1 | Proposal for virtual functionality
|
---|
2 |
|
---|
3 | There are two types of virtual inheritance in this proposal, relaxed
|
---|
4 | (implicit) and strict (explicit). Relaxed is the simpler case that uses the
|
---|
5 | existing trait system with the addition of trait references and vtables.
|
---|
6 | Strict adds some constraints and requires some additional notation but allows
|
---|
7 | for down-casting.
|
---|
8 |
|
---|
9 | Relaxed Virtual Inheritance:
|
---|
10 |
|
---|
11 | Imagine the following code :
|
---|
12 |
|
---|
13 | trait drawable(otype T) {
|
---|
14 | void draw(T* );
|
---|
15 | };
|
---|
16 |
|
---|
17 | struct text {
|
---|
18 | char* text;
|
---|
19 | };
|
---|
20 |
|
---|
21 | void draw(text*);
|
---|
22 |
|
---|
23 | struct line{
|
---|
24 | vec2 start;
|
---|
25 | vec2 end;
|
---|
26 | };
|
---|
27 |
|
---|
28 | void draw(line*);
|
---|
29 |
|
---|
30 | While all the members of this simple UI support drawing, creating a UI that
|
---|
31 | easily supports both these UI requires some tedious boiler-plate code:
|
---|
32 |
|
---|
33 | enum type_t { text, line };
|
---|
34 |
|
---|
35 | struct widget {
|
---|
36 | type_t type;
|
---|
37 | union {
|
---|
38 | text t;
|
---|
39 | line l;
|
---|
40 | };
|
---|
41 | };
|
---|
42 |
|
---|
43 | void draw(widget* w) {
|
---|
44 | switch(w->type) {
|
---|
45 | case text : draw(&w->text); break;
|
---|
46 | case line : draw(&w->line); break;
|
---|
47 | default : handle_error(); break;
|
---|
48 | }
|
---|
49 | }
|
---|
50 |
|
---|
51 | While this code will work as implemented, adding any new widgets or any new
|
---|
52 | widget behaviors requires changing existing code to add the desired
|
---|
53 | functionality. To ease this maintenance effort required CFA introduces the
|
---|
54 | concept of trait references.
|
---|
55 |
|
---|
56 | Using trait references to implement the above gives the following :
|
---|
57 |
|
---|
58 | trait drawable objects[10];
|
---|
59 | fill_objects(objects);
|
---|
60 |
|
---|
61 | while(running) {
|
---|
62 | for(drawable object : objects) {
|
---|
63 | draw(object);
|
---|
64 | }
|
---|
65 | }
|
---|
66 |
|
---|
67 | The keyword trait is optional (by the same rules as the struct keyword). This
|
---|
68 | is not currently supported in CFA and the lookup is not possible to implement
|
---|
69 | statically. Therefore we need to add a new feature to handle having dynamic
|
---|
70 | lookups like this.
|
---|
71 |
|
---|
72 | What we really want to do is express the fact that calling draw() on a trait
|
---|
73 | reference should find the underlying type of the given parameter and find how
|
---|
74 | it implements the routine, as in the example with the enumeration and union.
|
---|
75 |
|
---|
76 | For instance specifying that the drawable trait reference looks up the type
|
---|
77 | of the first argument to find the implementation would be :
|
---|
78 |
|
---|
79 | trait drawable(otype T) {
|
---|
80 | void draw(virtual T* );
|
---|
81 | };
|
---|
82 |
|
---|
83 | This could be implied in simple cases like this one (single parameter on the
|
---|
84 | trait and single generic parameter on the function). In more complex cases it
|
---|
85 | would have to be explicitly given, or a strong convention would have to be
|
---|
86 | enforced (e.g. implementation of trait functions is always drawn from the
|
---|
87 | first polymorphic parameter).
|
---|
88 |
|
---|
89 | Instances of a trait are created by wrapping an existing instance of a type
|
---|
90 | that implements that trait. This wrapper includes all the function pointers
|
---|
91 | and other values required to preform the dynamic look-up. These are chosen by
|
---|
92 | the normal look-up rules at the point of abstraction.
|
---|
93 |
|
---|
94 | One of the limitations of this design is that it does not support double
|
---|
95 | dispatching, which concretely means traits cannot have routines with more than
|
---|
96 | one virtual parameter. The program must have a single table to look up the
|
---|
97 | function on. Using trait references with traits with more than one parameter
|
---|
98 | is also restricted, initially forbidden, see extension.
|
---|
99 |
|
---|
100 | Ownership of the underlying structure is also a bit of a trick. Considering
|
---|
101 | the use cases for trait object, it is probably best to have the underlying
|
---|
102 | object be heap allocated and owned by the trait object.
