On Unix-like operating systems, the dos2unix and unix2dos commands convert plain text files from DOS or Mac format to Unix, and vice versa.
Description
In DOS/Windows text files, a line break, also known as newline, is a combination of two characters: a carriage return (CR) followed by a line feed (LF). In Unix text files a line break is a single character: the line feed (LF). In Mac text files, before macOS X, a line break was single carriage return (CR) character. Nowadays, macOS uses Unix style (LF) line breaks.
- Description
- Syntax
- Mac mode
- Conversion modes
- Unicode
- Recursive conversion
- Localization
- Return values
- Examples
- Linux commands help
Binary files are automatically skipped, unless conversion is forced.
Non-regular files, such as directories and FIFOs, are automatically skipped.
Symbolic links and their targets are by default kept untouched. Symbolic links can optionally be replaced, or the output can be written to the symbolic link target. Symbolic links on Windows are not supported. Windows symbolic links are always replaced, keeping the targets unchanged.
Dos2unix was modelled after dos2unix under SunOS/Solaris and has similar conversion modes.
Syntax
dos2unix [options] [FILE …] [-n INFILE OUTFILE …]
unix2dos [options] [FILE …] [-n INFILE OUTFILE …]
Options
Mac mode
In normal mode line breaks are converted from DOS to Unix and vice versa. Mac line breaks are not converted.
dos2unix – -foo Or in new file mode:
dos2unix -n – -foo out.txt
dos2unix: Only DOS line breaks are changed to two Unix line breaks. In Mac mode, only Mac line breaks are changed to two Unix line breaks.
unix2dos: Only Unix line breaks are changed to two DOS line breaks. In Mac mode, Unix line breaks are changed to two Mac line breaks.
The person who starts the conversion in new file (paired) mode will be the owner of the converted file. The read/write permissions of the new file will be the permissions of the original file minus the umask of the person who runs the conversion.
In old file (in-place) mode the converted file gets the same owner, group, and read/write permissions as the original file. Also, when the file is converted by another user who has write permissions on the file (e.g., user root). The conversion will be aborted when it is not possible to preserve the original values. Change of owner could mean that the original owner is not able to read the file any more. Change of group could be a security risk, the file could be made readable for persons for whom it is not intended. Preservation of owner, group, and read/write permissions is only supported on Unix.
In Mac mode, line breaks are converted from Mac to Unix and vice versa. DOS line breaks are not changed.
To run in Mac mode, use the command-line option “-c mac” or use the commands “mac2unix” or “unix2mac”.
Conversion modes
Unicode
Encodings
There exist different Unicode encodings. On Linux, Unicode files are often encoded in UTF-8 encoding. On Windows Unicode text files can be encoded in UTF-8, UTF-16, or UTF-16 big endian, but are mostly encoded in UTF-16 format.
When only option “-iso” is used dos2unix will try to determine the active code page. When this is not possible dos2unix will use default code page CP437, which is mainly used in the USA. To force a specific code page use options “-437” (US), “-850” (Western European), “-860” (Portuguese), “-863” (French Canadian), or “-865” (Nordic). Windows code page CP1252 (Western European) is also supported with option “-1252”. For other code pages use dos2unix in combination with iconv. iconv can convert between a long list of character encodings.
Never use ISO converion on Unicode text files. It will corrupt UTF-8 encoded files.
Some examples:
Convert from DOS default code page to Unix Latin-1:
dos2unix -iso -n in.txt out.txt
Convert from DOS CP850 to Unix Latin-1:
dos2unix -850 -n in.txt out.txt
Convert from Windows CP1252 to Unix Latin-1:
dos2unix -1252 -n in.txt out.txt
Convert from Windows CP1252 to Unix UTF-8 (Unicode):
iconv -f CP1252 -t UTF-8 in.txt | dos2unix > out.txt
Convert from Unix Latin-1 to DOS default code page:
unix2dos -iso -n in.txt out.txt
Convert from Unix Latin-1 to DOS CP850:
unix2dos -850 -n in.txt out.txt
Convert from Unix Latin-1 to Windows CP1252:
unix2dos -1252 -n in.txt out.txt
Convert from Unix UTF-8 (Unicode) to Windows CP1252:
unix2dos < in.txt | iconv -f UTF-8 -t CP1252 > out.txt
See also http://czyborra.com/charsets/codepages.html and http://czyborra.com/charsets/iso8859.html.
Conversion modes ascii, 7bit, and iso are similar to those of dos2unix/unix2dos under SunOS/Solaris.
Conversion
Unicode text files can have DOS, Unix or Mac line breaks, like regular text files.
