9   Other Chinese Features

This chapter describes features specific to the Chinese locale in Tru64 UNIX that are not described elsewhere.

9.1   Phrase Support in the VT382-D Terminal

You can download up to 100 phrase definitions into the built-in memory of the VT382-D traditional Chinese terminal. You can create a phrase definition file containing the definitions and then download the file to the terminal through the serial port.

Note

The information presented in this section is not applicable to the system Phrase Utility or the dxim Phrase input method discussed in Chapter 7.

9.1.1   Creating a Phrase Definition File

Each phrase definition file can contain up to 100 phrase definitions. You can create the file using any editor (such as vi) that allows you to edit Chinese data.

9.1.2   Syntax of Phrase Definitions

Phrase definitions have the following syntax:

Table 9-1 lists the parameters.

Table 9-1: Phrase Definitions

The following examples show phrase definition files for the 8-bit and 7-bit environments respectively:

• For the 8-bit environment:

• For the 7-bit environment:

  In this example, the second slash in each phrase definition is regarded as part of the phrase definition:

9.1.3   Phrase Downloading

The procedure for transferring phrase definitions from a disk file to the terminal is called downloading. The downloaded phrases are kept in the terminal memory as long as the terminal is powered on. Consequently, a phrase definition file needs to be downloaded only once in a terminal session.

To download a phrase definition file to a terminal, display the file onto a terminal using the Tru64 UNIX cat command. You can also download phrase definitions using the Phrase Utility.

The following situations may occur during downloading:

• If a phrase code is duplicated, the new phrase definition replaces the old one.
• If a phrase code contains more than eight characters, the phrase definition is not accepted.
• If the ST code is typed incorrectly, all the characters from the first slash to the correct ST code (with the exception of DCS) will be treated as part of the phrase. For example:

  <DCS>1 vPHONETIC/注音輸入法AT
  <DCS>0 vINTERNAL/內碼輸入法<ST>

  The AT at the end of the first line is incorrect. If you enter the phrase code "PHONETIC", the following string is input:

  注音輸入法AT0 vINTERNAL/內碼輸入法

• If there are more than 100 phrase definitions, the phrase definitions beyond the limit are ignored.

9.2   Sorting Utility

The methods for sorting Chinese characters are shown in Table 9-2 and Table 9-3.

Table 9-2: Traditional Chinese Sorting Methods

Table 9-3: Simplified Chinese Sorting Methods

You can sort Chinese data using the internationalized sort utility. This utility allows you to use one sorting method by selecting the respective locale as described in Chapter 3. In some cases, you might find that using one sorting method is insufficient to meet your needs. You may need to sort your data with multiple collating sequences. For instance, many characters can have the same number of strokes and you might want to sort these characters further according to their radicals.

To sort characters according to their radicals, Tru64 UNIX provides an extended sort utility, called asort, that you can use to sort or merge files containing Chinese characters according to specified collating sequences. The asort utility has the same syntax as that of the sort utility, but it provides two additional options:

-C"collate_sequence"	Defines the collating sequences where collate_sequence is a list of identifiers or abbreviations of the collating sequences for sorting or merging a file.
-v	Sorts Chinese data in breadth-first comparison, just like the behavior of the VMS/Hanyu or VMS/Hanzi sorting mechanism. By default, depth-first comparison is used.

9.2.1   Multiple Collating Sequences

The asort utility allows you to specify multiple collating sequences. By default, Chinese data is sorted by internal code. You can specify collating sequences with the -C option so that Chinese data can be sorted using other collation methods. For example, the following command sorts DEC Hanyu data files in the order of stroke, radical, and then phonetic:

Alternatively you can enter:

These commands first sort the input data file according to the number of strokes. If multiple characters have the same number of strokes, they are then sorted by radical. If multiple characters within this group start with the same radical, they will then be sorted by phonetic order.

Note

The asort utility is locale sensitive. You should first set the LANG environment variable to the required Chinese locales before using the asort utility.

9.2.2   Comparing Depth-First and Breadth-First Sorting

By default, the asort utility compares Chinese data according to the specified collating sequences using depth-first comparison. That is, each character in a sort field is compared using all the specified collating sequences until the collating order is resolved. When two characters have the same collating order, the next pair of characters is compared.

