Home

3 - Oracle Documentation

1. d The PROM monitor and OpenWindows 3 3 provide fonts for western Europe standardized as ISO 8859 1 also called ISO Latin 1 These ISO fonts use the full 8 bit address space of a byte The system is supposed to come up in 8 bit mode by default but in case it does not simply type stty cs8 istrip in a shell or command window When SunOS boots on a SPARC system it automatically recognizes the keyboard type If you plug an alternate keyboard into a running system SPARC hardware generates an interrupt and returns you to the PROM monitor After the PROM monitor s or ok prompt simply type c or go for continue Then choose Refresh from the OpenWindows Workspace Utilities menu and execute the 1 command in a shell or command window When SunOS boots on an x86 system the system is supposed to come up in 8 bit mode by default If it does not simply type stty pass ina shell or command window Support for Internationalization 17 18 Native Language Keyboards The SPARC Type 5 and Type 4 keyboards generate 8 bit characters or events System translation tables generate appropriate character codes based on these events The SunOS translation tables are in usr share lib keytables OpenWindows keyboard tables are SOPENWINHOME etc keytables The PC AT101 and PC AT102 keyboards generate 8 bit characters or events System translation tables generate appropriate characte
2. Common International Page Sizes Available SPARC Keyboards Available PC Keyboards Compose Key Sequences and Wide Character Representation International Library Routines Library Routines for Codeset Independence Language and Territory Names Compose Key Sequences ISO 8859 Standard Codesets ISO 8859 Standard Code Sets 14 16 16 20 28 29 42 70 71 75 78 ix Developer s Guide to Internationalization August 1994 Book Organization Preface This book is intended for people who need to write internationalized software for the Solaris environment Knowledge of the C programming language is assumed All operating system information pertains to SunOS while all window system information pertains to OpenWindows 3 3 The material in this book is organized as follows Chapter 1 introduces key notions of internationalization and localization Chapter 2 describes some of the cultural differences that global software design must resolve Chapter 3 describes internationalization features that the operating system offers to users and application developers Chapter 4 describes correct international coding practices and provides C language examples Chapter 5 discusses writing internationalized window system code Chapter 6 describes how
3. or 1 1 1 slot 2 SS2 1 or SS2 1 1 or SS2 1 1xxxxxxx 1xxxxxxx slot 3 SS3 1 or SS3 1 1xxxxxxx or SS3 1 1 1xxxxxxx Developer s Guide to Internationalization August 1994 Qo lll Whether code sets 1 2 and 3 are single byte double byte or triple byte depends on the locale Code set 1 could be used to represent ISO Latin 1 but this is not the usual practice in East Asian locales EUC divides the code set space into graphic and control characters Graphic characters are those that can be displayed Special characters include control characters unassigned characters and the space and delete characters Control characters are characters other than graphic characters whose occurrence may initiate modify or stop a control operation The following table indicates the single byte special characters Special Character EUC Representation Space 00100000 Delete 01111111 Control codes Primary 000 Control codes Supplementary 100 552 and 553 are examples of supplementary control codes Wide Characters EUC is intended primarily for external data storage and its encoding schemes provide reasonably compact representations for data storage However EUC is not very convenient for internal processing its variable length nature compl
4. Word Delimiters Word Order xu sp DS pde Sort Order 55229 cea ta alt Character Sets aet PVP VA epa Codesets TOP X96 utopia coe ee TERT epp E ET EEEE Keyboard Differences Other oso u ow uwa SS VAR we Ala Punctuation i kk mL E X Xe ea Symbol duns sasaqa pip Pat ea Measurements P ABeSSIZeS nuo ae ee Re eae E mam 3 Support for Internationalization Keyboards and Keyboards c cin pd EE Lada Generating Characters Not U S Keyboard Using the Floating Accent Keys Modems qasqa Dumb Terminals Developer s Guide to Internationalization August 1994 AO AO goaren ent Su Ao 25 Character and Code 25 Extended UNIX Code EUC 26 Wide Characters 27 Multi byte Library Routines 28 jor wu id CE ete oad 34 Whatisa Locale epe oh 34 Locale Categories 35 4 Writing Internationalize
5. Object Size and Positioning International XView contains attributes to allow run time configuration of the size and position of objects making up an applications interface Size and positioning information is retrieved from an Xwindows style resource database The attributes associated with sizing and positioning objects are XV_INSTANCE_NAME Provides a mechanism to give each XView object a unique name so that it can be queried from the database XV_USE_DB Allows you to specify a null terminated list of attributes value pairs Each attribute s value will be looked up in the database if not found the specified value is used as the default XV_LOCALE_DIR An xv init attribute tells XView where the app defaults directory is located The resource database file resides in a directory called app defaults Like the scheme used by the message object files the parent of the app defaults directory is the name of the locale The name of the resource database file is the name of the application with a db suffix appended to it To illustrate the use of these attributes see how the following portion of code from the file my app c is modified xv init XV INIT ARGC PTR ARGV amp argc argv 0 frame xv create XV NULL FRAME XV NULL control xv create frame PANEL XV NULL button create control PANEL BUTTON PANEL LABEL STRING Quit XV X 10 XV NULL Assuming th
6. O acute O A O circumflex O O tilde O d O umlaut x x x multiply O O slash U E U grave U U acute U U circumflex U 9 U umlaut Y Y Y acute T H P Thorn s s 8 ess zed digraph s a a grave a acute AN circumflex Developer s Guide to Internationalization August 1994 Qo II Table 3 3 Compose Key Sequences Compose Key Sequence Result Description a a tilde a 4 umlaut 7 angstrom ae ligature O B e T 1 i i 1 i 1 4 i d n A 6 6 u u u c cedilla e grave e acute e circumflex e umlaut i grave i acute i circumflex i umlaut eth n tilde o grave acute o circumflex o tilde o umlaut divide o slash u grave u acute u circumflex u umlaut Support for Internationalization 23 24 Table 3 3 Compose Key Sequences Compose Key Sequence Result Description y i y acute t h thorn y d y y umlaut Decimal Representation A third method of generating characters is using their decimal representation Every character corresponds to a unique number Up to 256 different characters can be used although some terminals only support 128 When the COMPOSE key is used followed by three digits the character that is internally represente
7. and certain other countries UNIX and OPEN LOOK are registered trademarks of UNIX System Laboratories Inc a wholly owned subsidiary of Novell Inc PostScript and Display PostScript are trademarks of Adobe Systems Inc All other product names mentioned herein are the trademarks of their respective owners All SPARC trademarks including the SCD Compliant Logo are trademarks or registered trademarks of SPARC International Inc SPARCstation SPARCserver SPARCengine SPARCstorage SPARCware SPARCcenter SPARCclassic SPARCcluster SPARCdesign SPARC811 SPARCprinter UltraSPARC microSPARC SPARCworks and SPARCompiler are licensed exclusively to Sun Microsystems Inc Products bearing SPARC trademarks are based upon an architecture developed by Sun Microsystems Inc The OPEN LOOK and Sun Graphical User Interfaces were developed by Sun Microsystems Inc for its users and licensees Sun acknowledges the pioneering efforts of Xerox in researching and developing the concept of visual or graphical user interfaces for the computer industry Sun holds a non exclusive license from Xerox to the Xerox Graphical User Interface which license also covers Sun s licensees who implement OPEN LOOK GUIs and otherwise comply with Sun s written license agreements X Window System is a product of the Massachusetts Institute of Technology THIS PUBLICATION IS PROVIDED AS IS WITHOUT WARRANTY OF ANY KIND EITHER EXPRESS OR IMPLIED INCLUDING BUT NOT LIMITE
8. Table B 1 Compose Key Sequences 74 Compose Result Description T H Th Thorn s s amp Eszett digraph s a I a grave a a acute a circumflex a a tilde a i a umlaut a 5 a angstrom a e ae ae ligature c cedilla e e grave e t e acute e e circumflex e e umlaut i i i grave i d 1 1 acute 1 A 1 i circumflex i E i i umlaut d d eth n fi n tilde 9 grave acute AN o circumflex 6 o tilde E 6 o umlaut divide slash Developer s Guide to Internationalization August 1994 II Table B 1 Compose Key Sequences Compose Result Description u u grave u u acute u es a u circumflex u E u umlaut y y y acute t h th thorn y y y umlaut The previous table of Compose key sequences is for ISO 8859 1 only The ISO 8859 codesets are listed below Table B 2 ISO 8859 Standard Codesets Codeset Name Coverage Approved 8859 1 Latin 1 Western Europe 15 February 1987 8859 2 Latin 2 Eastern Europe 15 February 1987 8859 3 Latin 3 Maltese Catalan Galician Esperanto 15 April 1988 8859 4 Latin 4 Baltic and Nordic region 15 April 1988 8859 5 Cyrillic Slavic countries 1 December 1988 8859 6 Arabic Arab countries 15 August 1987 8859 7 Greek Greece 15 November 1987 8859 8 Hebrew Israel 1 June 1988 8859 9 Latin 5 8859 1 minus Iceland plus Turkey 15 May 1989 Compose Key
9. higher Localization costs can be high particularly for translation but often repay themselves quickly in higher sales Conforming to Standards Many standards bodies are developing guidelines for internationalized software Practices described in this book conform to various ANSI IEEE ISO and X Open standards International standards are still evolving so not all interfaces can be guaranteed forever 2 Developer s Guide to Internationalization August 1994 II Internationalization Levels SunSoft defines four levels of internationalization described below Levels are not necessarily hiearchical but do indicate difficulty of implementation Level 1 Text and Codesets Software that is level 1 compliant is 8 bit clean and can therefore use the ISO 8859 1 also called ISO Latin 1 codeset Historically many programmers assumed their application needed only the ASCII character set Because the ASCII codeset employs only seven bits out of an 8 bit byte the most significant bit was often used to store information about the character For example setting the most significant bit on might indicate that the character is highlighted The ISO Latin 1 codeset employs all eight bits Software that uses the most significant bit for its own purposes is not level 1 compliant Level 2 Formats and Collation Software is level 2 compliant if its formatting and collation methods are locale sensitive Many different form
10. letter isupper c capital letter islower c lower case letter isdigit c digit from 0 9 isxdigit c hexadecimal digit from 0 isalnum c alphanumeric letter or digit isspace c white space character ispunct c punctuation mark isprint c printable character iscntrl c control character isascii c 7 bit character isgraph c visible graphics character Avoid Managing the Keyboard Type 5 Type 4 PC AT101 and PC AT102 keyboards have provisions for typing all the ISO Latin 1 characters If programs use system and window services to read characters they need only be 8 bit clean But if programs read dev kbd or perform keystroke mapping they are managing the keyboard If that s the case make sure to support these input methods for all keyboard layouts Entering any character using a single keystroke Note that German keyboards switch Y with Z while French keyboards switch Q and W with A and Z Developer s Guide to Internationalization August 1994 4 Using dead key accent followed by a given keystroke to produce any valid ISO 8859 1 character Pressing the Compose key followed by two additional keystrokes to produce any valid ISO character The most obvious way to test software that manages the keyboard is to type all the Compose sequences in Table 3 1 and to type all keystrokes on several country kit keyboards If at all possible use OpenWindows to manage the keyboard for you This sa
11. lock gettext find type flg gettext page gettext record The German message catalog would then appear as follows Developer s Guide to Internationalization August 1994 HS II Other Languages msgid Unable to s the s n msgstr Das Programm kann 2Ss nicht 1 s n msgid lock msgstr sperren msgid find msgstr finden msgid page msgstr die Seite msgid record msgstr den Rekord This example might not work on other vendors systems because of multiple gettext calls per line Please consider carefully the effects of dynamic messaging You might have to reposition parameters during translation Often this fact isn t recognized until translation actually begins by which time it s already too late the software would have to be laboriously re released SunOS 5 4 provides a get text 1 command to retrieve translated messages from a catalog This command reads the TEXTDOMAIN environment variable for domain name and the TEXTDOMAINDIR environment variable for the path name to the message database For PostScript programmers the NeWS toolkit TNT 3 3 provides some internationalization facilities but NeWS itself does not Summary of Requirements This chapter presented the standard X Open and SunOS library routines for internationalizing software Use standard X Open routines whenever possible but use the SunOS gettext function instead of the X Open catgets
