\input texinfo
@c %**start of header
@setfilename m4.info
@settitle GNU macro processor
@finalout
@c %**end of header
@include version.texi
@ifinfo
@set Francois Franc,ois
@end ifinfo
@tex
@set Francois Fran\noexpand\ptexc cois
@end tex
@ifinfo
@format
START-INFO-DIR-ENTRY
* m4: (m4). A powerful macro processor.
END-INFO-DIR-ENTRY
@end format
@end ifinfo
@ifinfo
This file documents the GNU @code{m4} utility.
Copyright (C) 1989, 1990, 1991, 1992, 1993, 1994, 2004 Free Software
Foundation, Inc.
Permission is granted to make and distribute verbatim copies of
this manual provided the copyright notice and this permission notice
are preserved on all copies.
@ignore
Permission is granted to process this file through TeX and print the
results, provided the printed document carries copying permission
notice identical to this one except for the removal of this paragraph
(this paragraph not being relevant to the printed manual).
@end ignore
Permission is granted to copy and distribute modified versions of this
manual under the conditions for verbatim copying, provided that the entire
resulting derived work is distributed under the terms of a permission
notice identical to this one.
Permission is granted to copy and distribute translations of this manual
into another language, under the above conditions for modified versions,
except that this permission notice may be stated in a translation approved
by the Foundation.
@end ifinfo
@titlepage
@title GNU m4, version @value{VERSION}
@subtitle A powerful macro processor
@subtitle Edition @value{EDITION}, @value{UPDATED}
@author by Ren@'e Seindal
@page
@vskip 0pt plus 1filll
Copyright @copyright{} 1989, 90, 91, 92, 93, 94 Free Software Foundation, Inc.
Permission is granted to make and distribute verbatim copies of
this manual provided the copyright notice and this permission notice
are preserved on all copies.
Permission is granted to copy and distribute modified versions of this
manual under the conditions for verbatim copying, provided that the entire
resulting derived work is distributed under the terms of a permission
notice identical to this one.
Permission is granted to copy and distribute translations of this manual
into another language, under the above conditions for modified versions,
except that this permission notice may be stated in a translation approved
by the Foundation.
@end titlepage
@ifinfo
@node Top, Preliminaries, (dir), (dir)
@top GNU @code{m4}
@c @item @b{@code{m4}} @value{hfillkludge} (UtilD, UtilT, SrcCD)
@c
GNU @code{m4} is an implementation of the traditional UNIX macro
processor. It is mostly SVR4 compatible, although it has some
extensions (for example, handling more than 9 positional parameters
to macros). @code{m4} also has builtin functions for including
files, running shell commands, doing arithmetic, etc. Autoconf needs
GNU @code{m4} for generating @file{configure} scripts, but not for
running them.
GNU @code{m4} was originally written by Ren@'e Seindal, with
subsequent changes by @value{Francois} Pinard and other volunteers
on the Internet. All names and email addresses can be found in the
file @file{THANKS} from the GNU @code{m4} distribution.
This is release @value{VERSION}. It is now to be considered stable,
future releases are only meant to fix bugs, increase speed, or improve
documentation. However@dots{}
An experimental feature, which would improve @code{m4} usefulness,
allows for changing the syntax for what is a @dfn{word} in @code{m4}.
You should use:
@comment ignore
@example
./configure --enable-changeword
@end example
@noindent
if you want this feature compiled in. The current implementation
slows down @code{m4} considerably and is hardly acceptable. So, it
might go away, do not count on it yet.
@menu
* Preliminaries:: Introduction and preliminaries
* Syntax:: Lexical and syntactic conventions
* Macros:: How to invoke macros
* Definitions:: How to define new macros
* Conditionals:: Conditionals and loops
* Debugging:: How to debug macros and input
* Input Control:: Input control
* File Inclusion:: File inclusion
* Diversions:: Diverting and undiverting output
* Text handling:: Macros for text handling
* Arithmetic:: Macros for doing arithmetic
* UNIX commands:: Macros for running UNIX commands
* Miscellaneous:: Miscellaneous builtin macros
* Frozen files:: Fast loading of frozen states
* Compatibility:: Compatibility with other versions of m4
* Concept index:: Index for many concepts
* Macro index:: Index for all m4 macros
--- The Detailed Node Listing ---
Introduction and preliminaries
* Intro:: Introduction to @code{m4}
* History:: Historical references
* Invoking m4:: Invoking @code{m4}
* Bugs:: Problems and bugs
* Manual:: Using this manual
Lexical and syntactic conventions
* Names:: Macro names
* Quoted strings:: Quoting input to m4
* Other tokens:: Other kinds of input tokens
* Comments:: Comments in m4 input
How to invoke macros
* Invocation:: Macro invocation
* Inhibiting Invocation:: Preventing macro invocation
* Macro Arguments:: Macro arguments
* Quoting Arguments:: On Quoting Arguments to macros
* Macro expansion:: Expanding macros
How to define new macros
* Define:: Defining a new macro
* Arguments:: Arguments to macros
* Pseudo Arguments:: Pseudo arguments to macros
* Undefine:: Deleting a macro
* Defn:: Renaming macros
* Pushdef:: Temporarily redefining macros
* Indir:: Indirect call of macros
* Builtin:: Indirect call of builtins
Conditionals, loops and recursion
* Ifdef:: Testing if a macro is defined
* Ifelse:: If-else construct, or multibranch
* Loops:: Loops and recursion in m4
How to debug macros and input
* Dumpdef:: Displaying macro definitions
* Trace:: Tracing macro calls
* Debug Levels:: Controlling debugging output
* Debug Output:: Saving debugging output
Input control
* Dnl:: Deleting whitespace in input
* Changequote:: Changing the quote characters
* Changecom:: Changing the comment delimiters
* Changeword:: Changing the lexical structure of words
* M4wrap:: Saving input until end of input
File inclusion
* Include:: Including named files
* Search Path:: Searching for include files
Diverting and undiverting output
* Divert:: Diverting output
* Undivert:: Undiverting output
* Divnum:: Diversion numbers
* Cleardiv:: Discarding diverted text
Macros for text handling
* Len:: Calculating length of strings
* Index:: Searching for substrings
* Regexp:: Searching for regular expressions
* Substr:: Extracting substrings
* Translit:: Translating characters
* Patsubst:: Substituting text by regular expression
* Format:: Formatting strings (printf-like)
Macros for doing arithmetic
* Incr:: Decrement and increment operators
* Eval:: Evaluating integer expressions
Running UNIX commands
* Syscmd:: Executing simple commands
* Esyscmd:: Reading the output of commands
* Sysval:: Exit codes
* Maketemp:: Making names for temporary files
Miscellaneous builtin macros
* Errprint:: Printing error messages
* M4exit:: Exiting from m4
Compatibility with other versions of @code{m4}
* Extensions:: Extensions in GNU m4
* Incompatibilities:: Facilities in System V m4 not in GNU m4
* Other Incompat:: Other incompatibilities
@end menu
@end ifinfo
@node Preliminaries, Syntax, Top, Top
@chapter Introduction and preliminaries
This first chapter explains what is GNU @code{m4}, where @code{m4}
comes from, how to read and use this documentation, how to call the
@code{m4} program and how to report bugs about it. It concludes by
giving tips for reading the remainder of the manual.
The following chapters then detail all the features of the @code{m4}
language.
@menu
* Intro:: Introduction to @code{m4}
* History:: Historical references
* Invoking m4:: Invoking @code{m4}
* Bugs:: Problems and bugs
* Manual:: Using this manual
@end menu
@node Intro, History, Preliminaries, Preliminaries
@section Introduction to @code{m4}
@code{m4} is a macro processor, in the sense that it copies its
input to the output, expanding macros as it goes. Macros are either
builtin or user-defined, and can take any number of arguments.
Besides just doing macro expansion, @code{m4} has builtin functions
for including named files, running UNIX commands, doing integer
arithmetic, manipulating text in various ways, recursion, etc@dots{}
@code{m4} can be used either as a front-end to a compiler, or as a
macro processor in its own right.
The @code{m4} macro processor is widely available on all UNIXes.
Usually, only a small percentage of users are aware of its existence.
However, those who do often become commited users. The growing
popularity of GNU Autoconf, which prerequires GNU @code{m4} for
@emph{generating} the @file{configure} scripts, is an incentive
for many to install it, while these people will not themselves
program in @code{m4}. GNU @code{m4} is mostly compatible with the
System V, Release 3 version, except for some minor differences.
@xref{Compatibility}, for more details.
Some people found @code{m4} to be fairly addictive. They first use
@code{m4} for simple problems, then take bigger and bigger challenges,
learning how to write complex @code{m4} sets of macros along the way.
Once really addicted, users pursue writing of sophisticated @code{m4}
applications even to solve simple problems, devoting more time
debugging their @code{m4} scripts than doing real work. Beware that
@code{m4} may be dangerous for the health of compulsive programmers.
@node History, Invoking m4, Intro, Preliminaries
@section Historical references
The historical notes included here are fairly incomplete, and not
authoritative at all. Please knowledgeable users help us to more
properly write this section.
@code{GPM} has been an important ancestor of @code{m4}. See
C. Stratchey: ``A General Purpose Macro generator'', Computer Journal
8,3 (1965), pp. 225 ff. @code{GPM} is also succintly described into
David Gries classic ``Compiler Construction for Digital Computers''.
While @code{GPM} was @emph{pure}, @code{m4} was meant to deal more
with the true intricacies of real life: macros could be recognized
with being pre-announced, skipping whitespace or end-of-lines was
made easier, more constructs were builtin instead of derived, etc.
Originally, @code{m4} was the engine for Rational FORTRAN preprocessor,
that is, the @code{ratfor} equivalent of @code{cpp}.
@node Invoking m4, Bugs, History, Preliminaries
@section Invoking @code{m4}
The format of the @code{m4} command is:
@comment ignore
@example
@code{m4} [@var{option}@dots{}] [@var{macro-definitions}@dots{}] [@var{input-file}@dots{}]
@end example
@cindex command line, options
@cindex options, command line
All options begin with @samp{-}, or if long option names are used, with
a @samp{--}. A long option name need not be written completely, and
unambigous prefix is sufficient. @code{m4} understands the following
options:
@table @code
@item --version
Print the version number of the program on standard output, then
immediately exit @code{m4} without reading any @var{input-files}.
@item --help
Print an help summary on standard output, then immediately exit
@code{m4} without reading any @var{input-files}.
@item -G
@itemx --traditional
Suppress all the extensions made in this implementation, compared to the
System V version. @xref{Compatibility}, for a list of these.
@item -E
@itemx --fatal-warnings
Stop execution and exit @code{m4} once the first warning has been
issued, considering all of them to be fatal.
@item -d@var{flags}
@itemx --debug=@var{flags}
Set the debug-level according to the flags @var{flags}. The debug-level
controls the format and amount of information presented by the debugging
functions. @xref{Debug Levels}, for more details on the format and
meaning of @var{flags}.
@item -l@var{num}
@itemx --arglength=@var{num}
Restrict the size of the output generated by macro tracing. @xref{Debug
Levels}, for more details.
@item -o@var{file}
@itemx --error-output=@var{file}
Redirect debug and trace output to the named file. Error messages are
still printed on the standard error output. @xref{Debug Output}, for
more details.
@item -I@var{dir}
@itemx --include=@var{dir}
Make @code{m4} search @var{dir} for included files that are not found in
the current working directory. @xref{Search Path}, for more details.
@item -e
@itemx --interactive
Makes this invocation of @code{m4} interactive. This means that all
output will be unbuffered, and interrupts will be ignored.
@item -s
@itemx --synclines
Generate synchronisation lines, for use by the C preprocessor or other
similar tools. This is useful, for example, when @code{m4} is used as a
front end to a compiler. Source file name and line number information
is conveyed by directives of the form @samp{#line @var{linenum}
"@var{filename}"}, which are inserted as needed into the middle of the
input. Such directives mean that the following line originated or was
expanded from the contents of input file @var{filename} at line
@var{linenum}. The @samp{"@var{filename}"} part is often omitted when
the file name did not change from the previous directive.
