Giter VIP home page Giter VIP logo

authorize-rule's Introduction

NAME
    Authorize::Rule - Rule-based authorization mechanism

VERSION
    version 0.007

SYNOPSIS
    This is an extensive example, showing various options:

        my $auth = Authorize::Rule->new(
            rules => {
                dev => {
                    Payroll => [ [0] ], # always deny
                    ''      => [ [1] ], # default allow for unknown resources
                },

                tester => {
                    '' => [
                        # labeled rulesets
                        'check tester' => [
                            # all rules must apply
                            # key 'is_test' with value 1
                            # and keys test_name/test_id must exist
                            1, { is_test => 1 }, 'test_name', 'test_id'
                        ],
                        'default' => [0],
                    ]
                },

                admin => {
                    '' => [
                        # the admin does *not* have a passwordless ssh key
                        [ 1, { passwordless_ssh_key => undef } ],
                    ],
                },

                ceo => {
                    '' => [
                        [
                            # decide on the value of the return ourselves
                            # based on the resource
                            sub {
                                my $res = shift;
                                return 'Access Granted for ' . $res->{'resource'};
                            },

                            # a rule that is itself a subroutine
                            # with access to the parameters
                            sub { has_permission( $_[0]->{'resource'} ) },

                            # a rule with a key that matches the result of a sub
                            { now => sub { correct_relative_time() } },
                        ],
                    ],
                },

                biz_rel => {
                    Graphs    => [ [0] ],
                    Databases => [
                        # access to reservations table
                        [ 1, { table => 'Reservations' } ],
                    ],

                    Invoices => [
                        [ 0, 'user' ],
                        [ 1         ],
                    ],

                    Payroll => [ [1] ],
                    Revenue => [ [1] ],
                    ''      => [ [0] ],
                },

                support => {
                    Databases => [
                        [ 1, { table => 'Complaints' } ],
                    ],

                    Invoices => [ [1] ],
                    ''       => [ [0] ],
                },

                sysadmins => {
                    Graphs => [ [1] ],
                    ''     => [ [0] ],
                },
            },

            entity_groups => {
                sysadmins => [ qw<John Jim Goat> ],
            },

            resource_groups => {
                Graphs => [ 'ThisGraphs', 'ThoseGraphs' ],
            },
        );

    *(this example is not taken from any actual code)*

DESCRIPTION
    Authorize::Rule allows you to provide a set of rulesets, each containing
    rules, for authorizing access of entities to resources. This does not
    cover authentication, or fine-grained parameter checking.

    While authentication asks "who are you?", authorization asks "what are
    you allowed to do?"

    The system is based on decisions per entities, resources, and any
    optional parameters.

ALPHA CODE
    I can't promise some of this won't change in the next few versions.

    Stay tuned.

SPECIFICATION
    The specification covers several elements:

    *   Entity

    *   Resource

    *   Action

    *   Optional parameters

    *   Optional label

    The general structure is:

        {
            ENTITY => {
                RESOURCE => [
                    OPTIONAL_LABEL => [ ACTION, RULE_1, RULE_2, ...RULE_N ],
                ]
            }

        }

    Allowed rules are:

        # parameters must have this key with this value
        [ $action, { key  => 'value' } ]
        [ $action, { name => 'Marge' } ]

        # parameters must not have this key
        [ $action, { key => undef } ]

        # parameters must have these keys, values aren't checked
        [ $action, 'key1', 'key2', ... ]

        # they can be seamlessly mixed
        [ $action, { Company => 'Samsung' }, { Product => 'Phone' }, 'model_id' ]

        # and yes, this is the equivalent of:
        [ $action, { Company => 'Samsung', Product => 'Phone' }, 'model_id' ]

        # a mix of keys with expected values and keys expected not to exist
        [ $action, { name => 'Marge', holding_knife => undef } ]

        # labels can be applied to rulesets:
        'verifying test account' => [ $action, { username => 'tester' } ]

        # rules can be a subroutine
        [ $action, sub { my $params = shift; ... } ]

        # keys can match to a subroutine result
        [ $action, { Company => sub { get_company( $_[0]->{'company'} ) } } ]

        # and lastly, actions can be subroutines
        [ sub { my $result_hash = shift; return 'OK' if ... }, { %params } ]

    An action is either true, false, or a code reference which returns one.
    The recommended values for true or false are 1 or 0. Traditionally these
    will be 1 or 0:

        [ 1, RULES... ]
        [ 0, RULES... ]
        [ 'FAILURE', RULES... ]

        my $result = $auth->is_allowed( $entity, $resource );
        if ( $result eq 'FAILURE' ) {
            ...
        }

    Or as a code reference:

        [ sub {...}, RULES... ]

    The code reference receives the entire result hash as a parameter:

        [ sub {
            my $result = shift;

            # $result = {
            #     ruleset_idx => 1,
            #     params      => $PARAMS,
            #     entity      => $ENTITY,
            #     resource    => $RESOURCE,
            # }
        }, RULES... ]

    Rules and rulesets are read consecutively, so you might want to position
    your rules in order to exit early. When a ruleset is matches, the
    execution or rulesets stops and the action is returned.