12. void setlocale LC CTYPE 37 38 Among other things this ensures that European accented characters such as are correctly identified with an isalpha library call Note that the empty string argument indicates that the application should set its codeset according to the environment variable LC_ALL LC_CTYPE or LANG in that order of precedence If none of these environment variables is set the default locale is C which results in old style UNIX behavior Internally this call changes values in the __ct ype array of the C library This in turn affects the behavior of various ct ype 3 library routines The LC_CTYPE locale category may also affect other functions including wide character handling In most cases library packages should rely on the programmer to call setlocale inside the application If not the call is re entrant in the sense that it affects static data structures Applications that fail to call set locale would simply fail to get international features To set all the above locale categories at the same time use the LC ALL argument to setlocale instead of just LC_CTYPE In practice most applications should set the LC_ALL category once and for all Make Software 8 bit Clean Programs shouldn t alter the most significant bit of a char The computer industry used this bit for parity many years ago but it didn t work out well data got corrupted because software ignored the parit
13. PostScript TeX to PostScript regular text to PostScript Tektronix 4014 to PostScript Diablo 630 to PostScript plot 5 to PostScript All these conversions are 8 bit clean Be aware that standard paper size around the world varies widely Internationalized applications do not assume any particular set of page sizes SunOS provides no support for tracking the output page size this is the responsibility of the application program itself Character and Code Sets The ISO Latin 1 character set is used to represent European characters sets ISO Latin 1 uses eight bits one byte to represent each character allowing for 255 characters It is compatible with the 7 bit ASCII code set in that all ASCII characters have identical encodings in the ISO Latin 1 when the most significant bit in ISO Latin 1 is set to 0 ISO Latin 1 can be thought of as a superset of ASCII for purposes of text representation East Asian characters however cannot fit into a single byte Consequently Chinese Japanese and Korean all require multi byte code sets for language processing SunOS uses the EUC encoding scheme to represent multi byte characters With EUC one to four bytes may be used to store a character EUC characters are not necessarily convenient to process by standard functions because of their variable length nature so SunOS provides functions to convert EUC characters to wide characters In SunOS wide characters are four bytes long and
14. Steps Internationalization An internationalized application s executable image is portable between languages and regions To internationalize software you Use the interfaces described in this book to create software whose environment can be modified dynamically without the software needing to be recompiled Separate all printable and displayable messages that the user sees from the executable image Keep these message strings in a message database Message strings are translated for a language and region called a locale as part of the localization process Related databases that specify formats for time currency and numbers are translated at the same time To use a localized version of a product users set an environment variable The product then displays user messages in their translated form and also formats date time and currency according to the locale specific conventions Thus users gain control of their software s language and behavior Advantages of Internationalization Creating internationalized software automatically expands the market for your product If you follow the steps outlined in this book your product will be compatible with all the languages and cultures supported by Solaris Many computer firms including Sun Microsystems and Digital Equipment obtain around half of their revenue from outside the United States Profit margins are often higher abroad so contribution to net income can be even
15. This sentence is easy to translate N because it is included in one printf statement n Translation problems can arise with compound messages especially when more than one sentence could be produced at run time Here is some code that would be very difficult to translate poor practice multi part compound message printf s Unable to s d data s s 55 func alloc flg allocate free count file flg file structure count 1 5 perror Quite apart from being poor programming practice this fragment of code would be much clearer to the reader and much easier to translate if it were split into separate print statements inside an if else block that would select the correct message at run time if alloc_flg if file flg printf Unable to allocate d file n count else printf Unable to allocate d structure n count else if file flg printf Unable to free d file n count else printf Unable to free d structure n count The issue of making the objects plural is not addressed in this example because in many languages pluralization involves more than adding s to the end of a word Writing Internationalized Code 53 54 Dynamic Messaging Dynamic messaging is used when the exact content or order of a message is not known until run time Unless done carefully dynamic messaging causes translation problems If the positional dependence
16. can be processed using wide character library routines The next sections examine EUC and wide characters in detail Support for Internationalization 25 26 Extended UNIX Code E LIC SunOS has adopted the EUC from USL s Multi National Language Supplement MNLS EUC is used primarily for storing data in files EUC is comprised of four code sets three of which may be multi byte and two of which must be announced by 8 bit control codes known as single shift characters EUC s primary code set code set 0 is used for ASCII The three supplementary code sets code sets 1 2 and 3 can be assigned to different code sets by the locale administrator Code set 0 is single byte with the most significant bit set to zero The supplementary code sets can be single or multi byte with the most significant bit set to one Code sets 2 and 3 have a preceding single shift character known as 552 and 553 respectively where 552 Ox8E 10001110 and 553 0x8F 10001111 There is no 551 Differentiating between code sets is done as follows If the high bit is 0 the code set is ASCII If the high bit is 1 the byte is checked for SS2 or SS3 to determine code set The length in bytes of characters from that code set is retrieved from the LC_CTYPE locale database governing character classification associated with the current locale Code set EUC Representation slot 0 slot 1 1 or 1 1
17. controll buttonl xv y 10 If the layout has changed for a specific locale an interface object may need to be repositioned For example let s say that for the Spanish locale the interface looks more presentable if the above button is positioned at position 20 11 rather than position 10 10 the developer needs to do is to change the x and y values in the db file and rerun the application without recompiling In this case the above database lines would be changed to Developer s Guide to Internationalization August 1994 O1 II Input Method my_app base_frame controll buttonl xv_x 20 my app base frame controll buttonl xv y 11 The button would then appear at position 20 11 Phonetic based languages such as English are easy to type on a keyboard However typing is much more complicated for East Asian languages based on ideograms When entering Japanese for example the typist must select between Kanji Hiragana Katakana and Romaji Roman letters used phonetically to produce Japanese words There must also be a way to convert phonetic characters into Kanji Input is further complicated because a given syllable can frequently be represented by several different Kanji characters the user must select the Kanji character that is appropriate for the specific context Here s how you d type Japanese under OpenWindows with a Japanese keyboard 1 Set the LANG environment to Japanese 2 The initial input mode sim
18. first time a reference to the manual page for the command is often given with the section number in parentheses command section For example grep 1 is the grep command described in section 1 of the manual pages Developer s Guide to Internationalization August 1994 Introduction 1 What Is Internationalization Internationalization is a way of designing and producing software that can easily be adapted to local markets Internationalized products can be localized or adapted to different languages and cultures with minimal effort How Is Localization Different Internationalization is the process of making software portable between languages or regions while localization is the process of adapting software for specific languages or regions International software can be developed using interfaces that modify program behavior at run time in accordance with specific cultural requirements Localization involves establishing on line information to support a language or region called a locale Unlike software that must be completely rewritten before it can work with different native languages and customs internationalized software does not require rewriting It can be ported from one locale to another without change Solaris is internationalized providing the infrastructure and interfaces you need to create internationalized software Chapters 3 and 4 describe what facilities are available and how to use them II Basic
19. function because gettext has an easier programming interface than catgets If you re writing window based software consult Chapter 5 for further information on how to internationalize window objects Some messages produced by routines in libc are localized The application must be dynamically linkedbe be sure that the localized messages are displayed Writing Internationalized Code 55 56 Checklists for Internationalization Do Call setlocale to initialize language cultural conventions Make software 8 bit clean by verifying that it doesn t alter the most significant bit of 8 bit bytes Watch for sign extension problems since by default Sun C compilers treat all characters as signed values Use standard code sets such as ISO Latin 1 or EUC Use ctype 3 library routines to identify character ranges Remember that many countries use a comma for the decimal separator both printf and scanf have been changed accordingly Use strftime not ctime Use the 1 1 call if you need to obtain local currency formats Replace calls to strcmp with calls to strcoll Surround user message strings with gettext calls and create separate message files for easy translation Plan for size changes in user messages after translation Use explicit functions to center and align text Allow punctuation characters within words Provide translation notes where possible as in code comments D
20. is not recommended and is not used in Solaris Whichever mechanism is used for messages translation and only one method should be used in any one application level 3 compliant software should strive to avoid compound messages messages consisting of separately composed parts because word order is different in many languages Level 4 Asian Language Support Level 4 compliant software provides support for East Asian languages which often require multi byte codesets because of their large character inventory Solaris provides such support with the Extended Unix Code EUC This is a method for switching between multiple codesets three of which may in turn be multi byte codesets Developer s Guide to Internationalization August 1994 Formats and Conventions Overview 2 Different countries in world use completely different conventions for writing date time numbers currency delimiting words and phrases and quoting material You should not code these conventions directly into your application an internationalized product in conjunction with a localization package will be sensitive to the appropriate formats Chapters 4 and 5 provide guidance on coding practices used to achieve these formats Formatting Differences Time Formats The following table shows some of the ways to write 11 59 p m Table 2 1 International Time Formats Locale Format Canadian 23 59 Finnish 23 59 German 23 59 Uhr Norwegian KI 23 59 U
