Commit Graph

2332 Commits

Author SHA1 Message Date
Fedor Isakov 5b1449fa82 Update all usages of moved concepts. 2020-04-13 17:58:34 +02:00
Fedor Isakov a0bc96b3e0 Move all concepts related to term to j.m.coderules. 2020-04-13 17:58:34 +02:00
Fedor Isakov 204d1ff7e4 Move migration to termlist type to j.m.coderules. 2020-04-13 17:58:33 +02:00
Fedor Isakov 2ccee6e805 Move all term-related generator templates to j.m.coderules. 2020-04-13 17:58:33 +02:00
Fedor Isakov 0ce5e4813c Drop all deprecated concepts from j.m.logic lang. 2020-04-13 17:58:33 +02:00
Fedor Isakov f55a377528 Truncate migrations that are using deprecated stuff.
All deprecated concepts from j.m.logic are to be deleted.
Concepts related to terms are to be moved to j.m.closures.
2020-04-13 17:58:33 +02:00
Fedor Isakov 296b3e9271 Re-apply all rerunnable migrations. 2020-04-13 17:58:33 +02:00
Fedor Isakov a1c98b5bf0 Apply all outstanding migrations. 2020-04-13 17:58:33 +02:00
Fedor Isakov 52bca36d32 Rebuild project. 2020-04-13 17:58:33 +02:00
Fedor Isakov 4845e120e9 Introduce termlist type, rename multilogical type to logical[]. 2020-04-13 17:58:33 +02:00
Fedor Isakov e8a19249e6 Migration to replace spliced LogicalItemExpression on a list node.
Usually one uses Splice with LogicalItemExpression wrapping a logical variable.
Using LogicalDataForm wrapping same variable makes Splice unnecessary.
In order to keep code readable these are wrapped again in a cons list.
In general, one should avoid using bare LogicalDataForm, as these are
indistiguishable from normal LogicalVariable.
2020-04-13 17:58:33 +02:00
Fedor Isakov 065aac819c Introduce construct to directly insert term constructor in data form.
Usually one uses Splice with LogicalItemExpression wrapping a constructor.
The new TermConstructorDataForm provides a shortcut.
2020-04-13 17:58:33 +02:00
Fedor Isakov 2af087428d Document j.m.logic language. Rearrange virtual packages. 2020-04-13 17:58:33 +02:00
Fedor Isakov 922fe7ef2e Prevent bad migrations from being re-applied.
Replacing of == with = is not always justified.
ListNodeCons is now deprecated and already replaced with
new concept.
2020-04-13 17:58:33 +02:00
Fedor Isakov 9cacd25e53 Drop unused class from reactor API. 2020-04-13 17:58:33 +02:00
Fedor Isakov 1109471f9a Replace usages of MultiMetaLogical with MetaLogicalArray.
The new class MetaLogicalArray does not inherit from MetaLogical,
which should help catch subtle bugs related to errors in
generator code.
2020-04-13 17:58:33 +02:00
Fedor Isakov 73beb56834 Introduce ConsListNode to replace a deprecated concept. Migration.
Prefer aggregation over inheritance: ListNodeCons extending ListNode
is not a good design.
Also drop LogicalUnrapContext, as it has no real use.
2020-04-13 17:58:33 +02:00
Fedor Isakov cf77688a71 Replace instances of deprecated ListRole with ChildRole. 2020-04-13 17:58:33 +02:00
Fedor Isakov def1d9fbc6 Optimize generator, drop usages of ListRole. 2020-04-13 17:58:33 +02:00
Fedor Isakov b3d71ad13d Drop unused method from ListNode, simplify dataform-logical bridging. 2020-04-13 17:58:33 +02:00
Fedor Isakov 7413e9e387 Force save all models. 2020-04-13 17:58:33 +02:00
Fedor Isakov 021853b996 Documenting LogicalUtil, minor refactorings.
MetaLogical are accessed through MetaLogicalFactory in generated code.
2020-04-13 17:58:33 +02:00
Fedor Isakov cac5573cb9 Drop instances of deprecated concepts in baseLangExt
.
2020-04-13 17:58:33 +02:00
Fedor Isakov d22abb3835 Drop instances of deprecated concepts in lambdacalc. 2020-04-13 17:58:33 +02:00
Fedor Isakov 749f60e665 Deprecate a bunch of obsolete concepts. 2020-04-13 17:58:33 +02:00
Fedor Isakov 4a60bc36f6 Migrate test lang to new macro system. 2020-04-13 17:58:33 +02:00
Fedor Isakov ace1303a96 Fix analysis to support new pseudo constraints. 2020-04-13 17:58:32 +02:00
Grigorii Kirgizov 145aa79753 Regenerate build.xml for mpscore 2020-04-13 15:10:37 +03:00
Grigorii Kirgizov bd87873160 Move baseLanguage.tuplesExt lang to 'samples' and upd build script 2020-04-13 14:36:37 +03:00
Grigorii Kirgizov d00b55e679 Provide more robust way to discover a single aspect clique for nodes in TypecheckingActions
Several aspect licques can be found, some irrelevant
(e.g. for typechecking.annotation lang), so a relevant
clique must be chosen.
2020-04-13 12:41:12 +03:00
Grigorii Kirgizov 6348f89d34 Achieve regeneration of rules from deep changes to "imported" data
requiredHash of a rule now includes hashes not only of SNode & SNodeRef
but also of anything except Coderules rt types (e.g. MetaLogical & DataForm).
The motivation is that if something else is used -- then it must come
from AST, so the rule must be regenerated on any changes to that data.
requiredHash is renamed to usedDataHash to reflect that fact.

