Skip to content

Latest commit

 

History

History
91 lines (73 loc) · 2.52 KB

kd2_db_schema.rst

File metadata and controls

91 lines (73 loc) · 2.52 KB
Raw

KANJIDIC2 Database Schema

Original XML Schema

kanjidic2
- header
  - file_version
  - database_version
  - date_of_creation
* character   # what is meaning of * after list of child elements?  (a, b, c, ...)*
  - literal
  - codepoint
    + cp_value
  - radical
    + rad_value
  - misc
    ? grade
    + stroke_count
    + variant
    ? freq
    * rad_name
    ? jlpt
  ? dic_number
    + dic_ref
  ? query_code
    + q_code
  ? reading_meaning
    * rmgroup
      * reading
      * meaning
    * nanori

SQLite3 Schema

CREATE TABLE header (file_version TEXT, database_version TEXT, date_of_creation TEXT);

character: -> CharacterTable
id INTEGER, literal TEXT, grade INTEGER, freq INTEGER, jlpt INTEGER
codepoint, radical, variant: -> TypeValueTable
id INTEGER, fk INTEGER, type TEXT, value TEXT
stroke_count: -> StrokeCountTable
id INTEGER, fk INTEGER, count INTEGER
rad_name, nanori: -> KeyValueTable
id INTEGER, fk INTEGER, value TEXT
dic_number: (First revision)
id INTEGER, fk INTEGER, type TEXT, m_vol TEXT, m_page TEXT - Might remove moro pieces and put in separate table to save space... - Alternatively could merge m_vol/m_page into the moro code directly
query_code:

id INTEGER, fk INTEGER, type TEXT, skip_misclass TEXT, value TEXT - Might move skip miscodes into a separate table... - Might re-code SKIP values separately based upon integers... but

the same could be argued about all other codes... probably shouldn't waste time on this.
rmgroup:
id INTEGER, fk INTEGER
reading:
id INTEGER, fk INTEGER, type TEXT, on_type TEXT, r_status TEXT, value TEXT
meaning:
id INTEGER, fk INTEGER, lang TEXT, value TEXT

Notes

  • Initial tests show indices to be seemingly unnecessary for KANJIDIC2 reading/meaning/nanori searches. I do not know if this is true on slower systems, but for the time being the default "%xxx%" pattern search is adequate.

  • KANJIDIC2 uses special notation for kunyomi readings. Specifically, - is used to note prefixes/suffixes, and . is used to separate okurigana.

    Likely it would be beneficial to make a special index table for looking up kunyomi readings quickly. The prefix/suffix marker doesn't cause results to be dropped, but hte okurigana marker does, and the user shouldn't need to supply (or know about) such details.