21. strings require some effort The one line initialization statement char greeting Hello below must be converted into the two lines char greeting greeting gettext Hello If strings must be stored in an array be sure to declare arrays large enough to hold all possible translations Then call strcpy as follows char greeting BUFSIZ greeting gettext Hello sizeof greeting Use bindtextdomain Many applications do not require root permission for installation and thus cannot place their messages usr lib locale Moreover most applications need messages in their own directory hierarchy to simplify export across a network So most applications should use bindtextdomain to associate a path name with a message domain This routine is available in SunOS Here s a sample invocation char buf BUFSIZ strcpy buf getenv OPENWINHOME bindtextdomain xview strcat buf locale textdomain xview printf gettext Cannot find font If the LANG environment had been set to fr in SunOS this would obtain a translation from SOPENWINHOME locale fr LC_MESSAGES xviewmo Passing null pointer as the second argument causes bindtextdomain to return the path name associated with the first argument s message domain Here s how the path name to message files is constructed under SunOS pathname SLANG LC MESSAGES 1 domain mo Developer s Gu
22. D TO THE IMPLIED WARRANTIES OF MERCHANTABILITY FITNESS FOR A PARTICULAR PURPOSE OR NON INFRINGEMENT THIS PUBLICATION COULD INCLUDE TECHNICAL INACCURACIES OR TYPOGRAPHICAL ERRORS CHANGES ARE PERIODICALLY ADDED TO THE INFORMATION HEREIN THESE CHANGES WILL BE INCORPORATED IN NEW EDITIONS OF THE PUBLICATION SUN MICROSYSTEMS INC MAY MAKE IMPROVEMENTS AND OR CHANGES IN THE PRODUCT S AND OR THE PROGRAM S DESCRIBED IN THIS PUBLICATION AT ANY TIME ion 7 Contents 1 e e e e e e e How Is Localization Different Basic Steps Internationalization Advantages of Internationalization Conforming to Internationalization Levels Level 1 Text and Codesets Level 2 Formats and Collation Level 3 Messages and Text Presentation Level 4 Asian Language Formats and Conventions OvervieW Formatting Differences Time Formats Date Formats OV QI H Ww CQ G 02 N NY P eae iii aS CUTE casos tutuuy s u xam Word and Letter
23. DevelopersGuide to Intemationalization 2550 Garcia Avenue S Mountain View CA 94043 U S A 8 655 Microsystems Inc Busin 1994 Sun Microsystems Inc 2550 Garcia Avenue Mountain View California 94043 1100 U S A Allrights reserved This product and related documentation are protected by copyright and distributed under licenses restricting its use copying distribution and decompilation No part of this product or related documentation may be reproduced in any form by any means without prior written authorization of Sun and its licensors if any Portions of this product may be derived from the UNIX and Berkeley 4 3 BSD systems licensed from UNIX System Laboratories Inc wholly owned subsidiary of Novell Inc and the University of California respectively Third party font software in this product is protected by copyright and licensed from Sun s font suppliers RESTRICTED RIGHTS LEGEND Use duplication or disclosure by the United States Government is subject to the restrictions set forth in DFARS 252 227 7013 c 1 ii and FAR 52 227 19 The product described in this manual may be protected by one or more U S patents foreign patents or pending applications TRADEMARKS Sun theSun logo Sun Microsystems Sun Microsystems Computer Corporation SunSoft the SunSoft logo Solaris SunOS OpenWindows DeskSet ONC and NFS are trademarks or registered trademarks of Sun Microsystems Inc in the U S
24. K 11 59 PM Time is represented by both a 12 hour clock and 24 hour clock sometimes known as railroad time The hour and minute separator can be either a colon or a period Some countries attach letters to the time indicating that this is a time but these are not strictly necessary Time zone splits occur between and within countries Although a time zone can be described in terms of how many hours it is ahead of or behind Greenwich Mean Time GMT this number is not always an integer For example Newfoundland is in a time zone that is half an hour different from the adjacent time zone Daylight Savings Time DST starts and ends on different dates that can vary from country to country Date Formats This table shows some of the date formats used around the world Of course there is a good deal of variation even within countries so these formats are not the final truth Table 2 2 Imternational Date Formats Locale Convention Example Canadian English 1989 08 13 Canadian French yyyy mm dd 1989 08 13 Danish dd mm yy 13 08 89 Finnish dd mm yyyy 13 08 1989 French dd mm yy 13 08 89 German dd mm yy 13 08 89 Italian dd mm yy 13 08 89 Norwegian dd mm yy 13 08 89 Spanish dd mm yy 13 08 89 Swedish yyyy mm dd 1989 08 13 UK English dd mm yy 13 08 89 US English mmc dd yy 08 13 89 Developer s Guide to Internationalization August 1994 No lll Numbers Decimal and Thousands S
25. Many of the features above will not work for statically linked programs 15 3 Keyboards and Peripherals Keyboards The Type 5 and Type 4 keyboards are currently available in 18 versions 15 for Roman alphabets and three for Asian languages Table 3 1 Available SPARC Keyboards Belgium France Denmark Netherlands Spain Switzerland German Japan Canada Germany Norway Sweden Finland United Kingdom Korea Canada French Italy Portugal Switzerland French United States Taiwan International keyboards are normally delivered with country kits but are also available separately The PC AT101 and PC AT102 keyboards are currently available in Table 3 2 Available PC Keyboards Belgium Norway Sweden United Kingdom Canada Italy Portugal Switzerland French United States Denmark Netherlands Spain Switzerland German 16 Developer s Guide to Internationalization August 1994 3 The lower half of the ISO Latin 1 code set contains all characters from the ASCII code set while the upper half includes accented and special characters shown in the figure below Figure 3 1 Upper Half of ISO 8859 1 shelltool bin csh D CO s Oe O O eemm R a O 2 Ceo Ae C O CH 4 LT X
26. Sequences 75 76 Developer s Guide to Internationalization August 1994
27. ar Arabic bg Bulgarian ca Catalan co Corsican cs Czech cy Welsh da Danish de German de_CH Swiss German el Greek en English en_UK UK English en_US US English Esperanto es Spanish eu Basque fa Persian fi Finnish French fr_BE Belgian French fr_CA Canadian French frcH Swiss French fy Frisian ga Irish gd Scots Gaelic hu Hungarian is Icelandic it Italian iw Hebrew ja Japanese ji Yiddish kl Greenlandic ko Korean lv Latvian nl Dutch no Norwegian pl Polish pt Portuguese ro Romanian ru Russian sh Serbo Croatian Sk Slovak Su Serbian sv Swedish th Thai tr Turkish zh Chinese zh TW Chinese in Taiwan Developer s Guide to Internationalization August 1994 Compose Key Sequences B SPARC and x86 keyboards can produce all characters in the standard ISO 8859 1 Latin 1 codeset by using the Compose key On a SPARC keyboard pressing the Compose key should illuminate the light on the key On an x86 keyboard the Compose key is control shift F1 hold the Control and Shift keys down and press the F1 key Localized keyboards do not all generate the same key codes for ISO Latin 1 characters However system translation tables take care of producing the right character code SunOS finds its tables in usr share lib keytables Table B 1 shows how to produce special ISO Latin 1 characters using Compose key sequences Table B 1 Compose Key Sequences Compose Result Description space space space i inv
28. ards Similarly other keyboards often contain many characters not visible on the US keyboard However the Compose key can be used to produce any character in the ISO Latin 1 code set on any keyboard that supports it See Appendix B for a list of different keyboard layouts Both the position and the type of punctuation symbols can vary between languages In Spanish and j appear at the beginnings of sentences while in Finnish colons can occur inside words Commonly used symbols in one culture often have no meaning in another culture Because the common US rural mailbox for example does not exist in other countries it would not make a universal MailTool icon Developer s Guide to Internationalization August 1994 No II Measurements Gender While most countries now use the metric system of measurement the United States parts of Canada and the United Kingdom albeit unofficially still use the imperial system The symbols for feet and inches are not understood in all countries The spelling of adjectives articles and nouns are gender dependent in some languages In French for example un petit gamin and une petite gamine both mean a cute kid The first expression however refers to a boy and the second expression to a girl Also neuter objects in English a computer for example have gender in other languages un ordinateur is a masculine noun in French Titles an
29. at is reversed in the equivalent Spanish phrase uno gato negro And in French the negatives ne and pas surround the word they negate as in the phrase I do not speak which in French is Je ne parle pas Sorting order for particular characters is not the same in all languages For example the character sorts with the ordinary o in Germany but sorts separately in Sweden where it is the last letter of the alphabet Character Sets Number of Characters While the English alphabet contains only 26 characters some languages contain many more characters Japanese for example can contain over 40 000 characters Chinese even more Western European Alphabets The alphabets of most western European countries are similar to the standard 26 character alphabet used in English speaking countries but there are often some additional basic characters some marked or accented characters and some ligatures Formats and Conventions Overview 9 10 Japanese Japanese text is composed of three different scripts mixed together Kanji ideographs derived from Chinese and two phonetic scripts or syllabaries Hiragana and Katakana Although each character in Hiragana has an equivalent in Katakana Hiragana is the most common script with cursive rather than block like letter forms Kanji characters are used to write root words Katakana is mostly used to represent foreign words words im
30. ative integer index into the C library __ctype table This is because the functions are actually macros that generate stubs of in line code which assume the argument is an integer and propagate the sign bit accordingly char ch isascii ch The code above could be written like this unsigned char ch isascii ch Watch for the use of unadorned chars Unfortunately they have probably been used extensively throughout most code It is therefore a non trivial task to change all char data to unsigned char especially as this might garner some lint or compiler warnings So char ch ch 0 0 is better written as unsigned char ch ch 0 0 On the other hand char cp while isspace cp is written as char cp while isspace unsigned char cp Although all this may sound like a lot of work in many cases existing code executes correctly in 8 bit mode without any changes to the code You are primarily on the lookout for lazy coding habits that assume ASCII is the only Developer s Guide to Internationalization August 1994 4 form of character encoding available When you fix problems they are usually easy to test using the Compose key of the Type 4 Type 5 PC AT101 and PC AT102 keyboard Note that the C compiler does not support 8 bit or multi byte characters in object names that is names of routines variables and so forth although it does allow you to initialize 8 bit or multi by
31. ats are employed throughout the world to represent date time currency numbers and units Also some alphabets have more letters than others and the order in which letters are sorted within national alphabets varies from language to language Programs that leave the format design and sorting order to the localization center in a particular country are considered level 2 compliant Level 3 Messages and Text Presentation User visible text in a level 3 compliant application should be easily translatable into the languages of various target markets User visible text includes help text error messages property sheets buttons text on icons and so forth A common way to provide message translation is to put text to be translated into a separate file with messages indexed either by string contents the POSIX UniForum method or by number the XPG 3 method Software using separate message files for messaging rather than encapsulating user messages in the binary is considered level 3 compliant Solaris provides two different but incompatible methods of translating text The first and preferred method is the POSIX and UniForum proposed standard using the gettext function Introduction 3 The second method of message translating is XPG 3 style message catalogs is the X Open Portability Guide XPG 3 provides the catgets function to obtain translated message strings from a message database The XPG message translation standard
32. d 37 Dinkins co Np IR C D o REV E DR ERE 37 Jexband assa cec e be ee es 37 Callsetlocale yu ua r uva 37 Make Software 8 bit Clean 38 Watch for Sign Extension Problems 39 Employ Standard Code Sets 41 Generating 1 41 Use ctype Library Routines 42 Avoid Managing the Keyboard 42 Formats and Collation 43 Time and Date Formats 43 Monetary Formats avec OR s oe ECC Ra eet Papae 44 Replace strcmp With strcoll 45 User Messages and Text Presentation 47 Localized Text Handling 48 Contents v Vi Where Do Messages Reside Using gettext e E ett Surround Strings with gettext Use Changing the Text Domain Create Separate Message Files Text Length and Height May Vary Try to Avoid Compound Messages Dynamic Other Lang g S ku Summary Of Requirements Ch
