clean up some test stuff
[gitmo/Moose.git] / TODO
CommitLineData
5dbb47c8 1== Todo for 2.0400
85910734 2
1b1eb1ca 3=== RT Tickets
4
c78a59d7 5RT#67731 - Union types reports spurious parent types -
6https://rt.cpan.org/Ticket/Display.html?id=67731
7
8RT#59478/RT#63000 - 0+ overload causes NV conversion on == on perls before
95.14 - this causes comparisons to fail when the number can't fit in an NV
10without precision loss. I'd like to fix this in a more general way (forcing
11anyone else who might be using == on tc objects to do weird things isn't very
12good), although it's hard to test to see what actually works.
13
14RT#69839 - UNIVERSAL methods should show up as methods, so things like method
15modifiers work
1b1eb1ca 16
b927b1d4 17=== Register implicitly created class/role types
18
19When you do has foo => (isa => 'Bar'), it returns a class_type for Bar, but
20doesn't register it. This means that later you can declare "subtype 'Bar', as
21'Str', ..." and it'll work, and later instances of the 'Bar' type will use that
22one. We should register the implicitly created ones so that trying to redefine
23it after it's used throws an error.
24
25== Todo for 2.0600
26
85910734 27=== Revise MetaRole API to reunify class/role metaroles:
28
29 apply_metaroles(
30 for => $meta,
31 roles => {
32 attribute => [...],
33 class => [...],
34 role_attribute => [ ... ],
35 }
36 );
37
38If the $meta is a class, we apply the roles to the class. If it's a role, we
39hold onto them and apply them as part of applying the role to a class.
40
41To make this all work nicely, we'll probably want to track the original role
42where a method was defined, just like we do with attributes currently. We'll
43also need to store method modifiers with their original role, which may mean
44adding some sort of Moose::Meta::Role::MethodModifier class.
45
46For each role-specific thing (methods, attributes, etc.) we should allow a
47role_attribute, role_method, etc. key. The common case will be that the
48metaroles are intended for the consuming class, but we should allow for
49metaroles on the role's metaobjects as well.
50
ef0801e7 51=== Deprecate old-style Moose extensions
52
53Moose extensions that work by calling Moose->init_meta(metaclass =>
54'Some::Custom::Metaclass', ...) during their own init_meta should be
55deprecated, so they can be removed later (this should fix the issues with
56init_meta generation in Moose::Exporter, see RT51561)
57
d8fed7e3 58This needs to wait until the previous fix gets in, since it will hopefully eliminate the need to write custom init_meta methods entirely.
59
24b9aca7 60=== Attributes in roles need to be able to participate in role composition
61
62Right now, this fails with no decent workaround:
63
64 package R1;
65 use Moose::Role;
66 has foo => (is => 'ro');
67
68 package R2;
69 use Moose::Role;
70 with 'R1';
71 requires 'foo';
72
73 package C;
74 use Moose;
75 with 'R2';
76
77Role attributes really need to be able to participate in role-role combination.
78This should also fix "with 'Role1', 'Role2'" being broken when Role1 implements
79a method as an accessor and Role2 requires that method, but at least in that
80case you can split it into two 'with' statements with minimal loss of
81functionality.
82
83=== Method modifiers in roles should silently add 'requires' for them
84
85This shouldn't be a functionality change, just a better error message (and
86better introspectability). This shouldn't happen if the role already contains a
87method by that name, so it'll depend on the previous fix going in (so "has foo
88=> (is => 'ro'); around foo => sub { }" doesn't produce a 'requires' entry).
89
90=== Add overloading support
91
92or at least, don't break existing overloading support
93
94This shouldn't treat the overloading stuff as actual methods, since that's just
95an implementation detail, but we should provide an API for add_overload,
96get_overload, get_overload_list, etc. In particular, this would allow
97namespace::autoclean to not break things.
98
99Also, MooseX::Role::WithOverloading should probably be cored.
100
d8fed7e3 101This should probably also wait for the metarole unification fix, to avoid the
102::WithOverloading stuff being too insane.
103
24b9aca7 104
105== Todo for later
106
3091d476 107=== has +foo in roles
108
109There's no actual reason for this not to work, and it gets asked often enough
110that we really should just do it at some point.
