breaktext is a small program that uses libunibreak (formerly liblinebreak) to break the lines of input text. One may type breaktext to see the usage.

Some usage examples follow.

Modern Unix systems (including Linux and macOS):

• breaktext input.txt output.txt breaks a UTF-8 text file with no explicit language info

Windows:

• breaktext input.txt output.txt breaks a UTF-16 text file with no explicit language info
• breaktext -LChinese_China.936 -lzh - < input.txt > output.txt breaks a Chinese text file encoded in CP936

The ‘native’ wide character type wchar_t is used in I/O routines, which causes this platform-dependent behaviour. On POSIX-compliant systems, the environment variables LANG, LC_ALL, and LC_CTYPE control the locale/encoding (unless overridden with the -L option), and UTF-8 will probably be used by default on modern systems. On Windows, the encoding is dependent on whether stdin/stdout is used for I/O: console I/O will be automatically converted to/from wchar_t (which is UTF-16) according to the system locale setting (overridable with -L), but files (excepting the stdin/stdout case) will always be in just wchar_t (UTF-16).