33. d Addresses Paper Sizes Mr Miss Mrs and Ms are common titles in the US but are not used in many other countries Address formats differ from country to country In many countries the postal code includes letters as well as numbers Within each country a small number of paper sizes are commonly used normally with one of those sizes being much more common than the others Most countries follow ISO Standard 216 Writing paper and certain classes of printed matter Trimmed sizes A and B series Formats and Conventions Overview 13 14 Summary Internationalized applications should not make assumptions about the page sizes available to them Solaris provides no support for tracking output page size this is the responsibility of the application program itself Table 2 5 Common International Page Sizes Paper Type Dimensions Countries ISO A4 21 0 cm 29 7cm Everywhere except US ISO A5 14 8 cm 21 0cm Everywhere except US JIS B4 25 9 cm by 36 65 cm JIS B5 18 36 cm 25 9cm Japan US Letter 8 5 inch by 11 inches US and Canada US Legal 8 5 inch by 14 inches US and Canada Standard paper trays distributed with LaserWriter and LaserWriter II support US letter US legal and A4 paper sizes The SPARCprinter s paper tray supports all these in addition to B5 International code does not contain any implicit cultural assumptions Instead it uses standardized interfaces that make use of
34. d by the three digit number in decimal is generated This feature is also derived from the DOS system Press the COMPOSE key sequence followed by 065 and an A appears on the screen 65 is the decimal value used by computers to store the uppercase letter A Press the COMPOSE key sequence followed by 136 and the letter appears If you type pemapkeys d all deadkeys and compose sequences are disabled Using the Floating Accent Keys Modems On some keyboards certain keys appear with an empty box Q underneath the diacritical mark These are referred to as floating accent keys They allow you to type in a composite character without using the Compose key Type the floating accent key first followed by the key for the letter to be accented Modems set to 8 bit no parity mode will work with 8 bit data Dumb Terminals Dumb terminals set to 8 bit space parity mode will work with 8 bit data although they are unlikely to display the proper characters Developer s Guide to Internationalization August 1994 Qo lll Printers Support in the SunOS system for native language printing includes Transmission of 8 bit characters by 1 The serial line must also be 8 bit clean and the printer must support the ISO Latin 1 character set for the characters to come out properly The SunOS 1p subsystem will spool PostScript files It can also translate the following formats to PostScript for spooling and printing troff to
35. d message identifiers can contain special characters and escape sequences as defined in the following table Description Symbol lt newline gt n lt tab gt t lt vertical tab gt v lt backspace gt b lt carriage return gt r lt form feed gt f lt backslash gt lt double quote gt octal bit pattern Nddd hex bit pattern xDD Translating Messages 67 68 The escape sequence ddd consists of a backslash followed by 1 2 or 3 octal digits specifying the value of the desired character The escape sequence NxDD is similar but introduces a hexadecimal value If the character following backslash is not specified in the table above the effect is undefined The Message Object File Running a file in portable message format through the msgfmt utility produces a binary version of the file which can be installed in the locale directory for example APPHOME 1ocale LANG LC MESSAGES This file can then be accessed by gettext at run time After you have saved the edited po file run the msg mt command 5 msgfmt demotext po The output from msgfmt is a message object file whose name is the domain name suffixed with mo The msgfmt utility generates one mo file for each text domain in the po file s In the above example executing msgfmt on the file messages po produces a binary file name messages mo Install this file into the appropriate LC MESSAGES sub directory renaming it to the
36. dims void xv get font FONT STRING DIMS string amp dims The dims structure now contains the string s width as its first element and the string s height as its second element XView application layout may be implicit or explicit Implicit layout relies on the toolkit to position objects Explicit layout means that the programmer specifies pixel values or row and column values both for position x y and dimension width height and gaps If you create a series of controls without specifying location information the toolkit will position items based on its built in spacing rules Implicit layout will very likely make localization easier since control objects will often be repositioned for you as required So if you have time try using implicit layout whenever possible DeskSet 3 3 tools use a custom library for implicit layout The XView toolkit provides a limited implicit layout facility Basically control area objects such as buttons are positioned based on default spacing rules However if button width increases so that the right most button is outside the control area the toolkit wraps that button to the next line rather than expanding the control area DeskSet 3 3 tools such as Mail Tool and Calendar Manager re layout gracefully because developers have added code to properly center and size objects based on current window size as well as font size 60 Developer s Guide to Internationalization August 1994 O1 II
37. e are about 7000 commonly used Hanzi characters although emerging standards number Hanzi Developer s Guide to Internationalization August 1994 2 in the tens of thousands In the Republic of China ROC or Taiwan current standards require more than 13 000 characters 6000 others have been recently standardized but are considered rare If a character is not a root character it usually consists of two or more parts two being most common In two part characters one part generally represents meaning and the other represents pronunciation Occasionally both parts represent meaning The radical is the most important element and characters are traditionally arranged by radical of which there are several hundred The same sound can be represented by many different characters which are not interchangeable in usage Some characters are more appropriate than others in a given context the appropriate one is distinguished phonetically by the use of tones By contrast spoken Japanese and Korean lack tones There are several phonetic systems for representing Chinese In mainland China the most common is pinyin which uses Roman characters and is widely employed in the west for place names such as Beijing The Wade Giles system is an older phonetic system formerly used for place names such as Peking In Taiwan zhuyin or bopomofo an extensive phonetic alphabet with unique letter forms is often used instead Commercial applications part
38. e local short form of date and time or C for the local long form Also x produces the local date form numeric and X yields the local time form If you try out the above program your results will look something like this setenv LC_TIME de Montag 16 Marz 1992 19 19 19 Uhr PST setenv LC TIME fr a out lundi 16 mars 1992 19 19 20 PST Unfortunately many often used combinations of date and time are missing from the standard Neither short nor long form of the local date is available and there is no abbreviation for time without seconds or time zone Monetary Formats Use localeconv 3 function to obtain currency formats It reads formatting conventions of the current locale to populate an 1conv structure then returns a pointer to the filled in object Unfortunately this gives you a data structure but not a string A library routine is needed that converts a floating point number to a string containing the appropriate monetary format Standards committees are working on this Until they agree on a solution here is a highly simplified method for printing currency in a locale independent manner include lt locale h gt include libintl h include libintl h include lt stdio h gt char pcurrency amount double amount char string 512 Developer s Guide to Internationalization August 1994 HS II struct lconv lconv localeconv lconv localeconv sprintf string s s 2f s
39. e message file to create a mo message object file This is a data file containing a binary message tree for quick machine access When a program calls gettext or dgettext the message routines find the appropriate translated string Your message object file should be called domain mo where domain is the name of the text domain you selected 4 Install the message object file domain mo into the LC MI ESSAGI ES directory This might be under usr lib locale LANG but more likely will be the directory specified by your program s bindtextdom ain call 65 66 The Portable Message File First run the xgettext command to extract gettext strings xgettext demotext demotext c The output from xgettext is a portable message suffix po file for each domain specified For the simple Hello world example presented in Chapter 4 you would execute xgettext on the source code demotext resulting an output file called messages po In this example messages is the default domain name This file would contain msgid Hello World Nn msgstr msgid Goodbye Nn msgstr A translator provides the translations by editing this file appending the target language text to the msgstr lines If the target language were French the portable object file might then look like this msgid Hello World Nn msgstr Bonjour tout le msgid Goodbye Nn msg
40. e resource database db file for the Spanish locale resides in home locale spanish app defaults here s how to modify the code Internationalizing Open Windows 61 62 bindtextdomain my_app_strings home locale xv_init XV_INIT_ARGC_PTR_ARGV amp argc argv XV_LOCALE_DIR home locale XV_USE_LOCALE TRUE 0 frame xv_create XV_NULL FRAME XV INSTANCE NAME base frame XV NULL control xv create frame PANEL XV INSTANCE NAME controll XV NULL button create control PANEL BUTTON XV INSTANCE NAME buttonl PANEL LABEL STRING dgettext my app strings Quit USE DB XV X 10 XV Y 10 XV NULL XV NULL When the application runs it builds a path using the directory pointed to by XV LOCALE DIR the name of the current locale and app defaults This path is used to search for the resource database db file Any attribute specified in the XV USE DB attribute list whose parent is given a name with XV INSTANCE NAME will have its valued looked up in the file If the attribute is found the value in the database is used if the attribute isn t found the value specified in the source code is used For the example above the entries in the resource database file my app db would look like this my app base frame controll buttonl xv x 10 my app base frame
41. ecklists for Internationalization Using X Open Message 5 Internationalizing OpenWindows Window Object Layout Object Size and Input Method aa Ute la MO To a uma Font Handling 5053 cte te bets C etre 6 Translating Creating Message File The Portable Message File The Message Object A Language and Territory Names Language and Territory Developer s Guide to Internationalization August 1994 48 49 50 51 51 52 52 54 55 55 56 57 59 60 61 63 64 65 65 66 68 69 69 Standard Locale Names 69 B Compose Key Sequences 71 uu 77 Contents Vil viii Developer s Guide to Internationalization August 1994 Tables Table 2 1 Table 2 2 Table 2 3 Table 2 4 Table 2 5 Table 3 1 Table 3 2 Table 3 3 Table 3 4 Table 3 5 Table 4 1 Table A 1 Table B 1 Table B 2 Table 6 1 International Time Formats International Date Formats International Numeric Conventions International Monetary
42. eparators The United Kingdom and the United States are two of the few places in the world that use a period to indicate the decimal place Many other countries use a comma instead The decimal separator is also called the radix character Likewise while the UK and US use a comma to separate thousands groups many other countries use a period for this instead and some countries separate thousands groups with a thin space This table shows some commonly used numeric formats Table 2 3 International Numeric Conventions Locale Large Number Canadian French 4 294 967 295 00 Canadian English 4 294 967 295 00 Danish 4 294 967 295 00 Finnish 4 294 967 295 00 French 4 294 967 295 00 German 4 294 967 295 00 Italian 4 294 967 295 00 Norwegian 4 294 967 295 00 Spanish 4 294 967 295 00 Swedish 4 294 967 295 00 UK English 4 294 967 295 00 US English 4 294 967 295 00 Data files containing locale specific formats will be misinterpreted when transferred to a system in a different locale For example a file containing numbers in a French format will not be useful to a UK specific program Formats and Conventions Overview 7 List Separators There are no particular locale conventions that specify how to separate numbers in a list They are sometimes comma delimited in the UK and the US but often spaces and semicolons are used Certainly international software never hardwires any number delimiter Currenc
