mirror of
git://git.code.sf.net/p/cdesktopenv/code
synced 2025-02-13 19:52:20 +00:00
710 lines
29 KiB
HTML
710 lines
29 KiB
HTML
<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 4.0 Transitional//EN" "http://www.w3.org/TR/REC-html40/loose.dtd">
|
|
<HTML>
|
|
<HEAD>
|
|
<META name="generator" content="mm2html (AT&T Research) 2012-01-11">
|
|
<TITLE> www/ksh/builtins.mm mm document </TITLE>
|
|
<META name="author" content="gsf+dgk+kpv">
|
|
<STYLE type="text/css">
|
|
div.FI { padding-left:2em; text-indent:0em; }
|
|
div.HI { padding-left:4em; text-indent:-2em; }
|
|
dt { float:left; clear:both; }
|
|
dd { margin-left:3em; }
|
|
</STYLE>
|
|
</HEAD>
|
|
<BODY bgcolor=white link=slateblue vlink=teal >
|
|
<TABLE border=0 align=center width=96%>
|
|
<TBODY><TR><TD valign=top align=left>
|
|
<!--INDEX--><!--/INDEX-->
|
|
<P>
|
|
<CENTER>
|
|
<H3><CENTER><FONT color=red><FONT face=courier>Guidelines for writing <TT>ksh-93</TT> built-in commands</FONT></FONT></CENTER></H3>
|
|
<BR>David G. Korn
|
|
<P><I></I>
|
|
</CENTER>
|
|
<P>
|
|
<CENTER><FONT color=red><FONT face=courier><H3 align=center><A name="Abstract">Abstract</A></H3></FONT></FONT></CENTER>
|
|
One of the features of <TT>ksh93</TT>, the latest version of <TT>ksh</TT>,
|
|
is the ability to add built-in commands at run time.
|
|
This feature only works on operating systems that have the ability
|
|
to load and link code into the current process at run time.
|
|
Some examples of the systems that have this feature
|
|
are Linux, System V Release 4, Solaris, Sun OS, HP-UX Release 8 and above,
|
|
AIX 3.2 and above, and Microsoft Windows systems.
|
|
<P>
|
|
This memo describes how to write and compile programs
|
|
that can be loaded into <TT>ksh</TT> at run time as built-in
|
|
commands.
|
|
<P>
|
|
<P><HR><CENTER><FONT color=red><FONT face=courier><H3><A name="INTRODUCTION">INTRODUCTION</A></H3></FONT></FONT></CENTER>
|
|
A built-in command is executed without creating a separate process.
|
|
Instead, the command is invoked as a C function by <TT>ksh</TT>.
|
|
If this function has no side effects in the shell process,
|
|
then the behavior of this built-in is identical to that of
|
|
the equivalent stand-alone command. The primary difference
|
|
in this case is performance. The overhead of process creation
|
|
is eliminated. For commands of short duration, the effect
|
|
can be dramatic. For example, on SUN OS 4.1, the time to
|
|
run <TT>wc</TT> on a small file of about 1000 bytes, runs
|
|
about 50 times faster as a built-in command.
|
|
<P>
|
|
In addition, built-in commands may have side effects on the
|
|
shell environment.
|
|
This is usually done to extend the application domain for
|
|
shell programming. For example, there is a group of X-windows extension
|
|
built-ins that make heavy use of the shell variable namespace.
|
|
These built-ins are added at run time and
|
|
result in a windowing shell that can be used to write
|
|
X-windows applications.
|
|
<P>
|
|
While there are definite advantages to adding built-in
|
|
commands, there are some disadvantages as well.
|
|
Since the built-in command and <TT>ksh</TT> share the same
|
|
address space, a coding error in the built-in program
|
|
may affect the behavior of <TT>ksh</TT>; perhaps causing
|
|
it to core dump or hang.
|
|
Debugging is also more complex since your code is now
|
|
a part of a larger entity.
|
|
The isolation provided by a separate process
|
|
guarantees that all resources used by the command
|
|
will be freed when the command completes.
|
|
Resources used by a built-in must be meticulously maintained and freed.
|
|
Also, since the address space of <TT>ksh</TT> will be larger when built-in are loaded,
|
|
it may increase the time it takes <TT>ksh</TT> to fork() and
|
|
exec() non-built-in commands.
|
|
It makes no sense to add a built-in command that takes
|
|
a long time to run or that is run only once, since the performance
|
|
benefits will be negligible.
|
|
Built-ins that have side effects in the current shell
|
|
environment have the disadvantage of increasing the
|
|
coupling between the built-in and <TT>ksh</TT>, making
|
|
the overall system less modular and more monolithic.
