\input texinfo   @c -*-texinfo-*-

@c %**start of header
@setfilename argparser.info
@settitle argparser
@afourpaper
@documentencoding UTF-8
@documentlanguage en
@finalout
@c %**end of header


@dircategory Development
@direntry
* argparser: (argparser).            Command line argument parser library
@end direntry


@copying
Copyright @copyright{} 2013 Mattias Andrée

@quotation
Permission is granted to copy, distribute and/or modify this document
under the terms of the GNU Free Documentation License, Version 1.3 or
any later version published by the Free Software Foundation; with no
Invariant Sections, with no Front-Cover Texts, and with no Back-Cover
Texts. A copy of the license is included in the section entitled
``GNU Free Documentation License''.
@end quotation
@end copying

@ifnottex
@node Top
@top argparser -- Command line argument parser library
@insertcopying
@end ifnottex

@titlepage
@title argparser
@subtitle Command line argument parser library
@author by Mattias Andrée (maandree)

@page
@vskip 0pt plus 1filll
@insertcopying
@end titlepage

@contents



@menu
* Overview::                        Brief overview of @command{argparser}.
* Python version::                  Using the Python version.
* Java version::                    Using the Java version.
* C version::                       Using the C version.
@c* Bash version::                    Using the Bash version.
* GNU Free Documentation License::  Copying and sharing this manual.
@end menu



@node Overview
@chapter Overview

Command line argument parser library for multiple languages with the
capability of both @code{-} and @code{+} short options and both
@code{--} and @code{++} long options. It supports variadic options,
colours and @code{--} and temporary @code{--} using @code{++}.

A variadic option is a option that takes all arguments placed after
it as an argument associated with it, rather than just the directly
following argument as an argumented option does.

Options that are missing descriptions are not printed by the help
functions.



@node Python version
@chapter Python version

To use argparser you need to import @code{ArgParser} from @code{argparser}
using the instruction @code{from argparser import *}.

The class @code{ArgParser} is used for parsing argument and requires an
instance. @code{ArgParser}'s constructor takes two manditory arguments
and three optional arguments. The first argument, @var{description}, is
a short, single-line, description of the program. The second argument,
@var{usage}, is multi-line usage text that is recommended to contain ANSI
colour escape sequences (for portability ESC 39m, ESC 49m and colours
beyond the lower 8 are not not recommended to be used); @code{None} can
be used if you do not have a usage descriptor.
@code{ArgParser}'s constructors there optional arguments are:
@var{longdescription}, @var{program} and @var{usestderr}.
@var{longdescription} is a long, multi-line, description of the program;
@code{None} can be used if you do not have a long description.
@var{program} is the program part of the executed command, but if you
rather, you can provide an exact name. You can also use the static function
@code{ArgParser.parent_name} to get the program part of the executed command
of a parent proces, it takes two optional arguments: @var{levels}, the
number of parent levels, by default 1; and @var{hasinterpretor}, whether
to get the name from an invocation of Python, by default false.
@var{usestderr} is by default @code{False} which means that printing is
done to stdout, otherwise, printing is done to stderr.

Before parsing arguments you need to populate the @code{ArgParser}
instance with valid options. There are three methods for this,
@code{add_argumentless}, @code{add_argumented} and @code{add_variadic},
that adds argumentless options, argumented options and variadic options
respectively. With the exception that @code{add_argumentless} does not
have the parameter @var{arg}, these three arguments has one manditory
argument, @var{alternatives}, and three optional arguments: @var{default},
@var{arg}, @var{help}. @var{default}'s deault value is zero, and either
the primary alternative or its index in @var{alternatives}.
@var{alternatives} is a list of alternatives, for example, a help option
could have @code{['-h', '--help']}. @var{arg} is a descriptive short name
of a argument associated with the option, and @var{help} is description
of the option and may span multiple lines but should only do so if the
lines below the first is just extra details.

When you have populated your @code{ArgParser} with options, it is time
to parse arguments, it is done with the method @code{parses} that optional
takes and list of arguments. If you choose to use a list of arguments
rather than letting @code{ArgParser} use arguments used to start the
program, the first ement will not be parsed as it is assumed to be the
executable.