43. erted exclamation c cents 1 pounds sterling o x a currency symbol y y yen broken 5 5 section 71 72 Table B 1 Compose Key Sequences Compose Result Description x ji umlaut diaeresis o copyright a E feminine ordinal left guillemet not sign soft hyphen r o registered a B macron A 0 degree plus minus 2 2 superscript 2 A 3 sup 3 superscript 3 N N prime acute accent u mu mu micro P q pilcro paragraph middle dot cedilla A 1 sup 1 superscript 1 x S masculine ordinal gt gt right guillemet 1 4 1 4 quarter 1 2 1 2 half 3 4 3 4 three quarters 2 inverted question A A grave A 7 A acute A s A A circumflex Developer s Guide to Internationalization August 1994 II Table B 1 Compose Key Sequences Compose Result Description A A A tilde A E umlaut angstrom ligature C C cedilla E E E grave E E acute E A E circumflex E d E umlaut I 2 I I grave I I I acute I I circumflex I I I umlaut D D Eth N N N tilde O grave O acute A O circumflex O tilde O O umlaut x x x multiply O slash U U grave U U acute U A U circumflex U U U umlaut 7 X Y acute Compose Key Sequences 73
44. et server XV LC DISPLAY LANG NULL Additionally there are two new text functions to handle wide characters XwcDrawString is the wide character equivalent of XDrawString and XwcTextWidth is the wide character equivalent of XTextWidth Note Earlier releases of OpenWindows displayed ISO Latin 1 characters by default even if the LC CTYPE environment variable was not set Because this release of OpenWindows supports multibyte characters and EUC Latin 1 characters not displayed unless the CTYPE environment variable is set to iso 8859 1 Developer s Guide to Internationalization August 1994 Translating Messages O This chapter explains how to create and install message catalogs for your internationalized application Other aspects of localization such as formatting and collation are provided to you along with any SunOS localization packages you have installed Your translated user messages together with a localization package allow your internationalized application to function properly for a given locale Creating a Message File Follow these steps to create a message catalog 1 Use xgettext 1 to generate a po portable message file This is a readable text file intended for editing 2 Translate each msgid string in the portable message file placing the translation on the msgstr line directly below it 3 Run the msgfmt 1 program on the portabl
45. functions for wide characters and all the str functions by ws functions for wide character string operations When making applications multi byte capable you should increase buffer size four fold in order to preserve efficiency since wchar_t is four bytes long Table 3 5 International Library Routines Library Routine Description Locale Management setlocale set or query language or locale nl langinfo obtain various language or locale information Character Type isalnum is letter or digit isalpha is a letter isascii is 7 bit ASCII character iscntrl is control code isdigit is a digit isgraph is visible islower is lower case isprint is printable ispunct is a punctuation mark Support for Internationalization 29 30 Table 3 5 International Library Routines Library Routine isspace isupper isxdigit toascii tolower toupper String Collation strcoll strxfrm Date and Time strftime Formatted Output printf fprintf sprintf Formatted Input scanf fscanf sscanf Monetary Format localeconv SunOS Messaging bindtextdomain textdomain gettext Description is white space is upper case is a hexadecimal digit convert to ASCII convert to lower case convert to upper case compare two strings transform string for comparison convert date and time to string print formatted string format stri
46. ha isprint and so on Software should have been using these functions all along as they were always needed for portability to IBM s EBCDIC codeset SunOS also provides wide character equivalents such as iswalpha and iswprint Fix code that assumes characters fall in the range 0 127 by extending the range of such tables static int hashtable 127 bad practice For example the above declaration would be better coded as follows include lt limits h gt static int hashtable UCHAR_MAX UCHAR MAX is defined lt limits h gt on all ANSI C conforming systems Watch for Sign Extension Problems One issue that is sometimes invisible to the programmer is the way the C compilers default to using signed for all fundamental data types This can sometimes cause substantial problems in both application and library code Code that casts char to other lengths may be dirty Because the char data type is signed in SunOS when a char variable holds an 8 bit character that has the most significant bit set sign extension takes place during assignment Needless to say a negative integer might cause problems later on Writing Internationalized Code 39 40 ris char c 0xa0 1 i is now negative Do not pass raw characters to functions that require short int or long arguments This is bad practice because of the sign extension problem For example the following code is incorrect as it produces a neg
47. heir message strings Note Sun recommends that programmers use gettext instead even though it is not an official standard SunOS and OpenWindows 3 3 both use gettext exclusively If source portability is a concern be advised that gettext source is freely available for integration into your product Writing Internationalized Code 57 58 Developer s Guide to Internationalization August 1994 Internationalizing OpenWindows D For the first time in Solaris Asian OpenWindows has been integrated into mainstream OpenWindows 3 3 As a result international software no longer requires separate releases for Asia it can be released as a single binary Both the XView and OLIT toolkits support level 3 and level 4 internationalization providing both messaging and multibyte characters For information on international XView see the internationalization portions of the XView Developer s Notes For information on international OLIT see the internationalization chapter of the OLIT Reference Manual Note that 8 bit characters for western European languages no longer appear by default LC CTYPE must be set to iso 8859 1 first East Asian languages all require their own Feature Package which includes appropriate font sets input manager server and translated messages You may want to develop applications with OpenWindows Developer s Guide 3 3 a window tool designed specifically for building user interfaces Usually De
48. icates constructing homogenous character arrays for example To assist convenient internal processing SunOS provides a wide character format plus a collection of library functions for operating on wide characters Additional functions are available to convert from EUC format to wide characters and from wide characters back to EUC format Support for Internationalization 27 28 Wide characters are the ANSI C data type wchar_t defined in SunOS as typedef long code sets with one two or three bytes get mapped to wide characters as shown below Four bytes are enough to represent the entire Chinese character set defined by the Chinese National Standard CNS 11643 86 with space left over for user defined characters Table 3 4 EUC and Wide Character Representation Code set EUC Representation Wide Character Representation 0 00000000 00000000 00000000 1 1 1 00110000 00000000 00 2 552 1 1 12222222 00010000 000 2222222 3 553 1 1 1zzzzzzz 00100000 000 xyyyyyyy yzzzzzzz Wide characters provide a standard character size and are useful for indexing interprocess communication memory management and other tasks that use character counts and known array sizes Wide characters are stateless and unambiguous within a given locale Note that with the S
49. icularly those that deal with people s names need to consider the impact of code set expansion Many people in the ROC have names containing characters that do not exist in any standard code set Space needs to be provided in unassigned code sets to deal with this issue Codesets for x86 The default codeset on Solaris for x86 is ISO 8859 1 IBM DOS 437 codeset is provided as an option in text mode however it is only provided at internationalization level 1 That is if you choose to download IBM DOS 437 codeset by typing loadfont c 437 pcmapkeys usr share lib keyboards 437 en US Formats and Conventions Overview 11 Keyboard Differences Other Differences 12 Punctuation Symbols there will be no support for non standard U S date time currency numbers units and collation There will be no support for non English message and text presentation and no multi byte character support Therefore non Windows users should only use IBM DOS 437 codeset in the default C locale You must be in the text mode to download the IBM codeset not the graphics mode If you are not using the standard U S PC keyboard replace en_US with the keyboard map related to your keyboard To download the default codeset in text mode type loadfont c 8859 pcmapkeys usr share lib keyboards 8859 en US See the loadfont 1 and pcmapkeys 1 manual pages Not all characters on the US keyboard appear on other keybo
50. ide to Internationalization August 1994 4 By default the pathname is usr lib locale although this be changed with bindtextdomain SLANG comes from the user s environment and domain is supplied by textdomain Changing the Text Domain The following two examples retrieve the same strings but have different effects on the text domain The first example does not change the current text domain The second example changes the current text domain to library_error_strings then retrieves the alternate language string of wrongbutton message dgettext library error strings wrongbutton or textdomain library error strings message gettext wrongbutton After writing an application program create a text domain by extracting gettext strings and placing them in a file with the alternate language equivalent The following section demonstrates this process Create Separate Message Files Once you have enclosed all user visible strings inside gettext wrappers you can run the xgettext command on your C source files to create a message file This produces a readable file the portable object for editing by translators Running msgfmt on this file produces a binary mo file the message object which should be installed under the LC MESSAGES directory Here s a sample interaction on demotext c xgettext m TRNSLT demotext c cat messages po domain demotext msgid Hello world m
51. installed localization packages Read Chapters 4 and 5 to learn what these standard interfaces are and how you can use them to create internationalized products Developer s Guide to Internationalization August 1994 Supportfor Internationalization 3 This chapter describes features of Solaris that provide the foundation for internationalization support Chapter 4 gives examples of international coding practices and Chapter 5 discusses window system specifics Here are some of the ways that the SunOS operating system supports international software applications Data paths are 8 bit clean in order to support ISO 8859 code sets and so that multi byte characters can survive intact Keyboard drivers and mapping tables are provided for a variety of code sets This allows software to cope with many European and Asian languages Printers modems and terminals are supported that can handle ASCII ISO Latin 1 and EUC code sets System locales are included for French German Italian Swedish and Japanese offering level 2 internationalization for these locales Standard C library routines provide support for writing software that can be easily localized Wide character library routines provide programming support for Asian language applications Applications can use either the X Open or the gettext message system for access to text that needs translation Note that only dynamically linked libraries provide international support
52. intf statements with calls to gettext a routine that searches the message database for key strings returning the corresponding translated string if it can locate one Make sure to compile with cc demotext c lintl o demotext Surround Strings with gettext Fortunately gettext is much easier to use than catgets All you really have to do is go through your programs enclosing literal strings inside gettext calls Here s an example of an error message before printf s Too few arguments Nn argv 0 and after printf gettext s Too few arguments n argv 0 Some rearrangement is required here but not much Library products should use dgettext in place of gettext since calling sequence cannot be guaranteed and different domains may be mixed together at random The library developer chooses the domain name printf dgettext xview Cannot find font in The call is equivalent to combined textdomain and gettext calls Writing Internationalized Code 49 50 In general you only need to message strings that users see Do not message strings containing system commands or file names such as sort or dev tty Be careful when messaging strings inside sprintf which is often used to build up path names or command lines You probably don t need to message strings used only for debugging Because integers and decimal numbers are not strings they don t need messaging either Initialized
