On Wed, May 05, 2004 at 10:49:05PM +0200, Björn Persson wrote: > What is the canonical way for a program to find out which character > encoding to use for standard input, standard output, command line > parameters, environment variables et cetera? One could look at the LANG > environment variable, but if no encoding is given there? Is ISO 8859-1 > the default for all languages? ISO 8859-1 is frequently not the default. Typically you'll use a combination of setlocale(3) and nl_langinfo(3) to learn about the current locale. The APPLICATION NOTES section of the utf-8(7) man page covers this in somewhat more detail. HTH, Nalin