In particular, fixes dependence of IndexedTupleAccess rule on idx
from `idx = tuple.index.getCompileTimeConstant()`.
2020-04-12 16:11:44 +03:00
Grigorii Kirgizov 8dfcff1844 Add typechecking for indexed tuples (new language .tuplesExt and tests) 2020-04-08 23:24:55 +03:00
Grigorii Kirgizov 27f26047d3 Fix aspect lookup for typechecking actions and coerceTo query
Not all used languages were collected in case of buliding
aspect clique by nodes -- use model of nodes instead.
2020-04-08 19:55:25 +03:00
Fedor Isakov 4e12f7bf3b Migrate baseLanguageExt sample to new expand/call macros. 2020-03-31 18:44:07 +02:00
Fedor Isakov 6a431fcb49 Rebuild project. 2020-03-31 18:44:07 +02:00
Fedor Isakov 9a8b3585d7 Fix runtime and generator to support queries and call macros.
Expand macro supports dispatching by concept.
Logical variables declared on macro's body are init'ed in apply method.
Enable the hack to ensure dependencies on macro input are tracked.
2020-03-31 18:43:22 +02:00
Fedor Isakov ffcc74459f Migrate fitch samples to new expand/call macros. 2020-03-31 11:43:30 +02:00
Fedor Isakov 538632a8a3 Migrate lambdacalc sample to new expand/call macros. 2020-03-31 11:43:30 +02:00
Fedor Isakov f261002feb Introduce extends clause in macro table. Drop name from expand macro.
Macro table can extend another macro table. All prototype template
declarations are imported automatically.
The editor now substitutes default values when entering new
expand macro/constraint filling the slots provided by the prototype.
Expand macro does not require name.
Minor editor improvements.
2020-03-31 11:27:09 +02:00
Fedor Isakov a2b369b9c3 Specify logicals on every expand template. Minor editor improvements.
Expand macro template must have declaration of logicals matching
exactly the logical declarations in the prototype.
Can't refer to logicals declared directly in the prototype.
2020-03-31 11:27:09 +02:00
Fedor Isakov 28d21f5349 Cleanup in macro runtime, expand macro parameter access.
Parameters can be specified for call macro only.
Logicals can be specified for expand and call macro.
Expand macro params are treated exactly as "context"
parameters.
2020-03-31 11:27:09 +02:00
Fedor Isakov 4260346cb9 Introduce expand/call macro pseudo constraints.
The old expand/call constructs are to be replaced by these.
Expand macro constraint refers to a named prototype, which
is specified in a macro table for matching concept.
Call macro constraint refers directly to a macro template.
Contex parameters are to be retired and explicit
parameters to expand macro constraint are to be used instead.
2020-03-31 11:27:09 +02:00
Fedor Isakov 6672ab21e8 Runtime support for new expand and call macro constraints. 2020-03-31 11:27:09 +02:00
Fedor Isakov a72327630e Implement runtime support for call and expand macro templates. 2020-03-31 11:27:09 +02:00
Fedor Isakov ea670422f1 Introduce ExpandMacro template and prototype.
ExpandMacroTemplate to be used for dispatching macro invocations
based on the concept of the specified node.
Prototype is an abstraction to be implemented by
specific templates.
2020-03-31 11:27:09 +02:00
Fedor Isakov d5ea8e8006 Introduce CallMacroTemplate, necessary refactorings. MPSCR-25
Call macros are to be parameterized with logicals and other parameters.
To replace combined expand/call macros causing confusion.
Context parameters to be passed explicitly.
2020-03-31 11:27:09 +02:00
Grigorii Kirgizov 79952036cd Fix ternary operator typechecking 2020-03-26 18:17:06 +03:00
Grigorii Kirgizov 13e1e10f59 Disable assertion failing for lambdacalc, related to proper parent match tracking
Problem lies in that journal doesn't track all justifications
for reactivated occurrences. For a parent of such occurrence
its original activating match is considered, while
justifications from the Reactivating match aren't tracked.
So, this information about immediate reactivating parent is lost.

lambdacalc lang typesystem heavily uses reactivation, so it fails.
2020-03-26 13:40:42 +03:00
Grigorii Kirgizov 654031c05c Extend docs for Justified.justifiedBy relation, add assertion for its antisymmetric property 2020-03-26 13:40:42 +03:00
Grigorii Kirgizov a2675351b5 Track justifications for all occurrences, but weaken information they carry.
Now Evidence isn't unique for each Chunk: Occurrence Chunks fully
share evidence and justifications with their activating match.
So, less collections of justifications are created, approx. 2 times less.
2020-03-26 13:40:41 +03:00