Synchronisation directives are always given on complete lines per
themselves. When a synchronisation discrepancy occurs in the middle of
an output line, the associated synchronisation directive is delayed
until the beginning of the next generated line.
@item -P
@itemx --prefix-builtins
Internally modify @emph{all} builtin macro names so they all start with
the prefix @samp{m4_}. For example, using this option, one should write
@samp{m4_define} instead of @samp{define}, and @samp{m4___file__}
instead of @samp{__file__}.
@item -W@var{REGEXP}
@itemx --word-regexp=@var{REGEXP}
Use an alternative syntax for macro names. This experimental
option might not be present on all GNU @code{m4} implementations.
(@pxref{Changeword}).
@item -H@var{n}
@itemx --hashsize=@var{n}
Make the internal hash table for symbol lookup be @var{n} entries big.
The number should be prime. The default is 509 entries. It should not
be necessary to increase this value, unless you define an excessive
number of macros.
@item -L@var{n}
@itemx --nesting-limit=@var{n}
Artificially limit the nesting of macro calls to @var{n} levels,
stopping program execution if this limit is ever exceeded. When not
specified, nesting is limited to 250 levels.
The precise effect of this option might be more correctly associated
with textual nesting than dynamic recursion. It has been useful
when some complex @code{m4} input was generated by mechanical means.
Most users would never need this option. If shown to be obtrusive,
this option (which is still experimental) might well disappear.
This option does @emph{not} have the ability to break endless
rescanning loops, while these do not necessarily consume much memory
or stack space. Through clever usage of rescanning loops, one can
request complex, time-consuming computations to @code{m4} with useful
results. Putting limitations in this area would break @code{m4} power.
There are many pathological cases: @w{@samp{define(`a', `a')a}} is
only the simplest example (but @pxref{Compatibility}). Expecting GNU
@code{m4} to detect these would be a little like expecting a compiler
system to detect and diagnose endless loops: it is a quite @emph{hard}
problem in general, if not undecidable!
@item -Q
@itemx --quiet
@itemx --silent
Suppress warnings about missing or superflous arguments in macro calls.
@item -B
@itemx -S
@itemx -T
These options are present for compatibility with System V @code{m4}, but
do nothing in this implementation.
@item -N@var{n}
@itemx --diversions=@var{n}
These options are present only for compatibility with previous
versions of GNU @code{m4}, and were controlling the number of possible
diversions which could be used at the same time. They do nothing,
because there is no fixed limit anymore.
@end table
@cindex macro definitions, on the command line
@cindex command line, macro definitions on the
Macro definitions and deletions can be made on the command line, by
using the @samp{-D} and @samp{-U} options. They have the following
format:
@table @code
@item -D@var{name}
@itemx -D@var{name}=@var{value}
@itemx --define=@var{name}
@itemx --define=@var{name}=@var{value}
This enters @var{name} into the symbol table, before any input files are
read. If @samp{=@var{value}} is missing, the value is taken to be the
empty string. The @var{value} can be any string, and the macro can be
defined to take arguments, just as if it was defined from within the
input.
@item -U@var{name}
@itemx --undefine=@var{name}
This deletes any predefined meaning @var{name} might have. Obviously,
only predefined macros can be deleted in this way.
@item -t@var{name}
@itemx --trace=@var{name}
This enters @var{name} into the symbol table, as undefined but traced.
The macro will consequently be traced from the point it is defined.
@item -F@var{file}
@itemx --freeze-state @var{file}
Once execution is finished, write out the frozen state on the specified
@var{file} (@pxref{Frozen files}).
@item -R@var{file}
@itemx --reload-state @var{file}
Before execution starts, recover the internal state from the specified
frozen @var{file} (@pxref{Frozen files}).
@end table
@cindex command line, filenames on the
@cindex filenames, on the command line
The remaining arguments on the command line are taken to be input file
names. If no names are present, the standard input is read. A file
name of @file{-} is taken to mean the standard input.
The input files are read in the sequence given. The standard input can
only be read once, so the filename @file{-} should only appear once on
the command line.
@node Bugs, Manual, Invoking m4, Preliminaries
@section Problems and bugs
If you have problems with GNU @code{m4} or think you've found a bug,
please report it. Before reporting a bug, make sure you've actually
found a real bug. Carefully reread the documentation and see if it
really says you can do what you're trying to do. If it's not clear
whether you should be able to do something or not, report that too; it's
a bug in the documentation!
Before reporting a bug or trying to fix it yourself, try to isolate it
to the smallest possible input file that reproduces the problem. Then
send us the input file and the exact results @code{m4} gave you. Also
say what you expected to occur; this will help us decide whether the
problem was really in the documentation.
Once you've got a precise problem, send e-mail to (Internet)
@file{bug-gnu-utils@@prep.ai.mit.edu} or (UUCP)
@file{mit-eddie!prep.ai.mit.edu!bug-gnu-utils}. Please include the
version number of @code{m4} you are using. You can get this information
with the command @samp{m4 --version}.
Non-bug suggestions are always welcome as well. If you have questions
about things that are unclear in the documentation or are just obscure
features, please report them too.
@node Manual, , Bugs, Preliminaries
@section Using this manual
This manual contains a number of examples of @code{m4} input and output,
and a simple notation is used to distinguish input, output and error
messages from @code{m4}. Examples are set out from the normal text, and
shown in a fixed width font, like this
@comment ignore
@example
This is an example of an example!
@end example
To distinguish input from output, all output from @code{m4} is prefixed
by the string @samp{@result{}}, and all error messages by the string
@samp{@error{}}. Thus
@comment ignore
@example
Example of input line
@result{}Output line from m4
@error{}and an error message
@end example
As each of the predefined macros in @code{m4} is described, a prototype
call of the macro will be shown, giving descriptive names to the
arguments, e.g.,
@comment ignore
@example
regexp(@var{string}, @var{regexp}, opt @var{replacement})
@end example
All macro arguments in @code{m4} are strings, but some are given special
interpretation, e.g., as numbers, filenames, regular expressions, etc.
The @samp{opt} before the third argument shows that this argument is
optional---if it is left out, it is taken to be the empty string. An
ellipsis (@samp{...}) last in the argument list indicates that any
number of arguments may follow.
This document consistently writes and uses @dfn{builtin}, without an
hyphen, as if it were an English word. This is how the @code{builtin}
primitive is spelled within @code{m4}.
@node Syntax, Macros, Preliminaries, Top
@chapter Lexical and syntactic conventions
@cindex input tokens
@cindex tokens
As @code{m4} reads its input, it separates it into @dfn{tokens}. A
token is either a name, a quoted string, or any single character, that
is not a part of either a name or a string. Input to @code{m4} can also
contain comments.
@menu
* Names:: Macro names
* Quoted strings:: Quoting input to m4
* Other tokens:: Other kinds of input tokens
* Comments:: Comments in m4 input
@end menu
@node Names, Quoted strings, Syntax, Syntax
@section Names
@cindex names
A name is any sequence of letters, digits, and the character @kbd{_}
(underscore), where the first character is not a digit. If a
name has a macro definition, it will be subject to macro expansion
(@pxref{Macros}).
Examples of legal names are: @samp{foo}, @samp{_tmp}, and @samp{name01}.
@node Quoted strings, Other tokens, Names, Syntax
@section Quoted strings
@cindex quoted string
A quoted string is a sequence of characters surrounded by the quotes
@kbd{`} and @kbd{'}, where the number of start and end quotes within the
string balances. The value of a string token is the text, with one
level of quotes stripped off. Thus
@comment ignore
@example
`'
@end example
is the empty string, and
@comment ignore
@example
`@w{}`quoted'@w{}'
@end example
is the string
@comment ignore
@example
`quoted'
@end example
The quote characters can be changed at any time, using the builtin macro
@code{changequote}. @xref{Changequote}, for more information.
@node Other tokens, Comments, Quoted strings, Syntax
@section Other tokens
Any character, that is neither a part of a name, nor of a quoted string,
is a token by itself.
@node Comments, , Other tokens, Syntax
@section Comments
@cindex comments
Comments in @code{m4} are normally delimited by the characters @samp{#}
and newline. All characters between the comment delimiters are ignored,
but the entire comment (including the delimiters) is passed through to
the output---comments are @emph{not} discarded by @code{m4}.
Comments cannot be nested, so the first newline after a @samp{#} ends
the comment. The commenting effect of the begin comment character
can be inhibited by quoting it.
The comment delimiters can be changed to any string at any time, using
the builtin macro @code{changecom}. @xref{Changecom}, for more
information.
@c FIXME: more examples would be useful here --ADR
@node Macros, Definitions, Syntax, Top
@chapter How to invoke macros
This chapter covers macro invocation, macro arguments and how macro
expansion is treated.
@menu
* Invocation:: Macro invocation
* Inhibiting Invocation:: Preventing macro invocation
* Macro Arguments:: Macro arguments
* Quoting Arguments:: On Quoting Arguments to macros
* Macro expansion:: Expanding macros
@end menu
@node Invocation, Inhibiting Invocation, Macros, Macros
@section Macro invocation
@cindex macro invocation
Macro invocations has one of the forms
@comment ignore
@example
name
@end example
@noindent
which is a macro invocation without any arguments, or
@comment ignore
@example
name(arg1, arg2, ..., arg@var{n})
@end example
@noindent
which is a macro invocation with @var{n} arguments. Macros can have any
number of arguments. All arguments are strings, but different macros
might interpret the arguments in different ways.
The opening parenthesis @emph{must} follow the @var{name} directly, with
no spaces in between. If it does not, the macro is called with no
arguments at all.
For a macro call to have no arguments, the parentheses @emph{must} be
left out. The macro call
@comment ignore
@example
name()
@end example
@noindent
is a macro call with one argument, which is the empty string, not a call
with no arguments.
@node Inhibiting Invocation, Macro Arguments, Invocation, Macros
@section Preventing macro invocation
An innovation of the @code{m4} language, compared to some of its
predecessors (like Stratchey's @code{GPM}, for example), is the ability
to recognize macro calls without resorting to any special, prefixed
invocation character. While generally useful, this feature might
sometimes be the source of spurious, unwanted macro calls. So, GNU
@code{m4} offers several mechanisms or techniques for inhibiting the
recognition of names as macro calls.
First of all, many builtin macros cannot meaningfully be called
without arguments. For any of these macros, whenever an opening
parenthesis does not immediately follow their name, the builtin macro
call is not triggered. This solves the most usual cases, like for
@samp{include} or @samp{eval}. Later in this document, the sentence
``This macro is recognized only when given arguments'' refers to this
specific provision.
There is also a command call option (@code{--prefix-builtins}, or
@code{-P}) which requires all builtin macro names to be prefixed
by @samp{m4_} for them to be recognized. The option has no effect
whatsoever on user defined macros. For example, with this option,
one has to write @code{m4_dnl} and even @code{m4_m4exit}.
If your version of GNU @code{m4} has the @code{changeword} feature
compiled in, there it offers far more flexibility in specifying the
syntax of macro names, both builtin or user-defined. @xref{Changeword},
for more information on this experimental feature.
Of course, the simplest way to prevent a name to be interpreted
as a call to an existing macro is to quote it. The remainder of
this section studies a little more deeply how quoting affects macro
invocation, and how quoting can be used to inhibit macro invocation.
Even if quoting is usually done over the whole macro name, it can also
be done over only a few characters of this name. It is also possible
to quote the empty string, but this works only @emph{inside} the name.
For example:
@comment ignore
@example
`divert'
`d'ivert
di`ver't
div`'ert
@end example
@noindent
all yield the string @samp{divert}. While in both:
@comment ignore
@example
`'divert
divert`'
@end example
@noindent
the @code{divert} builtin macro will be called.
The output of macro evaluations is always rescanned. The following
example would yield the string @samp{de}, exactly as if @code{m4}
has been given @w{@samp{substr(abcde, 3, 2)}} as input:
@comment ignore
@example
define(`x', `substr(ab')
define(`y', `cde, 3, 2)')
x`'y
@end example
Unquoted strings on either side of a quoted string are subject to
being recognized as macro names. In the following example, quoting the
empty string allows for the @code{dnl} macro to be recognized as such:
@comment ignore
@example
define(`macro', `di$1')
macro(v)`'dnl
@end example
@noindent
Without the quotes, this would rather yield the string @samp{divdnl}
followed by an end of line.