|
|
<P>
|
|
Despite these drawbacks, in many cases extending
|
|
<TT>ksh</TT> by adding built-in
|
|
commands makes sense and allows reuse of the shell
|
|
scripting ability in an application specific domain.
|
|
This memo describes how to write <TT>ksh</TT> extensions.
|
|
<P>
|
|
<P><HR><CENTER><FONT color=red><FONT face=courier><H3><A name="WRITING BUILT-IN COMMANDS">WRITING BUILT-IN COMMANDS</A></H3></FONT></FONT></CENTER>
|
|
There is a development kit available for writing <TT>ksh</TT>
|
|
built-ins as part of the AST (AT&T Software Technology) Toolkit.
|
|
The development kit has three directories,
|
|
<TT>include</TT>, <TT>lib</TT>, and <TT>bin</TT>.
|
|
It is best to set the value of the environment variable
|
|
<TT>PACKAGE_ast</TT> to the pathname of the directory
|
|
containing the development kit.
|
|
The <TT>include</TT> directory contains a sub-directory
|
|
named <TT>ast</TT> that contains interface prototypes
|
|
for functions that you can call from built-ins. The <TT>lib</TT>
|
|
directory contains the <TT>ast</TT> library
|
|
and a library named <TT>cmd</TT> that contains a version
|
|
of several of the standard POSIX<FONT SIZE=-6>[1]</FONT>
|
|
utilities that can be made run time built-ins.
|
|
The <TT>lib/ksh</TT> directory contains shared libraries
|
|
that implement other <TT>ksh</TT> built-ins.
|
|
The <TT>bin</TT> directory contains build tools such as <TT>nmake</TT><FONT SIZE=-6>[2]</FONT>.
|
|
To add built-ins at runtime, it is necessary to build a shared library
|
|
containing one or more built-ins that you wish to add.
|
|
The built-ins are then added by running <TT>builtin -f</TT> <EM>shared_lib</EM>.
|
|
Since the procedure for building share libraries is system dependent,
|
|
it is best to use
|
|
<TT>nmake</TT>
|
|
using the sample nmake makefile below as a prototype.
|
|
The AST Toolkit also contains some examples of built-in libraries under
|
|
the <TT>src/cmd/kshlib</TT> directory.
|
|
<P>
|
|
There are two ways to code adding built-ins. One method is to replace
|
|
the function <TT>main</TT> with a function
|
|
<TT>b_</TT><EM>name</EM>, where <EM>name</EM> is the name
|
|
of the built-in you wish to define.
|
|
A built-in command has a calling convention similar to
|
|
the <TT>main</TT> function of a program,
|
|
<TT>int main(int argc, char *argv[])</TT>.
|
|
except that it takes a third argument of type <TT>Shbltin_t*</TT> which can
|
|
be passed as <TT><FONT SIZE=-1>NULL</FONT></TT> if it is not used. The definition for
|
|
<TT>Shbltin_t*</TT> is in <TT><ast/shcmd.h></TT>.
|
|
Instead of <TT>exit</TT>, you need to use <TT>return</TT>
|
|
to terminate your command.
|
|
The return value will become the exit status of the command.
|
|
The <TT>open</TT> built-in, installed in <TT>lib/ksh</TT> in the AST Toolkit, uses this method.
|
|
The <TT>Shbltin_t</TT> structure contains a field named <TT>shp</TT> which is
|
|
a pointer the the shell data that is needed for <TT>shell</TT> library callbacks.
|
|
It also contains the fields, <TT>shrun</TT>, <TT>shtrap</TT>, <TT>shexit</TT>,
|
|
and <TT>shbltin</TT>
|
|
that are function pointers to the <TT>shell</TT> library functions <TT>sh_run</TT>, <TT>sh_trap</TT>
|
|
<TT>sh_exit</TT>, and <TT>sh_addbuiltin</TT>, respectively. These functions
|
|
can be invoked without the need for runtime symbol lookup when the
|
|
shell is statically linked with <TT>libshell</TT>.
|
|
<P>
|
|
The alternative method is to create a function <TT>lib_init</TT> and
|
|
use the <TT>Shbltin_t.shbltin()</TT> function to add one or more built-ins.
|
|
The <TT>lib_init</TT> function will be called with two arguments. The
|
|
first argument will be 0 when the library is loaded and the second
|
|
argument will be of type <TT>Shbltin_t*</TT>.
|
|
The <TT>dbm_t</TT> and <TT>dss</TT> shell built-ins use this method.
|
|
<P>
|
|
No matter which way you add built-ins you should add the line
|
|
<TT>SHLIB(</TT><EM>identifier</EM><TT>)</TT> as the last line of one
|
|
of the built-in source file, where <EM>identifier</EM> is any C identifier.
