漢字字碼與資料庫國際研討會，京都‧東京 1996年10月4日
從缺字問題，談漢字交換碼的重新設計──第二部分

A Descriptive Method for Re-engineering Hanzi
Information Interchange Codes

3. Design Issues

(9)An Example

For illustration of the descriptive coding method, an example is given in 〔Table 7〕. In 〔Table 7〕 , 50 characters/glyphs of the禸 family from glyph data base are listed. 40 of them are non-kernel characters. Therefore , only 10 kernel glyphs need numerical coding space.

For comparison, the same family derived by Professor周何 is listed in 〔Table 8〕 . These more than 80 characters can all be described by the kernel of our glyph data base. 〔Table 7〕is a result of graphics decomposition, while Table is derived form traditional Wen-zi-xue. It is clear that they may be merged into one structure by adding proper data structures to the present glyph data base.

(10)Replacement Code

The user defined coding space of numerical coding is no longer needed while descriptive extension is added to the existing code. Therefore, it can be used for other purpose. One possible way is assign this space for replacement codes. The replacement codes are used for replacing some rare used glyphs that happen to be frequently appeared in some specific text. To do so, some markups are needed to specify the replacement. For example, when code xxxx is used to replace a component expression of descriptive coding, the replace relation needs to be specified in the text file. So, a DTD and associated tags of Hanzi file can thus be designed accordingly to facilitate efficiency and information sharing .This is an undergoing topic of our project.