If you now want to use any option alternative rather than just the
primary (using just the primary is good to keep your code consistent)
invoke the nulladic method @code{support_alternatives}.

Before using your options you should so some checks that the combination
of options and arguments are valid. There are some methods provided to do
this. @code{test_exclusiveness} checks that no confliction options are
used, as the first argument, a set of options provided from which at most
one may be used; as a optional second argument, a return code can be provided
if you want the program to exit if there are option conflicts.
@code{test_allowed} checks that only allowed options are used, as the first
argument, a set of options provided in which all used arguments must exist;
as a optional second argument, a return code can be provided if you want the
program to exit if there are out of context option.
@code{test_files} checks that the number of arguments not associated with
an option is within an acccepted range, it takes three option arguments,
@var{min}, @var{max} and @var{exit_value}. @var{min} is the minimum count,
@var{max} is the maximum count, @code{None} if unlimited, and @var{exit_value}
is a return code can be provided if you want the program to exit if there
are out of context option. Remember that you should also check that the
number of times an option is used is acceptable.

After running @code{parses}, your @code{ArgParser} has five attributes:
@var{unrecognisedCount}, the number of unrecognised options; and
@var{message}, the join of @var{files} which is all arguments not
associated with an option, @var{arguments} the parsed arguments, and
@var{argcount}, the number of arguments in @var{arguments}.

All valid options are stored in your @code{ArgParser}'s @var{opts},
it is a dictionary from the option to either @code{None} if the option
has not been used, or a list of all used values, in order. A variadic
option without any argumnt will have a empty list and a argumentless
option will have list filled with @code{None}.

To print a help page based on the constructor arguments and populated
options invoke the nulladic method @code{help}.



@node Java version
@chapter Java version

The JAR for ArgParser, @file{ArgParser.jar}, is located in @file{/usr/lib}.
When using it you need to import @code{argparser.ArgParser}.

@code{ArgParser} has six constructors, conforming to:
@code{description, usage, [longDescription, [program]], [useStderr]}.
These correspond to those in the Python version.

@code{ArgParser.parent_name} from the Python version is named
@code{ArgParser.parentName} in the Java version and is identical
except it parses Java invocation rather then Python invocation.
Further @code{support_alternatives}, @code{test_exclusiveness},
@code{test_allowed}, @code{test_files} and @code{help} are named
@code{supportAlternatives}, @code{testExclusiveness}, @code{testAllowed},
@code{testFiles} and @code{help} respectively and @code{testFiles}
without @var{min} is named @code{testFilesMax} and @code{testFiles}
without @var{max} is named @code{testFilesMin}.

@code{parse} takes an array of arguments, which excludes the executable.

@var{message} is a @code{String}, @var{files} is a @code{ArrayList<String>},
@var{unrecognisedCount} is a @code{int}, @var{arguments} is a @code{String[]},
and @var{opts} is a @code{HashMap<String, String[]>}.
@var{argcount} does not exist.

To populate the @code{ArgParser} with valid options, use the method
@code{add} with the option and optional, before or after, the description.
An option is created using a constructor for either @code{Argumentless},
@code{Argumented} or @code{Variadic}.
@code{Argumentless}'s constructor takes an @code{String[]} of alternatives
and optional, before or after, the index of the primary alternative. If
the index is not after the alternative array, the alternatives parameter
is variadic and each alternative can be an argument.
@code{Argumentless}'s constructor and @code{Variadic}'s constructor also
takes a descriptive short name of a argument associated with the option
optioned anywhere, and the alterantive array is variadic if places last.



@node C version
@chapter C version

To use ArgParser, compile with the linker option @code{-largparser}
and use the include @code{#include <argparser.h>}. ArgParser uses the
namespace @code{args}. When you are done using ArgParser you should
free its resources using function @code{args_dispose}.

@code{args_init} corresponds the the @code{ArgParser} constructor in
the Python version, but all parameters are manditory.