OpenVMS/Hanyu or Open VMS/Hanzi use a slightly different sorting mechanisms. The HSORT utility provided with OpenVMS/Hanyu sorts characters in the whole sort field using the first collating method. The second collating method applies only if the collating order of the two sorting fields are identical. This is called breadth-first comparison. If you want your sorting results to be compatible with that generated by OpenVMS/Hanyu or OpenVMS/Hanzi, you can specify the -v option:

9.2.3   User-Defined Characters

The asort utility supports the sorting of user-defined characters with the collating values defined through the cedit utility. If required, the asort utility looks up the collating values from the User-Defined Character (UDC) database and sorts the data accordingly. The mechanism for sorting UDCs is totally transparent to you.

9.3   Hanyu and Hanzi DECterm

Hanyu DECterm is a VT382-D terminal emulator; Hanzi DECterm is a VT382-C terminal emulator. This section describes the Chinese features that are specific to the Hanyu and Hanzi DECterm. For details on the common internationalization features provided by DECterm, see Writing Software for the International Market.

9.3.1   Creating a Hanyu or Hanzi DECterm

The terminal type that DECterm emulates is sensitive to the session language.

To create a Hanyu DECterm through the Session Manager, set the session language to one of the traditional Chinese locales, for example, Chinese Taiwan, and then select DECterm from the Applications menu of Session Manager. To create a Hanzi DECterm, select one of the simplified Chinese locales, for example, Chinese China.

Alternatively, you can use the -xnllanguage qualifier to specify the terminal type of the DECterm to create. For example, you can use zh_CN.dechanzi as the value for -xnllanguage to create a Hanzi DECterm:

If you specify an unknown value for -xnllanguage, then ISO Latin-1 DECterm is assumed. If no Chinese font exists, it default to ISO Latin-1 DECterm.

The user interface language of Hanyu and Hanzi DECterm always follows the terminal type. The language is independent of the language selection.

9.3.2   Customizing DECterm

Except for customization of NRCS character sets, all customization features applicable to the ISO Latin-1 DECterm window can also be applied to any Hanyu or Hanzi DECterm window.

Customized features can be saved and restored in the same way as in ISO Latin-1 DECterm.

9.3.3   Font Sizes

By choosing the Big Font or Little Font option from the Window... item of the Options menu you can choose either the 24 x 24 or 16 x 18 Chinese fonts.

9.3.4   Terminal ID

By choosing the General... item from the Options menu you can change the general features, such as the terminal type, for the Hanyu and Hanzi DECterm from a dialog box. You can also choose the VT382 ID from the dialog box.

9.3.5   Interaction Style

By choosing the Input Method... item from the Options menu you can select the interaction style for Hanyu and Hanzi DECterm. For example, if you want to select the Root window interaction style, you can click on the Root window button and then apply the change. If you click on the ISO Latin 1 Input button, Hanyu and Hanzi DECterm disable the input of Chinese data until another style is selected.

9.3.6   Input Server

By choosing the Input Method... item from the Options menu you can switch to use another input server for Hanyu and Hanzi DECterm. By default, the traditional Chinese input server is used for Hanyu DECterm, and simplified Chinese input server is used for Hanzi DECterm. To select another input server, you can click on the Other button and then enter the input server name on the input field.

For Hanyu DECterm, you can enter DECCN to switch to the simplified Chinese input server. For Hanzi DECterm, you can enter DECTW to switch to the traditional Chinese input server. For details about these input servers, see Chapter 7.

9.3.7   Copying Information

You can use the Edit menu to copy information within or between DECterm windows. The cut-and-paste operation is enhanced to handle mixed ASCII and Chinese characters. Beyond this, conversion between traditional and simplified Chinese data is performed when data is copied between Hanyu DECterm and simplified Chinese applications, and between Hanzi DECterm and traditional Chinese applications, through the cut-and-paste or quick copying operation.

9.3.8   Default Chinese Character Set

Hanyu DECterm supports CNS 11643 (first and second planes), DTSCS, and all character sets supported by the ISO Latin-1 DECterm. Hanzi DECterm supports GB2312 and all character sets supported by the ISO Latin-1 DECterm.

ISO Latin-1 DECterm uses ISO 8859-1 (Latin-1) as the default character set. You can override this setting by choosing another option from the General... item on the Options menu. For Hanyu DECterm, the default character set for 8-bit data is the Hanyu character set (CNS 11643 and DTSCS). For Hanzi DECterm, the default character set for 8-bit data is the Hanzi character set GB2312.

In general, Hanyu and Hanzi DECterm cannot display mixed accented Latin-1 characters and Chinese characters. If you want to achieve this, you can output the data together with the designated character set escape sequences.