111
24b9aca7 112=== Actual API for metaclass extensions
113
114Right now, the only way to bundle multiple metaclass traits is via
115Moose::Exporter. This is unhelpful if you want to apply the extension to a
116metaclass object rather than a class you're actually writing. We should come up
117with an API for doing this.
118
119=== MooseX::NonMoose in core
120
121I think all of the actual issues are solved at this point. The only issue is
122the (necessary) implementation weirdness - it sets up multiple inheritance
123between the non-Moose class and Moose::Object, and it installs a custom
124constructor method at 'extends' time (although perhaps this could be solved by
125moving some of the logic back into Moose::Object::new?). Other than that, it
126handles everything transparently as far as I can tell.
127
128=== Fix attribute and method metaclass compatibility
129
130So i got this wrong when rewriting it last year - right now, metaclass compat
131checks the default attribute and method metaclasses, which is wrong. This means
132that if a parent class does "use MooseX::FollowPBP", then attributes declared
133in a subclass will get PBP-style accessors, which is quite surprising.
134
135On the other hand, sometimes metaclasses might need to be able to say "I'm
136going to assume that all of my attributes at least inherit from this custom
137class", so we might need to split it into "default specified by the user" and
138"default specified by the metaclass" and only do compat checking on the second?
139I'm not actually sure this is a valid use case though.
140
141Something that probably should be taken into account though is attributes and
142methods that extend existing attributes or methods from a superclass should
143inherit the metaclass of the existing one. Also not sure if this is correct,
144but something to think about.
145
146=== Rename a bunch of the public API methods
147
148Right now the public API is kind of a mess - we have things like get_method vs
149find_method_by_name (you almost always want to use the latter), there being no
150has_method equivalent that checks superclasses, get_method_list being public
151but only returning method names, while _get_local_methods is private (returning
152method objects), and yet neither of those looks at superclasses, and basically
153none of this naming follows any kind of consistent pattern.
154
155What we really need is a consistent and easy to remember API where the method
156that people would think to use first is the method that they actually mean.
157Something like renaming find_method_by_name to find_method, and get_method to
158find_local_method or something along those lines.
159
6ddfb832 160=== use Sub::Identify instead of doing our own thing with get_code_info
161
162No idea why we stopped using Sub::Identify in the past, but there's no reason
163not to do this. We have a bug fix in our version (the isGV_with_GP thing), so
164this should be submitted to Sub::Identify first.
165
c762df57 166=== Move method modifiers out to an external module
167
168Class::Method::Modifiers uses a different method for doing method modifiers,
169which I'm not sure why we aren't using in Moose right now. Optionally using
170Class::Method::Modifiers::Fast would be even better - it uses Data::Util to
171implement XS method modifiers, which could help things a lot.
172
173=== Move type constraints out to an external module
174
175There's nothing about our type constraint system that requires being tied to
176Moose - it's conceptually an entirely separate system that Moose just happens
177to use. Splitting it out into its own thing (that Moose could extend to add
178things like role types) would make things conceptually a lot cleaner, and would
179let people interested in just the type system have that.
180
181=== Merge Class::MOP and Moose
182
183This is a long term goal, but would allow for a lot of things to be cleaned up.
184There's a bunch of stuff that's duplicated, and other stuff that's not
185implemented as well as it could be (Class::MOP::Method::Wrapped should be a
186role, for instance).
187
24b9aca7 188
189== Things to contemplate
190
191=== Does applying metaroles really need to reinitialize the metaclass?
192
193Seems like the logic that's actually necessary is already contained in
194rebless_instance, and not reinitializing means that existing attributes and
195methods won't be blown away when metaroles are applied.
196
197=== Do we want to core namespace::autoclean behavior somehow?
198
199This would add Variable::Magic as a required XS dep (not a huge deal at the
200moment, since Sub::Name is also a required XS dep, but it'd be nice for Moose
201to be able to be pure perl again at some point in the future, and I'm not sure
202what the relative chances of Sub::Name vs Variable::Magic making it into core
203are). If we enabled it by default, this would also break things for people who
204have introduced Moose into legacy-ish systems where roles are faked using
205exporters (since those imported methods would be cleaned).
206
207If we decide we want this, we may want to core it as an option first ("use
208Moose -clean" or so), and move to making it the default later.
