3 === Revise MetaRole API to reunify class/role metaroles:
10 role_attribute => [ ... ],
14 If the $meta is a class, we apply the roles to the class. If it's a role, we
15 hold onto them and apply them as part of applying the role to a class.
17 To make this all work nicely, we'll probably want to track the original role
18 where a method was defined, just like we do with attributes currently. We'll
19 also need to store method modifiers with their original role, which may mean
20 adding some sort of Moose::Meta::Role::MethodModifier class.
22 For each role-specific thing (methods, attributes, etc.) we should allow a
23 role_attribute, role_method, etc. key. The common case will be that the
24 metaroles are intended for the consuming class, but we should allow for
25 metaroles on the role's metaobjects as well.
27 == Old todo (does anyone look at this?)
29 -------------------------------------------------------------------------------
31 -------------------------------------------------------------------------------
33 -------------------------------------------------------------------------------
35 -------------------------------------------------------------------------------
41 => where { Encode::is_utf8( $_[0] ) or $_[0] !~ m/[^0x00-0x7F]/x }
42 => optimize_as { defined($_[0]) && !ref($_[0]) };
46 => where { !Encode::is_utf8( $_[0] ) }
47 => optimize_as { defined($_[0]) && !ref($_[0]) };
51 Add support for doing it with Classes which do not have
52 a type constraint yet created
58 - inherited slot specs
60 'does' can be added to,.. but not changed
61 (need type unions for this)
65 a proxied attribute is an attribute
66 which looks like an attribute,
67 talks like an attribute, smells
68 like an attribute,.. but if you
69 look behind the curtain,.. its
70 over there.. in that other object
72 (... probably be a custom metaclass)
76 [13:16] mst stevan: slight problem with coerce
77 [13:16] mst I only get to declare it once
78 [13:17] mst so if I'm trying to declare it cast-style per-source-class rather than per-target-class
79 [13:17] mst I am extremely screwed
81 [13:17] stevan they are not class specific
82 [13:18] stevan they are attached to the type constraint itself
83 [13:18] * stevan ponders anon-coercion-metaobjects
84 [13:18] mst yes, that's fine
85 [13:19] mst but when I declare a class
86 [13:19] mst I want to be able to say "this class coerces to X type via <this>"
87 [13:19] stevan yeah something like that
88 [13:19] stevan oh,.. hmm
89 [13:20] stevan sort of like inflate/deflate?
90 [13:20] stevan around the accessors?
91 [13:25] * bluefeet has quit (Remote host closed the connection)
93 [13:27] mst nothing like that
94 [13:27] mst like a cast
95 [13:31] mst stevan: $obj->foo($bar); where 'foo' expects a 'Foo' object
96 [13:31] mst stevan: is effectively <Foo>$bar, right?
97 [13:32] mst stevan: I want to be able to say in package Bar
98 [13:32] mst stevan: coerce_to 'Foo' via { ... };
102 -----------------------------------------------------------
103 -- Type Constraints refactor
104 -----------------------------------------------------------
106 - add support for locally scoped TC
108 This would borrow from MooseX::TypeLibrary to prefix the TC with the name
109 of the package. It would then be accesible from the outside as the fully
110 scoped name, but the local attributes would use it first. (this would need support
111 in the registry for this).
113 - look into sugar extensions
115 Use roles as sugar layer function providers (ala MooseX::AttributeHelpers). This
116 would allow custom metaclasses to provide roles to extend the sugar syntax with.
118 (NOTE: Talk to phaylon a bit more on this)
120 - allow a switch of some kind to optionally turn TC checking off at runtime
122 The type checks can get expensive and some people have suggested that allowing
123 the checks to be turned off would be helpful for deploying into performance
124 intensive systems. Perhaps this can actually be done as an option to make_immutable?
128 * make the errors for TCs use ->message
129 * look into localizing the messages too
130 * make ANON TCs be lazy, so they can possibly be subsituted for the real thing later
131 * make ANON TCs more introspectable
139 where { (/[a-z][a-z0-9]+/i or fail('Invalid character(s)'))
140 and (length($_) >= 5 or fail('Too short (less than 5 chars)'))
142 on_fail { MyException->throw(value => $_[0], message => $_[1]) };
144 # fail() will just return false unless the call is made via
145 $tc->check_or_fail($value);
147 * and then something like this:
149 subtype Foo => as Bar => where { ... } => scoped => -global;
150 subtype Foo => as Bar => where { ... } => scoped => -local;
154 subtype Foo => as Bar => where { ... } => in __PACKAGE__ ;
156 # or (not sure if it would be possible)
158 my $Foo = subtype Bar => where { ... };
162 [17:10] <autarch> stevan: it should do it if I pass coerce => 1 as part of the attribute definition
163 [17:12] <stevan> autarch: what I am not 100% sure of is how to tell it to deep coerce and when to not
164 [17:13] <stevan> cause a basic coerce is from A to B
165 [17:13] <autarch> hmm
166 [17:13] <stevan> which is valid for collection types too
167 [17:13] <stevan> deep coercion is what you are asking for
168 [17:13] <autarch> yeah
169 [17:13] <stevan> so perhaps we add deep_coerce => 1
170 [17:13] <stevan> which will do it
171 [17:13] <autarch> that's fine for me
174 coerce_deeply => 1 # reads better
176 -------------------------------------------------------------------------------
178 -------------------------------------------------------------------------------
180 - rationalize all the get_X methods for classes (and roles)
182 We have get_attribute, get_attributes_list, get_all_attributes,
183 etc. First, we need to make the method names consistent. If something
184 returns an attribute vs a name, that needs to be clear from the method
185 name. We also need to make sure that local vs. "entire inheritance
186 chain" is clear from the name.
188 This is mostly a CMOP change.
190 - Metaclass constructors
192 There's a _lot_ of different conventions in here. Some things to consider:
195 * allowing new( 'name', %args ) vs ( name => 'name', %args )
196 * Method->wrap vs Method->new
198 - Moose::Meta::TypeConstraint::Parameter{izable,ized}
200 The relationship between these two classes is very odd. In particular,
201 this line in Parameterized is insane:
203 foreach my $type (Moose::Util::TypeConstraints::get_all_parameterizable_types()) {
205 Why does it need to loop through all parameterizable types? Shouldn't
206 it know which parameterizable type it "came from"?
208 - Moose::Util::TypeConstraints vs Moose::Meta::Type{Coercion,Constraint}
210 The Util module has _way_ too much functionality. It needs to be
211 refactored so it's a thin sugar layer on top of the meta API. As it
212 stands now, it does things like parse type names (and determine if
213 they're valid), manage the registry, and much more.
215 - Anything with a _(meta)?class method
217 Every method that returns a class name needs to become a rw attribute
218 that can be set via the constructor.
220 - The Moose::Error stuff
222 This is sort of half-implemented. We still use Carp directly, and the
223 internals can't decide how to throw an error (is it
224 Moose->throw_error, __PACKAGE__->throw_error, what?).
226 The internals need to be made consistent before we expose this to the
229 -------------------------------------------------------------------------------
231 -------------------------------------------------------------------------------
233 - Moose "strict" mode
235 use Moose 'strict'; This would allow us to have all sort of expensive tests
236 which can be turned off in prod.
238 - Moose::Philosophy.pod
240 To explain Moose from a very high level
244 We certainly have enough meta-information to make pretty complete POD docs.