Seiyuu Database Format Specifications

This document will describe the formats of the various seiyuu databases that exist in my seiyuu WWW pages. (http://www.win.ne.jp/~doi/seiyuu/seiyuu.html)

specification version 0.5 by Hitoshi Doi, 1996.01.21
specification version 1.0 by Hitoshi Doi, 1996.03.18
(updated Title Format to version 2)

There are basically three formats: 1) title format, 2) VA format, and 3) DB format. The seiyuu's role information contained in each format is the same, but they are just presented in a different order and layout.

Syntax of the Title Format (version 2)

The title format (named this way as the data is sorted by title) has the following syntax.

TITLE
 URL URL
 CHARACTER :: SEIYUU
 CHARACTER :: SEIYUU {EPISODE}
 CHARACTER :: SEIYUU -- DEBUT
 CHARACTER :: SEIYUU {EPISODE} -- DEBUT

The keyword URL has been added for Title Format version 2. This specifies the URL of the title, and this line is optional.

Syntax of the VA Format

The VA (voice actor) format (named this way as the data is sorted by seiyuu) has the following syntax.

SEIYUU
 /* COMMENTS */
 TITLE
  CHARACTER
  CHARACTER {EPISODE}
  CHARACTER -- DEBUT
 TITLE
  CHARACTER {EPISODE} -- DEBUT

Syntax of the DB Format

The DB (database) format (named this way as the data is comprised of complete lines of character, title, and seiyuu name) has the following syntax.

CHARACTER - TITLE - SEIYUU
CHARACTER - TITLE {EPISODE} - SEIYUU
[DEBUT] CHARACTER - TITLE - SEIYUU
[DEBUT] CHARACTER - TITLE {EPISODE} - SEIYUU

The data in format is usually generated by programs from the other formats, and these files will not be edited by humans. This format allows for easy grepping to find relevant information.

Definition of Tokens

Some details and conventions about the tokens are described below.

The {EPISODE} and -- DEBUT are optional.
If one seiyuu did more than one role, those character names should be put on different lines.
If one character goes by different names (for example Tsukino Usagi and Sailor Moon), those names should be put on the same line.
There are some standard fields in the COMMENTS section that the valookup program recognizes.

Database Processing Software

The following perl scripts exist to process the various seiyuu database formats.

valookup
dbconv

The valookup program reads in the VA format database, and produces various types of formatted output. Some of the output that valookup produces are:

partial HTML format output for the individual seiyuu WWW pages.
a seiyuu birthday listing.
role listing by seiyuu.
role listing by title.
role listing by character name.
NOTE: Some of these features are not needed any more because of the existance of the DB format and Title format databases.

The dbconv program reads in the Title format database, and produces various types of formatted output. Some of the output that dbconv produces are:

DB format database.
VA format database.
partial HTML format output for the individual seiyuu WWW pages.