Icon The Kermit Project   |   Now hosted by Panix.com
New York City USA   •   kermit@kermitproject.org
…since 1981
C-Kermit 9.0   Using C-Kermit

C-Kermit 9.0 Locale Features

Frank da Cruz
15 October 2013
Last update: Fri Feb 5 20:24:53 2016 Eastern USA time

Beginning in C-Kermit 9.0.304 Dev.06, C-Kermit has some internationalization features in its command language. These are based on the POSIX locale APIs and definitions. For starters, these features are available only in C-Kermit for Unix; let's see how they work out before trying to add them to C-Kermit for VMS.

Contents:   [About locales]   [How to set them]   [Controlling them in Kermit]   [Using them in Kermit]

What Is a Locale?

POSIX-compliant operating systems including Unix variants such as Linux, Mac OS X, FreeBSD, NetBSD, HP-UX, AIX, Solaris, etc, as well as VMS (a.k.a. OpenVMS) include internationalization features that are implemented in a standard cross-platform manner. Collectively, these features are known as the locale. To see if locales are supported on your Unix computer, type locale at the shell prompt; you should get a response something like this:
$ locale
LANG=en_US.ISO8859-1
LC_CTYPE="en_US.ISO8859-1"
LC_NUMERIC="en_US.ISO8859-1"
LC_TIME="en_US.ISO8859-1"
LC_COLLATE="en_US.ISO8859-1"
LC_MONETARY="en_US.ISO8859-1"
LC_MESSAGES="en_US.ISO8859-1"
LC_ALL=en_US.ISO8859-1
$

The locale is selected in Unix by environment variables and in VMS by logical names, which are approximately the same thing: variables that can be queried (or set) by any program, or by the user at the Unix shell or VMS DCL prompt. The variables that control localization are:

Table 1. Locale Variables
Variable Description
LC_COLLATE Used in sorting and collating
LC_CTYPE Controls character set handling.
LC_MONETARY Controls display input of money amounts.
LC_NUMERIC Controls display input of numbers.
LC_TIME Controls display input of dates and times.
LC_MESSAGES Controls language an character set of messages.
LANG Specifies the language to use.
LC_ALL Used to set all the above at once.

These are set to values that specify language, country, and character set. The character set you choose should be the one that your terminal window or emulator uses (see this list for Kermit). Here are some examples:

Table 2. Locale Examples
Locale Description
en_US.US-ASCII English, United States, US ASCII character set
es_DO.ISO8859-1 Spanish, Dominican Republic, ISO Latin Alphabet 1
de_AT.UTF-8 German, Austria, Unicode UTF-8
ru_SU.KOI8-R Russian, Soviet Union, KOI8-R character set
pt_BR.ISO8859-1 Portuguese, Brazil, ISO Latin Alphabet 1
C The locale that disables all locale processing
POSIX Synonym for C

The format of these strings is:

Table 3. Locale String Elements
Item Defined in
1. 2-letter language code (lowercase) ISO 639-1
2. Underscore character (_)  
3. 2-letter country code (uppercase) ISO 3166-1
4. Period character (.)  
5. Character encoding MIME character set names

Unfortunately there is no definitive list of character-set names, and in fact, no standard spelling for them. For example the name used for ISO 8859-1 Latin Alphabet 1 might be ISO8859-1 or ISO88591 or ISO-8859-1. Furthermore, the character-set name (and the preceding period) can be omitted, in which case a default character set is used that is defined for the particular computer.

Different computers support different sets of locales. In Unix, “locale -a” lists all the locales supported on the computer where the command is given. Even when a locale is supported, the scope of the support can vary. For example, it might cover dates and times but not messages or numbers.

Read more about locales here.

How to Set Your Locale

In Unix versions that support locales, a default or global locale is set for you. You can see what it is by typing locale at the shell prompt. If you would like to use a different locale, you can set the LC_ALL environment variable or any of the other variables from Table 1 individually. The way to set environment variables depends on (a) your shell, and (b) whether you want the definitions to be in effect all the time, or only in the shell, or only when running a particular command or program.

To establish an environment that you want to use all the time, put a command like this:

export LC_ALL=pt_BR.ISO8859-1
in your shell profile, whose name depends on which shell it is: .profile, .bash_profile, .login, etc. Some shells might not support this construct, in which case you'll need to do:
LC_ALL=pt_BR.ISO8859-1
export LC_ALL
For the exact format and spelling of the value (the part on the right side of the equals (=) sign), consult the links in Table 3 and then type “locale -a” to see what's available on your computer. If you wish you can also define the variables separately.
export LANG=pt_BR.ISO8859-1
export LC_TIME=en_US.ISO8859-1
If you would like to switch among different locales on the same computer, you can define aliases if your shell supports them, as bash does:
alias es="export LC_ALL=es_ES.ISO8859-1"
alias fr="export LC_ALL=fr_FR.ISO8859-1"
alias de="export LC_ALL=de_DE.ISO8859-1"
alias en="export LC_ALL=en_US.ISO8859-1"
(In bash, these would go in your .bashrc file; every shell has its own conventions.) Then just type "es" at the shell prompt to switch to Spanish, "fr" to switch to French, etc.

Controlling C-Kermit's Locale Features

Unix programs do not use locales unless they have been explicitly written do so. They do not automatically inherit the locale shown by the locale command at the shell prompt. Until now, this has been the case for C-Kermit.