|
---|
103 |
|
---|
104 | Extension: Multi-parameter Virtual Traits:
|
---|
105 |
|
---|
106 | This implementation can be extended to traits with multiple parameters if
|
---|
107 | one is called out as being the virtual trait. For example :
|
---|
108 |
|
---|
109 | trait iterator(otype T, dtype Item) {
|
---|
110 | Maybe(Item) next(virtual T *);
|
---|
111 | }
|
---|
112 |
|
---|
113 | iterator(int) generators[10];
|
---|
114 |
|
---|
115 | Which creates a collection of iterators that produce integers, regardless of
|
---|
116 | how those iterators are implemented. This may require a note that this trait
|
---|
117 | is virtual on T and not Item, but noting it on the functions may be enough.
|
---|
118 |
|
---|
119 |
|
---|
120 | Strict Virtual Inheritance:
|
---|
121 |
|
---|
122 | One powerful feature relaxed virtual does not support is the idea of down
|
---|
123 | casting. Once something has been converted into a trait reference there is
|
---|
124 | very little we can do to recover and of the type information, only the trait's
|
---|
125 | required function implementations are kept.
|
---|
126 |
|
---|
127 | To allow down casting strict virtual requires that all traits and structures
|
---|
128 | involved be organized into a tree. Each trait or struct must have a unique
|
---|
129 | position on this tree (no multiple inheritance).
|
---|
130 |
|
---|
131 | This is declared as follows :
|
---|
132 |
|
---|
133 | trait error(otype T) virtual {
|
---|
134 | const char * msg(T *);
|
---|
135 | }
|
---|
136 |
|
---|
137 | trait io_error(otype T) virtual error {
|
---|
138 | FILE * src(T *);
|
---|
139 | }
|
---|
140 |
|
---|
141 | struct eof_error virtual io_error {
|
---|
142 | FILE * fd;
|
---|
143 | };
|
---|
144 |
|
---|
145 | So the trait error is the head of a new tree and io_error is a child of it.
|
---|
146 |
|
---|
147 | Also the parent trait is implicitly part of the assertions of the children,
|
---|
148 | so all children implement the same operations as the parent. By the unique
|
---|
149 | path down the tree, we can also uniquely order them so that a prefix of a
|
---|
150 | child's vtable has the same format as its parent's.
|
---|
151 |
|
---|
152 | This gives us an important extra feature, runtime checking of the parent-child
|
---|
153 | relationship with virtual cast, where a pointer (and maybe a reference) to
|
---|
154 | a virtual type can be cast to another virtual cast. However the cast is
|
---|
155 | dynamicly check and only occurs if the underlying type is a child of the type
|
---|
156 | being cast to. Otherwise null is returned.
|
---|
157 |
|
---|
158 | (virtual TYPE)EXPR
|
---|
159 |
|
---|
160 | As an extention, the TYPE may be ommitted if it can be determained from
|
---|
161 | context, for instance if the cast occurs on the right hand side of an
|
---|
162 | assignment.
|
---|
163 |
|
---|
164 | Function look-up follows the same rules as relaxed (behavioural) inheritance.
|
---|
165 | Traits can be upcast and down cast without losing information unless the
|
---|
166 | trait is cast down to a structure. Here there are two options.
|
---|
167 |
|
---|
168 | Abstraction Time Binding: The more efficient and consistant with other parts
|
---|
169 | of CFA. Only the trait types use dynamic look-up, if converveted back into a
|
---|
170 | structure the normal static look-up rules find the function at compile time.
|
---|
171 | Casting down to a structure type can then result in the loss of a set of
|
---|
172 | bindings.
|
---|
173 | Construction Time Binding: For more consistant handling of the virtual
|
---|
174 | structs, they are always considered wrapped. Functions are bound to the
|
---|
175 | instance the moment it is constructed and remain unchanged throughout its
|
---|
176 | lifetime, so down casting does not lose information.
|
---|
177 |
|
---|
178 | (We will have to decide between one of these two.)
|
---|
179 |
|
---|
180 | Extension: Multiple Parents
|
---|
181 |
|
---|
182 | Although each trait/struct must have a unique position on each tree, it could
|
---|
183 | have positions on multiple trees. All this requires is the ability to give
|
---|
184 | multiple parents, as here :
|
---|
185 |
|
---|
186 | trait region(otype T) virtual drawable, collider;
|
---|
187 |
|
---|
188 | The restriction being, the parents must come from different trees. This
|
---|
189 | object (and all of its children) can be cast to either tree. This is handled
|
---|
190 | by generating a separate vtable for each tree the structure is in.
|
---|
191 |
|
---|
192 | Extension: Multi-parameter Strict Virtual
|
---|
193 |
|
---|
194 | If a trait has multiple parameters then one must be called out to be the one
|
---|
195 | we generate separate vtables for, as in :
|
---|
196 |
|
---|
197 | trait example(otype T, otype U) virtual(T) ...