EXAMPLES
  All resources
    Cats think they can do everything, and they can:

        my $rules = {
            Cat => {
                # default rule for any unmatched resource
                '' => [
                    # only 1 ruleset with no actual rules, just an action
                    [1]
                ],
            }
        }

        my $auth = Authorize::Rule->new( rules => $rules );
        $auth->is_allowed( cats => 'kitchen' ); # 1, success
        $auth->is_allowed( cats => 'bedroom' ); # 1, success

    If you don't like the example of cats (what's wrong with you?), try to
    think of a department (or person) given all access to all resources in
    your company:

        $rules = {
            Sysadmins => {
                '' => [ [1] ],
            },

            CEO => {
                '' => [ [1] ],
            },
        }

  Per resource
    Dogs, however, provide less of a problem. Mostly if you tell them they
    aren't allowed somewhere, they will comply. Dogs can't get on the table.
    Other than the table, we do want them to have access everywhere.

        $rules = {
            Cat => {
                '' => [ [1] ],
            },

            Dog => {
                Table => [ [0] ], # can't go on the table
                ''    => [ [1] ], # otherwise, allow everything
            },
        }

    A corporate example might refer to some departments (or persons) having
    access to some resources while denied everything else, or a certain
    resource not available to some while all others are.

        $rules = {
            CEO => {
                Payrolls => [ [0] ], # no access to Payrolls
                ''       => [ [1] ], # access to everything else
            },

            Support => {
                UserPreferences      => [ [1] ], # has access to this
                UserComplaintHistory => [ [1] ], # and this
                ''                   => [ [0] ], # but that's it
            },
        }

  Per resource and per conditions
    This is the most extensive control you can have. This allows you to set
    permissions based on conditions, such as specific parameters per
    resource.

    Suppose we have no problem for the dogs to walk on that one table we
    don't like?

        my $rules => {
            Dog => {
                Table => [
                    # if the table is owned by someone else, it's okay
                    [ 1, { owner => 'someone-else' } ],

                    # otherwise, no
                    [0],
                ],

                '' => [ [1] ], # but generally dogs can go everywhere
            }
        };

        my $auth = Authorize::Rule->new( rules => $rules );
        $auth->is_allowed( Dog => 'Table', { owner => 'me' } ); # 0, fails

    You can also specify just the existence (and "define"ss) of keys:

        my $rules = {
            Support => {
                ClientTable => [
                    [ 1, 'user_id' ], # must have a user id to access the table
                    [0],              # otherwise, access denied
                ]
            }
        };

  OR conditions
    If you want to create an OR condition, all you need is to provide
    another ruleset:

        my $rules = {
            Dog => {
                Table => [
                    [ 1, { carer => 'Jim'  } ], # if Jim takes care of the dog
                    [ 1, { carer => 'John' } ], # or if John does
                    [0],                        # otherwise, no
                ]
            }
        };

        $auth->is_allowed( Dog => 'Table', { owner => 'me'   } ); # 0, fails
        $auth->is_allowed( Dog => 'Table', { owner => 'Jim'  } ); # 1, succeeds
        $auth->is_allowed( Dog => 'Table', { owner => 'John' } ); # 1, succeeds

  AND conditions
    If you want to create an AND condition, just add more rules to the
    ruleset:

        my $rules = {
            Dog => {
                Table => [
                    [
                        1,                     # allow if...
                        { carer => 'John'   }, # john is the carer
                        { day   => 'Sunday' }, # it's Sunday
                        { clean => 1        }, # you're clean
                        'tag_id',              # and you have a tag id
                    # otherwise, no
                    [0],
                ]
            }
        };

    As shown in other examples above, any hash rules can be put in the same
    hash, so this is equivalent:

        my $rules = {
            Dog => {
                Table => [
                    [
                        1,                     # allow if...
                        {
                            carer => 'John',   # john is the carer
                            day   => 'Sunday', # it's Sunday
                            clean => 1,        # you're clean
                        },
                        'tag_id',              # and you have a tag id
                    # otherwise, no
                    [0],
                ]
            }
        };

    The order of rules does not change anything, except how quickly it might
    mismatch. If you have insane amounts of rules and conditions, it could
    make a difference, but unlikely.

  labeling
    Optional labels can be applied in order to help structure rulesets and
    understand which ruleset matched.

        my $rules = {
            Tester => {
                # Tester's rulesets for any resource
                '' => [
                    # regular ruleset
                    [ 1, 'test_mode' ], # if we're in test_mode

                    # labeled ruleset
                    'has test ID' => [ 1, 'test_id' ], # has a test ID
                ],
            },
        };

    Labeled and unlabeled rulesets can be interchanged freely.