All versions of dos2unix and unix2dos can convert UTF-8 encoded files, because UTF-8 was designed for backward compatibility with ASCII.
dos2unix and unix2dos with Unicode UTF-16 support can read little and big endian UTF-16 encoded text files. To see if dos2unix was built with UTF-16 support type “dos2unix -V”.
The Windows versions of dos2unix and unix2dos convert UTF-16 encoded files always to UTF-8 encoded files. Unix versions of dos2unix/unix2dos convert UTF-16 encoded files to the locale character encoding when it is set to UTF-8. Use the locale command to find out what the locale character encoding is.
Because UTF-8 formatted text files are well supported on both Windows and Unix, dos2unix, and unix2dos have no option to write UTF-16 files. All UTF-16 characters can be encoded in UTF-8. Conversion from UTF-16 to UTF-8 is without loss. UTF-16 files will be skipped on Unix when the locale character encoding is not UTF-8, to prevent accidental loss of text. When an UTF-16 to UTF-8 conversion error occurs, for instance when the UTF-16 input file contains an error, the file will be skipped.
ISO and 7-bit mode conversion do not work on UTF-16 files.
Byte Order Mark
On Windows, Unicode text files often have a Byte Order Mark (BOM), because many Windows programs (including Notepad) add BOMs by default. See also https://en.wikipedia.org/wiki/Byte_order_mark.
On Unix, Unicode files often don’t have a BOM. It is assumed that text files are encoded in the locale character encoding.
dos2unix can only detect if a file is in UTF-16 format if the file has a BOM. When an UTF-16 file doesn’t have a BOM, dos2unix sees the file as a binary file.
Use dos2unix in combination with iconv to convert an UTF-16 file without BOM.
Dos2unix never writes a BOM in the output file, unless you use option “-m”.
Unix2dos writes a BOM in the output file when the input file has a BOM, or when option “-m” is used.
Examples
Convert from Windows UTF-16 (with BOM) to Unix UTF-8:
dos2unix -n in.txt out.txt
Convert from Windows UTF-16 (without BOM) to Unix UTF-8:
iconv -f UTF-16 -t UTF-8 in.txt | dos2unix > out.txt
Convert from Unix UTF-8 to Windows UTF-8 with BOM:
unix2dos -m -n in.txt out.txt
Convert from Unix UTF-8 to Windows UTF-16:
unix2dos < in.txt | iconv -f UTF-8 -t UTF-16 > out.txt
Recursive conversion
Use dos2unix in combination with the find and xargs commands to recursively convert text files in a directory tree structure. For instance to convert all .txt files in the directory tree under the current directory type:
find . -name *.txt | xargs dos2unix
Localization
LANG
The primary language is selected with the environment variable LANG. The LANG variable consists out of several parts. The first part is in small letters the language code. The second is optional and is the country code in capital letters, preceded with an underscore. There is also an optional third part: character encoding, preceded with a dot. A few examples for POSIX standard type shells:
For a complete list of language and country codes see the gettext manual: https://www.gnu.org/software/gettext/manual/gettext.html#Language-Codes
On Unix systems, you can use to command locale to get locale-specific information.
LANGUAGE
With the LANGUAGE environment variable you can specify a priority list of languages, separated by colons. Dos2unix gives preference to LANGUAGE over LANG. For instance, first Dutch and then German: “LANGUAGE=nl:de”. You have to first enable localization, by setting LANG (or LC_ALL) to a value other than “C”, before you can use a language priority list through the LANGUAGE variable. See also the gettext manual: https://www.gnu.org/software/gettext/manual/gettext.html#The-LANGUAGE-variable
If you select a language that’s not available, the standard English messages are shown.
DOS2UNIX_LOCALEDIR
With the environment variable DOS2UNIX_LOCALEDIR, the LOCALEDIR set during compilation can be overruled. LOCALEDIR is used to find the language files. The GNU default value is “/usr/local/share/locale”. Option –version displays the LOCALEDIR that is used.
Example (POSIX shell):
export DOS2UNIX_LOCALEDIR=$HOME/share/locale
Return values
On success, zero is returned. When a system error occurs the last system error will be returned. For other errors 1 is returned.
The return value is always zero in quiet mode, except when wrong command-line options are used.
Examples
dos2unix
Get input from stdin and write output to stdout.
dos2unix a.txt b.txt
dos2unix -o a.txt b.txt
Both of the above commands will do the same thing: convert, and replace, both a.txt and b.txt in one command.
dos2unix -k a.txt
Converts and replaces a.txt while keeping original date stamp.