2 library/gallery migration
3 -------------------------
5 - check that examples/tests/cvc4-models.mlw can use arrays
6 instead of ref map (commit dfc90d)
8 - remove the temporary --type-only from bench/bench (commit 02be08f)
10 - DISCUSS: why3 prove theories/int.why fails because of a "variant"
11 clause in one of let functions: this clause requires int.Int
12 for the order relation, which is not loaded automatically
13 (program modules are expected to do "use import int.Int")
14 and, of course, cannot be used explicitly inside int.why.
16 - expression { e with f2 = e2 } is syntactic sugar for
17 { f1 = e.f1; f2 = e2; f3 = e.f3 }
18 We would get better error messages if we would produce instead
19 { f2 = e2; f1 = e.f1; f3 = e.f3 }
20 that is, with user-given fields first, then fields from e
21 See for instance bench/programs/bad-typing/with1.mlw
23 - the error message for
24 val foo (_x : int) : int
25 val f (x:int) : unit writes { foo.contents }
26 is "unbound symbol 'foo'", which is not correct
27 (is should rather be that "foo" has the wrong type)
29 - bench/programs/bad-to-keep/{at1.mlw,old2.mlw}
30 should at least produce a warning (meaningless 'old')
32 - no warning on lemma functions
33 cf examples/tests/lemma_functions.mlw
35 - "old" and "at" in program code gives a syntax error
36 (bench/programs/bad-to-keep/old3.mlw), we can do better
41 - rename the languages to "Why" and "WhyML". "Why3ML" is a horrible name.
46 - let modules register new command-line options. This would deprecate
47 the "stop flags" in Debug, and serve, for instance, for --type-only,
48 --parse-only, ocaml code extraction, printing of modules, etc.
50 - introduce a dedicated buffer-backed formatter for warnings.
51 The contents of the buffer would be shown on stdout or in a window
52 at selected points of program execution. Demote non-critical errors
53 (e.g. unnamed type variables) to warnings.
55 - should we create a common [exception Why.Error of exn] to facilitate
56 integration of the library? This would require a special [raise] call:
57 why_raise e = raise (Why.Error e)
59 - the drivers cannot deal with theories defined in the .mlw files.
60 Otherwise why3 would report an error. Is it acceptable?
65 - introduce logic predicates for program type invariants:
66 predicate <type_name>_invariant (result : <type_name>) = ...
67 Should this be just the top invariant, as written by the user,
68 or should we also include the invariants for the fields?
70 - allow to specify type invariants below a type definition,
71 provided there are no program declarations in between.
72 This allows us to define auxiliary logical functions and
73 predicates that depend on the (pure) type and can be used in
74 the invariant. However, the parser must know from the start
75 that the type has an invariant, what's the best syntax?
77 - type invariants are now assumed/asserted at the function call
78 boundaries. We can add a binary flag to Ityapp to allow open
79 types in function signatures (must be careful with reg_match!).
80 The type cast can then play the role of the "close" instruction.
81 Do we need it? What's the good syntax for open types?
83 - we check the context type invariants around Eany and after Eabstr.
84 It might be strange that Eabstr post-ensures the invariants that
85 didn't necessarily hold before its execution. Also, what about
86 return/raise invariants, should Eany and Eabstr enforce them?
87 (at the moment, they do)
89 - currently every unhandled exception has the postcondition "true".
90 "false" would be a poor choice, as it could introduce inconsistency
91 in the WPs of the caller. Should we allow unhandled exceptions at all?
93 - current WP does not respect the lexical scope. In the program
104 the last assert will be proven if the same let_defn [let v = !r]
105 and therefore the same pvsymbol v is used in both places (which
106 can be done using API). One possible solution is to ensure the
107 one-time usage of let_defns and rec_defns in programs.
109 - are mutable values worth it? They can only appear as pattern
110 variables standing for mutable fields, and allow us to have
111 mutable fields in algebraic types. On the other hand, they
112 require tedious checks pretty much everywhere in the code,
113 and they cannot be translated to OCaml.
115 - accept match e1 with pat -> e2 end as non-ghost when e1 is ghost
123 - infix symbols as constructors, e.g.
125 type list 'a = Nil | (::) 'a (list 'a)
127 - constants in patterns, e.g.
129 match ... with 0 :: r -> ... | ...
134 - should split_goal provide a "right-hand side only split"?
136 - produce reparsable tasks in Why3 format: how to preserve information about
137 the origins of symbols to be able to use drivers after reparsing?
139 - open (et échouer si "open A" et "open B" avec A et B déclarant un symbole
145 - save the output of the prover
147 - escape the string in the xml
149 - the filenames in the location inside a session should be relative
152 - use the new restore_path for the metas in session?
157 - the tools should verify that the provers have the same version
158 than reported in the configuration
159 Andrei: isn't this done?
161 - Maybe : make something generic for the dialog box with memory.
167 - allow other realizations for arithmetic, such as Zarith or GMP
168 (currently this is Num)
170 - avoid conversion to/from int in the for-loop
173 - %Exit -> Pervasives.Exit
178 - PVS: use a better name for PVS theory when printing a task, e.g.
179 file_theory_goal. Solution: do that when we have idents with origin
180 information (necessary for parsing a task).