C-Kermit 9.0.304 has been changed to inherit the user's prevailing locale unless told not to by:

If you never want C-Kermit to use locales, define the environment variable in your shell profile. If you want to inhibit locales for a particular instantiation of C-Kermit, do either one of the following (assuming C-Kermit is installed in your PATH with the name “kermit”):
kermit --nolocale
K_NOLOCALE=1 kermit
If you want to use a particular locale even though Kermit has been started with kermit --nolocale or K_NOLOCALE=1, you can use the SET LOCALE command to choose the desired locale.

Once C-Kermit has started, you can use the SHOW LOCALE command to see what locale C-Kermit is using:

C-Kermit> show locale
Locale enabled:
  LC_COLLATE="C"
  LC_CTYPE="de_DE.ISO8859-1"
  LC_MONETARY="de_DE.ISO8859-1"
  LC_MESSAGES="es_ES.ISO8859-1"
  LC_NUMERIC="de_DE.ISO8859-1"
  LANG="de_DE.ISO8859-1"
C-Kermit>
You can also change the locale in C-Kermit with the SET LOCALE command:
C-Kermit> set locale fr_FR.ISO8859-1
C-Kermit>
The SET LOCALE command sets LC_ALL (i.e. all the other LC_ items) to the given locale. Note that some or all of this operation can fail; for example if the underlying locale database does not have messages for the given language, even though it might have day and month names. Or it might not have them in the given character set. You can check the results of SET LOCAL by giving a SHOW LOCALE command afterwards.

SET LOCALE with no argument disables locale processing (it is the same as selecting the “C” locale).

There is no way in C-Kermit to set different locale items separately (e.g. German day names, Finnish month names, and French messages) but this could be added if there was any demand.

C-Kermit's Locale-Specific Features

Enabling locales in C-Kermit causes certain commands and functions to behave differently from before. It does not affect Kermit's built-in character-set conversions that it uses in file transfer, terminal connection, etc, but here are some things you might notice (and this is not necessarily and exhaustive list):

For UTF-8 encodings, the most likely outcome is that \fupper() and friends work the same way they do when C-Kermit has disabled locales (GRüßE AUS KöLN).

Here are the first two locale-aware features in Kermit:

The dayname(date,fc) function
Given a date (in practically any format at all), this function returns the day of the week in the language and character set specified by the current locale. If the second argument (fc, function code) is included and is a nonzero integer, the result is spelled out in full otherwise it is abbreviated as customary for the locale. The date can also be a digit 1-7, standing for Monday-Sunday. It can also be omitted, which means to use the current date.

The monthname(date,fc) function
Same arguments as dayname(date,fc), but returns the name of the month.

Another function exhibits some new behavior (starting in C-Kermit 9.0.304 Dev.19):
The cvtdate(date,fc) function
This function accepts a free-format date-time string as its first argument and a function code (number) as its second, which specified the desired date-time format for the result. In some of these, three-letter English month-name abbreviations would be used, as in 15-Feb-2016. Now, if a non-default Locale is set, the month names are in the current locale. There is also a new function code: 6. This produces a date-time string in the following format:
dd monthname yyyy hh:mm:ss
where monthname is the full name of the month in given locale, for example:
C-Kermit>set locale es_ES.ISO8859-1
C-Kermit>for \%i 1 12 1 { echo \fcvtdate(2016-\%i-01,6) }
1 enero 2016 00:00:00
1 febrero 2016 00:00:00
1 marzo 2016 00:00:00
1 abril 2016 00:00:00
1 mayo 2016 00:00:00
1 junio 2016 00:00:00
1 julio 2016 00:00:00
1 agosto 2016 00:00:00
1 septiembre 2016 00:00:00
1 octubre 2016 00:00:00
1 noviembre 2016 00:00:00
1 diciembre 2016 00:00:00
C-Kermit>for \%i 1 6 1 { echo \%i. \fcvtdate(20160801,\%i) }
1. 2016-ago-01 00:00:00
2. 01-ago-2016 00:00:00
3. 20160801000000
4. Mon Aug  1 00:00:00 2016
5. 2016:08:01:00:00:00
6. 1 agosto 2016 00:00:00
C-Kermit>

Here's a little script that illustrates locale switching. The “touch /x” command is intended to elicit an access-denied error message:

define xx {                   # Macro to change locale and try some things.
   echo \%2
   .country := \fupper(\%1)
   if equ \%1 "en" .country = US
   set locale \%1_\m(country).ISO8859-1
   echo \v(year) \fmonthname(1,0) \fdayname(1,0)
   touch /x
   echo "-------------------"
}
xx fr French
xx es Spanish
xx de German
xx it Italian
xx pt Portuguese
xx en English
exit

And here are the results:

French
2013 janvier Lundi
?TOUCH /x: Autorisation refusée
-------------------
Spanish
2013 enero lunes
?TOUCH /x: Permiso denegado
-------------------
German
2013 Januar Montag
?TOUCH /x: Zugriff verboten
-------------------
Italian
2013 Gennaio Lunedì
?TOUCH /x: Permission denied
-------------------
Portuguese
2013 Janeiro Segunda Feira
?TOUCH /x: Permission denied
-------------------
English
2013 January Monday
?TOUCH /x: Permission denied

This illustrates how locale support can vary. Notice that although the Italian and Portuguese locales know the day and month names, they do not have localized messages.

C-Kermit 9.0 Locales / The Kermit Project / kermit@kermitproject.org