209
210=== Fix the error system
211
212oh god it's terrible
213
214More specifically, we really want exception objects.
215
216=== Should using -excludes with a role add 'requires' for excluded methods?
217
218It seems to make sense, since otherwise you're violating the role's API
219contract.
220
221
85910734 222== Old todo (does anyone look at this?)
223
8b59f8d6 224-------------------------------------------------------------------------------
7af2c1d2 225BUGS
d03bd989 226-------------------------------------------------------------------------------
43d599e5 227
228-------------------------------------------------------------------------------
7af2c1d2 229FEATURES
8b59f8d6 230-------------------------------------------------------------------------------
231
d4967760 232- DDuncan's Str types
233
d03bd989 234subtype 'Str'
235 => as 'Value'
236 => where { Encode::is_utf8( $_[0] ) or $_[0] !~ m/[^0x00-0x7F]/x }
d4967760 237 => optimize_as { defined($_[0]) && !ref($_[0]) };
238
d03bd989 239subtype 'Blob'
240 => as 'Value'
241 => where { !Encode::is_utf8( $_[0] ) }
d4967760 242 => optimize_as { defined($_[0]) && !ref($_[0]) };
243
8b59f8d6 244- type unions
245
d03bd989 246Add support for doing it with Classes which do not have
8b59f8d6 247a type constraint yet created
248
249- type intersections
250
251Mostly just for Roles
252
253- inherited slot specs
254
db1ab48d 255'does' can be added to,.. but not changed
256(need type unions for this)
8b59f8d6 257
8b59f8d6 258- proxy attributes
259
db1ab48d 260a proxied attribute is an attribute
d03bd989 261which looks like an attribute,
262talks like an attribute, smells
263like an attribute,.. but if you
264look behind the curtain,.. its
db1ab48d 265over there.. in that other object
266
267(... probably be a custom metaclass)
8b59f8d6 268
f90e052d 269- local coerce
270
271[13:16] mst stevan: slight problem with coerce
272[13:16] mst I only get to declare it once
273[13:17] mst so if I'm trying to declare it cast-style per-source-class rather than per-target-class
274[13:17] mst I am extremely screwed
275[13:17] stevan yes
d03bd989 276[13:17] stevan they are not class specific
f90e052d 277[13:18] stevan they are attached to the type constraint itself
278[13:18] * stevan ponders anon-coercion-metaobjects
279[13:18] mst yes, that's fine
280[13:19] mst but when I declare a class
281[13:19] mst I want to be able to say "this class coerces to X type via <this>"
282[13:19] stevan yeah something like that
283[13:19] stevan oh,.. hmm
284[13:20] stevan sort of like inflate/deflate?
285[13:20] stevan around the accessors?
286[13:25] * bluefeet has quit (Remote host closed the connection)
287[13:27] mst no
288[13:27] mst nothing like that
289[13:27] mst like a cast
290[13:31] mst stevan: $obj->foo($bar); where 'foo' expects a 'Foo' object
291[13:31] mst stevan: is effectively <Foo>$bar, right?
292[13:32] mst stevan: I want to be able to say in package Bar
293[13:32] mst stevan: coerce_to 'Foo' via { ... };
294[13:32] mst etc.
d03bd989 295[13:53] stevan hmm
296
f3dc2e80 297-----------------------------------------------------------
298-- Type Constraints refactor
299-----------------------------------------------------------
300
301- add support for locally scoped TC
302
303This would borrow from MooseX::TypeLibrary to prefix the TC with the name
304of the package. It would then be accesible from the outside as the fully
305scoped name, but the local attributes would use it first. (this would need support
306in the registry for this).
307
308- look into sugar extensions
309
310Use roles as sugar layer function providers (ala MooseX::AttributeHelpers). This
311would allow custom metaclasses to provide roles to extend the sugar syntax with.
312
313(NOTE: Talk to phaylon a bit more on this)
314
315- allow a switch of some kind to optionally turn TC checking off at runtime
316
317The type checks can get expensive and some people have suggested that allowing
318the checks to be turned off would be helpful for deploying into performance
319intensive systems. Perhaps this can actually be done as an option to make_immutable?
