######################################################################
## File: $Id: README 3442 2005-05-14 14:14:01Z spadkins $
######################################################################
1. What is the App-Options distribution?
You might say App-Options distribution is "yet another command line
option processor." However, it was created for maximum ease of use
with the needs of the Perl 5 Enterprise Environment in mind
(more on that later).
The distribution consists of one Perl module and one shell script.
App::Options - a perl module which combines command line parameters,
environment variables, and configuration files to produce
a hash of option values.
prefix - a shell script which works for ksh (Korn shell) or
bash (Bourne Again Shell) which allows you to change
a family of environment variables (PATH, LD_LIBRARY_PATH,
MANPATH, etc.) necessary for running programs out of a
directory in which software has been installed.
(This script is useful but not necessary, and non-Unix users
may safely ignore it.)
2. What are the features?
FEATURES OF App::Options
o Flexible command line syntax
Parse long (--whatever) and short (-w) command line options
(with [--verbose=3] or without [--verbose] arguments)
and just put the values in %App::options (or some other hash
that you may specify).
o Automatic boolean command line switches
An option like "--help" is equivalent to "--help=1".
o Combines options with Environment Variables
The --whatever=value option can be supplied with the
APP_WHATEVER environment variable.
o Combines options with variables in global and local config files
The --whatever=value option can be supplied in config files with
the "whatever = value" statement. Initialization searches
"$HOME/.app/$prog.conf", "$HOME/.app/app.conf",
"$progdir/$prog.conf", "$progdir/app.conf",
"$prefix/etc/app/$prog.conf", "$prefix/etc/app/app.conf", and
"/etc/app.conf" in order to find the option values.
o Import other files with "import = filename" statement
Config files can, in essence, include other config files.
o Stop importing other files with "flush_imports = 1" statement
A config file can use this statement to tell the program
to stop searching for other config files it was planning
on searching.
o Option values undergo variable expansion
This means that some values may rely on other values.
i.e. "prefix = /usr/mycompany/3.2.1" and "logdir = ${prefix}/log"
o Config files can have conditional sections
A section specifier "[cleanup]" begins a section only valid
for the program "cleanup" (or "cleanup.exe" or cleanup.pl, etc.).
A section specifier "[ALL]" (or just "[]") begins another
unconditional section. In fact, "[cleanup]" is just a synonym
for "[app=cleanup]". Sections can be made conditional on the
current values of any variables. i.e. "[country=US;city=LAX]"
would begin a section of variables only valid if the two
specified variables have the required values.
o Config files can have conditional lines
Any section specifier can apply only to a single line by
putting a statement after it. "[city=LAX] state = CA"
o Modify @INC variable with the "perlinc = path1,path2" statement
This allows the person deploying the software to set up the Perl
include path so that a particular version of the software
installed on the system is used. Subsequent uses of "use" and
"require" will load modules from the configured locations.
o Validate that "required" options are provided
Certain options can be identified as required. Otherwise the
program will not run.
o Validate option values against types
Certain options can be identified as to their type or pattern
(integer, int, float, number, date, datetime, string, regexp).
If the option value is provided and it does not match the
type or pattern, the program will not run.
o Provide automatic "-?" and "--help" messages
Adds user-friendliness to programs with no extra effort.
3. What were the design goals that make this distribution unique?
#1 Support multiple installations of a complex suite of software.
This is useful for development and deployment of large systems,
where multiple versions may be in varying states of development,
testing, or production.
#2 Support a suite of programs all using a common set of configuration
files. Large systems have many programs. We don't want to
repeat the database name, username, and password in a separate
config file for each program.
#3 Make it so easy to use that a developer would be silly not to
use it in even his smallest and simplest of scripts. However,
it should be powerful enough to support advanced features as
the developer needs them.
#4 Support conditions where the environment variables are not
easy to control. i.e. CGI programs or cron jobs.
These were important design goals to support the App-Context variant
of the Perl 5 Enterprise Environment (P5EE). See the P5EE website
(http://www.officevision.com/pub/p5ee) for more details.
4. Why another module? Why not use Getopt::Long or others?
There are many configuration modules on CPAN. See
http://search.cpan.org/modlist/Option_Parameter_Config_Processing
The most important feature was to be able to run it within a
BEGIN block to modify the Perl include path (@INC). This is
most of design goal #1. This means very few dependencies and
only on core modules.
I started by writing a few lines of code in a BEGIN block.
Then it got to be a lot of lines of code in a BEGIN block.
Then I moved it out to a module so I could reuse it easily.
Then it grew into a full-fledged command line, environment
variable, and config file value option processor.
I did try Getopt::Long, but it wasn't that easy to use, you had
to code your own "--help" feature, and it didn't incorporate
environment variables or config files. I wanted something
more high-level and full-featured, so I wrote App::Options.
The description of the AppConfig distribution sounds similar
to what is described here. However, the following are the key
differences.
* App::Options does its option processing in the BEGIN block.
This allows for the @INC variable to be modified in time
for subsequent "use" and "require" statements.
This met design goal #1.
* App::Options consults a cascading set of option files.
These files include those which are system global, project
global, and user private. This allows for system
administrators, project developers, and in individual
users to all have complementary roles on defining
the configuration values. This met design goal #2.
* App::Options "sections" (i.e. "[cleanup]") are conditional.
It is conditional in App::Options, allowing you to use one
set of option files to configure an entire suite of programs
and scripts. In AppConfig, the section name is simply a
prefix which gets prepended to subsequest option names.
This also helped to meet design goal #2.
* App::Options is not a toolkit but a standardized way of
doing option processing. With AppConfig, you still have
to decide where to put config files, you still have to
code the "--help" feature, etc. With App::Options, you simply
"use App::Options;" and all the hard work is done.
Advanced options can be added later as necessary as args
to the "use" statement. This met design goal #3.
Design goal #4 required the autodetection of the ${prefix}
variable. Thus, the App::Options module can be integrated with
the discipline of choosing a root directory for your
software installation (i.e. PREFIX=/usr/mycompany/2.0.12).
5. You say it's so easy. Show me.
#!/usr/bin/perl
use App::Options;
# now use the %App::options hash
print "yada yada\n" if ($App::options{verbose});
It's that easy.
And it automatically has "-?" and "--help" support without
you having to code a thing.
Read the man page (or the code) if you want more power.
6. How do I install it?
To install this module, cd to the directory that contains this README
file and type the following (as usual).
perl Makefile.PL
make
make test
make install