Quoting may prevent recognizing as a macro name the concatenation of a
macro expansion with the surrounding characters. In this example:
@comment ignore
@example
define(`macro', `di$1')
macro(v)`ert'
@end example
@noindent
the input will produce the string @samp{divert}. If the quote was
removed, the @code{divert} builtin would be called instead.
@node Macro Arguments, Quoting Arguments, Inhibiting Invocation, Macros
@section Macro arguments
@cindex macros, arguments to
@cindex arguments to macros
When a name is seen, and it has a macro definition, it will be expanded
as a macro.
If the name is followed by an opening parenthesis, the arguments will be
collected before the macro is called. If too few arguments are
supplied, the missing arguments are taken to be the empty string. If
there are too many arguments, the excess arguments are ignored.
Normally @code{m4} will issue warnings if a builtin macro is called
with an inappropriate number of arguments, but it can be suppressed with
the @samp{-Q} command line option. For user defined macros, there is no
check of the number of arguments given.
Macros are expanded normally during argument collection, and whatever
commas, quotes and parentheses that might show up in the resulting
expanded text will serve to define the arguments as well. Thus, if
@var{foo} expands to @samp{, b, c}, the macro call
@comment ignore
@example
bar(a foo, d)
@end example
is a macro call with four arguments, which are @samp{a }, @samp{b},
@samp{c} and @samp{d}. To understand why the first argument contains
whitespace, remember that leading unquoted whitespace is never part
of an argument, but trailing whitespace always is.
@node Quoting Arguments, Macro expansion, Macro Arguments, Macros
@section Quoting macro arguments
@cindex quoted macro arguments
@cindex macros, quoted arguments to
@cindex arguments, quoted macro
Each argument has leading unquoted whitespace removed. Within each
argument, all unquoted parentheses must match. For example, if
@var{foo} is a macro,
@comment ignore
@example
foo(() (`(') `(')
@end example
is a macro call, with one argument, whose value is @samp{()@w{ }(()@w{ }(}.
It is common practice to quote all arguments to macros, unless you are
sure you want the arguments expanded. Thus, in the above
example with the parentheses, the `right' way to do it is like this:
@comment ignore
@example
foo(`() (() (')
@end example
It is, however, in certain cases necessary to leave out quotes for some
arguments, and there is nothing wrong in doing it. It just makes life a
bit harder, if you are not careful.
@node Macro expansion, , Quoting Arguments, Macros
@section Macro expansion
@cindex macros, expansion of
@cindex expansion of macros
When the arguments, if any, to a macro call have been collected, the
macro is expanded, and the expansion text is pushed back onto the input
(unquoted), and reread. The expansion text from one macro call might
therefore result in more macros being called, if the calls are included,
completely or partially, in the first macro calls' expansion.
Taking a very simple example, if @var{foo} expands to @samp{bar}, and
@var{bar} expands to @samp{Hello world}, the input
@comment ignore
@example
foo
@end example
@noindent
will expand first to @samp{bar}, and when this is reread and
expanded, into @samp{Hello world}.
@node Definitions, Conditionals, Macros, Top
@chapter How to define new macros
@cindex macros, how to define new
@cindex defining new macros
Macros can be defined, redefined and deleted in several different ways.
Also, it is possible to redefine a macro, without losing a previous
value, which can be brought back at a later time.
@menu
* Define:: Defining a new macro
* Arguments:: Arguments to macros
* Pseudo Arguments:: Pseudo arguments to macros
* Undefine:: Deleting a macro
* Defn:: Renaming macros
* Pushdef:: Temporarily redefining macros
* Indir:: Indirect call of macros
* Builtin:: Indirect call of builtins
@end menu
@node Define, Arguments, Definitions, Definitions
@section Defining a macro
@findex define
The normal way to define or redefine macros is to use the builtin
@code{define}:
@comment ignore
@example
define(@var{name} [, @var{expansion}])
@end example
@noindent which defines @var{name} to expand to @var{expansion}. If
@var{expansion} is not given, it is taken to be empty.
The expansion of @code{define} is void.
The following example defines the macro @var{foo} to expand to the text
@samp{Hello World.}.
@example
define(`foo', `Hello world.')
@result{}
foo
@result{}Hello world.
@end example
The empty line in the output is there because the newline is not
a part of the macro definition, and it is consequently copied to
the output. This can be avoided by use of the macro @code{dnl}.
@xref{Dnl}, for details.
The macro @code{define} is recognized only with parameters.
@node Arguments, Pseudo Arguments, Define, Definitions
@section Arguments to macros
@cindex macros, arguments to
@cindex Arguments to macros
Macros can have arguments. The @var{n}th argument is denoted by
@code{$n} in the expansion text, and is replaced by the @var{n}th actual
argument, when the macro is expanded. Here is a example of a macro with
two arguments. It simply exchanges the order of the two arguments.
@example
define(`exch', `$2, $1')
@result{}
exch(arg1, arg2)
@result{}arg2, arg1
@end example
This can be used, for example, if you like the arguments to
@code{define} to be reversed.
@example
define(`exch', `$2, $1')
@result{}
define(exch(``expansion text'', ``macro''))
@result{}
macro
@result{}expansion text
@end example
@xref{Quoting Arguments}, for an explanation of the double quotes.
@cindex GNU extensions
GNU @code{m4} allows the number following the @samp{$} to consist of one
or more digits, allowing macros to have any number of arguments. This
is not so in UNIX implementations of @code{m4}, which only recognize
one digit.
As a special case, the zero'th argument, @code{$0}, is always the name
of the macro being expanded.
@example
define(`test', ``Macro name: $0'')
@result{}
test
@result{}Macro name: test
@end example
If you want quoted text to appear as part of the expansion text,
remember that quotes can be nested in quoted strings. Thus, in
@example
define(`foo', `This is macro `foo'.')
@result{}
foo
@result{}This is macro foo.
@end example
@noindent
The @samp{foo} in the expansion text is @emph{not} expanded, since it is
a quoted string, and not a name.
@node Pseudo Arguments, Undefine, Arguments, Definitions
@section Special arguments to macros
@cindex special arguments to macros
@cindex macros, special arguments to
@cindex arguments to macros, special
There is a special notation for the number of actual arguments supplied,
and for all the actual arguments.
The number of actual arguments in a macro call is denoted by @code{$#}
in the expansion text. Thus, a macro to display the number of arguments
given can be
@example
define(`nargs', `$#')
@result{}
nargs
@result{}0
nargs()
@result{}1
nargs(arg1, arg2, arg3)
@result{}3
@end example
The notation @code{$*} can be used in the expansion text to denote all
the actual arguments, unquoted, with commas in between. For example
@example
define(`echo', `$*')
@result{}
echo(arg1, arg2, arg3 , arg4)
@result{}arg1,arg2,arg3 ,arg4
@end example
Often each argument should be quoted, and the notation @code{$@@} handles
that. It is just like @code{$*}, except that it quotes each argument.
A simple example of that is:
@example
define(`echo', `$@@')
@result{}
echo(arg1, arg2, arg3 , arg4)
@result{}arg1,arg2,arg3 ,arg4
@end example
Where did the quotes go? Of course, they were eaten, when the expanded
text were reread by @code{m4}. To show the difference, try
@example
define(`echo1', `$*')
@result{}
define(`echo2', `$@@')
@result{}
define(`foo', `This is macro `foo'.')
@result{}
echo1(foo)
@result{}This is macro This is macro foo..
echo2(foo)
@result{}This is macro foo.
@end example
@noindent
@xref{Trace}, if you do not understand this.
A @samp{$} sign in the expansion text, that is not followed by anything
@code{m4} understands, is simply copied to the macro expansion, as any
other text is.
@example
define(`foo', `$$$ hello $$$')
@result{}
foo
@result{}$$$ hello $$$
@end example
If you want a macro to expand to something like @samp{$12}, put a pair
of quotes after the @code{$}. This will prevent @code{m4} from
interpreting the @code{$} sign as a reference to an argument.
@node Undefine, Defn, Pseudo Arguments, Definitions
@section Deleting a macro
@cindex macros, how to delete
@cindex deleting macros
@cindex undefining macros
@findex undefine
A macro definition can be removed with @code{undefine}:
@comment ignore
@example
undefine(@var{name})
@end example
@noindent
which removes the macro @var{name}. The macro name must necessarily be
quoted, since it will be expanded otherwise.
The expansion of @code{undefine} is void.
@example
foo
@result{}foo
define(`foo', `expansion text')
@result{}
foo
@result{}expansion text
undefine(`foo')
@result{}
foo
@result{}foo
@end example
It is not an error for @var{name} to have no macro definition. In that
case, @code{undefine} does nothing.
The macro @code{undefine} is recognized only with parameters.
@node Defn, Pushdef, Undefine, Definitions
@section Renaming macros
@cindex macros, how to rename
@cindex renaming macros
@findex defn
It is possible to rename an already defined macro. To do this, you need
the builtin @code{defn}:
@comment ignore
@example
defn(@var{name})
@end example
@noindent
which expands to the @emph{quoted definition} of @var{name}. If the
argument is not a defined macro, the expansion is void.
If @var{name} is a user-defined macro, the quoted definition is simply
the quoted expansion text. If, instead, @var{name} is a builtin, the
expansion is a special token, which points to the builtin's internal
definition. This token is only meaningful as the second argument to
@code{define} (and @code{pushdef}), and is ignored in any other context.
Its normal use is best understood through an example, which shows how to
rename @code{undefine} to @code{zap}:
@example
define(`zap', defn(`undefine'))
@result{}
zap(`undefine')
@result{}
undefine(`zap')
@result{}undefine(zap)
@end example
In this way, @code{defn} can be used to copy macro definitions, and also
definitions of builtin macros. Even if the original macro is removed,
the other name can still be used to access the definition.
The macro @code{defn} is recognized only with parameters.
@node Pushdef, Indir, Defn, Definitions
@section Temporarily redefining macros
@cindex macros, temporary redefinition of
@cindex temporary redefinition of macros
@cindex redefinition of macros, temporary
It is possible to redefine a macro temporarily, reverting to the
previous definition at a later time.
@findex popdef
@findex pushdef
This is done with the builtins @code{pushdef} and @code{popdef}:
@comment ignore
@example
pushdef(@var{name} [, @var{expansion}])
popdef(@var{name})
@end example
@noindent
which are quite analogous to @code{define} and @code{undefine}.
These macros work in a stack-like fashion. A macro is temporarily
redefined with @code{pushdef}, which replaces an existing definition of
@var{name}, while saving the previous definition, before the new one is
installed. If there is no previous definition, @code{pushdef} behaves
exactly like @code{define}.
If a macro has several definitions (of which only one is accessible),
the topmost definition can be removed with @code{popdef}. If there is
no previous definition, @code{popdef} behaves like @code{undefine}.
@example
define(`foo', `Expansion one.')
@result{}
foo
@result{}Expansion one.
pushdef(`foo', `Expansion two.')
@result{}
foo
@result{}Expansion two.
popdef(`foo')
@result{}
foo
@result{}Expansion one.
popdef(`foo')
@result{}
foo
@result{}foo
@end example
If a macro with several definitions is redefined with @code{define}, the
topmost definition is @emph{replaced} with the new definition. If it is
removed with @code{undefine}, @emph{all} the definitions are removed,
and not only the topmost one.
@example
define(`foo', `Expansion one.')
@result{}
foo
@result{}Expansion one.
pushdef(`foo', `Expansion two.')
@result{}
foo
@result{}Expansion two.
define(`foo', `Second expansion two.')
@result{}
foo
@result{}Second expansion two.
undefine(`foo')
@result{}
foo
@result{}foo
@end example
It is possible to temporarily redefine a builtin with @code{pushdef}
and @code{defn}.
The macros @code{pushdef} and @code{popdef} are recognized only with
parameters.
@node Indir, Builtin, Pushdef, Definitions
@section Indirect call of macros
@cindex indirect call of macros
@cindex call of macros, indirect
@cindex macros, indirect call of
@cindex GNU extensions
@findex indir
Any macro can be called indirectly with @code{indir}:
@comment ignore
@example
indir(@var{name}, ...)
