11.6 Substituting text by regular expression

Global substitution in a string is done by patsubst:

Builtin: patsubst (string, regexp, [replacement])

Searches string for matches of regexp, and substitutes replacement for each match. The syntax for regular expressions is the same as in GNU Emacs (see Regexp).

The parts of string that are not covered by any match of regexp are copied to the expansion. Whenever a match is found, the search proceeds from the end of the match, so a character from string will never be substituted twice. If regexp matches a string of zero length, the start position for the search is incremented, to avoid infinite loops.

When a replacement is to be made, replacement is inserted into the expansion, with ‘\n’ substituted by the text matched by the nth parenthesized sub-expression of patsubst, for up to nine sub-expressions. The escape ‘\&’ is replaced by the text of the entire regular expression matched. For all other characters, ‘\’ treats the next character literally. A warning is issued if there were fewer sub-expressions than the ‘\n’ requested, or if there is a trailing ‘\’.

The replacement argument can be omitted, in which case the text matched by regexp is deleted.

The macro patsubst is recognized only with parameters.

patsubst(`GNUs not Unix', `^', `OBS: ')
⇒OBS: GNUs not Unix
patsubst(`GNUs not Unix', `\<', `OBS: ')
⇒OBS: GNUs OBS: not OBS: Unix
patsubst(`GNUs not Unix', `\w*', `(\&)')
⇒(GNUs)() (not)() (Unix)()
patsubst(`GNUs not Unix', `\w+', `(\&)')
⇒(GNUs) (not) (Unix)
patsubst(`GNUs not Unix', `[A-Z][a-z]+')
⇒GN not 
patsubst(`GNUs not Unix', `not', `NOT\')
error→m4:stdin:6: Warning: trailing \ ignored in replacement
⇒GNUs NOT Unix

Here is a slightly more realistic example, which capitalizes individual words or whole sentences, by substituting calls of the macros upcase and downcase into the strings.

Composite: upcase (text)

Composite: downcase (text)

Composite: capitalize (text)

Expand to text, but with capitalization changed: upcase changes all letters to upper case, downcase changes all letters to lower case, and capitalize changes the first character of each word to upper case and the remaining characters to lower case.

First, an example of their usage, using implementations distributed in m4-1.4.19/examples/capitalize.m4.

$ m4 -I examples
include(`capitalize.m4')
⇒
upcase(`GNUs not Unix')
⇒GNUS NOT UNIX
downcase(`GNUs not Unix')
⇒gnus not unix
capitalize(`GNUs not Unix')
⇒Gnus Not Unix

Now for the implementation. There is a helper macro _capitalize which puts only its first word in mixed case. Then capitalize merely parses out the words, and replaces them with an invocation of _capitalize. (As presented here, the capitalize macro has some subtle flaws. You should try to see if you can find and correct them; or see Answers).

$ m4 -I examples
undivert(`capitalize.m4')dnl
⇒divert(`-1')
⇒# upcase(text)
⇒# downcase(text)
⇒# capitalize(text)
⇒#   change case of text, simple version
⇒define(`upcase', `translit(`$*', `a-z', `A-Z')')
⇒define(`downcase', `translit(`$*', `A-Z', `a-z')')
⇒define(`_capitalize',
⇒       `regexp(`$1', `^\(\w\)\(\w*\)',
⇒               `upcase(`\1')`'downcase(`\2')')')
⇒define(`capitalize', `patsubst(`$1', `\w+', `_$0(`\&')')')
⇒divert`'dnl

While regexp replaces the whole input with the replacement as soon as there is a match, patsubst replaces each occurrence of a match and preserves non-matching pieces:

define(`patreg',
`patsubst($@)
regexp($@)')dnl
patreg(`bar foo baz Foo', `foo\|Foo', `FOO')
⇒bar FOO baz FOO
⇒FOO
patreg(`aba abb 121', `\(.\)\(.\)\1', `\2\1\2')
⇒bab abb 212
⇒bab

Omitting regexp evokes a warning, but still produces output; contrast this with an empty regexp argument.

patsubst(`abc')
error→m4:stdin:1: Warning: too few arguments to builtin `patsubst'
⇒abc
patsubst(`abc', `')
⇒abc
patsubst(`abc', `', `\\-')
⇒\-a\-b\-c\-