53. milarly on a terminal The only difference is that when is pressed nothing happens until e is pressed after which the character appears on the screen Developer s Guide to Internationalization August 1994 3 In text mode a utility that can be used to assign deadkeys pcmapkeys is supplied This utility is used to do everything discussed in this section To define as a deadkey and try the other examples listed below type the command pemapkeys f usr share lib keyboards dead circumflex In OpenWindows the utility Xmodmap may be used to remap keys See the Xmodmap 1 manual page for additional information Now when you press nothing appears on the screen When an e is typed next the letter appears To use the character alone press first and then the spacebar If a sequence of two characters is typed that does not make sense no character is sent to the application that is currently being used and the machine beeps to indicate that an erroneous combination was typed Using the Compose Key SPARC and x86 Compose Key on SPARC The regular SPARC Type 5 or Type 4 keyboard can produce all characters in the standard ISO 8859 1 Latin 1 code set by means of the Compose key Such characters are typically composite characters that include diacritical marks To produce a composite character on the US English keyboard first press the Compose key Next press the key for the desired diacritical mark and then the key f
54. monic form This is the rationale for the messaging interface using gettext which is based on a proposal made by the UniForum Technical Subcommittee on Internationalization This extension can be run in conjunction with the existing schemes or can be used Writing Internationalized Code 47 as the sole technique for messaging SunOS application Applications that call gettext must include the header file 1ibintl h and compile with the 11061 linker option Localized Text Handling When creating international applications developers usually write text strings error messages text for buttons and menus and so forth in their native language for later translation into other languages SunOS lets you define any language as the native language and any other language as the alternate language The steps to localize text handling are 1 Verify that source code uses textdomain and gettext or else dgettext These functions accept native language strings as arguments and return equivalent foreign language strings 2 Extract native language text strings from the gettext functions and store them with their foreign language equivalents in a portable message file This extraction can be done by hand or with the xgettext program Where Do Messages Reside Under SunOS system messages for libraries and utilities reside in usr lib locale language LC MESSAGES domain mo where language is the specific language fr fo
55. n 33 II Qo Naming Rules What Is a Locale 34 In SunOS the following objects must be composed of ASCII characters User name group name and passwords System name Names of printers and special devices Names of terminals dev tty Process ID numbers Message queues semaphores and shared memory labels The following may be composed of ISO Latin 1 or EUC characters File names Directory names Command names Shell variables and environment variable names Mount points for file systems NIS key names and domain names The names of NFS shared files should be composed of ASCII characters Although files and directories may have names and contents composed of characters from supplementary code sets using only the ASCII code set allows NFS mounting across any machine regardless of localization The key concept for application programs is that of a program s locale The locale is an explicit model and definition of a native language environment The notion of a locale is explicitly defined and included in the library definitions of the ANSI C Language standard The locale consists of a number of categories for which there are language dependent formatting or other specifications A program s locale defines its code sets date and time formatting conventions monetary conventions decimal formatting conventions and collation order A locale name is comprised of language territory and possibly code set although te
56. ng from right get length of wide char string return display width of wide char string Support for Internationalization 31 32 Table 3 5 International Library Routines Library Routine wsspn wscspn wspbrk wstok Wide Formatting wsprintf wsscanf Wide Numbers wstol wstod Wide Strings wscasecmp wsncasecmp wscoll wsxfrm Wide Standard I O fgetwc getwchar fgetws getws fputwc putwchar fputws putws ungetwc Wide Ctype Description return span of one wide char string in another return span of one wide char string not in another return pointer to one wide char string in another move token through wide char string generate wide char string according to format interpret wide char string according to format convert wide char string to long integer convert wide char string to double precision compare wide char strings ignore case differences compare wide char strings to length n ignore case collate wide char strings transform wide char string for comparison get multi byte char from stream convert to wide char get multi byte char from stdin convert to wide char get multi byte string from stream convert to wide char get multi byte string from stdin convert to wide char convert wide char to multi byte char put to stream convert wide char to multi byte char put to stdin convert wide char to multi byte string put to stream conve
57. ng to file stream format string in memory scan formatted string scan string from file stream scan string in memory returns structure containing monetary format associate path name with message domain open message catalog domain get message from catalog Developer s Guide to Internationalization August 1994 Qo II Table 3 5 International Library Routines Library Routine dgettext X Open Messaging catopen catgets catclose Regular Expressions regexpr 3G Multi byte Handling mblen mbtowc wctomb mbstowcs wcstombs Wide Characters wscat wsncat wsdup wscmp wsncmp wscpy wsncpy wschr wsrchr wslen wscol Description get message from catalog domain open message catalog X Open get message from catalog X Open close message catalog X Open regular expression handler is EUC multibyte capable but no wide character interface has been provided get length of multi byte character multi byte to wide character wide character to multi byte character multi byte string to wide character string wide character string to multi byte string concatenate wide char strings concatenate wide char strings to length n duplicate wide char string compare wide char strings compare wide char strings to length copy wide char strings copy wide char strings to length find character in wide char string find character in wide char stri
58. of keywords is hard coded into a program code needs to be changed before messages can be successfully translated Obviously this defeats the purpose of internationalization X Open defines an extension to the printf family that permits changing the order of parameter insertion SunOS also supports this extension For example the conversion format 1 s inserts parameter one as a string while 255 inserts parameter two The entire format string is parameter zero Here s a small example of how these extensions can be used This printf statement has position dependent keywords because the verb must come before the object poor practice position dependent keywords printf Unable to s the s n lock flg 2 lock find type flg page record This could produce any of four messages in English Unable to lock the page Unable to find the page Unable to lock the record Unable to find the record Here are those four messages translated into German Note that the auxiliary verb must follow not precede the object Das Programm kann die Seite nicht sperren Das Programm kann die Seite nicht finden Das Programm kann den Rekord nicht sperren Das Programm kann den Rekord nicht finden German syntax requires different word order so the program s keywords must be reversed Here is that printf statement properly written for dynamic messaging printf gettext Unable to s the s n lock flg gettext
59. on t Split user message strings write out whole messages instead Embed graphics in code Use jargon abbreviations and acronyms Make assumptions about word order in strings Assume numbers of bytes or characters Test for alphabetic characters by comparing specific characters A Z for instance Hard code the decimal separator in parsing or arithmetic calculations Hard code or limit the size of currency fields Assume the size of paper on which application output will be printed Use single letter commands to represent the first letter of commands such as p for print Make assumptions about what might be in particular byte locations Developer s Guide to Internationalization August 1994 HS II Reference global external variables used by system libraries such as char _ _1 1 unsigned char _ctype unsigned char _numeric char time struct wctype wcptr int lflag Use both catgets and gettext Using X Open Message Catalogs Although catgets is the standard X Open function its programming interface is not as convenient as gettext You must pass catgets two extra numeric parameters one for the catalog descriptor and another for the message set number Moreover catgets requires you to assign a numeric message identifier to each message which can make code confusing and difficult to maintain Message numbers can easily get out of sync with t
60. or the desired alphabetical character You may type the diacritical mark and the alphabetical character in either order For example to produce press the Compose key then type a and order doesn t matter When testing software make sure to try all these combinations on your keyboard If your software manages the keyboard directly you should also try all special keys on the 13 European SPARC keyboards and perhaps the Asian SPARC keyboards as well Compose Key on x86 On x86 systems the default COMPOSE key sequence for Solaris for x86 is CTRL SHIFT F1 Many MS DOS DOS users will be familiar with it When in COMPOSE mode the system expects two more characters to be typed by the user to generate a character Press CTRL SHIFT F1 followed by n to produce Support for Internationalization 19 20 the Spanish the n in ma ana on the screen If you press the COMPOSE key sequence followed by pressing twice an inverted exclamation sign appears on the screen In text mode both the value of the COMPOSE key and the list of COMPOSE key sequences and the characters they generate can be specified in a file that is then processed by the pcmapkeys command see 1 In text mode the following tables are only valid for the ISO 8859 1 codeset and not for the optional IBM DOS 437 codeset In OpenWindows these tables are valid because OpenWindows only supports the ISO 8859 1 codeset Compose Key Sequences He
61. ported from languages other than Japanese There are tens of thousands of Kanji characters but the number commonly used has been declining steadily over the years Now only about 3500 are frequently used although the average Japanese writer has a vocabulary of merely 2000 Kanji characters Nonetheless computer systems must support more than 7000 because that is what the Japan Industry Standard JIS requires In addition there are about 170 Hiragana and Katakana characters On average 55 of Japanese text is Hiragana 35 Kanji and 10 Katakana Arabic numerals and Roman letters are also present in Japanese text Although it is possible to avoid the use of Kanji completely most Japanese readers find text containing Kanji easier to understand Korean Korean is similar to Japanese in that Chinese based ideograms called Hanja are mixed together with a phonetic alphabet Hangul Hanja is used mostly to avoid confusion when Hangul would be ambiguous Hangul characters are formed by combining ten basic vowels and fourteen consonants two to five of which compose one syllable Hangul characters are often arranged in a square like the four on a pair of dice so that the group takes up the same space as a Hanja character Korean requires over 6000 Hanja characters plus about 96 Hangul characters Chinese Chinese usually consists entirely of characters from the ideographic script called Hanzi In the People s Republic of China PRO ther
62. proper textdomain as necessary Developer s Guide to Internationalization August 1994 Language and Territory Names A Language and Territory An internationalized program makes no assumptions about the language and format of text it is designed to handle It must work equally well in any locale for data generated internally text read from or written to files and messages presented to the user To determine locale specific conventions at run time programs query system databases installed in usr 1lib locale for cultural data Applications should identify the proper locale using ISO standard names presented below Standard Locale Names A locale name is always composed of a language name sometimes with a territory name appended SunSoft s list of accepted names for languages and territories follows below Language names all lower case came from the ISO 639 standard Territory names all upper case came from the ISO 3166 standard Language is separated from territory by an underscore and territory is optional Some languages have longer more mnemonic names For example JLE accepts the locale name japanese KLE accepts the locale name korean HLE accepts the locale name t chinese and CLE accepts the locale name chinese Solaris 2 3 continues to support these long names 69 70 Table A 1 Language and Territory Names Locale Language Territory Locale Language Territory Default C locale
