1 Name usage types: as parameters, as base classes, as callables. This potentially restricts
2 attribute usage effects because names mentioned as base classes are not propagated and
3 made freely available for use in attribute accesses.
4
5 Low-Level Instructions and Macro Instructions
6 =============================================
7
8 Have contexts and values stored separately in memory. This involves eliminating DataValue
9 and storing attributes using two words.
10
11 Migrate macro instructions such as the *Index instructions to library code implemented
12 using low-level instructions.
13
14 Consider introducing classic machine level instructions (word addition, subtraction, and
15 so on) in order to implement all current RSVP instructions.
16
17 Move common code sequences to a library routine, such as the context checking that occurs
18 in functions and methods.
19
20 Dataflow Optimisations
21 ======================
22
23 Assignments, particularly now that no result register exists, may cause StoreTemp/LoadTemp
24 instruction pairs to be produced and these could be eliminated.
25
26 Class and Module Attribute Assignment
27 =====================================
28
29 Verify that the context information is correctly set, particularly for the unoptimised
30 cases.
31
32 Update docs/assignment.txt.
33
34 Prevent assignments within classes, such as method aliasing, from causing the source of an
35 assignment from being automatically generated. Instead, only external references should be
36 registered.
37
38 Prevent "from <module> import ..." statements from registering references to such local
39 aliases such that they cause the source of each alias to be automatically generated.
40
41 Consider attribute assignment observations, along with the possibility of class and module
42 attribute assignment.
43
44 (Note direct assignments as usual, indirect assignments via the attribute usage
45 mechanism. During attribute collection and inference, add assigned values to all
46 inferred targets.)
47
48 (Since class attributes can be assigned, StoreAttrIndex would no longer need to reject
49 static attributes, although this might still be necessary where attribute usage analysis
50 has not been performed.)
51
52 Potentially consider changing static attribute details to use object-relative offsets in
53 order to simplify the instruction implementations. This might allow us to eliminate the
54 static attribute flag for attributes in the object table, at least at run-time.
55
56 Dynamic Attribute Access
57 ========================
58
59 Consider explicit accessor initialisation:
60
61 attr = accessor("attr")
62 getattr(C, attr)
63
64 Attribute Usage
65 ===============
66
67 To consider: is it useful to distinguish between attribute name sets when the same names
68 are mentioned, but where one path through the code sets different values on attributes
69 than another? The _attrtypes collapses observations in order to make a list of object
70 types for a name, and the final set of names leading to such type deductions might be a
71 useful annotation to be added alongside _attrcombined.
72
73 Interface/Type Generalisation
74 -----------------------------
75
76 Consolidate interface observations by taking all cached table accesses and determining
77 which usage patterns lead to the same types. For example, if full usage of {a, b} and
78 {a, b, c} leads to A and B in both cases, either {a, b} can be considered as partial usage
79 of the complete interface {a, b, c}, or the latter can be considered as an
80 overspecification of the former.
81
82 Consider type deduction and its consequences where types belong to the same hierarchy
83 and where a guard could be generated for the most general type.
84
85 Consider permitting multiple class alternatives where the attributes are all identical.
86
87 Support class attribute positioning similar to instance attribute positioning, potentially
88 (for both) based on usage observations. For example, if __iter__ is used on two classes,
89 the class attribute could be exposed at a similar relative position to the class (and
90 potentially accessible using a LoadAttr-style instruction).
91
92 **** Constant attribute users need not maintain usage since they are already resolved. ****
93
94 Self-related Usage
95 ------------------
96
97 Usage of self to restrict attribute usage observations and coverage.
98
99 Perform attribute usage on attributes of self as names, potentially combining observations
100 across methods.
101
102 Additional Guards
103 -----------------
104
105 Consider handling branches of values within namespaces in order to support more precise value usage.
106
107 Loop entry points and other places where usage becomes more specific might be used as
108 places to impose guards. See tests/attribute_access_type_restriction_loop_list.py for an
109 example. (Such information is already shown in the reports.)
110
111 Strict Interfaces/Types
112 -----------------------
113
114 Make the gathering of usage parameterisable according to the optimisation level so that a
115 choice can be made between control-flow-dependent observations and the simple collection
116 of all attributes used with a name (producing a more static interface observation).