|
|
This line provides version information to the shell <TT>builtin</TT> command
|
|
that it uses to verify compatibility between the built-in and <TT>ksh</TT>
|
|
implementation versions. <TT>builtin</TT> fails with a diagnostic on version
|
|
mismatch. The diagnostic helps determine whether <TT>ksh</TT> is out of
|
|
date and requires an upgrade or the built-in is out of date and requires
|
|
recompilation.
|
|
<P>
|
|
The steps necessary to create and add a run time built-in are
|
|
illustrated in the following simple example.
|
|
Suppose you wish to add a built-in command named <TT>hello</TT>
|
|
which requires one argument and prints the word hello followed
|
|
by its argument. First, write the following program in the file
|
|
<TT>hello.c</TT>:
|
|
<DIV class=FI>
|
|
<PRE>
|
|
#include <stdio.h>
|
|
int b_hello(int argc, char *argv[], void *context)
|
|
{
|
|
if(argc != 2)
|
|
{
|
|
fprintf(stderr,"Usage: hello arg\n");
|
|
return(2);
|
|
}
|
|
printf("hello %s\n",argv[1]);
|
|
return(0);
|
|
}
|
|
SHLIB(hello)
|
|
</DIV>
|
|
</PRE>
|
|
<P>
|
|
Next, the program needs to be compiled.
|
|
If you are building with AT&T <TT>nmake</TT> use the following <TT>Makefile</TT>:
|
|
<DIV class=FI>
|
|
<PRE>
|
|
:PACKAGE: --shared ast
|
|
hello plugin=ksh :LIBRARY: hello.c
|
|
</DIV>
|
|
</PRE>
|
|
and run <TT>nmake install</TT> to compile, link, and install the built-in shared library
|
|
in <TT>lib/ksh/</TT> under <TT>PACKAGE_ast</TT>.
|
|
If the built-in extension uses several <TT>.c</TT> files, list all of these on
|
|
the <TT>:LIBRARY:</TT> line.
|
|
<P>
|
|
Otherwise you will have to compile <TT>hello.c</TT> with an option
|
|
to pick up the AST include directory
|
|
(since the AST <TT><stdio.h></TT> is required for <TT>ksh</TT> compatibility)
|
|
and options required for generating shared libraries.
|
|
For example, on Linux use this to compile:
|
|
<DIV class=FI>
|
|
<PRE>
|
|
cc -fpic -I$PACKAGE_ast/include/ast -c hello.c
|
|
</DIV>
|
|
</PRE>
|
|
and use the appropriate link line.
|
|
It really is best to use <TT>nmake</TT> because the 2 line Makefile above
|
|
will work on all systems that have <TT>ksh</TT> installed.
|
|
<P>
|
|
If you have several built-ins, it is desirable
|
|
to build a shared library that contains them all.
|
|
<P>
|
|
The final step is using the built-in.
|
|
This can be done with the <TT>ksh</TT> command <TT>builtin</TT>.
|
|
To load the shared library <TT>libhello.so</TT> from the current directory
|
|
and add the built-in <TT>hello</TT>, invoke the command,
|
|
<DIV class=FI>
|
|
<PRE>
|
|
builtin -f ./libhello.so hello
|
|
</DIV>
|
|
</PRE>
|
|
The shared library prefix (<TT>lib</TT> here) and suffix (<TT>.so</TT> here) be omitted;
|
|
the shell will add an appropriate suffix
|
|
for the system that it is loading from.
|
|
If you install the shared library in <TT>lib/ksh/</TT>, where <TT>../lib/ksh/</TT> is
|
|
a directory on <STRONG>$PATH</STRONG>, the command
|
|
<DIV class=FI>
|
|
<PRE>
|
|
builtin -f hello hello
|
|
</DIV>
|
|
</PRE>
|
|
will automatically find, load and install the built-in on any system.
|
|
Once this command has been invoked, you can invoke <TT>hello</TT>
|
|
as you do any other command.
|
|
If you are using <TT>lib_init</TT> method to add built-ins then no arguments
|
|
follow the <TT>-f</TT> option.
|
|
<P>
|
|
It is often desirable to make a command <EM>built-in</EM>
|
|
the first time that it is referenced. The first
|
|
time <TT>hello</TT> is invoked, <TT>ksh</TT> should load and execute it,
|
|
whereas for subsequent invocations <TT>ksh</TT> should just execute the built-in.