63. r French for example and domain is the specific text domain for that application Message files should usually be installed in the same directory hierarchy as the application software Using SunOS applications can associate a directory with a message domain by calling bindtextdomain Although the message file need not be named after its application maintenance is easier when the names are similar The catopen and textdomain library routines actually open the message file If a message file is missing users get the untranslated English or Unixese message which is a string contained in the catgets or gettext call Developer s Guide to Internationalization August 1994 HS II Using gettext It is recommended that programmers use gettext when writing applications for Solaris even though gettext is not an official standard SunOS and OpenWindows 3 3 both use gettext exclusively Here is a short program demotext that uses both the gettext and catgets functions include lt locale h gt include libintl h include lt stdio h gt main demotext c setlocale LC_ALL textdomain demotext printf Ss n gettext Hello world printf Ss n gettext Goodbye The second line sets the message domain with text domain specifying the message domain which is identical to the message file name The other lines merely surround literal strings in pr
64. r codes based on these events The SunOS translation tables are in usr share lib keyboards OpenWindows keyboard tables are OPENWINHOME etc keytables If at all possible use the operating system or window system s keyboard translation tables Applications that read events directly from the keyboard must perform their own key mappings which isn t easy Moreover such applications need to be revised every time new keyboards are devised Generating Characters Not ona U S Keyboard Although non English characters like the German 4 or the French are not present on a keyboard designed for use in American English most of these characters can be generated This allows users to write French letters on American systems for example There are three ways to generate characters for which there are no keycaps explicit symbols on the keyboard Deadkeys x86 systems only Compose sequences SPARC and x86 systems The decimal representation of the character x86 systems only Deadkeys x86 The deadkey was invented by typewriter manufacturers For example imagine you need the French character A French typewriter does not have a key for this character but it has keys for both and When the key is pressed a circumflex is printed but the typewriter carriage does not move When the e key is then pressed the letter e is printed on the same spot as the circumflex and an is formed This technique works very si
65. re is a table shows how to produce special ISO Latin 1 characters using Compose key sequences Table 3 3 Compose Key Sequences Compose Key Sequence Result Description space space no break space i inverted exclamation cents 1 pounds sterling a currency symbol y broken bar s 8 section umlaut diaeresis c copyright a 6 feminine ordinal lt lt lt left guillemet 5 not sign soft hyphen r o registered Developer s Guide to Internationalization August 1994 Qo II Table 3 3 Compose Key Sequences Compose Key Sequence Result Description d M macron e 0 j degree plus minus 2 2 superscript 2 3 3 superscript 3 N prime acute accent u u mu micro P q pilcro paragraph 7 5 cedilla A middle dot A 1 1 superscript 1 o masculine ordinal gt gt right guillemet 1 4 y quarter 1 2 15 half 3 4 3 4 three quarters inverted question A A grave A A acute A A circumflex A A A tilde A 8 umlaut T angstrom ligature C C cedilla E E grave Support for Internationalization 21 22 Table 3 3 Compose Key Sequences Compose Key Sequence Result Description E i E acute E D E circumflex E E umlaut I i 1 I grave I I acute I A i I circumflex I u I umlaut D D Eth N N N tilde O grave
66. rritory is dropped when not needed Code set is usually assumed For example German is de an abbreviation for Deutsch while Swiss German is de CH CH being an abbreviation for Confoederatio Helvetica See Appendix A for a list of accepted locale names Developer s Guide to Internationalization August 1994 3 Generally the locale name is specified by the LANG environment variable Locale categories are subordinate to LANG but may be set separately in which case they override LANG If LC_ALL is set it overrides not only LANG but all the separate locale categories as well Locale Categories The locale categories are as follows LC_CTYPE A directory whose files control the behavior of character handling functions The LC_CTYPE ctype file specifies character types for lt ctype h gt LC_TIME A readable file that specifies date and time formats including month names days of the week and common full and abbreviated representations LC_MONETARY A binary file that specifies monetary formats Very few SunOS commands or library routines actually use this database LC_NUMERIC A tiny file that specifies the decimal separator or radix character and the thousands separator LC_COLLATE A directory containing files that specify sorting order for a locale and string conversions required to attain this ordering LC_MESSAGES A directory containing message catalogs user message transla
67. rt wide char to multi byte string put to stdin push a wide char back into input stream Developer s Guide to Internationalization August 1994 Qo II Table 3 5 International Library Routines Library Routine iswalpha iswupper iswlower iswdigit iswxdigit iswalnum iswspace iswpunct iswprint iswgraph iswcntrl iswascii isphonogram isideogram isenglish isnumber isspecial towupper towlower Codeset Info getwidth euclen euccol eucscol csetlen csetcol Description is wide character letter is wide character upper case is wide character lower case is wide character digit is wide character hex digit is wide character alphanumeric is wide character white space is wide character punctuation is wide character printable is wide character graphic is wide character control is wide character ASCII is wide character phonogram is wide character ideogram is wide char in English alphabet from sup code set is wide character digit from supplementary code set is special wide character from sup code set convert wide character to upper case convert wide character to lower case get code set information on EUC and screen width get EUC byte length get EUC character display width get EUC string display width return number of bytes for an EUC code set return columns needed to display EUC code set Support for Internationalizatio
68. sNn lconv gt p_cs_precedes lconv gt currency_symbol lconv p sep by space amount lconv p sep by space lconv p cs precedes lconv currency symbol return string main double amount void setlocale LC ALL scanf 1f amp amount printf 5 pcurrency amount Replace st romp With 11 Alphabetic ordering varies from one language to another For example in Spanish immediately follows n and digraphs ch and ll immediately follow c and 1 respectively In German the ligature f is collated as if it were ss Swedish has additional unique characters following z Danish and Norwegian have additional characters following z The traditional library routine for comparing strings st rcmp remains unchanged Because it uses ASCII order strcmp places after Z even in English This ordering is often unacceptable By contrast the new library routines strcoll and strxfrm can produce any sort order you want Use strcoll to compare strings or strxfrm to transform strings to ones that collate correctly Fortunately strcoll takes the same parameters and returns the same values as strcmp Unfortunately strcoll does a lot more work and is consequently slower To speed up applications that compare strings frequently use strxfrm to store transformed strings into arrays that collate more efficiently Wri
69. sgstr TRNSLT Hello world msgid Goodbye msgstr TRNSLT Goodbye msgfmt demotext po su Password mv demotext mo usr lib locale test LC MESSAGES Writing Internationalized Code 51 52 In the portable object the msgstr keyword indicates the search string passed to gettext while the msgid keyword indicates the translated string Use the m flag with xgettext to produce test messages by prepending some obvious string If gettext can t find an appropriate string in the message catalog it returns the string you passed as its first parameter This means code won t fail if message files are missing since by default they get indexed source strings However you should always deliver message files with an application because translators may not have access to your source code If you anticipate difficulty translating msgid insert a comment or hand edit msgstr to make things clear See Chapter 6 for more information on creating message files Text Length and Height May Vary Be aware that translated messages may be of different length and height than the original messages Translation into certain languages such as German often produces longer messages than in English Some language translations may yield shorter messages East Asian language ideographs are usually taller than Roman characters Window system resource files OpenWindows 3 3 specify height and width of panel bu
70. str Au revoir Nn Lines beginning with an sharp are comments Empty lines are ignored Statement lines contain a directive and value A directive always starts at the beginning of line and is separated from its value by white space If your application program calls textdomain or dgettext to set the message domain then xgettext produces a textdomain mo file for each text domain in the source code The following statements are recognized Developer s Guide to Internationalization August 1994 II domain domainname States that all following target strings until another domain directive are contained in domainname which can be any string up to 255 bytes containing any character in the portable character set but excluding slash Until the first domain directive all target strings belong to the default domain messages msgid message_identifier Specifies the value of the message identifier associated with the following msgstr directive The message_identifier string identifies a target string at retrieval time Usually this is the untranslated string msgstr message_string Specifies the target string associated with the message_identifier string declared in the immediately preceding msgid directive Every statement containing a msgstr directive must be immediately preceded by a msgid statement Usually message_string is the translated message Any other form of input line syntax is illegal Message strings an
71. t count printf s gt 1 treeprint p gt right User Messages and Text Presentation One of the most critical tasks in software internationalization is providing messages that can be translated easily Messages are what users see first help text button labels menu choices usage summaries error diagnostics and so forth The ease of message localization can vary greatly In a well designed application non technical people can translate message files into their native languages In a non international application engineers fluent in a language must translate every string inside a program then recompile the code There should be no explicit strings in an international application except those passed to gettext Two similar but incompatible methods for international messaging in SunOS catgets from the X Open standard and gettext from the POSIX 1b and UniForum proposals Both routines provide an interface to message catalogs text databases that are easy to compose translate and access Because the contents of a message catalog are separate from application code text can be selected by locale at run time without altering the code itself SunOS supports the existing SVR4 messaging schemes This includes both the X Open XPG3 message catalog scheme using catgets and the SVR4 private scheme using gett xt However both schemes are inflexible when it comes to handling messages identified by mne
72. te data in strings Employ Standard Code Sets In certain locales OpenWindows 3 3 provides support for the ISO 8859 1 standard codeset also known as ISO Latin 1 Test your software by typing all the Compose sequences in Table 3 1 If your software can display all characters in ISO Latin 1 it can also display all characters on European keyboards For Asian locales use code sets supported in the Asian Feature Sets Generating PostScript Code that generates PostScript must use the nnn octal form for characters above hexadecimal 7F since the PostScript interpreter cannot read characters with the eighth bit set If you are using TranScript 2 1 1 you can obtain the encoding for ISO Latin 1 by using the name ISOLatinlEncoding to re order the font encoding Note that ISOLatinlEncoding is available in TranScript 2 1 1 but not in previous releases of TranScript PostScript Level Two interpreters have the name ISOLatinlEncoding asa defined name See the PostScript Language Reference Manual 2nd edition Writing Internationalized Code 41 Use ct ype Library Routines As mentioned above text processing software must avoid hard coded character ranges Upper and lower case letters punctuation marks numeric digits and spaces should be defined using library routines under lt ctype h gt rather than with hard coded character ranges Table 4 1 Library Routines for Codeset Independence Routine Character is a isalpha c