320
321- misc. minor bits
322
323* make the errors for TCs use ->message
324* look into localizing the messages too
325* make ANON TCs be lazy, so they can possibly be subsituted for the real thing later
326* make ANON TCs more introspectable
327* add this ...
328
329#
330# Type Definition
331#
332subtype 'Username',
333 from 'Str',
334 where { (/[a-z][a-z0-9]+/i or fail('Invalid character(s)'))
335 and (length($_) >= 5 or fail('Too short (less than 5 chars)'))
336 }
337on_fail { MyException->throw(value => $_[0], message => $_[1]) };
338
339# fail() will just return false unless the call is made via
340$tc->check_or_fail($value);
341
342* and then something like this:
343
344subtype Foo => as Bar => where { ... } => scoped => -global;
345subtype Foo => as Bar => where { ... } => scoped => -local;
346
347# or
348
349subtype Foo => as Bar => where { ... } => in __PACKAGE__ ;
350
351# or (not sure if it would be possible)
352
353my $Foo = subtype Bar => where { ... };
354
355# ----------
356
357[17:10] <autarch> stevan: it should do it if I pass coerce => 1 as part of the attribute definition
358[17:12] <stevan> autarch: what I am not 100% sure of is how to tell it to deep coerce and when to not
359[17:13] <stevan> cause a basic coerce is from A to B
360[17:13] <autarch> hmm
361[17:13] <stevan> which is valid for collection types too
362[17:13] <stevan> deep coercion is what you are asking for
363[17:13] <autarch> yeah
364[17:13] <stevan> so perhaps we add deep_coerce => 1
365[17:13] <stevan> which will do it
366[17:13] <autarch> that's fine for me
367[17:13] <stevan> k
368
369coerce_deeply => 1 # reads better
370
7af2c1d2 371-------------------------------------------------------------------------------
372INTERNALS
373-------------------------------------------------------------------------------
374
375- rationalize all the get_X methods for classes (and roles)
376
377We have get_attribute, get_attributes_list, get_all_attributes,
378etc. First, we need to make the method names consistent. If something
379returns an attribute vs a name, that needs to be clear from the method
380name. We also need to make sure that local vs. "entire inheritance
381chain" is clear from the name.
382
7af2c1d2 383This is mostly a CMOP change.
384
385- Metaclass constructors
386
387There's a _lot_ of different conventions in here. Some things to consider:
388
389* new vs _new
390* allowing new( 'name', %args ) vs ( name => 'name', %args )
391* Method->wrap vs Method->new
392
7af2c1d2 393- Moose::Meta::TypeConstraint::Parameter{izable,ized}
394
395The relationship between these two classes is very odd. In particular,
396this line in Parameterized is insane:
397
398 foreach my $type (Moose::Util::TypeConstraints::get_all_parameterizable_types()) {
399
400Why does it need to loop through all parameterizable types? Shouldn't
401it know which parameterizable type it "came from"?
402
403- Moose::Util::TypeConstraints vs Moose::Meta::Type{Coercion,Constraint}
404
405The Util module has _way_ too much functionality. It needs to be
406refactored so it's a thin sugar layer on top of the meta API. As it
407stands now, it does things like parse type names (and determine if
408they're valid), manage the registry, and much more.
409
7af2c1d2 410- Anything with a _(meta)?class method
411
412Every method that returns a class name needs to become a rw attribute
413that can be set via the constructor.
d03bd989 414
cac484fa 415- The Moose::Error stuff
416
417This is sort of half-implemented. We still use Carp directly, and the
418internals can't decide how to throw an error (is it
419Moose->throw_error, __PACKAGE__->throw_error, what?).
420
421The internals need to be made consistent before we expose this to the
422rest of the world.
423
8b59f8d6 424-------------------------------------------------------------------------------
425TO PONDER
426-------------------------------------------------------------------------------
427
428- Moose "strict" mode
429
430use Moose 'strict'; This would allow us to have all sort of expensive tests
d03bd989 431which can be turned off in prod.
432
8b59f8d6 433- Moose::Philosophy.pod
434
435To explain Moose from a very high level
436
687e52bb 437- moosedoc
8b59f8d6 438
687e52bb 439We certainly have enough meta-information to make pretty complete POD docs.
d03bd989 440
441
442