|
|
This can be done by creating a file named <TT>hello</TT>
|
|
with the following contents:
|
|
<DIV class=FI>
|
|
<PRE>
|
|
function hello
|
|
{
|
|
unset -f hello
|
|
builtin -f hello hello
|
|
hello "$@"
|
|
}
|
|
</DIV>
|
|
</PRE>
|
|
This file <TT>hello</TT> needs to be placed in a directory that is
|
|
in your <STRONG><FONT SIZE=-1>FPATH</FONT></STRONG> variable, and the built-in shared library
|
|
should be installed in <TT>lib/ksh/</TT>, as described above.
|
|
<P>
|
|
<P><HR><CENTER><FONT color=red><FONT face=courier><H3><A name="CODING REQUIREMENTS AND CONVENTIONS">CODING REQUIREMENTS AND CONVENTIONS</A></H3></FONT></FONT></CENTER>
|
|
As mentioned above, the entry point for built-ins must either be of
|
|
the form <TT>b_</TT><EM>name</EM> or else be loaded from a function named
|
|
<TT>lib_init</TT>.
|
|
Your built-ins can call functions from the standard C library,
|
|
the <TT>ast</TT> library, interface functions provided by <TT>ksh</TT>,
|
|
and your own functions.
|
|
You should avoid using any global symbols beginning with
|
|
<STRONG>sh_</STRONG>,
|
|
<STRONG>nv_</STRONG>,
|
|
and
|
|
<STRONG>ed_</STRONG>
|
|
since these are used by <TT>ksh</TT> itself.
|
|
<TT>#define</TT> constants in <TT>ksh</TT> interface
|
|
files use symbols beginning with <TT>SH_</TT> and <TT>NV_</TT>,
|
|
so avoid using names beginning with these too.
|
|
<P>
|
|
<H4><A name="Header Files">Header Files</A></H4>
|
|
The development kit provides a portable interface
|
|
to the C library and to libast.
|
|
The header files in the development kit are compatible with
|
|
K&R C<FONT SIZE=-6>[3]</FONT>,
|
|
ANSI-C<FONT SIZE=-6>[4]</FONT>,
|
|
and C++<FONT SIZE=-6>[5]</FONT>.
|
|
<P>
|
|
The best thing to do is to include the header file <TT><shell.h></TT>.
|
|
This header file causes the <TT><ast.h></TT> header, the
|
|
<TT><error.h></TT> header and the <TT><stak.h></TT>
|
|
header to be included as well as defining prototypes
|
|
for functions that you can call to get shell
|
|
services for your builtins.
|
|
The header file <TT><ast.h></TT>
|
|
provides prototypes for many <STRONG>libast</STRONG> functions
|
|
and all the symbol and function definitions from the
|
|
ANSI-C headers, <TT><stddef.h></TT>,
|
|
<TT><stdlib.h></TT>, <TT><stdarg.h></TT>, <TT><limits.h></TT>,
|
|
and <TT><string.h></TT>.
|
|
It also provides all the symbols and definitions for the
|
|
POSIX<FONT SIZE=-6>[6]</FONT>
|
|
headers <TT><sys/types.h></TT>, <TT><fcntl.h></TT>, and
|
|
<TT><unistd.h></TT>.
|
|
You should include <TT><ast.h></TT> instead of one or more of
|
|
these headers.
|
|
The <TT><error.h></TT> header provides the interface to the error
|
|
and option parsing routines defined below.
|
|
The <TT><stak.h></TT> header provides the interface to the memory
|
|
allocation routines described below.
|
|
<P>
|
|
Programs that want to use the information in <TT><sys/stat.h></TT>
|
|
should include the file <TT><ls.h></TT> instead.
|
|
This provides the complete POSIX interface to <TT>stat()</TT>
|
|
related functions even on non-POSIX systems.
|
|
<P>
|
|
<P>
|
|
<H4><A name="Input/Output">Input/Output</A></H4>
|
|
<TT>ksh</TT> uses <STRONG>sfio</STRONG>,
|
|
the Safe/Fast I/O library<FONT SIZE=-6>[7]</FONT>,
|
|
to perform all I/O operations.
|
|
The <STRONG>sfio</STRONG> library, which is part of <STRONG>libast</STRONG>,
|
|
provides a superset of the functionality provided by the standard
|
|
I/O library defined in ANSI-C.
|
|
If none of the additional functionality is required,
|
|
and if you are not familiar with <STRONG>sfio</STRONG> and
|
|
you do not want to spend the time learning it,
|
|
then you can use <TT>sfio</TT> via the <TT>stdio</TT> library
|
|
interface. The development kit contains the header <TT><stdio.h></TT>
|
|
which maps <TT>stdio</TT> calls to <TT>sfio</TT> calls.
|
|
In most instances the mapping is done
|
|
by macros or inline functions so that there is no overhead.
|
|
The man page for the <TT>sfio</TT> library is in an Appendix.
