QA: run_qa v1.6 form + ExplicitImports#88
Merged
ChrisRackauckas merged 4 commits intoJul 3, 2026
Merged
Conversation
Convert the hand-rolled JET-only QA group (test/qa/jet_tests.jl) to the SciMLTesting 1.6 run_qa form with ExplicitImports enabled. test/qa/qa.jl: - run_qa(PDEBase; explicit_imports = true) — Aqua + JET + ExplicitImports. - Aqua: treat_as_own = [PDEBase.symbolic_discretize] (PDEBase extends SciMLBase.symbolic_discretize on a PDESystem as its discretization entry point); deps_compat check_extras = false (folder-model test deps live in test/qa/Project.toml, root [extras] are vestigial). - ExplicitImports ignore-lists for re-exporter/non-public names owned by ModelingToolkitBase / TermInterface / SymbolicUtils / Symbolics / SciMLBase. - ei_broken = (:no_implicit_imports,): PDEBase's surface rests on heavy `using ModelingToolkit / Symbolics / SymbolicUtils / SciMLBase / DomainSets`; making each implicitly used name explicit is a large refactor tracked separately. src/PDEBase.jl: - Remove 8 stale (unused) explicit imports flagged by check_no_stale_explicit_imports: parameters, varmap_to_vars, solve_for, expand_derivatives, diff2term, variable, Postwalk, symtype. - Remove the phantom export `pde_substitute_filter` (exported but never defined — Aqua undefined_exports failure; only pde_substitute exists). test/qa/Project.toml: - Add Aqua as a direct dep (Aqua.test_ambiguities spawns a child process that `import Aqua`s, so it must be a direct dep, not transitive via SciMLTesting). - Drop Symbolics / SymbolicUtils (were only used by the old jet_tests.jl). - SciMLTesting compat -> "1.6". Verified locally on Julia LTS against released SciMLTesting 1.6.0: QA group 16 Pass / 1 Broken (no_implicit_imports) / 0 Fail / 0 Error; Core group alloc_tests 12/12 pass. Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com> Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…aries The QA conversion to `run_qa` runs full `JET.report_package(mode=:typo)` instead of the old 5 hand-picked `report_call`s, which never traversed `parse_boundaries.jl`. On Julia 1.12 this surfaced 9 JET reports: - `isperiodic`: `issequal` is not defined — a genuine typo for `isequal` (present since the package's first commit). Calling `isperiodic` would throw `UndefVarError`. Fixed to `isequal`. - `parse_bcs`: `local u_, u__, x_, x__` left the bindings unassigned on paths JET cannot prove are unreachable, yielding 8 "may be undefined" reports. Initializing them to `nothing` makes them definitely-assigned without changing behavior (they are only read inside `if isinterface`, which is set true only in the same branches that assign real values). After the fix, `JET.report_package(PDEBase; mode=:typo)` reports 0 errors and the QA group passes on Julia 1.12.6 (16 Pass | 1 Broken | 0 Fail). Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com> Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…lic) SciMLTesting 1.7 runs the public-API ExplicitImports checks (all_qualified_accesses_are_public / all_explicit_imports_are_public) only on Julia >= 1.11, and the base-lib make-public releases (SciMLBase 3.24, CommonSolve 0.2.9, Symbolics 7.28, SymbolicUtils 4.36, ModelingToolkit 11.29, ArrayInterface 7.26) made many previously-internal names public. Empty both public-API ignore-lists, then restore only the names still flagged as non-public on Julia 1.12 against the registered releases: - all_qualified_accesses_are_public: keep ExtraVariablesSystemException (ModelingToolkit); drop Arr, NullParameters, unwrap, value (now public). - all_explicit_imports_are_public: keep AbstractDiscretization, AbstractDiscretizationMetadata (SciMLBase); Chain, Prewalk, maketerm, metadata (SymbolicUtils); ProblemTypeCtx (ModelingToolkit); rename, setname (Symbolics). Drop the 12 get_* names and unwrap (now public). Verified: Julia 1.12 QA group green (16 pass, 1 broken=no_implicit_imports, 0 fail/error) with the minimal lists; Julia 1.10 QA group green (14 pass, 1 broken, 0 fail/error) with public checks skipped. The *_via_owners lists and ei_broken(no_implicit_imports) are unchanged. Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com> Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
…non-public re-exports) Co-Authored-By: Chris Rackauckas <accounts@chrisrackauckas.com>
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Please ignore until reviewed by @ChrisRackauckas.
Converts the hand-rolled JET-only QA group (
test/qa/jet_tests.jl) to the SciMLTesting 1.6run_qaform with ExplicitImports enabled (folder-discoveryrun_testsmodel,test/qa/qa.jl).test/qa/qa.jlrun_qa(PDEBase; explicit_imports = true)— runs Aqua + JET + ExplicitImports.piracies = (; treat_as_own = [PDEBase.symbolic_discretize])— PDEBase extendsSciMLBase.symbolic_discretize(::PDESystem, ::AbstractDiscretization)as its discretization entry point; the function is conceptually owned by the discretization stack PDEBase implements.deps_compat = (; check_extras = false)— folder-model test deps live intest/qa/Project.toml; the root[extras](SafeTestsets,Test) are vestigial.test_packagemode (target_modules = (PDEBase,),mode = :typo) — clean, 0 reports (strictly stronger than the old 5 hand-pickedreport_calls, all of which were also clean).ExplicitImports findings (6 checks)
src/PDEBase.jl:parameters,varmap_to_vars,solve_for,expand_derivatives,diff2term,variable,Postwalk,symtype. (Verified each has no non-import use insrc/.)ei_broken = (:no_implicit_imports,): PDEBase's public surface rests on heavyusing ModelingToolkit / Symbolics / SymbolicUtils / SciMLBase / DomainSets; making each of the ~20 implicitly used names explicit is a large, risky refactor deferred and tracked separately (auto-flags once resolved).ProblemTypeCtx,get_*,varmap_to_vars,ExtraVariablesSystemException(owner ModelingToolkitBase, via ModelingToolkit),maketerm/metadata(owner TermInterface, via SymbolicUtils),unwrap(owner SymbolicUtils, via Symbolics).AbstractDiscretization/AbstractDiscretizationMetadata/NullParameters; SymbolicUtilsChain/Prewalk/maketerm/metadata; ModelingToolkitget_*/ProblemTypeCtx/ExtraVariablesSystemException; Symbolicsunwrap/value/Arr/rename/setname). Each ignore documents its source package.Other src fix
pde_substitute_filter(exported but never defined anywhere — Aquaundefined_exportsfailure; onlypde_substituteexists).test/qa/Project.tomlAquaas a direct dep (Aqua.test_ambiguitiesspawns a child process thatimport Aquas, so Aqua must be a direct dep, not transitive via SciMLTesting).Symbolics/SymbolicUtils(only used by the oldjet_tests.jl).SciMLTestingcompat →"1.6".Verification (Julia LTS, released SciMLTesting 1.6.0, no dev-from-branch)
QA/qa.jl | 16 Pass | 1 Broken (no_implicit_imports) | 0 Fail | 0 Error.Core/alloc_tests.jl | 12 Pass | 0 Fail; "PDEBase tests passed".🤖 Generated with Claude Code