|
---|
198 |
|
---|
199 | This can generate a separate vtable for each U for which all the T+U
|
---|
200 | implementations are provided. These are then separate nodes in the tree (or
|
---|
201 | the root of different trees) as if each was created individually. Providing a
|
---|
202 | single unique instance of these nodes would be the most difficult aspect of
|
---|
203 | this extension, possibly intractable, though with sufficient hoisting and
|
---|
204 | link-once duplication it may be possible.
|
---|
205 |
|
---|
206 | Example:
|
---|
207 |
|
---|
208 | trait argument(otype T) virtual {
|
---|
209 | char short_name(virtual T *);
|
---|
210 | bool is_set(virtual T *);
|
---|
211 | };
|
---|
212 |
|
---|
213 | trait value_argument(otype T, otype U) virtual(T) argument {
|
---|
214 | U get_value(virtual T *);
|
---|
215 | };
|
---|
216 |
|
---|
217 | Extension: Structural Inheritance
|
---|
218 |
|
---|
219 | Currently traits must be the internal nodes and structs the leaf nodes.
|
---|
220 | Structs could be made internal nodes as well, in which case the child structs
|
---|
221 | would likely structurally inherit the fields of their parents.
|
---|
222 |
|
---|
223 |
|
---|
224 | Storing the Virtual Lookup Table (vtable):
|
---|
225 |
|
---|
226 | We have so far been silent on how the vtable is created, stored and accessed.
|
---|
227 | The vtables for the two types might be handled slightly differently and then
|
---|
228 | there is also the hierarchy data for virtual casts.
|
---|
229 |
|
---|
230 | The hierarchy data is simple conceptually. A single (exactly one copy) pointer
|
---|
231 | for each type can act as the identity for it. The value of the pointer is
|
---|
232 | its parent type, with the root pointer being NULL. Additional meta-data
|
---|
233 | can accompany the parent pointer, such as a string name or the vtable fields.
|
---|
234 |
|
---|
235 | They types of each vtable can be constructed from the definitions of the
|
---|
236 | traits (or internal nodes). The stand alone/base vtable is the same for both
|
---|
237 | kinds of inheritance. It may be argumented differently however (include parent
|
---|
238 | /this pointer in hierachal inheritance).
|
---|
239 |
|
---|
240 | Creation of the actual vtable is tricky. For classical single implementation
|
---|
241 | semantics we would assemble the functions and create one vtable at compile
|
---|
242 | time. However, not only does this not give CFA-like behaviour, it is
|
---|
243 | impossible generally because types can satify assertions in different ways at
|
---|
244 | different times and stop satifying them. A special set of harder rules could
|
---|
245 | be used, instead we have decided to try creating multiple vtables for each
|
---|
246 | type. The different vtables will all implement the same type but not always
|
---|
247 | in the same way.
|
---|
248 |
|
---|
249 | Storage has some issues from creation. If the contents of every vtable could
|
---|
250 | be determained at compile time they could all be created and stored
|
---|
251 | statically. However since thunks can be constructed on the stack and become
|
---|
252 | the best match, that isn't always possible. Those will have to be stored in
|
---|
253 | dynamic memory. Which means that all vtables must be stored dynamically or
|
---|
254 | there must be a way to determain which ones to free when the trait object is
|
---|
255 | destroyed.
|
---|
256 |
|
---|
257 | Access has two main options:
|
---|
258 |
|
---|
259 | The first is through the use of fat pointers, or a tuple of pointers. When the
|
---|
260 | object is converted to a trait reference, the pointers to its vtables are
|
---|
261 | stored along side it.
|
---|
262 |
|
---|
263 | This allows for compatibility with existing structures (such as those imported
|
---|
264 | from C) and is the default storage method unless a different one is given.
|
---|
265 |
|
---|
266 | The other is by inlining the vtable pointer as "intrusive vtables". This adds
|
---|
267 | a field to the structure to the vtable. The trait reference then has a single
|
---|
268 | pointer to this field, the vtable includes an offset to find the beginning of
|
---|
269 | the structure again.
|
---|
270 |
|
---|
271 | This is used if you specify a vtable field in the structure. If given in the
|
---|
272 | trait the vtable pointer in the trait reference can then become a single
|
---|
273 | pointer to the vtable field and use that to recover the original object
|
---|
274 | pointer as well as retrieve all operations.
|
---|
275 |
|
---|
276 | trait drawable(otype T) {
|
---|
277 | vtable drawable;
|
---|
278 | };
|
---|
279 |
|
---|
280 | struct line {
|
---|
281 | vtable drawable;
|
---|
282 | vec2 start;
|
---|
283 | vec2 end;
|
---|
284 | };
|
---|
285 |
|
---|
286 | This inline code allows trait references to be converted to plain pointers
|
---|
287 | (although they still must be called specially). The vtable field may just be
|
---|
288 | an opaque block of memory or it may allow user access to the vtable. If so
|
---|
289 | then there should be some way to retrieve the type of the vtable, which will be
|
---|
290 | autogenerated and often unique.