|
|
<P>
|
|
However, there are some very nice extensions and
|
|
performance improvements in <TT>sfio</TT>
|
|
and if you plan any major extensions I recommend
|
|
that you use it natively.
|
|
<P>
|
|
<H4><A name="Error Handling">Error Handling</A></H4>
|
|
For error messages it is best to use the <TT>ast</TT> library
|
|
function <TT>errormsg()</TT> rather that sending output to
|
|
<TT>stderr</TT> or the equivalent <TT>sfstderr</TT> directly.
|
|
Using <TT>errormsg()</TT> will make error message appear
|
|
more uniform to the user.
|
|
Furthermore, using <TT>errormsg()</TT> should make it easier
|
|
to do error message translation for other locales
|
|
in future versions of <TT>ksh</TT>.
|
|
<P>
|
|
The first argument to
|
|
<TT>errormsg()</TT> specifies the dictionary in which the string
|
|
will be searched for translation.
|
|
The second argument to <TT>errormsg()</TT> contains that error type
|
|
and value. The third argument is a <EM>printf</EM> style format
|
|
and the remaining arguments are arguments to be printed
|
|
as part of the message. A new-line is inserted at the
|
|
end of each message and therefore, should not appear as
|
|
part of the format string.
|
|
The second argument should be one of the following:
|
|
<DIV class=SH>
|
|
<DL>
|
|
<DT><TT>ERROR_exit(</TT><EM>n</EM><TT>)</TT>:<DD><BR>
|
|
If <EM>n</EM> is not-zero, the builtin will exit value <EM>n</EM> after
|
|
printing the message.
|
|
<DT><TT>ERROR_system(</TT><EM>n</EM><TT>)</TT>:<DD><BR>
|
|
Exit builtin with exit value <EM>n</EM> after printing the message.
|
|
The message will display the message corresponding to <TT>errno</TT>
|
|
enclosed within <TT>[ ]</TT> at the end of the message.
|
|
<DT><TT>ERROR_usage(</TT><EM>n</EM><TT>)</TT>:<DD><BR>
|
|
Will generate a usage message and exit. If <EM>n</EM> is non-zero,
|
|
the exit value will be 2. Otherwise the exit value will be 0.
|
|
<DT><TT>ERROR_debug(</TT><EM>n</EM><TT>)</TT>:<DD><BR>
|
|
Will print a level <EM>n</EM> debugging message and will then continue.
|
|
<DT><TT>ERROR_warn(</TT><EM>n</EM><TT>)</TT>:<DD><BR>
|
|
Prints a warning message. <EM>n</EM> is ignored.
|
|
</DL><P>
|
|
<H4><A name="Option Parsing">Option Parsing</A></H4>
|
|
The first thing that a built-in should do is to check
|
|
the arguments for correctness and to print any usage
|
|
messages on standard error.
|
|
For consistency with the rest of <TT>ksh</TT>, it is best
|
|
to use the <TT>libast</TT> functions <TT>optget()</TT> and
|
|
<TT>optusage()</TT>for this
|
|
purpose.
|
|
The header <TT><error.h></TT> includes prototypes for
|
|
these functions.
|
|
The <TT>optget()</TT> function is similar to the
|
|
System V C library function <TT>getopt()</TT>,
|
|
but provides some additional capabilities.
|
|
Built-ins that use <TT>optget()</TT> provide a more
|
|
consistent user interface.
|
|
<P>
|
|
The <TT>optget()</TT> function is invoked as
|
|
<DIV class=FI>
|
|
<PRE>
|
|
int optget(char *<EM>argv</EM>[], const char *<EM>optstring</EM>)
|
|
</DIV>
|
|
</PRE>
|
|
where <TT>argv</TT> is the argument list and <TT>optstring</TT>
|
|
is a string that specifies the allowable arguments and
|
|
additional information that is used to format <EM>usage</EM>
|
|
messages.
|
|
In fact a complete man page in <TT>troff</TT> or <TT>html</TT>
|
|
can be generated by passing a usage string as described
|
|
by the <TT>getopts</TT> command.
|
|
Like <TT>getopt()</TT>,
|
|
single letter options are represented by the letter itself,
|
|
and options that take a string argument are followed by the <TT>:</TT>
|
|
character.
|
|
Option strings have the following special characters:
|
|
<DIV class=SH>
|
|
<DL>
|
|
<DT><TT>:</TT><DD>
|
|
Used after a letter option to indicate that the option
|
|
takes an option argument.
|
|
The variable <TT>opt_info.arg</TT> will point to this
|
|
value after the given argument is encountered.
|
|
<DT><TT>#</TT><DD>
|
|
Used after a letter option to indicate that the option
|
|
can only take a numerical value.