9.3.9   Chinese Character Input and Output

You can enter Chinese characters in Hanyu and Hanzi DECterm by invoking any of the Chinese input modes as described in Table 7-4.

Mixed ASCII and Chinese characters can be displayed properly in Hanyu and Hanzi DECterm without any special settings.

9.3.10   Reconnecting the Input Server

The Chinese input server provides you the ability to input Chinese characters. If this process does not exist or terminates for some reason, one of the following messages is displayed:

• "Hanyu input method does not exist"
• "Hanzi input method does not exist"

You can restart the input server and then use the Reset Terminal option from the Commands menu to reconnect the Hanyu and Hanzi DECterm to the input server.

9.3.11   VT382-D and VT382-C Terminal Functions

The following functions of the VT382-D and VT382-C terminal are implemented in the Hanyu and Hanzi DECterm terminal emulator respectively:

• Display characteristics and capabilities
• Text capabilities

• Support for Terminal State Interrogation (TSI)
• Support for all of the Chinese input methods
• Support for the following character sets:

• Control Representation mode
• Support for sixel graphics
• Support for UDK editing function
• Chinese character display attributes: reverse, underline, bold, blink, double height/width

The following functions of the VT382-D terminal are implemented in the Hanyu DECterm terminal emulator:

• Display/Suppress leading code

• The escape sequence DECLCSM, that is, Leading Code Suppression Mode, is also supported in Hanyu DECterm.

For details about the VT382-D terminal functions, see the VT382-D Programming Reference Manual and VT382-D User's Manual. For details about the VT382-C terminal functions, see the VT382-C Programming Reference Manual and VT382-C User's Manual.

9.4   Phrase Conversion

Chapter 2 describes how Tru64 UNIX supports conversion between different codesets using the iconv utility. This utility can also be used for phrase conversion. When phrase conversion is activated, a phrase in traditional Chinese can be converted to a phrase in simplified Chinese, or the reverse. Phrase conversion does not apply to traditional Chinese encoded in Telecode, but this is the only exception.

To activate the phrase conversion option, you can define the ICONV_PHRCONV environment variable. If this environment variable is set to mark, the converted phrases are enclosed in brackets ([]) to highlight the conversion result for visual checking.

The phrase conversion databases in the /usr/share/phrdb directory are normal text files with the same file names as those of the algorithmic converters in /usr/lib/nls/loc/iconv/*. These phrase conversion databases contain entries for phrase conversion pairs.

9.5   Special Characters in nroff

The nroff utility has been internationalized to format text of various languages. When a Chinese document is formatted using nroff, its contents are handled according to Chinese formatting rules:

• Text line wrapping - - Chinese text, unlike English text, does not use space characters as separators between words or as break points for text line wrapping. Instead, a Chinese text line can be broken between any two consecutive Chinese characters, with the following exceptions:
  
  - Some Chinese characters cannot be placed at the beginning of a text line. They are called no-first characters. For traditional Chinese, the no-first characters include the following:
  

  

  For simplified Chinese, the no-first characters include the following:
  
  、。ˉˇ‥〃―～∥... ， " 〕〉》」』
  〗】∶！），．：；＞？］｝
  
  - Some Chinese characters cannot be placed at the end of a text line. They are called no-last characters. For traditional Chinese, the no-last characters include the following:
  

  

  For simplified Chinese, the no-last characters include the followings:
  
  ～ ' " 〔〈《「『〖【（［｛
  
  - Some English characters are handled similarly according to this rule. No-first English characters include the following:
  
  ! ) , . : ; > ? ] )
  
  
  No-last English characters include the following:
  
  ( < [ (
  

• Right justification - To justify the right margin of a paragraph, nroff inserts space characters at proper break points. For English, space characters are usually inserted at word breaks. For Chinese, space characters are inserted only at the following places:
  
  - Before a can-space-before character if it is not placed at the beginning of a text line.
  
  For traditional Chinese, the can-space-before characters include the following:
  

  

  For simplified Chinese, the can-space-before characters include the following:
  
  ' " 〔〈《「『〖（［｛
  
  - After a can-space-after character if it is not placed at the end of a text line.
  
  For traditional Chinese, the can-space-after characters include the following:
  

  

  For simplified Chinese, the can-space-after characters include the followings:
  
  、。 ' " 〕〉》」』〗】！），．＞？］｝