@end example
@noindent
which results in a call to the macro @var{name}, which is passed the
rest of the arguments. This can be used to call macros with ``illegal''
names (@code{define} allows such names to be defined):
@example
define(`$$internal$macro', `Internal macro (name `$0')')
@result{}
$$internal$macro
@result{}$$internal$macro
indir(`$$internal$macro')
@result{}Internal macro (name $$internal$macro)
@end example
The point is, here, that larger macro packages can have private macros
defined, that will not be called by accident. They can @emph{only} be
called through the builtin @code{indir}.
@c FIXME: Why indir does not require at least one parameter?
@node Builtin, , Indir, Definitions
@section Indirect call of builtins
@cindex indirect call of builtins
@cindex call of builtins, indirect
@cindex builtins, indirect call of
@cindex GNU extensions
@findex builtin
Builtin macros can be called indirectly with @code{builtin}:
@comment ignore
@example
builtin(@var{name}, ...)
@end example
@noindent
which results in a call to the builtin @var{name}, which is passed the
rest of the arguments. This can be used, if @var{name} has been given
another definition that has covered the original.
The macro @code{builtin} is recognized only with parameters.
@node Conditionals, Debugging, Definitions, Top
@chapter Conditionals, loops and recursion
Macros, expanding to plain text, perhaps with arguments, are not quite
enough. We would like to have macros expand to different things, based
on decisions taken at run-time. E.g., we need some kind of conditionals.
Also, we would like to have some kind of loop construct, so we could do
something a number of times, or while some condition is true.
@menu
* Ifdef:: Testing if a macro is defined
* Ifelse:: If-else construct, or multibranch
* Loops:: Loops and recursion in m4
@end menu
@node Ifdef, Ifelse, Conditionals, Conditionals
@section Testing macro definitions
@cindex conditionals
@findex ifdef
There are two different builtin conditionals in @code{m4}. The first is
@code{ifdef}:
@comment ignore
@example
ifdef(@var{name}, @var{string-1}, opt @var{string-2})
@end example
@noindent
which makes it possible to test whether a macro is defined or not. If
@var{name} is defined as a macro, @code{ifdef} expands to
@var{string-1}, otherwise to @var{string-2}. If @var{string-2} is
omitted, it is taken to be the empty string (according to the normal
rules).
@example
ifdef(`foo', ``foo' is defined', ``foo' is not defined')
@result{}foo is not defined
define(`foo', `')
@result{}
ifdef(`foo', ``foo' is defined', ``foo' is not defined')
@result{}foo is defined
@end example
The macro @code{ifdef} is recognized only with parameters.
@node Ifelse, Loops, Ifdef, Conditionals
@section Comparing strings
@cindex comparing strings
@findex ifelse
The other conditional, @code{ifelse}, is much more powerful. It can be
used as a way to introduce a long comment, as an if-else construct, or
as a multibranch, depending on the number of arguments supplied:
@comment ignore
@example
ifelse(@var{comment})
ifelse(@var{string-1}, @var{string-2}, @var{equal}, opt @var{not-equal})
ifelse(@var{string-1}, @var{string-2}, @var{equal}, ...)
@end example
@noindent
Used with only one argument, the @code{ifelse} simply discards it and
produces no output. This is a common @code{m4} idiom for introducing a
block comment, as an alternative to repeatedly using @code{dnl}. This
special usage is recognized by GNU @code{m4}, so that in this case, the
warning about missing arguments is never triggered.
If called with three or four arguments, @code{ifelse} expands into
@var{equal}, if @var{string-1} and @var{string-2} are equal (character
for character), otherwise it expands to @var{not-equal}.
@example
ifelse(foo, bar, `true')
@result{}
ifelse(foo, foo, `true')
@result{}true
ifelse(foo, bar, `true', `false')
@result{}false
ifelse(foo, foo, `true', `false')
@result{}true
@end example
@cindex multibranches
However, @code{ifelse} can take more than four arguments. If given more
than four arguments, @code{ifelse} works like a @code{case} or @code{switch}
statement in traditional programming languages. If @var{string-1} and
@var{string-2} are equal, @code{ifelse} expands into @var{equal}, otherwise
the procedure is repeated with the first three arguments discarded. This
calls for an example:
@example
ifelse(foo, bar, `third', gnu, gnats, `sixth', `seventh')
@result{}seventh
@end example
Naturally, the normal case will be slightly more advanced than these
examples. A common use of @code{ifelse} is in macros implementing loops
of various kinds.
The macro @code{ifelse} is recognized only with parameters.
@node Loops, , Ifelse, Conditionals
@section Loops and recursion
@cindex recursive macros
@cindex macros, recursive
There is no direct support for loops in @code{m4}, but macros can be
recursive. There is no limit on the number of recursion levels, other
than those enforced by your hardware and operating system.
@cindex loops
Loops can be programmed using recursion and the conditionals described
previously.
@findex shift
There is a builtin macro, @code{shift}, which can, among other things,
be used for iterating through the actual arguments to a macro:
@comment ignore
@example
shift(...)
@end example
@noindent
It takes any number of arguments, and expands to all but the first
argument, separated by commas, with each argument quoted.
@example
shift(bar)
@result{}
shift(foo, bar, baz)
@result{}bar,baz
@end example
An example of the use of @code{shift} is this macro, which reverses the
order of its arguments:
@example
define(`reverse', `ifelse($#, 0, , $#, 1, ``$1'',
`reverse(shift($@@)), `$1'')')
@result{}
reverse
@result{}
reverse(foo)
@result{}foo
reverse(foo, bar, gnats, and gnus)
@result{}and gnus, gnats, bar, foo
@end example
While not a very interesting macro, it does show how simple loops can be
made with @code{shift}, @code{ifelse} and recursion.
@cindex forloops
@cindex loops, counting
@cindex counting loops
Here is an example of a loop macro that implements a simple forloop. It
can, for example, be used for simple counting:
@comment ignore
@example
forloop(`i', 1, 8, `i ')
@result{}1 2 3 4 5 6 7 8
@end example
The arguments are a name for the iteration variable, the starting value,
the final value, and the text to be expanded for each iteration. With
this macro, the macro @code{i} is defined only within the loop. After
the loop, it retains whatever value it might have had before.
For-loops can be nested, like
@comment ignore
@example
forloop(`i', 1, 4, `forloop(`j', 1, 8, `(i, j) ')
')
@result{}(1, 1) (1, 2) (1, 3) (1, 4) (1, 5) (1, 6) (1, 7) (1, 8)
@result{}(2, 1) (2, 2) (2, 3) (2, 4) (2, 5) (2, 6) (2, 7) (2, 8)
@result{}(3, 1) (3, 2) (3, 3) (3, 4) (3, 5) (3, 6) (3, 7) (3, 8)
@result{}(4, 1) (4, 2) (4, 3) (4, 4) (4, 5) (4, 6) (4, 7) (4, 8)
@result{}
@end example
The implementation of the @code{forloop} macro is fairly
straightforward. The @code{forloop} macro itself is simply a wrapper,
which saves the previous definition of the first argument, calls the
internal macro @code{_forloop}, and re-establishes the saved definition of
the first argument.
The macro @code{_forloop} expands the fourth argument once, and tests
to see if it is finished. If it has not finished, it increments
the iteration variable (using the predefined macro @code{incr},
@pxref{Incr}), and recurses.
Here is the actual implementation of @code{forloop}:
@comment ignore
@example
define(`forloop',
`pushdef(`$1', `$2')_forloop(`$1', `$2', `$3', `$4')popdef(`$1')')
define(`_forloop',
`$4`'ifelse($1, `$3', ,
`define(`$1', incr($1))_forloop(`$1', `$2', `$3', `$4')')')
@end example
Notice the careful use of quotes. Only three macro arguments are
unquoted, each for its own reason. Try to find out @emph{why} these
three arguments are left unquoted, and see what happens if they are
quoted.
Now, even though these two macros are useful, they are still not robust
enough for general use. They lack even basic error handling of cases
like start value less than final value, and the first argument not being
a name. Correcting these errors are left as an exercise to the reader.
@node Debugging, Input Control, Conditionals, Top
@chapter How to debug macros and input
When writing macros for @code{m4}, most of the time they woould not
work as intended (as is the case with most programming languages).
There is a little support for macro debugging in @code{m4}.
@menu
* Dumpdef:: Displaying macro definitions
* Trace:: Tracing macro calls
* Debug Levels:: Controlling debugging output
* Debug Output:: Saving debugging output
@end menu
@node Dumpdef, Trace, Debugging, Debugging
@section Displaying macro definitions
@cindex displaying macro definitions
@cindex macros, displaying definitions
@cindex definitions, displaying macro
@findex dumpdef
If you want to see what a name expands into, you can use the builtin
@code{dumpdef}:
@comment ignore
@example
dumpdef(...)
@end example
@noindent
which accepts any number of arguments. If called without any arguments,
it displays the definitions of all known names, otherwise it displays
the definitions of the names given. The output is printed directly on
the standard error output.
The expansion of @code{dumpdef} is void.
@example
define(`foo', `Hello world.')
@result{}
dumpdef(`foo')
@error{}foo: `Hello world.'
@result{}
dumpdef(`define')
@error{}define: <define>
@result{}
@end example
The last example shows how builtin macros definitions are displayed.
@xref{Debug Levels}, for information on controlling the details of the
display.
@node Trace, Debug Levels, Dumpdef, Debugging
@section Tracing macro calls
@cindex tracing macro expansion
@cindex macro expansion, tracing
@cindex expansion, tracing macro
@findex traceon
@findex traceoff
It is possible to trace macro calls and expansions through the builtins
@code{traceon} and @code{traceoff}:
@comment ignore
@example
traceon(...)
traceoff(...)
@end example
@noindent
When called without any arguments, @code{traceon} and @code{traceoff}
will turn tracing on and off, respectively, for all defined macros.
When called with arguments, only the named macros are affected.
The expansion of @code{traceon} and @code{traceoff} is void.
Whenever a traced macro is called and the arguments have been collected,
the call is displayed. If the expansion of the macro call is not void,
the expansion can be displayed after the call. The output is printed
directly on the standard error output.
@example
define(`foo', `Hello World.')
@result{}
define(`echo', `$@@')
@result{}
traceon(`foo', `echo')
@result{}
foo
@error{}m4trace: -1- foo -> `Hello World.'
@result{}Hello World.
echo(gnus, and gnats)
@error{}m4trace: -1- echo(`gnus', `and gnats') -> ``gnus',`and gnats''
@result{}gnus,and gnats
@end example
The number between dashes is the depth of the expansion. It is one most
of the time, signifying an expansion at the outermost level, but it
increases when macro arguments contain unquoted macro calls.
@xref{Debug Levels}, for information on controlling the details of the
display.
@node Debug Levels, Debug Output, Trace, Debugging
@section Controlling debugging output
@cindex controlling debugging output
@cindex debugging output, controlling
The @samp{-d} option to @code{m4} controls the amount of details
presented, when using the macros described in the preceding sections.
The @var{flags} following the option can be one or more of the
following:
@table @code
@item t
Trace all macro calls made in this invocation of @code{m4}.
@item a
Show the actual arguments in each macro call. This applies to all macro
calls if the @samp{t} flag is used, otherwise only the macros covered by
calls of @code{traceon}.
@item e
Show the expansion of each macro call, if it is not void. This applies
to all macro calls if the @samp{t} flag is used, otherwise only the
macros covered by calls of @code{traceon}.
@item q
Quote actual arguments and macro expansions in the display with the
current quotes.
@item c
Show several trace lines for each macro call. A line is shown when the
macro is seen, but before the arguments are collected; a second line
when the arguments have been collected and a third line after the call
has completed.
@item x
Add a unique `macro call id' to each line of the trace output. This is
useful in connection with the @samp{c} flag above.
@item f
Show the name of the current input file in each trace output line.
@item l
Show the the current input line number in each trace output line.
@item p
Print a message when a named file is found through the path search
mecanism (@pxref{Search Path}), giving the actual filename used.