|
|
The variable <TT>opt_info.num</TT> will contain this
|
|
value after the given argument is encountered.
|
|
<DT><TT>?</TT><DD>
|
|
Used after a <TT>:</TT> or <TT>#</TT> (and after the optional <TT>?</TT>)
|
|
to indicate the the
|
|
preceding option argument is not required.
|
|
<DT><TT>[</TT>...<TT>]</TT><DD><BR>
|
|
After a <TT>:</TT> or <TT>#</TT>, the characters contained
|
|
inside the brackets are used to identify the option
|
|
argument when generating a <EM>usage</EM> message.
|
|
<DT><EM>space</EM><DD><BR>
|
|
The remainder of the string will only be used when generating
|
|
usage messages.
|
|
</DL>
|
|
</DIV>
|
|
<P>
|
|
The <TT>optget()</TT> function returns the matching option letter if
|
|
one of the legal option is matched.
|
|
Otherwise, <TT>optget()</TT> returns
|
|
<DIV class=SH>
|
|
<DL>
|
|
<DT><TT>':'</TT><DD>
|
|
If there is an error. In this case the variable <TT>opt_info.arg</TT>
|
|
contains the error string.
|
|
<DT><TT>0</TT><DD>
|
|
Indicates the end of options.
|
|
The variable <TT>opt_info.index</TT> contains the number of arguments
|
|
processed.
|
|
<DT><TT>'?'</TT><DD>
|
|
A usage message has been required.
|
|
You normally call <TT>optusage()</TT> to generate and display
|
|
the usage message.
|
|
</DL>
|
|
</DIV>
|
|
<P>
|
|
The following is an example of the option parsing portion
|
|
of the <TT>wc</TT> utility.
|
|
<DIV class=FI>
|
|
<PRE>
|
|
#include <shell.h>
|
|
while(1) switch(n=optget(argv,"xf:[file]"))
|
|
{
|
|
case 'f':
|
|
file = opt_info.arg;
|
|
break;
|
|
case ':':
|
|
error(ERROR_exit(0), opt_info.arg);
|
|
break;
|
|
case '?':
|
|
error(ERROR_usage(2), opt_info.arg);
|
|
break;
|
|
}
|
|
</DIV>
|
|
</PRE>
|
|
<P>
|
|
<H4><A name="Storage Management">Storage Management</A></H4>
|
|
It is important that any memory used by your built-in
|
|
be returned. Otherwise, if your built-in is called frequently,
|
|
<TT>ksh</TT> will eventually run out of memory.
|
|
You should avoid using <TT>malloc()</TT> for memory that must
|
|
be freed before returning from you built-in, because by default,
|
|
<TT>ksh</TT> will terminate you built-in in the event of an
|
|
interrupt and the memory will not be freed.
|
|
<P>
|
|
The best way to to allocate variable sized storage is
|
|
through calls to the <STRONG>stak</STRONG> library
|
|
which is included in <STRONG>libast</STRONG>
|
|
and which is used extensively by <TT>ksh</TT> itself.
|
|
Objects allocated with the <TT>stakalloc()</TT>
|
|
function are freed when you function completes
|
|
or aborts.
|
|
The <STRONG>stak</STRONG> library provides a convenient way to
|
|
build variable length strings and other objects dynamically.
|
|
The man page for the <STRONG>stak</STRONG> library is contained
|
|
in the Appendix.
|
|
<P>
|
|
Before <TT>ksh</TT> calls each built-in command, it saves
|
|
the current stack location and restores it after
|
|
it returns.
|
|
It is not necessary to save and restore the stack
|
|
location in the <TT>b_</TT> entry function,
|
|
but you may want to write functions that use this stack
|
|
are restore it when leaving the function.
|
|
The following coding convention will do this in
|
|
an efficient manner:
|
|
<DIV class=FI>
|
|
<PRE>
|
|
<EM>yourfunction</EM>()
|
|
{
|
|
char *savebase;
|
|
int saveoffset;
|
|
if(saveoffset=staktell())
|
|
savebase = stakfreeze(0);
|
|
...
|
|
if(saveoffset)
|
|
stakset(savebase,saveoffset);
|
|
else
|
|
stakseek(0);
|
|
}
|
|
</DIV>
|
|
</PRE>
|
|
<P>
|
|
<P><HR><CENTER><FONT color=red><FONT face=courier><H3><A name="CALLING <TT>ksh</TT> SERVICES">CALLING <TT>ksh</TT> SERVICES</A></H3></FONT></FONT></CENTER>
|
|
Some of the more interesting applications are those that extend
|
|
the functionality of <TT>ksh</TT> in application specific directions.