73. ting Internationalized Code 45 46 This program reads standard input builds a binary tree in the correct order using strcoll to compare strings then prints out the binary tree This code may be used for tasks such as listing files in a subwindow include lt locale h gt include lt stdio h gt include lt string h gt struct tnode node of binary tree char line int count struct tnode left right main collate sort list of lines using strcoll struct tnode root tree char line BUFSIZ root NULL void setlocale LC ALL while fgets line BUFSIZ stdin root tree root line treeprint root struct tnode tree p line install line at or below tr pointer struct tnode p char line char cp malloc strcpy int cond if p NULL struct tnode malloc sizeof struct tnode if cp malloc strlen line 1 NULL strcpy cp line p gt line cp p gt count 1 p gt left p gt right NULL else if cond strcoll line p gt line 0 p gt count else if lt 0 p gt left tr p gt 1eft line else gt 0 p gt right tree p gt right line return p Developer s Guide to Internationalization August 1994 Hx II treeprint p print tree recursively starting at p struct tnode p if p NULL treeprint p gt left while p g
74. tions This locale directory would be empty until a localization package containing system message translations is installed Note that many application packages would have their own separate LC_MESSAGES directories All of these locale categories with the exception of LCI in both the X Open and ANSI standards LC_MI ESSAGES are defined ESSAGI ES is Sun specific Support for Internationalization 35 36 Developer s Guide to Internationalization August 1994 Linking Text and Code sets Writing Internationalized Code 4 This chapter describes some specific steps that you should take to internationalize applications The material is divided into four main topics text and code sets formatting and collation user messages and nonglobal locales Some internationalization components depend on dynamic linking to function correctly The default when compiling and linking in the Solaris environment is dynamic linking Take care not to specify static linking Call set locale The SunOS system support the POSIX ANSI C function set locale which initializes language and cultural conventions Most applications should set the locale category LC_CTYPE except those not concerned with character interpretation such as block I O to disk or network To control the dynamic handling of different code sets in an application add these lines to your code include lt locale h gt main
75. to create and install translated user messages for internationalized applications Appendix A lists accepted locale names for language and territory Appendix B illustrates localized keyboard layouts Appendix C is a glossary of special terms used throughout the book xi Conventions The following conventions are used in the procedures and examples throughout this guide System prompts and error messages are printed in listing font Information you type as a command or in response to prompts is shown in boldface listing font Type everything shown in boldface exactly as it appears in the text Parts of a command shown in italic text refer to a variable that you have to substitute from a selection It is up to you to make the correct substitution Dialogues between you and the system are enclosed in boxes pwd home machine scotty 5 Sections of program code enclosed boxes include lt stdio h gt main printf Hello 1 exit 0 Control characters shown word Control followed the appropriate character such as Control D To enter a control character hold down the key marked Control and press the appropriate key The standard prompt signs are the dollar sign or percent sign for ordinary users and the sharp when a command must be executed by root or the superuser When commands are mentioned in the text for the
76. ttons and such DevGuide 3 3 will employ these facilities In some cases it s best to use implicit object positioning letting the window system figure out where to place things See Chapter 5 for more details Try to Avoid Compound Messages Creating easily translated messages is an art form that involves more than just inserting gettext calls around strings Remember that word order varies from language to language so complex messages can be very difficult to translate properly A common sense guideline is to avoid compound messages with more than two s parts whenever possible There are two approaches to messaging static and dynamic Static messaging simply involves looking up strings in a message catalog with no reordering taking place Dynamic messaging also involves looking up strings in a message Developer s Guide to Internationalization August 1994 HS II catalog but those strings are reordered and assembled at run time International standards provide an ordering extension to printf for implementing dynamic messaging The advantage of static messaging is simplicity Use it whenever possible However avoid splitting strings across two printf statements which makes messages difficult to translate You shouldn t need to do this anyway since you can pass abbreviated strings to gettext to retrieve longer strings in your message catalog Alternatively you can use a backslash to hide the line break printf
77. ulates a Type 4 or a PC AT101 US keyboard To switch input modes press the Nihongo Japanese key 3 Type words using phonetic values in Romaji or press the Romaji Kana toggle key and type words in Hiragana Japanese engineers often prefer typing in Romaji but most clerks prefer Hiragana since it requires fewer keystrokes With either method the system echoes Hiragana letters in reverse video 4 If that Hiragana word is the one you want press the Commit key 5 To convert the word to Kanji press the Conversion key The system shows a Kanji spelling in the Pre edit Region If that s the spelling you want press the Commit key If not press the Conversion key again to see the next choice Internationalizing Open Windows 63 5 Font Handling 64 In OpenWindows 3 3 all 56 OpenFonts in all sizes include ISO Latin 1 characters Most other OpenWindows fonts do not Solaris supports EUC encoding as did releases of Asian OpenWindows Refer to Chapter 3 for more information on EUC To support languages with multiple character fonts XView applications can open a font set Here s an example of a call to open the Minchou font with Internationalized XView This is intended only as an example the actual interface for Japanese and Asian OpenWindows might be different Xv_Font_Set font_set font_set Xv_Font_Set xv_find frame FONT_SET FONT FAMILY FONT FAMILY MINCHOU FONT STYLE FONT STYLE NORMAL FONT LOCALE xv g
78. unOS model the single byte ISO Latin 1 character is represented in wide character form as follows 00110000 00000000 00000000 This is the only way in which a single byte character can lose its sense of single bytedness Multi byte Library Routines SunOS provides four library routines to convert characters and strings from EUC representation to wide character representation and back again SunOS also provides a set of library routines to perform standard operations on wide characters Developer s Guide to Internationalization August 1994 Qo lll Use mbtowc to convert EUC representation to a wide character and wctomb to convert a wide character to representation For strings arrays of characters use mbstowcs to convert representation strings to wide character strings and wcstombs to convert wide character strings to EUC representation strings All these are in libc SunOS provides library routines 1ibw to replace or supplement character and string routines in 1ibc Compile multi byte programs using the 1w option to the linker A wide character standard I O package is available its routines have the letter w in front of c for character based routines or s for string based routines There is the wsprintf function for wide character formatting and the wsscanf function for wide character input interpretation All the is functions are duplicated by isw
79. vGuide is available about one month after the associated release of OpenWindows DevGuide 3 3 offers many aids for localizing as well as internationalizing applications Specific approaches are discussed in the Internationalization appendix of the DevGuide 3 3 User s Manual Briefly GXV DevGuide s XView code generator helps you to internationalize your application by placing dgettext function calls around strings associated with the user interface Furthermore GXV has an option for adding the object size and position attributes discussed in the next section 59 II O1 Window systems have to take care of three things that operating systems don t have to worry about object layout input method and font handling The following sections describe how applications might deal with these issues Window Object Layout Window objects include push pins pop up menus control buttons and scrollbars Layout of objects containing strings such as menus and buttons can change because the dimensions of those strings often change when switching languages For example a control button in English is probably shorter than the equivalent in German Also a control button in Japanese is probably taller than the equivalent in English Even in the same language control buttons can grow and overlap when the font size increases In XView here s how you would determine the length and height of a text string extern Xv Font font Font string dims
80. ves you time and gives your software a uniform user interface Formats and Collation Many different formats are employed throughout the world to represent date time currency numbers and units These formats should not be hard wired into your code Instead programs should call set locale then the various locale specific format routines leaving format design to localization work for each country or language For string collation sort orders may vary for different languages Programs should use the strcoll or strxfrm library routine to perform string comparisons which use locale specific collation order Note Locale specific collation requires that the application be dynamically linked Time and Date Formats The secret to producing time and date formats valid in many locales is the strftime library routine First set the program clock by calling time then populate a tm structure by calling Localtime Pass this structure to strftime along with a format for date and time plus a holding buffer include lt locale h gt include lt libintl h gt include lt stdio h gt include lt time h gt main time_t clock time Writing Internationalized Code 43 4 struct tm tm localtime char buf 128 setlocale LC ALL clock time time t 0 tm localtime amp clock strftime buf sizeof buf C tm printf s n buf Recommended formats are c for th
81. y bit Now standards committees have decided to define 8 bit code sets which means you have to clean up your code now Here are some problems to look for Code that explicitly uses the most significant bit for its own purposes is said to be dirty There may be valid reasons for altering the most significant bit but dirty code often involves setting and clearing private flags define INVERSE 0x80 bad practice char c c INVERSE Find another way to encode this information A trick used several times in the operating system was to extend this data type to be unsigned short or unsigned int and later setting the top bit of the new data type Developer s Guide to Internationalization August 1994 4 Code that assumes characters are only seven bits long is dirty Here s example of masking off the most significant bit on the assumption it s just the parity bit c string i amp 0 7 bad practice A useful exercise is to search your code for constants like 0x80 0x7f 0200 0177 127 and 128 These constants often highlight problematic code immediately if such bit patterns are used in conjunction with character handling Code that assumes a particular character range needs fixing if gt amp amp lt z bad practice Rewrite this to if islower c Use codeset independent routines found in lt ctype h gt such as isalp
82. y units and presentation order vary greatly around the world This table shows monetary formats in some countries Table 2 4 International Monetary Conventions Locale Currency Example Canadian English Dollar 1 234 56 Canadian French Dollar 1 234 565 Danish Kroner kr kr 1 234 56 Finnish Markka mk 1 234 mk French Franc F F1 234 56 German Deutsche Mark DM 1 234 56DM Italian Lira L L1 234 56 Japanese Yen Y Y1 234 Norwegian Krone kr kr 1 234 56 Spanish Peseta Pts 1 234 56Pts Swedish Krona Kr 1234 56KR UK English Pound 1 234 56 US English Dollar 1 234 56 Note that local and international symbols for currency can differ For example the designation for the French Franc is F in France but this is often written as internationally to distinguish it from other Francs such as the Swiss Franc or the Polynesian Francs Be aware also that a converted currency amount may take up more or less space than the original amount To illustrate 1 000 can become L1 307 000 Developer s Guide to Internationalization August 1994 No II Word and Letter Differences Word Delimiters Word Order Sort Order Usually words are separated by a space character In Japanese and Thai however there is often no delimiter between words The order of words in phrases and sentences varies between languages For instance the order of the words cat and black in a black c

3 - Oracle Documentation

Contents

Download Pdf Manuals

Related Search

Related Contents