followup 3: rename Pre_Conds.* and I_Model.* (eliminate *_TEST as much as possible)
1 * MW: implement ML toplevel spans in official Isabelle, e.g.
2 with separator \<^medskip> and some tool support for automated
3 update (in "isabelle update" and/or PIDE);
5 * MW?: rename *.sml to *.ML, potentially with tool support;
7 * MW: clarify/eliminate Isabelle/Scala add-ons (presently unused)
9 diff -r /home/makarius/isabelle/repos-Isabelle2021/src/Pure/build-jars ./src/Pure/build-jars
11 > src/Tools/isac/BridgeJEdit/isac.scala
13 diff -r /home/makarius/isabelle/repos-Isabelle2021/src/Pure/Tools/scala_project.scala ./src/Pure/Tools/scala_project.scala
15 > "src/Tools/isac/etc" -> Path.explode("isabelle.isac"),
17 * MW: Skip_Proof.make_thm: theory -> term -> thm ? could now change signature to
18 : Proof.context -> term -> thm
19 * MW: Make $ISABELLE_ISAC_TEST work without sessions (please synchronise with WN)
20 * MW: ? implement a template, inserted in a proof in the line of metis;
21 hints are given in src/../BridgeJEdit/template.sml
22 and in test/../BridgeJEdit/template.sml
24 ***** some items for discussion
26 * Clarify symmetric rule: Thm.apply_attribute Calculation.symmetric thm context (!?);
29 * "fun pr_ord" is not required if used with @{make_string}, @{print}, @{print tracing};
30 takes only static arguments ----------------^^^^^^^^^^^^^^, not value of "hd_ord (f, g)"
31 ? are there better approximations to old output of (*1*) than with (*2*)
32 (*1*)val _ = tracing ("hd_ord (f, g) = " ^ ((pr_ord o hd_ord) (f, g)) );
33 (*2*)val _ = @{print tracing}{a = "hd_ord (f, g) = ", z = hd_ord (f, g)}(**)
36 ***** for the few items below WN asks MW for help
38 * strange ?ML?-error WN230404 (= pointer to respective test): Test_Isac.thy runs without errors except 3 files
39 "Knowledge/eqsystem-1.sml"
40 "Knowledge/eqsystem-2.sml"
41 "Knowledge/eqsystem-3.sml"
42 This error is strange, since these 3 fiels use the same src as others; the error remains although the
43 constructors of both functions, Pre_Conditionss.mk_env and I_Model.feedback_OLD_to_TEST,
44 have been made equal (Model_Def.Cor_TEST, etc).
46 * postpone: ?How accomplish two user-requirements by Outer_Syntax.command \<^command_keyword>\<open>Example\<close>
47 (1) start a Calculation with a CAS_Cmd
48 (2) start an Example from scratch, i.e. with (Formalize.empty, References.empty)
49 Proposals for a solution are in test/../Test_VSCode_Example.thy
50 in subsubsection ‹Start Example with a CAS_Cmd›
53 ***** priority of WN items is top down, most urgent/simple on top
55 * WN: ---broken elementwise input to lists---
56 * WN: undetected ERROR in autoCalculate --- due to Post_Conds.check? (_OLD .. dispels ?)
57 * WN: "review (descriptor, ts)"; ts : term list, because this supports element-wise input of lists.
58 ts = [_] is determined by <Model_Pattern.is_list_descr descriptor>.
59 Thus, as long (descriptor, ts) is together, ts never contains <Const ("List.list.Cons", _)> --
60 this holds for specify-phase internally, but not to the outside (presentation to user, transfer to solve-phase).
61 TODO: shift this as comment to structure Model_Def, type o_model_single + referenced by datatype i_model_feedback.
62 TODO: o_model, i_model etc
63 * WN: type descriptor is defind 5 times !!!
64 * WN: introduction of I_Model.T_TEST is a huge step, which is partitioned into smaller steps, starting with
65 changeset cb44eefd3c7b "repare 1 PIDE turn 11: .." and continued with "repare 1..n".