|
|
A prime example of this is the X-windows extension which adds
|
|
builtins to create and delete widgets.
|
|
The <STRONG>nval</STRONG> library is used to interface with the shell
|
|
name space.
|
|
The <STRONG>shell</STRONG> library is used to access other shell services.
|
|
<P>
|
|
<H4><A name="The nval library">The nval library</A></H4>
|
|
A great deal of power is derived from the ability to use
|
|
portions of the hierarchal variable namespace provided by <TT>ksh-93</TT>
|
|
and turn these names into active objects.
|
|
<P>
|
|
The <STRONG>nval</STRONG> library is used to interface with shell
|
|
variables.
|
|
A man page for this file is provided in an Appendix.
|
|
You need to include the header <TT><nval.h></TT>
|
|
to access the functions defined in the <STRONG>nval</STRONG> library.
|
|
All the functions provided by the <STRONG>nval</STRONG> library begin
|
|
with the prefix <TT>nv_</TT>.
|
|
Each shell variable is an object in an associative table
|
|
that is referenced by name.
|
|
The type <TT>Namval_t*</TT> is pointer to a shell variable.
|
|
To operate on a shell variable, you first get a handle
|
|
to the variable with the <TT>nv_open()</TT> function
|
|
and then supply the handle returned as the first
|
|
argument of the function that provides an operation
|
|
on the variable.
|
|
You must call <TT>nv_close()</TT> when you are finished
|
|
using this handle so that the space can be freed once
|
|
the value is unset.
|
|
The two most frequent operations are to get the value of
|
|
the variable, and to assign value to the variable.
|
|
The <TT>nv_getval()</TT> returns a pointer the the
|
|
value of the variable.
|
|
In some cases the pointer returned is to a region that
|
|
will be overwritten by the next <TT>nv_getval()</TT> call
|
|
so that if the value isn't used immediately, it should
|
|
be copied.
|
|
Many variables can also generate a numeric value.
|
|
The <TT>nv_getnum()</TT> function returns a numeric
|
|
value for the given variable pointer, calling the
|
|
arithmetic evaluator if necessary.
|
|
<P>
|
|
The <TT>nv_putval()</TT> function is used to assign a new
|
|
value to a given variable.
|
|
The second argument to <TT>putval()</TT> is the value
|
|
to be assigned
|
|
and the third argument is a <EM>flag</EM> which
|
|
is used in interpreting the second argument.
|
|
<P>
|
|
Each shell variable can have one or more attributes.
|
|
The <TT>nv_isattr()</TT> is used to test for the existence
|
|
of one or more attributes.
|
|
See the appendix for a complete list of attributes.
|
|
<P>
|
|
By default, each shell variable passively stores the string you
|
|
give with with <TT>nv_putval()</TT>, and returns the value
|
|
with <TT>getval()</TT>. However, it is possible to turn
|
|
any node into an active entity by assigning functions
|
|
to it that will be called whenever <TT>nv_putval()</TT>
|
|
and/or <TT>nv_getval()</TT> is called.
|
|
In fact there are up to five functions that can
|
|
associated with each variable to override the
|
|
default actions.
|
|
The type <TT>Namfun_t</TT> is used to define these functions.
|
|
Only those that are non-<TT>NULL</TT> override the
|
|
default actions.
|
|
To override the default actions, you must allocate an
|
|
instance of <TT>Namfun_t</TT>, and then assign
|
|
the functions that you wish to override.
|
|
The <TT>putval()</TT>
|
|
function is called by the <TT>nv_putval()</TT> function.
|
|
A <TT>NULL</TT> for the <EM>value</EM> argument
|
|
indicates a request to unset the variable.
|
|
The <EM>type</EM> argument might contain the <TT>NV_INTEGER</TT>
|
|
bit so you should be prepared to do a conversion if
|
|
necessary.
|
|
The <TT>getval()</TT>
|
|
function is called by <TT>nv_getval()</TT>
|
|
value and must return a string.
|
|
The <TT>getnum()</TT>
|
|
function is called by by the arithmetic evaluator
|
|
and must return double.
|
|
If omitted, then it will call <TT>nv_getval()</TT> and
|
|
convert the result to a number.
|
|
<P>
|
|
The functionality of a variable can further be increased
|
|
by adding discipline functions that
|
|
can be associated with the variable.
|
|
A discipline function allows a script that uses your
|
|
variable to define functions whose name is
|
|
<EM>varname</EM><TT>.</TT><EM>discname</EM>
|
|
where <EM>varname</EM> is the name of the variable, and <EM>discname</EM>
|
|
is the name of the discipline.