@item i
Print a message each time the current input file is changed, giving file
name and input line number.
@item V
A shorthand for all of the above flags.
@end table
If no flags are specified with the @samp{-d} option, the default is
@samp{aeq}. The examples in the previous two sections assumed the
default flags.
@cindex GNU extensions
@findex debugmode
There is a builtin macro @code{debugmode}, which allows on-the-fly control of
the debugging output format:
@comment ignore
@example
debugmode(opt @var{flags})
@end example
@noindent
The argument @var{flags} should be a subset of the letters listed above.
As special cases, if the argument starts with a @samp{+}, the flags are
added to the current debug flags, and if it starts with a @samp{-}, they
are removed. If no argument is present, the debugging flags are set to
zero (as if no @samp{-d} was given), and with an empty argument the flags
are reset to the default.
@node Debug Output, , Debug Levels, Debugging
@section Saving debugging output
@cindex saving debugging output
@cindex debugging output, saving
@cindex output, saving debugging
@cindex GNU extensions
@findex debugfile
Debug and tracing output can be redirected to files using either the
@samp{-o} option to @code{m4}, or with the builtin macro @code{debugfile}:
@comment ignore
@example
debugfile(opt @var{filename})
@end example
@noindent
will send all further debug and trace output to @var{filename}. If
@var{filename} is empty, debug and trace output are discarded and if
@code{debugfile} is called without any arguments, debug and trace output
are sent to the standard error output.
@node Input Control, File Inclusion, Debugging, Top
@chapter Input control
This chapter describes various builtin macros for controlling the input
to @code{m4}.
@menu
* Dnl:: Deleting whitespace in input
* Changequote:: Changing the quote characters
* Changecom:: Changing the comment delimiters
* Changeword:: Changing the lexical structure of words
* M4wrap:: Saving input until end of input
@end menu
@node Dnl, Changequote, Input Control, Input Control
@section Deleting whitespace in input
@cindex deleting whitespace in input
@findex dnl
The builtin @code{dnl} reads and discards all characters, up to and
including the first newline:
@comment ignore
@example
dnl
@end example
@noindent
and it is often used in connection with @code{define}, to remove the
newline that follow the call to @code{define}. Thus
@example
define(`foo', `Macro `foo'.')dnl A very simple macro, indeed.
foo
@result{}Macro foo.
@end example
The input up to and including the next newline is discarded, as opposed
to the way comments are treated (@pxref{Comments}).
Usually, @code{dnl} is immediately followed by an end of line or some
other whitespace. GNU @code{m4} will produce a warning diagnostic if
@code{dnl} is followed by an open parenthesis. In this case, @code{dnl}
will collect and process all arguments, looking for a matching close
parenthesis. All predictable side effects resulting from this
collection will take place. @code{dnl} will return no output. The
input following the matching close parenthesis up to and including the
next newline, on whatever line containing it, will still be discarded.
@node Changequote, Changecom, Dnl, Input Control
@section Changing the quote characters
@cindex changing the quote delimiters
@cindex quote delimiters, changing the
@findex changequote
The default quote delimiters can be changed with the builtin
@code{changequote}:
@comment ignore
@example
changequote(opt @var{start}, opt @var{end})
@end example
@noindent
where @var{start} is the new start-quote delimiter and @var{end} is the
new end-quote delimiter. If any of the arguments are missing, the default
quotes (@code{`} and @code{'}) are used instead of the void arguments.
The expansion of @code{changequote} is void.
@example
changequote([, ])
@result{}
define([foo], [Macro [foo].])
@result{}
foo
@result{}Macro foo.
@end example
If no single character is appropriate, @var{start} and @var{end} can be
of any length.
@example
changequote([[, ]])
@result{}
define([[foo]], [[Macro [[[foo]]].]])
@result{}
foo
@result{}Macro [foo].
@end example
Changing the quotes to the empty strings will effectively disable the
quoting mechanism, leaving no way to quote text.
@example
define(`foo', `Macro `FOO'.')
@result{}
changequote(, )
@result{}
foo
@result{}Macro `FOO'.
`foo'
@result{}`Macro `FOO'.'
@end example
There is no way in @code{m4} to quote a string containing an unmatched
left quote, except using @code{changequote} to change the current
quotes.
Neither quote string should start with a letter or @samp{_} (underscore),
as they will be confused with names in the input. Doing so disables
the quoting mechanism.
@node Changecom, Changeword, Changequote, Input Control
@section Changing comment delimiters
@cindex changing comment delimiters
@cindex comment delimiters, changing
@findex changecom
The default comment delimiters can be changed with the builtin
macro @code{changecom}:
@comment ignore
@example
changecom(opt @var{start}, opt @var{end})
@end example
@noindent
where @var{start} is the new start-comment delimiter and @var{end} is
the new end-comment delimiter. If any of the arguments are void, the
default comment delimiters (@code{#} and newline) are used instead of
the void arguments. The comment delimiters can be of any length.
The expansion of @code{changecom} is void.
@example
define(`comment', `COMMENT')
@result{}
# A normal comment
@result{}# A normal comment
changecom(`/*', `*/')
@result{}
# Not a comment anymore
@result{}# Not a COMMENT anymore
But: /* this is a comment now */ while this is not a comment
@result{}But: /* this is a comment now */ while this is not a COMMENT
@end example
@cindex comments, copied to output
Note how comments are copied to the output, much as if they were quoted
strings. If you want the text inside a comment expanded, quote the
start comment delimiter.
Calling @code{changecom} without any arguments disables the commenting
mechanism completely.
@example
define(`comment', `COMMENT')
@result{}
changecom
@result{}
# Not a comment anymore
@result{}# Not a COMMENT anymore
@end example
@node Changeword, M4wrap, Changecom, Input Control
@section Changing the lexical structure of words
@cindex lexical structure of words
@cindex words, lexical structure of
@findex changeword
@quotation
The macro @code{changeword} and all associated functionnality is
experimental. It is only available if the @code{--enable-changeword}
option was given to @code{configure}, at GNU @code{m4} installation
time. The functionnality might change or even go away in the future.
@emph{Do not rely on it}. Please direct your comments about it the
same way you would do for bugs.
@end quotation
A file being processed by @code{m4} is split into quoted strings, words
(potential macro names) and simple tokens (any other single character).
Initially a word is defined by the following regular expression:
@comment ignore
@example
[_a-zA-Z][_a-zA-Z0-9]*
@end example
Using @code{changeword}, you can change this regular expression. Relaxing
@code{m4}'s lexical rules might be useful (for example) if you wanted to
apply translations to a file of numbers:
@comment ignore
@example
changeword(`[_a-zA-Z0-9]+')
define(1, 0)
@result{}1
@end example
Tightening the lexical rules is less useful, because it will generally
make some of the builtins unavailable. You could use it to prevent
accidental call of builtins, for example:
@comment ignore
@example
define(`_indir', defn(`indir'))
changeword(`_[_a-zA-Z0-9]*')
esyscmd(foo)
_indir(`esyscmd', `ls')
@end example
Because @code{m4} constructs its words a character at a time, there
is a restriction on the regular expressions that may be passed to
@code{changeword}. This is that if your regular expression accepts
@samp{foo}, it must also accept @samp{f} and @samp{fo}.
@code{changeword} has another function. If the regular expression
supplied contains any bracketed subexpressions, then text outside
the first of these is discarded before symbol lookup. So:
@comment ignore
@example
changecom(`/*', `*/')
changeword(`#\([_a-zA-Z0-9]*\)')
#esyscmd(ls)
@end example
@code{m4} now requires a @samp{#} mark at the beginning of every
macro invocation, so one can use @code{m4} to preprocess shell
scripts without getting @code{shift} commands swallowed, and plain
text without losing various common words.
@code{m4}'s macro substitution is based on text, while @TeX{}'s is based
on tokens. @code{changeword} can throw this difference into relief. For
example, here is the same idea represented in @TeX{} and @code{m4}.
First, the @TeX{} version:
@comment ignore
@example
\def\a@{\message@{Hello@}@}
\catcode`\@@=0
\catcode`\\=12
@result{}@@a
@result{}@@bye
@end example
@noindent
Then, the @code{m4} version:
@comment ignore
@example
define(a, `errprint(`Hello')')
changeword(`@@\([_a-zA-Z0-9]*\)')
@result{}@@a
@end example
In the @TeX{} example, the first line defines a macro @code{a} to
print the message @samp{Hello}. The second line defines @key{@@} to
be usable instead of @key{\} as an escape character. The third line
defines @key{\} to be a normal printing character, not an escape.
The fourth line invokes the macro @code{a}. So, when @TeX{} is run
on this file, it displays the message @samp{Hello}.
When the @code{m4} example is passed through @code{m4}, it outputs
@samp{errprint(Hello)}. The reason for this is that @TeX{} does
lexical analysis of macro definition when the macro is @emph{defined}.
@code{m4} just stores the text, postponing the lexical analysis until
the macro is @emph{used}.
You should note that using @code{changeword} will slow @code{m4} down
by a factor of about seven.
@node M4wrap, , Changeword, Input Control
@section Saving input
@cindex saving input
@cindex input, saving
@findex m4wrap
It is possible to `save' some text until the end of the normal input has
been seen. Text can be saved, to be read again by @code{m4} when the
normal input has been exhausted. This feature is normally used to
initiate cleanup actions before normal exit, e.g., deleting temporary
files.
To save input text, use the builtin @code{m4wrap}:
@comment ignore
@example
m4wrap(@var{string}, ...)
@end example
@noindent
which stores @var{string} and the rest of the arguments in a safe place,
to be reread when end of input is reached.
@example
define(`cleanup', `This is the `cleanup' actions.
')
@result{}
m4wrap(`cleanup')
@result{}
This is the first and last normal input line.
@result{}This is the first and last normal input line.
^D
@result{}This is the cleanup actions.
@end example
The saved input is only reread when the end of normal input is seen, and
not if @code{m4exit} is used to exit @code{m4}.
It is safe to call @code{m4wrap} from saved text, but then the order in
which the saved text is reread is undefined. If @code{m4wrap} is not used
recursively, the saved pieces of text are reread in the opposite order
in which they were saved (LIFO---last in, first out).
@node File Inclusion, Diversions, Input Control, Top
@chapter File inclusion
@cindex file inclusion
@cindex inclusion, of files
@code{m4} allows you to include named files at any point in the input.
@menu
* Include:: Including named files
* Search Path:: Searching for include files
@end menu
@node Include, Search Path, File Inclusion, File Inclusion
@section Including named files
@findex include
@findex sinclude
There are two builtin macros in @code{m4} for including files:
@comment ignore
@example
include(@var{filename})
sinclude(@var{filename})
@end example
@noindent
both of which cause the file named @var{filename} to be read by
@code{m4}. When the end of the file is reached, input is resumed from
the previous input file.
The expansion of @code{include} and @code{sinclude} is therefore the
contents of @var{filename}.
It is an error for an @code{include}d file not to exist. If you do
not want error messages about non-existent files, @code{sinclude} can
be used to include a file, if it exists, expanding to nothing if it
does not.
@example
include(`no-such-file')
@result{}
@error{}30.include:2: m4: Cannot open no-such-file: No such file or directory
sinclude(`no-such-file')
@result{}
@end example
Assume in the following that the file @file{incl.m4} contains the lines:
@comment ignore
@example
Include file start
foo
Include file end
@end example
@noindent
Normally file inclusion is used to insert the contents of a file
into the input stream. The contents of the file will be read by
@code{m4} and macro calls in the file will be expanded:
@example
define(`foo', `FOO')
@result{}
include(`incl.m4')
@result{}Include file start
@result{}FOO
@result{}Include file end
@result{}
@end example
The fact that @code{include} and @code{sinclude} expand to the contents
of the file can be used to define macros that operate on entire files.
Here is an example, which defines @samp{bar} to expand to the contents
of @file{incl.m4}:
@example
define(`bar', include(`incl.m4'))
@result{}
This is `bar': >>>bar<<<
@result{}This is bar: >>>Include file start
@result{}foo
@result{}Include file end
@result{}<<<
@end example
This use of @code{include} is not trivial, though, as files can contain
quotes, commas and parentheses, which can interfere with the way the
@code{m4} parser works.