|
---|
291 |
|
---|
292 | It may be worth looking into a way to force the vtable pointer to be in a
|
---|
293 | particular location, which would save the storage to store the offset and
|
---|
294 | maybe the offset operation itself (offset = 0). However it may not be worth
|
---|
295 | introducing a new language feature for.
|
---|
296 | As of writing, exceptions actually use this system.
|
---|
297 |
|
---|
298 |
|
---|
299 | Keyword Usage:
|
---|
300 |
|
---|
301 | It may be desirable to add fewer new keywords than discussed in this proposal.
|
---|
302 | It is possible that "virtual" could replace both "vtable" above with
|
---|
303 | unambiguous contextual meaning. However, for purposes of clarity in the design
|
---|
304 | discussion it is beneficial to keep the keywords for separate concepts distinct.
|
---|
305 |
|
---|
306 |
|
---|
307 | Trait References and Operations:
|
---|
308 |
|
---|
309 | sizeof(drawable) will return the size of the trait object itself. However :
|
---|
310 |
|
---|
311 | line a_line;
|
---|
312 | drawable widget = a_line;
|
---|
313 | sizeof(widget);
|
---|
314 |
|
---|
315 | Will instead return the sizeof the underlying object, although the trait must
|
---|
316 | require that its implementation is sized for there to be a meaningful value
|
---|
317 | to return. You may also get the size of the trait reference with
|
---|
318 |
|
---|
319 | sizeof(&widget);
|
---|
320 |
|
---|
321 | Calling free on a trait reference will free the memory for the object. It will
|
---|
322 | leave the vtables alone, as those are (always?) statically allocated.
|
---|
323 |
|
---|
324 |
|
---|
325 | Special Traits:
|
---|
326 |
|
---|
327 | trait is_virtual_parent(dtype parent, dtype child) { ... };
|
---|
328 |
|
---|
329 | There are others but I believe this one to be the most important. The trait
|
---|
330 | holds if the parent type is a strict virtual ancestor (any number of levels)
|
---|
331 | of child. It will have to exist at least internally to check for upcasts and
|
---|
332 | it can also be used to optimize virtual casts into upcasts. Or a null value or
|
---|
333 | error if the cast would never succeed. Exporting it to a special trait allows
|
---|
334 | users to express that requirement in their own polymorphic code.
|
---|
335 |
|
---|
336 |
|
---|
337 | Implementation:
|
---|
338 |
|
---|
339 | Before we can generate any of the nessasary code, the compiler has to get some
|
---|
340 | additional information about the code that it currently does not collect.
|
---|
341 |
|
---|
342 | First it should establish all child->parent links so that it may travel up the
|
---|
343 | hierarchy to grab the nessasary information, and create the actual parent
|
---|
344 | pointers in the strict virtual tables. It should also maintain the connections
|
---|
345 | between the virtual type (structure or trait), the vtable type and the vtable
|
---|
346 | instance (or default instance for relaxed virtual if multiple are allowed). To
|
---|
347 | this end a sub-node should be created with the nessasary pointers. Traits and
|
---|
348 | structs with virtual can create an instance and store all the nessasary data.
|
---|
349 |
|
---|
350 | With the hierarchy in place it can generate the vtable type for each type,
|
---|
351 | it will generally have a function pointer field for each type assertion in
|
---|
352 | some consistant order. Strict virtual will also have a pointer to the parent's
|
---|
353 | vtable and intrusive vtables will also have the offset to recover the original
|
---|
354 | pointer. Sized types will also carry the size.
|
---|
355 |
|
---|
356 | Wheither the vtable is intrusive or not should also be save so that the trait
|
---|
357 | object/reference/pointer knows if it has to store 1 or 2 pointers. A wrapper
|
---|
358 | function will have to be generated for each type assertion so that they may
|
---|
359 | be called on the trait type, these can probably be inlined.
|
---|
360 |
|
---|
361 | The virtual parameter will also have to be marked (implicately or explicately)
|
---|
362 | until code generation so that the wrapper functions know where to go to get
|
---|
363 | the vtable for the function look up. That could probably be added as a
|
---|
364 | storageclass, although one that is only valid on type assertions.
|
---|
365 |
|
---|
366 | The generated vtable will than have to have a vtable instance created and
|
---|
367 | filled with all the approprate values. Stricter matching may have to be used
|
---|
368 | to ensure that the functions used are stable. It will also have to use
|
---|
369 | ".gnu.linkonce" or equilant to ensure only one copy exists in the final code
|
---|
370 | base.
|
---|