|
|
When the user defines such a function, the <TT>settrap()</TT>
|
|
function will be called with the name of the discipline and
|
|
a pointer to the parse tree corresponding to the discipline
|
|
function.
|
|
The application determines when these functions are actually
|
|
executed.
|
|
By default, <TT>ksh</TT> defines <TT>get</TT>,
|
|
<TT>set</TT>, and <TT>unset</TT> as discipline functions.
|
|
<P>
|
|
In addition, it is possible to provide a data area that
|
|
will be passed as an argument to
|
|
each of these functions whenever any of these functions are called.
|
|
To have private data, you need to define and allocate a structure
|
|
that looks like
|
|
<DIV class=FI>
|
|
<PRE>
|
|
struct <EM>yours</EM>
|
|
{
|
|
Namfun_t fun;
|
|
<EM>your_data_fields</EM>;
|
|
};
|
|
</DIV>
|
|
</PRE>
|
|
<P>
|
|
<H4><A name="The shell library">The shell library</A></H4>
|
|
There are several functions that are used by <TT>ksh</TT> itself
|
|
that can also be called from built-in commands.
|
|
The man page for these routines are in the Appendix.
|
|
<P>
|
|
The <TT>sh_addbuiltin()</TT> function can be used to add or delete
|
|
builtin commands. It takes the name of the built-in, the
|
|
address of the function that implements the built-in, and
|
|
a <TT>void*</TT> pointer that will be passed to this function
|
|
as the third argument whenever it is invoked.
|
|
If the function address is <TT>NULL</TT>, the specified built-in
|
|
will be deleted. However, special built-in functions cannot
|
|
be deleted or modified.
|
|
<P>
|
|
The <TT>sh_fmtq()</TT> function takes a string and returns
|
|
a string that is quoted as necessary so that it can
|
|
be used as shell input.
|
|
This function is used to implement the <TT>%q</TT> option
|
|
of the shell built-in <TT>printf</TT> command.
|
|
<P>
|
|
The <TT>sh_parse()</TT> function returns a parse tree corresponding
|
|
to a give file stream. The tree can be executed by supplying
|
|
it as the first argument to
|
|
the <TT>sh_trap()</TT> function and giving a value of <TT>1</TT> as the
|
|
second argument.
|
|
Alternatively, the <TT>sh_trap()</TT> function can parse and execute
|
|
a string by passing the string as the first argument and giving <TT>0</TT>
|
|
as the second argument.
|
|
<P>
|
|
The <TT>sh_isoption()</TT> function can be used to set to see whether one
|
|
or more of the option settings is enabled.
|
|
</DIV>
|
|
<P><HR><CENTER><FONT color=red><FONT face=courier><H3><A name="References">References</A></H3></FONT></FONT></CENTER>
|
|
<P>
|
|
<DL compact>
|
|
|
|
<DT>[1]<DD>
|
|
<EM>POSIX - Part 2: Shell and Utilities,</EM>
|
|
IEEE Std 1003.2-1992, ISO/IEC 9945-2:1993.
|
|
<DT>[2]<DD>
|
|
Glenn Fowler,
|
|
<EM>A Case for make</EM>,
|
|
Software - Practice and Experience, Vol. 20 No. S1, pp. 30-46, June 1990.
|
|
<DT>[3]<DD>
|
|
Brian W. Kernighan and Dennis M. Ritchie,
|
|
<EM>The C Programming Language</EM>,
|
|
Prentice Hall, 1978.
|
|
<DT>[4]<DD>
|
|
American National Standard for Information Systems - Programming
|
|
Language - C, ANSI X3.159-1989.
|
|
<DT>[5]<DD>
|
|
Bjarne Stroustroup,
|
|
<EM>C++</EM>,
|
|
Addison Wesley, xxxx
|
|
<DT>[6]<DD>
|
|
<EM>POSIX - Part 1: System Application Program Interface,</EM>
|
|
IEEE Std 1003.1-1990, ISO/IEC 9945-1:1990.
|
|
<DT>[7]<DD>
|
|
David Korn and Kiem-Phong Vo,
|
|
<EM>SFIO - A Safe/Fast Input/Output library,</EM>
|
|
Proceedings of the Summer Usenix,
|
|
pp. , 1991.
|
|
</DL>
|
|
<P>
|
|
<HR>
|
|
<TABLE border=0 align=center width=96%>
|
|
<TR>
|
|
<TD align=left></TD>
|
|
<TD align=center></TD>
|
|
<TD align=right>March 13, 2012</TD>
|
|
</TR>
|
|
</TABLE>
|
|
<P>
|
|
|
|
</TD></TR></TBODY></TABLE>
|
|
|
|
</BODY>
|
|
</HTML>
|