The builtin macros @code{include} and @code{sinclude} are recognized
only when given arguments.
@node Search Path, , Include, File Inclusion
@section Searching for include files
@cindex search path for included files
@cindex included files, search path for
@cindex GNU extensions
GNU @code{m4} allows included files to be found in other directories
than the current working directory.
If a file is not found in the current working directory, and the file
name is not absolute, the file will be looked for in a specified search
path. First, the directories specified with the @samp{-I} option will
be searched, in the order found on the command line. Second, if the
@samp{M4PATH} environment variable is set, it is expected to contain a
colon-separated list of directories, which will be searched in order.
If the automatic search for include-files causes trouble, the @samp{p}
debug flag (@pxref{Debug Levels}) can help isolate the problem.
@node Diversions, Text handling, File Inclusion, Top
@chapter Diverting and undiverting output
Diversions are a way of temporarily saving output. The output of
@code{m4} can at any time be diverted to a temporary file, and be
reinserted into the output stream, @dfn{undiverted}, again at a later
time.
Numbered diversions are counted from 0 upwards, diversion number 0
being the normal output stream. The number of simultaneous diversions
is limited mainly by the memory used to describe them, because GNU
@code{m4} tries to keep diversions in memory. However, there is a
limit to the overall memory usable by all diversions taken altogether
(512K, currently). When this maximum is about to be exceeded,
a temporary file is opened to receive the contents of the biggest
diversion still in memory, freeing this memory for other diversions.
So, it is theoretically possible that the number of diversions be
limited by the number of available file descriptors.
@c FIXME: need some explanation here why this is a useful feature, not
@c just how you use it.
@menu
* Divert:: Diverting output
* Undivert:: Undiverting output
* Divnum:: Diversion numbers
* Cleardiv:: Discarding diverted text
@end menu
@node Divert, Undivert, Diversions, Diversions
@section Diverting output
@cindex diverting output to files
@cindex output, diverting to files
@cindex files, diverting output to
@findex divert
Output is diverted using @code{divert}:
@comment ignore
@example
divert(opt @var{number})
@end example
@noindent
where @var{number} is the diversion to be used. If @var{number} is left
out, it is assumed to be zero.
The expansion of @code{divert} is void.
When all the @code{m4} input will have been processed, all existing
diversions are automatically undiverted, in numerical order.
@example
divert(1)
This text is diverted.
divert
@result{}
This text is not diverted.
@result{}This text is not diverted.
^D
@result{}
@result{}This text is diverted.
@end example
Several calls of @code{divert} with the same argument do not overwrite
the previous diverted text, but append to it.
If output is diverted to a non-existent diversion, it is simply
discarded. This can be used to suppress unwanted output. A common
example of unwanted output is the trailing newlines after macro
definitions. Here is how to avoid them.
@example
divert(-1)
define(`foo', `Macro `foo'.')
define(`bar', `Macro `bar'.')
divert
@result{}
@end example
This is a common programming idiom in @code{m4}.
@node Undivert, Divnum, Divert, Diversions
@section Undiverting output
@findex undivert
Diverted text can be undiverted explicitly using the builtin
@code{undivert}:
@comment ignore
@example
undivert(opt @var{number}, ...)
@end example
@noindent
which undiverts the diversions given by the arguments, in the order
given. If no arguments are supplied, all diversions are undiverted, in
numerical order.
@c FIXME: Explain what happens when undiverting all to else than 0.
The expansion of @code{undivert} is void.
@example
divert(1)
This text is diverted.
divert
@result{}
This text is not diverted.
@result{}This text is not diverted.
undivert(1)
@result{}
@result{}This text is diverted.
@result{}
@end example
Notice the last two blank lines. One of them comes from the newline
following @code{undivert}, the other from the newline that followed the
@code{divert}! A diversion often starts with a blank line like this.
When diverted text is undiverted, it is @emph{not} reread by @code{m4},
but rather copied directly to the current output, and it is therefore
not an error to undivert into a diversion.
When a diversion has been undiverted, the diverted text is discarded,
and it is not possible to bring back diverted text more than once.
@example
divert(1)
This text is diverted first.
divert(0)undivert(1)dnl
@result{}
@result{}This text is diverted first.
undivert(1)
@result{}
divert(1)
This text is also diverted but not appended.
divert(0)undivert(1)dnl
@result{}
@result{}This text is also diverted but not appended.
@end example
Attempts to undivert the current diversion are silently ignored.
@cindex GNU extensions
@cindex file inclusion
@cindex inclusion, of files
GNU @code{m4} allows named files to be undiverted. Given a non-numeric
argument, the contents of the file named will be copied, uninterpreted, to
the current output. This complements the builtin @code{include}
(@pxref{Include}). To illustrate the difference, assume the file
@file{foo} contains the word @samp{bar}:
@example
define(`bar', `BAR')
@result{}
undivert(`foo')
@result{}bar
@result{}
include(`foo')
@result{}BAR
@result{}
@end example
@node Divnum, Cleardiv, Undivert, Diversions
@section Diversion numbers
@cindex diversion numbers
@findex divnum
The builtin @code{divnum}:
@comment ignore
@example
divnum
@end example
@noindent
expands to the number of the current diversion.
@example
Initial divnum
@result{}Initial 0
divert(1)
Diversion one: divnum
divert(2)
Diversion two: divnum
divert
@result{}
^D
@result{}
@result{}Diversion one: 1
@result{}
@result{}Diversion two: 2
@end example
The last call of @code{divert} without argument is necessary, since the
undiverted text would otherwise be diverted itself.
@node Cleardiv, , Divnum, Diversions
@section Discarding diverted text
@cindex discarding diverted text
@cindex diverted text, discarding
Often it is not known, when output is diverted, whether the diverted
text is actually needed. Since all non-empty diversion are brought back
on the main output stream when the end of input is seen, a method of
discarding a diversion is needed. If all diversions should be
discarded, the easiest is to end the input to @code{m4} with
@samp{divert(-1)} followed by an explicit @samp{undivert}:
@example
divert(1)
Diversion one: divnum
divert(2)
Diversion two: divnum
divert(-1)
undivert
^D
@end example
@noindent
No output is produced at all.
Clearing selected diversions can be done with the following macro:
@example
define(`cleardivert',
`pushdef(`_num', divnum)divert(-1)undivert($@@)divert(_num)popdef(`_num')')
@result{}
@end example
It is called just like @code{undivert}, but the effect is to clear the
diversions, given by the arguments. (This macro has a nasty bug! You
should try to see if you can find it and correct it.)
@node Text handling, Arithmetic, Diversions, Top
@chapter Macros for text handling
There are a number of builtins in @code{m4} for manipulating text in
various ways, extracting substrings, searching, substituting, and so on.
@menu
* Len:: Calculating length of strings
* Index:: Searching for substrings
* Regexp:: Searching for regular expressions
* Substr:: Extracting substrings
* Translit:: Translating characters
* Patsubst:: Substituting text by regular expression
* Format:: Formatting strings (printf-like)
@end menu
@node Len, Index, Text handling, Text handling
@section Calculating length of strings
@cindex length of strings
@cindex strings, length of
@findex len
The length of a string can be calculated by @code{len}:
@comment ignore
@example
len(@var{string})
@end example
@noindent
which expands to the length of @var{string}, as a decimal number.
@example
len()
@result{}0
len(`abcdef')
@result{}6
@end example
The builtin macro @code{len} is recognized only when given arguments.
@node Index, Regexp, Len, Text handling
@section Searching for substrings
@findex index
Searching for substrings is done with @code{index}:
@comment ignore
@example
index(@var{string}, @var{substring})
@end example
@noindent
which expands to the index of the first occurrence of @var{substring} in
@var{string}. The first character in @var{string} has index 0. If
@var{substring} does not occur in @var{string}, @code{index} expands to
@samp{-1}.
@example
index(`gnus, gnats, and armadillos', `nat')
@result{}7
index(`gnus, gnats, and armadillos', `dag')
@result{}-1
@end example
The builtin macro @code{index} is recognized only when given arguments.
@node Regexp, Substr, Index, Text handling
@section Searching for regular expressions
@cindex regular expressions
@cindex GNU extensions
@findex regexp
Searching for regular expressions is done with the builtin
@code{regexp}:
@comment ignore
@example
regexp(@var{string}, @var{regexp}, opt @var{replacement})
@end example
@noindent
which searches for @var{regexp} in @var{string}. The syntax for regular
expressions is the same as in GNU Emacs. @xref{Regexps, , Syntax of
Regular Expressions, emacs, The GNU Emacs Manual}.
If @var{replacement} is omitted, @code{regexp} expands to the index of
the first match of @var{regexp} in @var{string}. If @var{regexp} does
not match anywhere in @var{string}, it expands to -1.
@example
regexp(`GNUs not Unix', `\<[a-z]\w+')
@result{}5
regexp(`GNUs not Unix', `\<Q\w*')
@result{}-1
@end example
If @var{replacement} is supplied, @code{regexp} changes the expansion
to this argument, with @samp{\@var{n}} substituted by the text
matched by the @var{n}th parenthesized sub-expression of @var{regexp},
@samp{\&} being the text the entire regular expression matched.
@example
regexp(`GNUs not Unix', `\w\(\w+\)$', `*** \& *** \1 ***')
@result{}*** Unix *** nix ***
@end example
The builtin macro @code{regexp} is recognized only when given arguments.
@node Substr, Translit, Regexp, Text handling
@section Extracting substrings
@cindex extracting substrings
@cindex substrings, extracting
@findex substr
Substrings are extracted with @code{substr}:
@comment ignore
@example
substr(@var{string}, @var{from}, opt @var{length})
@end example
@noindent
which expands to the substring of @var{string}, which starts at index
@var{from}, and extends for @var{length} characters, or to the end of
@var{string}, if @var{length} is omitted. The starting index of a string
is always 0.
@example
substr(`gnus, gnats, and armadillos', 6)
@result{}gnats, and armadillos
substr(`gnus, gnats, and armadillos', 6, 5)
@result{}gnats
@end example
The builtin macro @code{substr} is recognized only when given arguments.
@node Translit, Patsubst, Substr, Text handling
@section Translating characters
@cindex translating characters
@cindex characters, translating
@findex translit
Character translation is done with @code{translit}:
@comment ignore
@example
translit(@var{string}, @var{chars}, @var{replacement})
@end example
@noindent
which expands to @var{string}, with each character that occurs in
@var{chars} translated into the character from @var{replacement} with
the same index.
If @var{replacement} is shorter than @var{chars}, the excess characters
are deleted from the expansion. If @var{replacement} is omitted, all
characters in @var{string}, that are present in @var{chars} are deleted
from the expansion.
Both @var{chars} and @var{replacement} can contain character-ranges,
e.g., @samp{a-z} (meaning all lowercase letters) or @samp{0-9} (meaning
all digits). To include a dash @samp{-} in @var{chars} or
@var{replacement}, place it first or last.
It is not an error for the last character in the range to be `larger'
than the first. In that case, the range runs backwards, i.e.,
@samp{9-0} means the string @samp{9876543210}.
@example
translit(`GNUs not Unix', `A-Z')
@result{}s not nix
translit(`GNUs not Unix', `a-z', `A-Z')
@result{}GNUS NOT UNIX
translit(`GNUs not Unix', `A-Z', `z-a')
@result{}tmfs not fnix
@end example
The first example deletes all uppercase letters, the second converts
lowercase to uppercase, and the third `mirrors' all uppercase letters,
while converting them to lowercase. The two first cases are by far the
most common.
The builtin macro @code{translit} is recognized only when given
arguments.
@node Patsubst, Format, Translit, Text handling
@section Substituting text by regular expression
@cindex regular expressions
@cindex pattern substitution
@cindex substitution by regular expression
@cindex GNU extensions
@findex patsubst
Global substitution in a string is done by @code{patsubst}:
@comment ignore
@example
patsubst(@var{string}, @var{regexp}, opt @var{replacement})
@end example
@noindent
which searches @var{string} for matches of @var{regexp}, and substitutes
@var{replacement} for each match. The syntax for regular expressions
is the same as in GNU Emacs.