117
118 AttributeError
119 --------------
120
121 Consider attribute usage observations being suspended or optional inside blocks where
122 AttributeError may be caught (although this doesn't anticipate such exceptions being
123 caught outside a function altogether). For example:
124
125 y = a.y
126 try:
127 z = a.z # z is an optional attribute
128 except AttributeError:
129 z = None
130
131 Instantiation Deduction
132 -----------------------
133
134 Consider handling CallFunc in micropython.inspect in order to produce instances of specific classes.
135 Then, consider adding support for guard removal/verification where known instances are involved. For
136 example:
137
138 l = []
139 l.append(123) # type deductions are filtered using instantiation knowledge
140
141 Frame Optimisations
142 ===================
143
144 Stack frame replacement where a local frame is unused after a call, such as in a tail call
145 situation.
146
147 Local assignment detection plus frame re-use. Example: slice.__init__ calls
148 xrange.__init__ with the same arguments which are unchanged in xrange.__init__. There is
149 therefore no need to build a new frame for this call, although in some cases the locals
150 frame might need expanding.
151
152 Reference tracking where objects associated with names are assigned to attributes of other
153 objects may assist in allocation optimisations. Recording whether an object referenced by
154 a name is assigned to an attribute, propagated to another name and assigned to an
155 attribute, or passed to another function or method might, if such observations were
156 combined, allow frame-based or temporary allocation to occur.
157
158 Instantiation
159 =============
160
161 Specific instances could be produced, providing type information and acting somewhat like
162 classes during inspection.
163
164 Inlining
165 ========
166
167 Where a function or method call can always be determined, the body of the target could be
168 inlined - copied into place - within the caller. If the target is only ever called by a
169 single caller it could be moved into place. This could enhance deductions based on
170 attribute usage since observations from the inlined function would be merged into the
171 caller.
172
173 Function Specialisation
174 =======================
175
176 Specialisation of certain functions, such as isinstance(x, cls) where cls is a known
177 constant.
178
179 Structure and Object Table Optimisations
180 ========================================
181
182 Fix object table entries for attributes not provided by any known object, or provide an
183 error, potentially overridden by options. For example, the augmented assignment methods
184 are not supported by the built-in objects and thus the operator module functions cause
185 the compilation to fail. Alternatively, just supply the methods since something has to do
186 so in the builtins.
187
188 Consider attribute merging where many attributes are just aliases for the same underlying
189 definition.
190
191 Consider references to defaults as occurring only within the context of a particular
192 function, thus eliminating default value classes if such functions are not themselves
193 invoked.
194
195 Scope Handling
196 ==============
197
198 Consider merging the InspectedModule.store tests with the scope conflict handling.
199
200 Consider labelling _scope on assignments and dealing with the assignment of removed
201 attributes, possibly removing the entire assignment, and distinguishing between such cases
202 and unknown names.
203
204 Check name origin where multiple branches could yield multiple scope interpretations:
205
206 try:
207 set # built-in name
208 except NameError:
209 from sets import Set as set # local definition of name
210
211 set # could be confused by the local definition at run-time
212
213 Object Coverage
214 ===============
215
216 Support __init__ traversal (and other implicit names) more effectively.
217
218 Importing Modules
219 =================
220
221 Consider supporting relative imports, even though this is arguably a misfeature.
222
223 Other
224 =====
225
226 Consider a separate annotation phase where deductions are added to the AST for the
227 benefit of both the reporting and code generation phases.
228
229 Support self attribute visualisation in the reports and/or provide a function or
230 annotations which can provide the eventual optimisation directly to such components.
231
232 Check context_value initialisation (avoiding or handling None effectively).
233
234 Consider better "macro" support where new expressions need to be generated and processed.
235
236 Detect TestIdentity results involving constants, potentially optimising status-affected
237 instructions:
238
239 TestIdentity(x, y) # where x is always y
240 JumpIfFalse(...) # would be removed (never false)
241 JumpIfTrue(...) # changed to Jump(...)
242
243 Status-affected blocks could be optimised away for such constant-related results.