GlyphWiki logo
navigation
help
search

toolbox
languages
articlediscussionedit this pagehistory

GlyphWiki:NamingGuidelines

From GlyphWiki, the free glyph database

To start

On GlyphWiki, users giving glyphs names freely is tolerated, but some names given do have meanings to them. Please follow these naming guidelines.

Guidelines concerning indicating ISO/IEC 10646 code points

Glyphs named u####, u2#### (where # indicates a hexadecimal digit, 0–9, a–f; limited to CJK kanji code points) correspond to ISO/IEC 10646 code points.

The designs of character forms, as a general rule, are to be based on the “J” column of the standard (i.e., the Japanese style where applicable). Strictly speaking, this amounts to the recorded forms of characters specified in the latest edition of the Japanese national standard (JIS X 0213:2004). In the case of code points not in the Japanese standards, the forms listed in the J-column, as well as the close K- and T-columns, are recommended. Characters not in the Japanese standards are outside of the Jōyō Kanji, so please design them as kyūjitai (traditional form). Also, the forms carried in accordance with Morohashi Tetsuji’s “Dai Kanwa Jiten” would be much preferred as well.

Code points found only in the G-column (e.g. simplified hanzi) preferably should be designed as strictly Chinese; this is a point at which you may want to refrain from the decorative elements found in Japanese designs (e.g. designs where the first stroke of a “ム” pattern intersects the second), but GlyphWiki being Japanese, where the goal is to manage glyphs of Japanese design, please register Japanese-style design data even though these will become so-called false characters (i.e., they never existed in Japan).

Naming guidelines for the purpose of forms that change in the hen area

For example, the kanji “金” (kane), when made the left-hand-side radical kane-hen, becomes the u91d1-01 shape. On GlyphWiki, this is termed a “hen transformation” (偏化変形). As concerns some UCS codepoint kanji, in cases where a hen-transformed part does not have an existing code point, suffix codes have been prepared to indicate a hen transformation part.

Suffix codeMeaning
-01Left part of a left/right combination; left/middle part of a left/middle/right combination
-02Right part of a left/right combination; right part of a left/middle/right combination
-03Top part of a top/bottom combination; top/middle part of a top/middle/bottom combination
-04Bottom part of a top/bottom combination; bottom part of a top/middle/bottom combination
-05Outer part of an enclosure
-06Inner part of an enclosure
-07Not for any specific position‚ but used as a non-individual part
-08Used as a tall part (suitable as both -01 and -02 parts)
-09Used as a squat part (suitable as both -03 and -04 parts)
-10An enclosure’s outer part in which the density of the contents is higher than normal
-11An enclosure’s outer part in which the density of the contents is lower than normal

For example, in the case of u91d1-01, since the original “金” is u91d1, a name of “u91d1-01” would be desirable. However, with this particular part, there also happens to be a character code assigned to the part itself: “釒(u91d2)”, as is sometimes the case. In this situation, we endorse assigning the data to “釒(u91d2)” and furthermore setting up an alias at “u91d1-01” pointing to “釒(u91d2)”.

Also, for many parts, there are cases where the design differs for when it is used on its own as a character and when it is used as a part. In order to implement this, we recommend using the “-07” suffix code for parts where the location is not a given. Or perhaps, for balance-adjusted glyphs where the parts are used tall-and-skinny or short-and-fat, add the “-08” or “-09” suffixes.

At some point in the future, we intend to implement a function that automatically designs glyphs by describing the parts of a kanji and their relative positions. When that happens, you can search for a hen radical in the data (for example), and the results would come in the order of “-01”, “-08”, “-07”, “(nothing)”.

Naming Guidelines for the purposes of the five columns

In correspondence with the five-column table of UCS, when you want to register a glyph form other than that in the J-column (the Japanese standard), please use the following suffix codes. So other than the “Ucode” column on the table, the suffix letters are the same as the column title distinctions indicating the character’s source.

Suffix codeMeaning
No code(Slowly being abolished)
gG source (People's Republic of China and Singapore)
tT source (Taiwan)
jJ source (Japan)
kK source (South Korea)
vV source (Vietnam)
hH source (Hong Kong Special Administrative Region)
kp(Reserved for KP source) (North Korea)
uU source (U source in the ISO standards. (Originally this was "The Unicode Standard's character form" but this has been moved to "-us")
mM source (Macau)
usCharacter form in The Unicode Standard
iSIP (such as Ext. B) character forms in the ISO/IEC 10646:2003 standards (2011 is the same. 2012 is also written as "UCS2003")
jaan old J source before JIS X 0213 reflected it (Japanese sources of Extension A)
jsan old J source before JIS X 0213 reflected it (Japanese sources of JIS X 0212 supplementary kanji)
jvImaginary J source (see below)

About the imaginary J character form

So far, on GlyphWiki UCS code point glyphs without a designated source exist, but this is planned to be abolished. All UCS code points now need to specify a source. As a practical matter, about glyphs without a J source, "imaginary J character forms" are being attempted because design inconsistencies arise when glyphs that are based on non-J sources are used.

The guidelines for imaginary J character forms are still pending (GlyphWiki:ImaginaryJSourceGuidelines), but the general idea is as follows:

  • Does not look out of place next to kanji that have a J source
  • Specifically, they follow the style of a Heisei Minchou font
  • The fine point of the presence or absence of stroke connections is not to force unity (this is controversial)

When a glyph with a name corresponding to a UCS code is registered with a non-Japanese-style design

Due to visual errors in materials and such, there are times when a glyph that is not a Japanese design is registered as a glyph at a UCS code point. In these cases, it is necessary to first inspect the ISO/IEC 10646:2003 standard chart, and confirm the correct forms.