  Catch all
    You might ask *what if there is no last rule at the end for any other
    resource?*

    The answer is simple: the "default" clause will be used. You can find an
    explanation of it under *ATTRIBUTES*.

  Callbacks
   As rule
    A rule can be a callback:

        my $rules = {
            Marge => {
                '' => [
                    [ 1, sub {
                        my $params = shift;

                        time - $params->{'now'} < 10
                            and return 1;

                        return 0;
                    } ]
                ],
            }
        };

        $auth->is_allowed( 'Marge', 'Anywhere', { now => time } );

   As parameter
    You can compare a parameter value to the result of a callback:

        my $rules = {
            Marge => {
                '' => [
                    [ 1, { name => sub { get_name( $_[0]->{'entity'} ) } } ]
                ]
            }
        };

    This will compare the "name" value in the parameters to whatever will be
    returned by "get_name", which gets as a first argument the "entity" that
    is used - in this case, *Marge*.

   As action
    You can also set the action to be a callback, which allows to do two
    interesting things:

    *   Change the action

    *   Investigate the result hash

        my $rules = {
            Marge => {
                '' => [
                    [
                        sub {
                            my $result = shift;
                            return 'SucceededAt' . $result->{'resource'};
                        },
                        { time => 'now' }, # rule
                    ],
                ]
            }
        };

        my $auth   = Authorize::Rule->new( rules => $rules );
        my $action = $auth->is_allowed( 'Marge', 'Somewhere', %params );
        # $action = 'SucceededAtSomewhere'

    The result hash will contain information on the request and the
    matching, such as the ruleset which was matched.

ATTRIBUTES
  default
    In case there is no matching rule for the entity/resource/conditions,
    what would you like to do. The default is to deny (0), but you can
    change it to allow by default if there is no match.

        Authorize::Rule->new(
            default => 1, # allow by default
            rules   => {...},
        );

        Authorize::Rule->new(
            default => -1, # to make sure it's the catch-all
            rules   => {...},
        );

  entity_groups
    Entity groups allow you to group entities onto their own label. This
    means you can set up multiple entities at the same time, while still
    matching them by the entity name instead of group name.

        my $auth = Authorize::Rule->new(
            rules => {
                'My Group' => {
                    Desk => [ [1] ],
                },
            },

            entity_groups => {
                'My Group' => [ qw<Sawyer Mickey> ],
            },
        );

        # OK
        $auth->is_allowed( 'Sawyer', 'Desk' );

  resource_groups
    Resource groups allow you to group resources onto their own label, much
    like *entity_groups*. You can set up multiple resources at the same
    time, while still matching them by the resource name instead of the
    group name.

        my $auth = Authorize::Rule->new(
            rules => {
                Person => {
                    Home => [ [1] ],
                },
            },

            resource_groups => {
                Home => [ 'Bedroom', 'Living Room', ... ],
            },
        );

        # OK
        $auth->is_allowed( 'Person', 'Bedroom' );

  rules
    Rules can be either:

    *
         A hash reference of your permissions, defined by the specification explained
        above.

    *   A key name (string) indicating this key must exist with no
        restriction to the value other than it must be defined.

    *   A callback with a result that provides the success or fail in
        boolean context.

METHODS
  is_allowed
    Returns the action for the entity and resource.

    Effectively, this is the "action" key in the result coming from the
    "allowed" method described below.

  allowed
        my $result = $auth->allowed( $entity, $resource, $params );

    Returns an entire hash containing every piece of information that might
    be helpful:

    *   entity

    *   resource

    *   params

    *   action

    *   label

    *   ruleset_idx

        The index of the ruleset, starting from 1.

AUTHORS
    *   Sawyer X <[email protected]>

    *   Mickey Nasriachi <[email protected]>

COPYRIGHT AND LICENSE
    This software is copyright (c) 2014 by Sawyer X.

    This is free software; you can redistribute it and/or modify it under
    the same terms as the Perl 5 programming language system itself.

authorize-rule's People

Contributors

xsawyerx avatar

Watchers

 avatar  avatar  avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    ๐Ÿ–– Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. ๐Ÿ“Š๐Ÿ“ˆ๐ŸŽ‰

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google โค๏ธ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.