Multibyte Text Handling

If a Marine project has been identified as multibyte (PROJECT MBCHARSET in the Admin module), it is possible for users to enter multibyte element names and attributes in some Marine modules. This allows Far Eastern users to enter Kanji text, which is stored in Marine as multibyte text (two bytes per Kanji character). Everything in this section referring to Kanji also applies to Chinese characters. Marine uses the name 'FECs' (Far Eastern Characters) to indicate any multibyte character text. The byte pairs correspond to the appropriate ISO-IR registration. For example, for Kanji, the byte pairs correspond to ISO-IR 87, the Japanese standard registered with the International Standards organization.

FECs text can be input to or returned by DARs routines; for example:

Within DARs, multibyte text is input or returned as byte-pairs delimited by 'escape sequences'. These are the 'FECS-IN' and 'FECS-OUT' sequences:

These sequences were carefully selected and meet the following criteria:

• The FECS-IN sequence does not occur frequently in non-FECs usage

• The FECS-OUT sequence does not exist as a byte-pair

It is possible to freely mix single-byte and multi-byte text in any text string. For example, the text string 'abc&~4A;z & def' is interpreted as follows:

3 characters ‘abc’

2 byte pairs corresponding to '4A' and ';z' (meaning 'kanji')

4 characters ' def'

Genuine occurrences of '&~' are denoted by '&&~'; so that 'abc&&~&~4A;z & def' is interpreted as:

5 characters 'abc&~'

2 byte pairs corresponding to '4A' and ';z' (meaning 'kanji')

4 characters ' def'

It is possible for the byte-pair '&~' to occur within a FECs string. Thus, 'abc&~&~4A;z & def' is interpreted as:

3 characters 'abc'

3 byte pairs corresponding to '&~', '4A' and ';z'

4 characters ' def'

Using these simple rules, Far Eastern DARs programmers should be able to convert output FECs strings from PDMS format to the format corresponding to their host computer and its peripherals (for example, for output to a Kanji terminal or printer). It is not necessary for programmers or DARs users to know the layout of the ISO-IR 87 tabulation.

The following example program may assist programmers. It is a stand-alone program, using no DARs routines. The user is prompted to enter text strings containing FECS-IN and FECS-OUT sequences. The statement converting single-byte to multibyte is machine-specific.

The following is a DARs example, reading FECs names and text from a FECs project. Character function FECS is the same as that listed above.

The output from the above program appears as follows. Since Aveva has no Kanji printer, Kanji text in this example is indicated by '@' characters.