66 After changeset 94242a19b04c "prepare 4: narrow I_Model.T -- _TEST" it seems no more possible to make the test(s)
67 in test/../Test_Theory.thy run without braking too many tests. Thus
68 minimal code in src/* is changed and marked by (*T_TESTold*)xxx(*T_TEST* )xxx( *T_TESTnew*)
69 while the outcommenting shown is used for commits (where Test_Isac_Shorts runs without errors,
70 but Test_Theory.thy doen't work).
71 * WN: ad CS "remove sep_variants_envs":
72 test/../i-model.sml (*/----- reactivate with CS "remove sep_variants_envs"-----\* )
73 * WN: DEL get_simplifier: is only used in test/*
74 * WN: test/../pre-conditions.sml: there seems something wrong with env_subst-?-env_eval in check_TEST
75 thus we introduced check_TEST', compare --- Minisubplb/200-start-method-NEXT_STEP.sml ---
76 * WN: replace take in LibraryC with take from Library
77 * WN: Syntax.read_term @{context} --> Syntax.read_term\<^context> , see I_Model "UnIqE_tErM"
78 * WN: renaming in References
79 * WN: shift code from libraryC.sml to designated destination (ThyC, etc)
80 * WN: rename type Rule.rule -> Rule.T
81 * WN: improve naming in refine.sml, m-match.sml
82 * WN: review ?code ERROR? in Rewrite.eval_true, see WN230329
84 * WN: consider redesign of Pre_Conds.mk_envs <-?- Pre_Conds.environment together with Model_Pattern.split_descriptor
85 which could eliminate isalist2list and make handling of lists more convenient, see --- build Pre_Conds.check_TEST
86 Pre_Conds.environment is called seriously only once, "check (Tactic.Apply_Method" for preconds
87 !relate to instantiation of programs arguments!
88 * WN: correct design flaw in fun eval_*, see aa8760e4d987
89 * WN: unify get_ctxt in Solve_Step, Ctree (*DEPRECATED*)
90 * WN: review Prog_Tac:
91 (*+isa==isa2*)@{term "Substitution []"}; (*Free ("Subproblem",.. ALSO Subproblem, ?!*)
92 (*+isa==isa2*)@{term "Rewrite_Set ''norm_Rational''"}; (*Const ("Prog_Tac.Rewrite_Set",..*)
94 * WN: Step_Specify.initialise: remove hdl in return-value, replace Step_Specify.initialise' xxx'
95 ? which uses initialise !?
96 * WN: ? unify "no_met" with "empty_meth_id" from References.empty ?
98 * WN: proper ML antiquotations for "Tactical.Try" etc. --- be careful about unclear situations,
99 e.g. "Tactical.Try" vs. "Lucas_Interpreter.Try";
101 * WN: ? Rational.Cancel_p; extend use of \<^theory> to \<^theory_context>
103 * WN: Check/clarify Context.theory_name vs. Context.theory_long_name.
104 Context.theory_name seems to suffice after elimination of sessions.
106 * WN: more direct logical foundations wrt. Isabelle/HOL, eliminate many axiomatizations
107 - quite often "axiomatization ..." can be turned into "lemma ... by auto"
109 - sometimes this requires to use more specific types / type classes;
110 - sometimes this requires to use proper definitional mechanisms (e.g. 'primrec', 'fun');
111 - a few "hard" cases will remain, to be reconsidered eventually (e.g. differentiation);
113 * WN: rename Pos.* -- Pos.ints, Pos.spec, Pos.empty, Pos.ints_empty
115 * WN: redesign transition from Specification to Solution: how relate
116 - Formalise.model with variants (e.g. VSCode_Example)
117 reconsider separation of variants F_I, F_II, see MAWEN paper
118 - !?! I_Model of MethodC (fairly free sequence, dependent on Formalise.model)
119 - !?! formal arguments of program (fixed sequence)
121 * WN: avoid too many uses of "structure Data = Theory_Data", instead use approx. 1 data slot
122 per ML module, with type T = ... record or tuple (e.g. see @{apply} ML antiquotation)
124 * WN: rename *.sml to *.ML, potentially with tool support by MW;
125 * WN: rename src/Tools/isac/ to main/ (or something else);
126 rename src/Tools/isac/BridgeJEdit and test/Tools/isac/BridgeJEdit to */Tools/isac/BridgePIDE