The fold utility is a filter that shall fold lines from its input files, breaking the lines to have a maximum of width column positions (or bytes, if the -b option is specified). Lines shall be broken by the insertion of a <newline> such that each output line (referred to later in this section as a segment) is the maximum width possible that does not exceed the specified number of column positions (or bytes). A line shall not be broken in the middle of a character. The behavior is undefined if width is less than the number of columns any single character in the input would occupy.
If the <carriage-return>s, <backspace>s, or <tab>s are encountered in the input, and the -b option is not specified, they shall be treated specially:
The current count of line width shall be set to zero. The fold utility shall not insert a <newline> immediately before or after any <carriage-return>.
The fold utility shall conform to the Base Definitions volume of IEEE Std 1003.1-2001, Section 12.2, Utility Syntax Guidelines.
The following options shall be supported:
The following operand shall be supported:
The standard input shall be used only if no file operands are specified. See the INPUT FILES section.
If the -b option is specified, the input files shall be text files except that the lines are not limited to {LINE_MAX} bytes in length. If the -b option is not specified, the input files shall be text files.
The following environment variables shall affect the execution of fold:
The standard output shall be a file containing a sequence of characters whose order shall be preserved from the input files, possibly with inserted <newline>s.
The standard error shall be used only for diagnostic messages.
The following exit values shall be returned:
Default.
The following sections are informative.
The cut and fold utilities can be used to create text files out of files with arbitrary line lengths. The cut utility should be used when the number of lines (or records) needs to remain constant. The fold utility should be used when the contents of long lines need to be kept contiguous.
The fold utility is frequently used to send text files to printers that truncate, rather than fold, lines wider than the printer is able to print (usually 80 or 132 column positions).
An example invocation that submits a file of possibly long lines to the printer (under the assumption that the user knows the line width of the printer to be assigned by lp):
fold -w 132 bigfile | lp
Although terminal input in canonical processing mode requires the erase character (frequently set to <backspace>) to erase the previous character (not byte or column position), terminal output is not buffered and is extremely difficult, if not impossible, to parse correctly; the interpretation depends entirely on the physical device that actually displays/prints/stores the output. In all known internationalized implementations, the utilities producing output for mixed column-width output assume that a <backspace> backs up one column position and outputs enough <backspace>s to return to the start of the character when <backspace> is used to provide local line motions to support underlining and emboldening operations. Since fold without the -b option is dealing with these same constraints, <backspace> is always treated as backing up one column position rather than backing up one character.
Historical versions of the fold utility assumed 1 byte was one character and occupied one column position when written out. This is no longer always true. Since the most common usage of fold is believed to be folding long lines for output to limited-length output devices, this capability was preserved as the default case. The -b option was added so that applications could fold files with arbitrary length lines into text files that could then be processed by the standard utilities. Note that although the width for the -b option is in bytes, a line is never split in the middle of a character. (It is unspecified what happens if a width is specified that is too small to hold a single character found in the input followed by a <newline>.)
The tab stops are hardcoded to be every eighth column to meet historical practice. No new method of specifying other tab stops was invented.