Repository navigation
Preserve compound value types across user function calls - #406
Conversation
Resolve dotted builtin signatures consistently and preserve explicit strings, chart-period fallback and single evaluation. Reserve generated conversion names.
The existing symbol-clock anchor tests now check both calls in the emitted empty-argument conditional. Add pure-Python checks of branch direction and single evaluation, and make the temporary-name sweep reach the argument lambda. Product code is unchanged by this test update.
Lower missing reference values to their destination storage and retain drawing handle types through user functions and tuple returns. Support declared polyline types, fields and deletion arguments.
Read optional function metadata safely when inferring reference casts, so bare emitters retain matrix element types without raising an AttributeError.
Use the established integer storage and matrix alias in type-flow assertions. Check array specializations, mutations, matrix values and copy independence.
Share operand and value inference while an expression query is active, then release cached results before bindings can change. Preserve ordinary matrix member initialization when the declaration expression is still available.
Use the chart period for a generic missing timeframe argument and refuse numeric values for string parameters during analysis. Cover named arguments, helper calls, source locations and generated translation units.
Include only the stock-generated PF-E2095 definition and pin. Preserve all prior catalog entry bytes and their existing order.
Merge the current main while retaining the feature correction.
Merge the current main while retaining the feature correction.
Merge the current main while retaining the feature correction.
Use recognized Pine types in declared type slots to resolve equally literal diagnostic templates. Keep stable code order for other ties and preserve qualifier identities, raw arguments and rendering. Pin the timeframe type diagnostic and its named arguments with direct regression controls.
Reject non-string timeframe arguments through the intrinsic signature before C++ lowering, including helper parameters and boolean values. Preserve valid string calls and exercise complete native controls at distinct optimization levels. Declare the new diagnostic in the existing catalog delta fixture.
Keep generated state and callable clone identities distinct from authored members and use the same allocated identities at declarations and calls. Preserve noncolliding names and duplicate-state assertions.
Carry the generated argument-type entry and its pin into the member while preserving all other catalog bytes and order. Use the grammatical message and exact approved user presentation with the existing argument kinds. Retain the same numeric refusal cases and qualifier controls.
Use the argument occurrence within the call instead of an identically named helper parameter. Keep the exact line and column assertions and all source inputs, diagnostic identities and test rows unchanged.
Carry registered namespace types through the semantic query used for written helper calls while respecting lexical shadowing and version boundaries. Reuse the existing expression-start locator for timeframe type refusals.
Allocate composed callable and persistent-storage identities outside reserved helper families, and reserve all authored spellings before sibling-block qualification. Keep declarations, calls and history consumers on the same allocated identity. Exercise collision cases through the public pipeline.
Keep raw typed-method names on the function channel when they collide with authored data members. Continue allocating encoded names and collisions with generated storage so method definitions, calls and state stay distinct.
Sizing constants keep their Pine string type for the string consumers this change fixes, so a reassigned sizing variable reaches the risk host boundary as a string and is converted to its integer code there. Keeping the integer comparison would need integer storage for these variables, which the string consumers cannot use. The test still requires the call to read t after its write, never its initializer, and now also checks the code chosen for the assigned value.
Keep the established admission policy for non-nullable script selections while retaining exact compound value joins inside functions and methods. This avoids introducing collection history for previously unresolved global selections.
Resolve the analyzer type-query conflict by retaining both lexical namespace typing and callable compound return profiles. Preserve the corrected script selection history admission and both changelog entries. No test expectation or additional product behavior is changed in this merge resolution.
There was a problem hiding this comment.
Graphify reviewed this change.
Worth a look — the grounded gate found no coupling regressions or blocking issues, but 1 advisory finding(s) below merit a look before merge.
Not checked: tests were not run; formal verification proved 0 of 54 changed function(s) (2 sampled, not proven, 52 not verified).
Formal verification. PR-changed functions: 2/54 verified (0 proven, 2 may-equivalent, 0 distinguished) · 52 not verified (23 unsupported, 29 over cap).
Not verified on this run: \_\_init\_\_ (unsupported), \_analyze (unsupported), \_operator\_operand\_type (unsupported), \_propagate\_call\_site\_counts (unsupported), \_untyped\_param\_family\_conflicts (unsupported), \_visit\_FuncCall (unsupported), \_visit\_FuncDef (unsupported), \_visit\_MemberAccess (unsupported), \_visit\_TupleAssign (unsupported), \_visit\_VarDecl (unsupported), and 13 more; 29 changed function(s) beyond the cap and 0 skipped when the time budget ran out.