@var{args_arguments}, @var{args_arguments_count},
@var{args_unrecognised_count}, @var{args_message} and @var{args_files}
works as their Python version counterparts, and @var{args_files_count}
tells how many elements there are in @var{args_files}.

@code{args_parent_name} takes the number of levels as a manditory
parameters, and no other parameter. @code{args_help} and
@code{args_support_alternatives} takes no parameters, and @code{args_parse}
takes @var{argc} and @var{argv}, the arguments in your main method.

@code{args_test_files_min}, @code{args_test_files_max} and
@code{args_test_files} works are their Java version counterparts, except
that an return code may not be specified.
@code{args_test_allowed} and @code{args_test_exclusiveness} both takes
a string array followed by the number of strings in that array, and no
return code.

To popule ArgParser with valid options, use the method @code{args_add_option}
which takes the options followed by a description as its parameters, the
description can be @code{NULL}.
To create an options use either @code{args_new_argumentless},
@code{args_new_argumented} or @code{args_new_variadic}. With the exception
that @code{args_new_argumentless} does not have the first parameters, these
methods have three parameters: a descriptive short name of a argument
associated with the option, the index of the primary alternative, and
finally a variadic parameter with the alternatives, meaning that each
alternative is an argument.

To list all valid options, use the method @code{args_get_opts} which returns
a string array of all options, and @code{args_get_opts_count} which returns
the number of strings in the array returned by @code{args_get_opts}.
@code{args_opts_contains} with a option checks whether the option is available,
while @code{args_opts_used} checks whether it is used. @code{args_opts_new}
adds option and @code{args_opts_clear} removes all values stored for a
option. @code{args_opts_append} which takes a option and a value as
parameters store a value for a option and @code{args_opts_put} sets all
value for an option, it works the same way, except a string array of values
is used rather than just a string with a value, @code{args_opts_put_count}
stores the number of values. The get a number of values for a option, which
include @code{NULL} for argumentless occurrences, use @code{args_opts_get_count},
and to retrieve the array of values, use @code{args_opts_get}. Both takes
only the option as parameter.



@node Bash version
@chapter Bash version

The Bash version is written to be supported for older versions of Bash,
and is used by @code{source}:ing @file{/usr/lib/argparser.bash}. It works
similarly to the C version and @code{args_dispose} should be invoked when
done, optionally with the ID of the process that allocated the resouces.

Rather than a boolean, @code{args_init} takes a file descriptor as the
fifth argument, @code{1} for stdout (default) and @code{2} for stderr.

@code{args_parent_name} takes a boolean as a second parameter that tells
whether to return all arguments, on one line each, and defaults to false.
Booleans are true if and only if the value is @code{1}.

@code{args_test_exclusiveness} and @code{args_test_allowed} are variadic
and does not have a parameter for the number of options.

@code{args_test_files_min} and @code{args_test_files_max} does not exist,
but @code{args_test_files} takes the minimum count and optionally the
maxmimum count.

To add valid options use the methods @code{args_add_argumentless},
@code{args_add_argumented} and @code{args_add_variadic}. With the
except that @code{args_add_argumentless} does not take a name of
the argument, these methods takes the index of the primary alternative,
a one word name of the argument, a description (empty for none), and
then one argument per alternative.

@var{args_argcount} is the number of elements in @var{args_arguments}
and @var{args_files} does not have any such counterpart.
@var{args_has_message} is @code{1} if there is a message, that is,
if the length of @var{args_files} is not zero, and otherwise @code{0}.

To read the parsed arguments, the method @code{args_option} is used.
Its exit value is @code{0} if successful or if the return is true,
@code{1} if the return is false, and @code{2} on failure.
@code{args_option has} with a option name is used to check if a option
is used. @code{args_option count} with a option name echos the number
of values associated with the option. @code{args_option null} with a
option name and value index checks whether the value with that index
is null. @code{args_option get} with a option name echos all values,
one line each, but if any number of indices are also used only those
values are echoed.



@node GNU Free Documentation License
@appendix GNU Free Documentation License
@include fdl.texinfo

@bye