From: "Juan Manuel Guerrero" Organization: Darmstadt University of Technology To: djgpp-workers AT delorie DOT com Date: Tue, 16 May 2000 18:00:01 +0200 MIME-Version: 1.0 Content-type: text/plain; charset=US-ASCII Content-transfer-encoding: 7BIT Subject: gtxt035?.zip (beta2) uploaded X-mailer: Pegasus Mail for Windows (v2.54DE) Message-ID: <3EFCFF603C3@HRZ1.hrz.tu-darmstadt.de> Reply-To: djgpp-workers AT delorie DOT com The packages can be downloaded from: ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/beta/gtxt035s.zip ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/beta/gtxt035b.zip ftp://ftp.simtel.net/pub/simtelnet/gnu/djgpp/v2gnu/beta/gtxt035d.zip and mirrors. Changes against beta 1: 1) No function renaming. gettext() from libintl.a remains gettext(). 2) No modifications to gettext.m4. 3) Defaults for emacs, nm and ld in config.site. 4) .po file recoding. In my opinion the po file recoding is the job of the maintainer of the DJGPP package. This means it is one task more for config.bat. I have written a little shell script (recodepo.sh) that will be called from config.bat and that should do the job. The script will: 1) check if a po directory exists. If no po directory exists then the configure script will be executed with --disable-nls. 2) check if po/stamp-po file exists. If this file does not exist then the .po files must be recoded. If the package must be recoded then the script will first check for the required binaries: recode, msgfmt and xgettext. If they can be found the .po files are recoded and the configure script will be executed with --enable-nls else the configure script will be executed with --disable-nls. 3) All .po files will be recoded if stamp-po does not exist. The script will check which .po files exist (echo $XSRC/po/*.po). The number of the MSDOS code page that will be used is stored in the DosCodePage array. The country code is stored in CountryCode array. The UNIX charset that has been used to write the .po file is extracted from the .po file by the script. The main loop will will use the country code to select the right code page and recode the .po file from UNIX charset to MSDOS code page. It will also replace the UNIX charset string by the appropiate MSDOS code page string. Example: charset=ISO-8859-1 --> charset=cp850 The code pages that will be used have been choosen from MSDOS 6.22 COUNTRY.TXT file. This is the important quote: Character Country/Region Sets =========================================================================== Albania 852, 850 Area South (Arabic Countries/regions) 850, 864* Argentina 850, 437 Australia 437, 850 Austria 850, 437 Belgium 850, 437 Bosnia/Herzegovina 852, 850 Brazil 850, 437 Bulgaria 855, 850 Canada French 850, 863 Canada English 850, 863 Chile 850, 437 Colombia 850, 437 Croatia 852, 850 Czech Republic 852, 850 Denmark 850, 865 Ecuador 850, 437 Finland 850, 437 France 850, 437 Germany 850, 437 Greece 869, 737, 850 Hong Kong 437, ** Hungary 852, 850 Iceland 850, 861 India 437, ** International English 437, 850 Ireland 850, 437 Israel 850, 862* Italy 850, 437 Japan 437, 932* Korea 437, 934* Latin America 850, 437 Macedonia 855, 850 Malaysia 437, ** Mexico 850, 437 Netherlands 850, 437 New Zealand 437, 850 Norway 850, 865 People's Republic of China 437, 936* Poland 852, 850 Portugal 850, 860 Romania 852, 850 Russia 866, 855, 852, 850, 437 Serbia/Montenegro 855, 850 Singapore 437, ** Slovakia 852, 850 Slovenia 852, 850 South Africa 437, 850 Spain 850, 437 Sweden 850, 437 Switzerland 850, 437 Taiwan 437, 938* Turkey 857, 850 United Kingdom 850, 437 United States 437, 850 Venezuela 850, 437 Yugoslavia 852, 850 Yugoslavia Cyrillic 855, 850 The first column are the code page numbers that are used in the DosCodePage array. The code pages from the first column are those that will be selected during a standard MS-DOS installation process. These means that for a specific country, the appropiate code page from the first column will be set in autoexec.bat by the MS-DOS install program. IMHO, setting DosCodePage to exactly this values (first column) will meet the requirements of the majority of the average users. If for some reason the above selection should not meet the needs of a particluar user, he will have to recode the .po files manually. For this purpose the original (not recoded) .po files are also supplied in the po/po.orig sub directory. Because the DJGPP packages are usualy configured by the DJGPP maintainer, the average user will never be envolved with the recoding of the .po files. He will get always ready to use .po files and he don't need to install recode at all. If the user downloads a GNU package then he will have to reconfigure the package anyway. He must run config.bat and the recoding of the .po files will become one step more in the configuration process. Suggestions, objections, etc are welcome.