The parts of @var{string} that are not covered by any match of
@var{regexp} are copied to the expansion. Whenever a match is found, the
search proceeds from the end of the match, so a character from
@var{string} will never be substituted twice. If @var{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, @var{replacement} is inserted into
the expansion, with @samp{\@var{n}} substituted by the text matched by
the @var{n}th parenthesized sub-expression of @var{regexp}, @samp{\&}
being the text the entire regular expression matched.
The @var{replacement} argument can be omitted, in which case the text
matched by @var{regexp} is deleted.
@example
patsubst(`GNUs not Unix', `^', `OBS: ')
@result{}OBS: GNUs not Unix
patsubst(`GNUs not Unix', `\<', `OBS: ')
@result{}OBS: GNUs OBS: not OBS: Unix
patsubst(`GNUs not Unix', `\w*', `(\&)')
@result{}(GNUs)() (not)() (Unix)
patsubst(`GNUs not Unix', `\w+', `(\&)')
@result{}(GNUs) (not) (Unix)
patsubst(`GNUs not Unix', `[A-Z][a-z]+')
@result{}GN not
@end example
Here is a slightly more realistic example, which capitalizes individual
word or whole sentences, by substituting calls of the macros
@code{upcase} and @code{downcase} into the strings.
@example
define(`upcase', `translit(`$*', `a-z', `A-Z')')dnl
define(`downcase', `translit(`$*', `A-Z', `a-z')')dnl
define(`capitalize1',
`regexp(`$1', `^\(\w\)\(\w*\)', `upcase(`\1')`'downcase(`\2')')')dnl
define(`capitalize',
`patsubst(`$1', `\w+', `capitalize1(`\&')')')dnl
capitalize(`GNUs not Unix')
@result{}Gnus Not Unix
@end example
The builtin macro @code{patsubst} is recognized only when given
arguments.
@node Format, , Patsubst, Text handling
@section Formatted output
@cindex formatted output
@cindex output, formatted
@cindex GNU extensions
@findex format
Formatted output can be made with @code{format}:
@comment ignore
@example
format(@var{format-string}, ...)
@end example
@noindent
which works much like the C function @code{printf}. The first argument
is a format string, which can contain @samp{%} specifications, and the
expansion of @code{format} is the formatted string.
Its use is best described by a few examples:
@example
define(`foo', `The brown fox jumped over the lazy dog')
@result{}
format(`The string "%s" is %d characters long', foo, len(foo))
@result{}The string "The brown fox jumped over the lazy dog" is 38 characters long
@end example
Using the @code{forloop} macro defined in @xref{Loops}, this
example shows how @code{format} can be used to produce tabular output.
@comment ignore
@example
forloop(`i', 1, 10, `format(`%6d squared is %10d
', i, eval(i**2))')
@result{} 1 squared is 1
@result{} 2 squared is 4
@result{} 3 squared is 9
@result{} 4 squared is 16
@result{} 5 squared is 25
@result{} 6 squared is 36
@result{} 7 squared is 49
@result{} 8 squared is 64
@result{} 9 squared is 81
@result{} 10 squared is 100
@end example
The builtin @code{format} is modeled after the ANSI C @samp{printf}
function, and supports the normal @samp{%} specifiers: @samp{c},
@samp{s}, @samp{d}, @samp{o}, @samp{x}, @samp{X}, @samp{u}, @samp{e},
@samp{E} and @samp{f}; it supports field widths and precisions, and the
modifiers @samp{+}, @samp{-}, @samp{@w{ }}, @samp{0}, @samp{#}, @samp{h} and
@samp{l}. For more details on the functioning of @code{printf}, see the
C Library Manual.
@c FIXME: Why format does not require at least one argument?
@node Arithmetic, UNIX commands, Text handling, Top
@chapter Macros for doing arithmetic
@cindex arithmetic
@cindex integer arithmetic
Integer arithmetic is included in @code{m4}, with a C-like syntax. As
convenient shorthands, there are builtins for simple increment and
decrement operations.
@menu
* Incr:: Decrement and increment operators
* Eval:: Evaluating integer expressions
@end menu
@node Incr, Eval, Arithmetic, Arithmetic
@section Decrement and increment operators
@cindex decrement operator
@cindex increment operator
@findex incr
@findex decr
Increment and decrement of integers are supported using the builtins
@code{incr} and @code{decr}:
@comment ignore
@example
incr(@var{number})
decr(@var{number})
@end example
@noindent
which expand to the numerical value of @var{number}, incremented,
or decremented, respectively, by one.
@example
incr(4)
@result{}5
decr(7)
@result{}6
@end example
The builtin macros @code{incr} and @code{decr} are recognized only when
given arguments.
@node Eval, , Incr, Arithmetic
@section Evaluating integer expressions
@cindex integer expression evaluation
@cindex evaluation, of integer expressions
@cindex expressions, evaluation of integer
@findex eval
Integer expressions are evaluated with @code{eval}:
@comment ignore
@example
eval(@var{expression}, opt @var{radix}, opt @var{width})
@end example
@noindent
which expands to the value of @var{expression}.
Expressions can contain the following operators, listed in order of
decreasing precedence.
@table @code
@item -
Unary minus
@item **
Exponentiation
@item * / %
Multiplication, division and modulo
@item + -
Addition and subtraction
@item << >>
Shift left or right
@item == != > >= < <=
Relational operators
@item !
Logical negation
@item ~
Bitwise negation
@item &
Bitwise and
@item ^
Bitwise exclusive-or
@item |
Bitwise or
@item &&
Logical and
@item ||
Logical or
@end table
All operators, except exponentiation, are left associative.
Note that many @code{m4} implementations use @samp{^} as an alternate
operator for the exponentiation, while many others use @samp{^} for the
bitwise exclusive-or. GNU @code{m4} changed its behavior: it used to
exponentiate for @samp{^}, it now computes the bitwise exclusive-or.
Numbers without special prefix are given decimal. A simple @samp{0}
prefix introduces an octal number. @samp{0x} introduces an hexadecimal
number. @samp{0b} introduces a binary number. @samp{0r} introduces a
number expressed in any radix between 1 and 36: the prefix should be
immediately followed by the decimal expression of the radix, a colon,
then the digits making the number. For any radix, the digits are
@samp{0}, @samp{1}, @samp{2}, @dots{}. Beyond @samp{9}, the digits are
@samp{a}, @samp{b} @dots{} up to @samp{z}. Lower and upper case letters
can be used interchangeably in numbers prefixes and as number digits.
Parentheses may be used to group subexpressions whenever needed. For the
relational operators, a true relation returns @code{1}, and a false
relation return @code{0}.
Here are a few examples of use of @code{eval}.
@example
eval(-3 * 5)
@result{}-15
eval(index(`Hello world', `llo') >= 0)
@result{}1
define(`square', `eval(($1)**2)')
@result{}
square(9)
@result{}81
square(square(5)+1)
@result{}676
define(`foo', `666')
@result{}
eval(`foo'/6)
@error{}51.eval:14: m4: Bad expression in eval: foo/6
@result{}
eval(foo/6)
@result{}111
@end example
As the second to last example shows, @code{eval} does not handle macro
names, even if they expand to a valid expression (or part of a valid
expression). Therefore all macros must be expanded before they are
passed to @code{eval}.
If @var{radix} is specified, it specifies the radix to be used in the
expansion. The default radix is 10. The result of @code{eval} is
always taken to be signed. The @var{width} argument specifies a minimum
output width. The result is zero-padded to extend the expansion to the
requested width.
@example
eval(666, 10)
@result{}666
eval(666, 11)
@result{}556
eval(666, 6)
@result{}3030
eval(666, 6, 10)
@result{}0000003030
eval(-666, 6, 10)
@result{}-000003030
@end example
Take note that @var{radix} cannot be larger than 36.
The builtin macro @code{eval} is recognized only when given arguments.
@node UNIX commands, Miscellaneous, Arithmetic, Top
@chapter Running UNIX commands
@cindex executing UNIX commands
@cindex running UNIX commands
@cindex UNIX commands, running
@cindex commands, running UNIX
There are a few builtin macros in @code{m4} that allow you to run UNIX
commands from within @code{m4}.
@menu
* Syscmd:: Executing simple commands
* Esyscmd:: Reading the output of commands
* Sysval:: Exit codes
* Maketemp:: Making names for temporary files
@end menu
@node Syscmd, Esyscmd, UNIX commands, UNIX commands
@section Executing simple commands
@findex syscmd
Any shell command can be executed, using @code{syscmd}:
@comment ignore
@example
syscmd(@var{shell-command})
@end example
@noindent
which executes @var{shell-command} as a shell command.
The expansion of @code{syscmd} is void, @emph{not} the output from
@var{shell-command}! Output or error messages from @var{shell-command}
are not read by @code{m4}. @xref{Esyscmd}, if you need to process the
command output.
Prior to executing the command, @code{m4} flushes its output buffers.
The default standard input, output and error of @var{shell-command} are
the same as those of @code{m4}.
The builtin macro @code{syscmd} is recognized only when given arguments.
@node Esyscmd, Sysval, Syscmd, UNIX commands
@section Reading the output of commands
@findex esyscmd
@cindex GNU extensions
If you want @code{m4} to read the output of a UNIX command, use
@code{esyscmd}:
@comment ignore
@example
esyscmd(@var{shell-command})
@end example
@noindent
which expands to the standard output of the shell command
@var{shell-command}.
Prior to executing the command, @code{m4} flushes its output buffers.
The default standard input and error output of @var{shell-command} are
the same as those of @code{m4}. The error output of @var{shell-command}
is not a part of the expansion: it will appear along with the error
output of @code{m4}.
Assume you are positioned into the @file{checks} directory of GNU
@code{m4} distribution, then:
@example
define(`vice', `esyscmd(grep Vice ../COPYING)')
@result{}
vice
@result{} Ty Coon, President of Vice
@result{}
@end example
Note how the expansion of @code{esyscmd} has a trailing newline.
The builtin macro @code{esyscmd} is recognized only when given
arguments.
@node Sysval, Maketemp, Esyscmd, UNIX commands
@section Exit codes
@cindex exit code from UNIX commands
@cindex UNIX commands, exit code from
@cindex commands, exit code from UNIX
@findex sysval
To see whether a shell command succeeded, use @code{sysval}:
@comment ignore
@example
sysval
@end example
@noindent
which expands to the exit status of the last shell command run with
@code{syscmd} or @code{esyscmd}.
@example
syscmd(`false')
@result{}
ifelse(sysval, 0, zero, non-zero)
@result{}non-zero
syscmd(`true')
@result{}
sysval
@result{}0
@end example
@node Maketemp, , Sysval, UNIX commands
@section Making names for temporary files
@cindex temporary filenames
@cindex files, names of temporary
@findex maketemp
Commands specified to @code{syscmd} or @code{esyscmd} might need a
temporary file, for output or for some other purpose.
There is a builtin macro, @code{maketemp}, for making temporary file
names:
@comment ignore
@example
maketemp(@var{template})
@end example
@noindent
which expands to a name of a new, empty file, made from the string
@var{template}, which should end with the string @samp{XXXXXX}. The six
@code{X}'s are then replaced, usually with something that includes the
process id of the @code{m4} process, in order to make the filename unique.
@comment ignore
@example
maketemp(`/tmp/fooXXXXXX')
@result{}/tmp/fooa07346
@end example
The builtin macro @code{maketemp} is recognized only when given
arguments.
@node Miscellaneous, Frozen files, UNIX commands, Top
@chapter Miscellaneous builtin macros
This chapter describes various builtins, that do not really belong in
any of the previous chapters.
@menu
* Errprint:: Printing error messages
* M4exit:: Exiting from m4
@end menu
@node Errprint, M4exit, Miscellaneous, Miscellaneous
@section Printing error messages
@cindex printing error messages
@cindex error messages, printing
@cindex messages, printing error
@findex errprint
You can print error messages using @code{errprint}:
@comment ignore
@example
errprint(@var{message}, ...)
@end example
@noindent
which simply prints @var{message} and the rest of the arguments on the
standard error output.
The expansion of @code{errprint} is void.
