixes and suffixes are
262	  	marked with a "-". A second attribute r_status, if present,
263	  	will indicate with a value of "jy" whether the reading is
264	  	approved for a "Jouyou kanji".
265	-->
266<!ATTLIST reading r_status CDATA #IMPLIED>
267	<!--
268	See under ja_on and ja_kun above.
269	-->
270<!ELEMENT meaning (#PCDATA)>
271	<!--
272	The meaning associated with the kanji.
273	-->
274<!ATTLIST meaning m_lang CDATA #IMPLIED>
275	<!--
276	The m_lang attribute defines the target language of the meaning. It
277	will be coded using the two-letter language code from the ISO 639
278	standard. When absent, the value "en" (i.e. English) is implied. [{}]
279	-->
280] >
281<kanjidic2>
282</kanjidic2>
283