NetBSD/usr.bin/xstr/xstr.1

212 lines
5.6 KiB
Groff

.\" $NetBSD: xstr.1,v 1.17 2004/07/26 16:05:33 wiz Exp $
.\"
.\" Copyright (c) 1980, 1993
.\" The Regents of the University of California. All rights reserved.
.\"
.\" Redistribution and use in source and binary forms, with or without
.\" modification, are permitted provided that the following conditions
.\" are met:
.\" 1. Redistributions of source code must retain the above copyright
.\" notice, this list of conditions and the following disclaimer.
.\" 2. Redistributions in binary form must reproduce the above copyright
.\" notice, this list of conditions and the following disclaimer in the
.\" documentation and/or other materials provided with the distribution.
.\" 3. Neither the name of the University nor the names of its contributors
.\" may be used to endorse or promote products derived from this software
.\" without specific prior written permission.
.\"
.\" THIS SOFTWARE IS PROVIDED BY THE REGENTS AND CONTRIBUTORS ``AS IS'' AND
.\" ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
.\" IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE
.\" ARE DISCLAIMED. IN NO EVENT SHALL THE REGENTS OR CONTRIBUTORS BE LIABLE
.\" FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
.\" DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS
.\" OR SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION)
.\" HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
.\" LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY
.\" OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF
.\" SUCH DAMAGE.
.\"
.\" @(#)xstr.1 8.2 (Berkeley) 12/30/93
.\"
.Dd July 23, 2004
.Dt XSTR 1
.Os
.Sh NAME
.Nm xstr
.Nd "extract strings from C programs to implement shared strings"
.Sh SYNOPSIS
.Nm xstr
.Op Fl cv
.Op Fl l Ar array
.Op Fl
.Op Ar file ...
.Sh DESCRIPTION
.Nm
maintains a file
.Pa strings
into which strings in component parts of a large program are hashed.
These strings are replaced with references to this common area.
This serves to implement shared constant strings, most useful if they
are also read-only.
.Pp
Available options:
.Bl -tag -width XXlXarrayXX
.It Fl
.Nm
reads from the standard input.
.It Fl c
.Nm
will extract the strings from the C source
.Ar file
or the standard input
.Pq Fl ,
replacing
string references by expressions of the form (\*[Am]xstr[number])
for some number.
An appropriate declaration of
.Nm
is prepended to the file.
The resulting C text is placed in the file
.Pa x.c ,
to then be compiled.
The strings from this file are placed in the
.Pa strings
data base if they are not there already.
Repeated strings and strings which are suffixes of existing strings
do not cause changes to the data base.
.It Fl l Ar array
Specify the named array in program references to abstracted
strings.
The default array name is xstr.
.It Fl v
Be verbose.
.El
.Pp
After all components of a large program have been compiled, a file
.Pa xs.c
declaring the common
.Nm
space can be created by a command of the form:
.Pp
.Dl $ xstr
.Pp
The file
.Pa xs.c
should then be compiled and loaded with the rest
of the program.
If possible, the array can be made read-only (shared) saving
space and swap overhead.
.Pp
.Nm
can also be used on a single file.
The following command creates files
.Pa x.c
and
.Pa xs.c
as before, without using or affecting any
.Pa strings
file in the same directory:
.Pp
.Dl $ xstr name
.Pp
It may be useful to run
.Nm
after the C preprocessor if any macro definitions yield strings
or if there is conditional code which contains strings
which may not, in fact, be needed.
An appropriate command sequence for running
.Nm
after the C preprocessor is:
.Pp
.Bd -literal -offset indent
$ cc \-E name.c | xstr \-c \-
$ cc \-c x.c
$ mv x.o name.o
.Ed
.Pp
.Nm
does not touch the file
.Pa strings
unless new items are added, thus
.Xr make 1
can avoid remaking
.Pa xs.o
unless truly necessary.
.Sh FILES
.Bl -tag -width /tmp/xsxx* -compact
.It Pa strings
Data base of strings
.It Pa x.c
Massaged C source
.It Pa xs.c
C source for definition of array `xstr'
.It Pa /tmp/xs*
Temp file when `xstr name' doesn't touch
.Pa strings
.El
.Sh SEE ALSO
.Xr mkstr 1
.Sh HISTORY
The
.Nm
command appeared in
.Bx 3.0 .
.Sh BUGS
If a string is a suffix of another string in the data base,
but the shorter string is seen first by
.Nm
both strings will be placed in the data base, when just
placing the longer one there will do.
.Pp
.Nm
does not parse the file properly so it does not know not to process:
.Bd -literal
char var[] = "const";
.Ed
into:
.Bd -literal
char var[] = (\*[Am]xstr[N]);
.Ed
.Pp
These must be changed manually into an appropriate initialization for
the string, or use the following ugly hack.
.Pp
Also,
.Nm
cannot initialize structures and unions that contain strings.
Those can be fixed by changing from:
.Bd -literal
struct foo {
int i;
char buf[10];
} = {
1, "foo"
};
.Ed
to:
.Bd -literal
struct foo {
int i;
char buf[10];
} = {
1, { 'f', 'o', 'o', '\e0' }
};
.Ed
.Pp
The real problem in both cases above is that the compiler knows the size
of the literal constant so that it can perform the initialization required,
but when
.Nm
changes the literal string to a pointer reference, the size information is
lost.
It would require a real parser to do this right, so the obvious solution is
to fix the program manually to compile, or even better rely on the compiler
and the linker to merge strings appropriately.
.Pp
Finally,
.Nm
is not very useful these days because most of the string merging is done
automatically by the compiler and the linker, provided that the strings
are identical and read-only.