@example
errprint(`Illegal arguments to forloop
')
@error{}Illegal arguments to forloop
@result{}
@end example
A trailing newline is @emph{not} printed automatically, so it must be
supplied as part of the argument, as in the example. (BSD flavored
@code{m4}'s do append a trailing newline on each @code{errprint} call).
To make it possible to specify the location of the error, two
utility builtins exist:
@findex file
@findex line
@comment ignore
@example
__file__
__line__
@end example
@noindent
which expands to the quoted name of the current input file, and the
current input line number in that file.
@example
errprint(`m4:'__file__:__line__: `Input error
')
@error{}m4:56.errprint:2: Input error
@result{}
@end example
@node M4exit, , Errprint, Miscellaneous
@section Exiting from @code{m4}
@cindex exiting from @code{m4}
@findex m4exit
If you need to exit from @code{m4} before the entire input has been
read, you can use @code{m4exit}:
@comment ignore
@example
m4exit(opt @var{code})
@end example
@noindent
which causes @code{m4} to exit, with exit code @var{code}. If
@var{code} is left out, the exit code is zero.
@example
define(`fatal_error', `errprint(`m4: '__file__: __line__`: fatal error: $*
')m4exit(1)')
@result{}
fatal_error(`This is a BAD one, buster')
@error{}m4: 57.m4exit: 5: fatal error: This is a BAD one, buster
@end example
After this macro call, @code{m4} will exit with exit code 1. This macro
is only intended for error exits, since the normal exit procedures are
not followed, e.g., diverted text is not undiverted, and saved text
(@pxref{M4wrap}) is not reread.
@node Frozen files, Compatibility, Miscellaneous, Top
@chapter Fast loading of frozen states
@cindex fast loading of frozen files
@cindex frozen files for fast loading
@cindex initialization, frozen states
@cindex dumping into frozen file
@cindex reloading a frozen file
@cindex GNU extensions
Some bigger @code{m4} applications may be built over a common base
containing hundreds of definitions and other costly initializations.
Usually, the common base is kept in one or more declarative files,
which files are listed on each @code{m4} invocation prior to the
user's input file, or else, @code{include}'d from this input file.
Reading the common base of a big application, over and over again, may
be time consuming. GNU @code{m4} offers some machinery to speed up
the start of an application using lengthy common bases. Presume the
user repeatedly uses:
@comment ignore
@example
m4 base.m4 input.m4
@end example
@noindent
with a varying contents of @file{input.m4}, but a rather fixed contents
for @file{base.m4}. Then, the user might rather execute:
@comment ignore
@example
m4 -F base.m4f base.m4
@end example
@noindent
once, and further execute, as often as needed:
@comment ignore
@example
m4 -R base.m4f input.m4
@end example
@noindent
with the varying input. The first call, containing the @code{-F}
option, only reads and executes file @file{base.m4}, so defining
various application macros and computing other initializations. Only
once the input file @file{base.m4} has been completely processed, GNU
@code{m4} produces on @file{base.m4f} a @dfn{frozen} file, that is, a
file which contains a kind of snapshot of the @code{m4} internal state.
Later calls, containing the @code{-R} option, are able to reload
the internal state of @code{m4}'s memory, from @file{base.m4f},
@emph{prior} to reading any other input files. By this mean,
instead of starting with a virgin copy of @code{m4}, input will be
read after having effectively recovered the effect of a prior run.
In our example, the effect is the same as if file @file{base.m4} has
been read anew. However, this effect is achieved a lot faster.
Only one frozen file may be created or read in any one @code{m4}
invocation. It is not possible to recover two frozen files at once.
However, frozen files may be updated incrementally, through using
@code{-R} and @code{-F} options simultaneously. For example, if
some care is taken, the command:
@comment ignore
@example
m4 file1.m4 file2.m4 file3.m4 file4.m4
@end example
@noindent
could be broken down in the following sequence, accumulating the same
output:
@comment ignore
@example
m4 -F file1.m4f file1.m4
m4 -R file1.m4f -F file2.m4f file2.m4
m4 -R file2.m4f -F file3.m4f file3.m4
m4 -R file3.m4f file4.m4
@end example
Some care is necessary because not every effort has been made for
this to work in all cases. In particular, the trace attribute of
macros is not handled, nor the current setting of @code{changeword}.
Also, interactions for some options of @code{m4} being used in one call
and not for the next, have not been fully analyzed yet. On the other
end, you may be confident that stacks of @code{pushdef}'ed definitions
are handled correctly, so are @code{undefine}'d or renamed builtins,
changed strings for quotes or comments.
When an @code{m4} run is to be frozen, the automatic undiversion
which takes place at end of execution is inhibited. Instead, all
positively numbered diversions are saved into the frozen file.
The active diversion number is also transmitted.
A frozen file to be reloaded need not reside in the current directory.
It is looked up the same way as an @code{include} file (@pxref{Search
Path}).
Frozen files are sharable across architectures. It is safe to write
a frozen file one one machine and read it on another, given that the
second machine uses the same, or a newer version of GNU @code{m4}.
These are simple (editable) text files, made up of directives,
each starting with a capital letter and ending with a newline
(@key{NL}). Wherever a directive is expected, the character
@kbd{#} introduces a comment line, empty lines are also ignored.
In the following descriptions, @var{length}s always refer to
corresponding @var{string}s. Numbers are always expressed in decimal.
The directives are:
@table @code
@item V @var{number} @key{NL}
Confirms the format of the file. @var{number} should be 1.
@item C @var{length1} , @var{length2} @key{NL} @var{string1} @var{string2} @key{NL}
Uses @var{string1} and @var{string2} as the beginning comment and
end comment strings.
@item Q @var{length1} , @var{length2} @key{NL} @var{string1} @var{string2} @key{NL}
Uses @var{string1} and @var{string2} as the beginning quote and end quote
strings.
@item F @var{length1} , @var{length2} @key{NL} @var{string1} @var{string2} @key{NL}
Defines, through @code{pushdef}, a definition for @var{string1}
expanding to the function whose builtin name is @var{string2}.
@item T @var{length1} , @var{length2} @key{NL} @var{string1} @var{string2} @key{NL}
Defines, though @code{pushdef}, a definition for @var{string1}
expanding to the text given by @var{string2}.
@item D @var{number}, @var{length} @key{NL} @var{string} @key{NL}
Selects diversion @var{number}, making it current, then copy
@var{string} in the current diversion. @var{number} may be a negative
number for a non-existing diversion. To merely specify an active
selection, use this command with an empty @var{string}. With 0 as the
diversion @var{number}, @var{string} will be issued on standard output
at reload time, however this may not be produced from within @code{m4}.
@end table
@node Compatibility, Concept index, Frozen files, Top
@chapter Compatibility with other versions of @code{m4}
@cindex compatibility
This chapter describes the differences between this implementation of
@code{m4}, and the implementation found under UNIX, notably System V,
Release 3.
There are also differences in BSD flavors of @code{m4}. No attempt
is made to summarize these here.
@menu
* Extensions:: Extensions in GNU m4
* Incompatibilities:: Facilities in System V m4 not in GNU m4
* Other Incompat:: Other incompatibilities
@end menu
@node Extensions, Incompatibilities, Compatibility, Compatibility
@section Extensions in GNU @code{m4}
@cindex GNU extensions
This version of @code{m4} contains a few facilities, that do not exist
in System V @code{m4}. These extra facilities are all suppressed by
using the @samp{-G} command line option, unless overridden by other
command line options.
@itemize @bullet
@item
In the @code{$}@var{n} notation for macro arguments, @var{n} can contain
several digits, while the System V @code{m4} only accepts one digit.
This allows macros in GNU @code{m4} to take any number of arguments, and
not only nine (@pxref{Arguments}).
@item
Files included with @code{include} and @code{sinclude} are sought in a
user specified search path, if they are not found in the working
directory. The search path is specified by the @samp{-I} option and the
@samp{M4PATH} environment variable (@pxref{Search Path}).
@item
Arguments to @code{undivert} can be non-numeric, in which case the named
file will be included uninterpreted in the output (@pxref{Undivert}).
@item
Formatted output is supported through the @code{format} builtin, which
is modeled after the C library function @code{printf} (@pxref{Format}).
@item
Searches and text substitution through regular expressions are
supported by the @code{regexp} (@pxref{Regexp}) and @code{patsubst}
(@pxref{Patsubst}) builtins.
@item
The output of shell commands can be read into @code{m4} with
@code{esyscmd} (@pxref{Esyscmd}).
@item
There is indirect access to any builtin macro with @code{builtin}
(@pxref{Builtin}).
@item
Macros can be called indirectly through @code{indir} (@pxref{Indir}).
@item
The name of the current input file and the current input line number are
accessible through the builtins @code{__file__} and @code{__line__}
(@pxref{Errprint}).
@item
The format of the output from @code{dumpdef} and macro tracing can be
controlled with @code{debugmode} (@pxref{Debug Levels}).
@item
The destination of trace and debug output can be controlled with
@code{debugfile} (@pxref{Debug Output}).
@end itemize
In addition to the above extensions, GNU @code{m4} implements the
following command line options: @samp{-F}, @samp{-G}, @samp{-I},
@samp{-L}, @samp{-R}, @samp{-V}, @samp{-W}, @samp{-d},
@samp{-l}, @samp{-o} and @samp{-t}. @xref{Invoking m4}, for a
description of these options.
Also, the debugging and tracing facilities in GNU @code{m4} are much
more extensive than in most other versions of @code{m4}.
@node Incompatibilities, Other Incompat, Extensions, Compatibility
@section Facilities in System V @code{m4} not in GNU @code{m4}
The version of @code{m4} from System V contains a few facilities that
have not been implemented in GNU @code{m4} yet.
@itemize @bullet
@item
System V @code{m4} supports multiple arguments to @code{defn}. This is
not implemented in GNU @code{m4}. Its usefulness is unclear to me.
@end itemize
@node Other Incompat, , Incompatibilities, Compatibility
@section Other incompatibilities
There are a few other incompatibilities between this implementation of
@code{m4}, and the System V version.
@itemize @bullet
@item
GNU @code{m4} implements sync lines differently from System V @code{m4},
when text is being diverted. GNU @code{m4} outputs the sync lines when
the text is being diverted, and System V @code{m4} when the diverted
text is being brought back.
The problem is which lines and filenames should be attached to text that
is being, or has been, diverted. System V @code{m4} regards all the
diverted text as being generated by the source line containing the
@code{undivert} call, whereas GNU @code{m4} regards the diverted text as
being generated at the time it is diverted.
I expect the sync line option to be used mostly when using @code{m4} as
a front end to a compiler. If a diverted line causes a compiler error,
the error messages should most probably refer to the place where the
diversion were made, and not where it was inserted again.
@item
GNU @code{m4} makes no attempt at prohiting autoreferential definitions
like:
@comment ignore
@example
define(`x', `x')
define(`x', `x ')
@end example
There is nothing inherently wrong with defining @samp{x} to
return @samp{x}. The wrong thing is to expand @samp{x} unquoted.
In @code{m4}, one might use macros to hold strings, as we do for
variables in other programming languages, further checking them with:
@comment ignore
@example
ifelse(defn(`@var{holder}'), `@var{value}', @dots{})
@end example
@noindent
In cases like this one, an interdiction for a macro to hold its own
name would be a useless limitation. Of course, this leave more rope
for the GNU @code{m4} user to hang himself! Rescanning hangs may be
avoided through careful programming, a little like for endless loops
in traditional programming languages.
@item
@findex gnu
GNU @code{m4} without @samp{-G} option will define the macro
@code{__gnu__} to expand to the empty string.
@findex unix
On UNIX systems, GNU @code{m4} without the @samp{-G} option will define
the macro @code{__unix__}, otherwise the macro @code{unix}. Both will
expand to the empty string.
@end itemize
@node Concept index, Macro index, Compatibility, Top
@unnumbered Concept index
@printindex cp
@node Macro index, , Concept index, Top
@unnumbered Macro index
References are exclusively to the places where a builtin is introduced
the first time. Names starting and ending with @samp{__} have these
characters removed in the index.
@iftex
@sp 1
@end iftex
@printindex fn
@summarycontents
@contents
@bye