The standard chart is offered in PDF format at the following link:

http://std.dkuug.dk/jtc1/sc2/wg2/

Please confirm the correct forms, identify which region’s design belongs where for the currently registered glyph, and amend and/or move the glyph data to the correct place. Because there is no data-moving functionality at present, it is necessary to enter the edit mode for the original data to copy the data, and create the page with the new name to paste the data.

Combining both suffixes

When attaching both a regional code suffix and a hen transformation suffix, use the form “-(region)(hen transformation part)”. For example, a (hypothetical) left-side-only Taiwanese u725b would become “u725b-t01”.

Itaiji

For itaiji (kanji variants) other than the above cases, add the suffix "-var-" or "-itaiji-" to the UCS code, and then add a distinct three-digit base-10 number. "-var-" (variation) is used for the glyph unified to that code point, and "-itaiji-" (variant) is used for those not unified. "-var-" is expected to be applicable to most cases.

The three-digit number begins with "001". When the numbered glyph is deleted (blanked), that number is no longer used. Please do not reuse it for a different itaiji. Since "-var-" and "-itaiji-" are independent, those which share the same number do not mean related.

Please do not combine the suffix for the multi-column code chart with "-var-" or "-itaiji-" (u4e00-g-var-001 is invalid. Use u4e00-var-001).

Adobe-Japan1 (CID) glyphs

For the proprietary set developed at Adobe, there are two types: the Adobe-Japan1 collection, listing according to CID number, and the Unicode standard’s IVS ; without particularly preferring the one or the other, please use either of the following two forms:

  • aj1-##### (where ‘#’ is a base-10 digit) e.g. aj1-07765
  • (base UCS code)-(IVS code) e.g. u90a3-ue0101

Description according to IDS

It is also possible to use IDS descriptors. Please tie UCS codes together with “-” to describe the character (see the reference concerning IDS )。

Example: u2ffa-u98a8-u6625 (⿰, 風, 春)

Description of square characters

For square characters, place "kumimoji-" at the beginning of the glyph name, then the IDS description.

Example: kumimoji-u2ff1-u2ff0-u682a-u5f0f-u2ff0-u4f1a-u793e

Other character codes, dictionary indices, etc.

The below names are tacitly reserved, so please do not use them for anything but to target their corresponding glyphs. Furthermore, for the forms in the JIS standard listings below, we assume that they are in the J-column of the UCS. For the GL-area numbers, listings are in hexadecimal with the section and point numbers each having 32 (0x20) added unto them (e.g., section 53 point 21 translates to 5535; see GlyphWiki:GLAreaNumbers).

toki-########Toki unified character number. Eight base-10 digits; leading zeroes are not to be omitted.
koseki-######Koseki unified character number. Six base-10 digits; leading zeroes are not to be omitted.
juki-####Jūki Net (Residential Basic Book Network) unified character number. Four base-16 digits.
nyukan-####Nyūkan gaiji code. Four base-16 digits.
jmj-######MJ Mojizukeimei. Six base-10 digits; leading zeroes are not to be omitted.
gt-#####GT code
gt-k#####GT-k code
tron-##-####TRON code
cdp-####CDP
cbeta-#####CBETA
j78-####JIS X 0208:1978 standard listing illustrated form. Described as GL-area number.
j83-####JIS X 0208:1983 standard listing illustrated form. Described as GL-area number.
j90-####JIS X 0208:1990 (same as 1997) standard listing illustrated form. Described as GL-area number.
jsp-####JIS X 0212:1990 standard listing illustrated form. Described as GL-area number.
jx1-2000-####JIS X 0213:2000 standard listing’s first table’s illustrated form. Described as GL-area number.
jx1-2004-####JIS X 0213:2004 standard listing’s first table’s illustrated form. Described as GL-area number (use of UCS code point recommended).
jx2-####JIS X 0213:2000 (same as 2004) standard listing’s second table’s illustrated form. Described as GL-area number (use of UCS code point recommended).
k0-####KS X 1001 standard table illustrated form. Described as GL-area number.
c1-####CNS 11643 Plane 1
b-####Big5 code
jc3-####JEF-CHINA3 code
dkw-#####Number in Morohashi Tetsuji’s Dai Kanwa Jiten. For a prime (dash) symbol‚ add a ‘d’ to the end; for a double-prime symbol‚ add ‘dd’. For the supplemental volumes‚ use “dkw-h####”. Do not use “m-#####”.
kx-######Kāngxī Zìdiǎn(Kangxi Dictionary)
waseikanji-no-jiten-####Waseikanji no Jiten
kokuji-no-jiten-####Kokuji no Jiten
nihonjin-no-tsukutta-kanji-###Nihonjin no Tsukutta Kanji
zihai-######Zhōnghuá Zìhǎi
twedu-######-###(-#)Dictionary of Chinese Character Variants
baseparts-#####...Basic parts that compose the preliminary data from compiled kanji of JIS X 0208. (This is obsolete; please do not create anything under this pattern.)

About using Konjaku Mojikyō character codes

Concerning character numbers defined in the Konjaku Mojikyō font that the Mojikyo Institute distributes: due to its license , GlyphWiki cannot use them. Because of this, please use names based off of other sources. To avoid misunderstandings, please do not even use “mojikyo-” or “m-” names.