Graphify review — findings
Adds exact value-profile tracking for user-function call sites, so callee variants get the array, object and string parameter and return types the emitter actually uses, and conflicting structured argument types (e.g. vector<int>& vs vector<double>) split into separate variants. Tightens timeframe.in_seconds and timeframe.change so an evaluated empty string resolves to the chart period (argument evaluated once), while non-string arguments are refused before emission with PF-E2095. Keeps strategy sizing constants as strings until the host setting boundary, escapes int-to-float temporary names, and gives polyline and UDT na fields opaque typed handles, with the new string specialization at risk of contradicting an already-established return spec.
Worth a look
- Exact string specialization can contradict the established return spec —
pineforge_codegen/analyzer/call_handlers.py:2212· Escalate · medium- unverified hypothesis — survived the adversarial refute pass; needs human review
Review partial — this diff was larger than one review pass covers, so later files were not reviewed; some findings may be missing.
Analysis details — impact, health, verification
Impact & health
Graphify review
Impact — 6688 functions depend on the 2252 functions this change touches.
Health — this change adds coupling hotspots:
- worse:
Analyzer— 113 callers, 15 callees - worse:
CodeGen— 69 callers, 14 callees - worse:
_Fold— 4 callers, 8 callees - new:
get_variable_type_spec()— 4 callers, 3 callees - new:
argument_runs()— 0 callers, 8 callees - new:
test_numeric_timeframe_is_refused_during_analysis()— 0 callers, 7 callees - new:
test_open_numeric_domain_is_not_reported_as_a_guessed_float()— 0 callers, 7 callees - new:
test_non_string_timeframes_are_refused_by_semantic_type()— 0 callers, 7 callees - …and 1 more — each is listed as a finding
Verification — 6688 functions in the blast radius were not formally verified this run (proofs are advisory here).
Gate & verification
graphify gate
PASS — objectively clean (no health regressions, tests not run — proofs not run this pass (advisory)). Grounded, not self-assessed.
Advisory (not blocking):
- verification_scope: 5803 function(s) in the blast radius were not formally verified this run
Cross-repo impact
Cross-repo impact
This change impacts 2 symbol(s) across 1 peer repo(s) — blast radius over derived cross-repo edges:
- pineforge-4pass/pineforge-hpo —
python/pineforge_hpo/transpile.py(via import_moniker, heuristic),tests/python/test_transpile.py(via import_moniker, heuristic)
1 peer repo(s) in this group are not yet indexed — cross-repo impact may be incomplete.
Cross-repo linkage is derived (SCIP-moniker matches are sound; name/contract matches are heuristic) and ADVISORY — it never affects this check's conclusion.
Test selection
Test selection
536 of 536 test file(s) selected (100%) via static blast radius.
Escalated to a full run for safety (
noncode-change,uncovered-code) — the selection is not trustworthy on its own (see below). CI should run the whole suite.
tests/test_action_pins.py— full-run-safetytests/test_allow_entry_in_qualifier.py— impact, full-run-safetytests/test_allow_entry_in_request_const.py— impact, full-run-safetytests/test_analyzer.py— impact, full-run-safetytests/test_analyzer_matrix_inference.py— impact, full-run-safetytests/test_analyzer_ta_return_types.py— impact, full-run-safetytests/test_anchor_tf_change.py— impact, changed-test, full-run-safetytests/test_anchor_time.py— impact, full-run-safetytests/test_anchor_vwap.py— impact, changed-test, full-run-safetytests/test_array_bounds_insert_fill_slice.py— impact, full-run-safetytests/test_array_checked_access.py— impact, full-run-safetytests/test_array_empty_aggregates.py— impact, full-run-safetytests/test_array_history.py— impact, full-run-safetytests/test_array_missing_integer_index.py— impact, full-run-safetytests/test_array_new_color_fill_and_na_size.py— impact, full-run-safetytests/test_array_new_defaults_and_order.py— impact, full-run-safetytests/test_array_reference_semantics.py— impact, full-run-safetytests/test_array_retained_alias.py— impact, full-run-safetytests/test_array_slice_view.py— impact, full-run-safetytests/test_array_sort_indices.py— impact, full-run-safetytests/test_array_terminal_returns.py— impact, full-run-safetytests/test_arrseg_coded_run_stops.py— impact, full-run-safetytests/test_arrseg_review_fixes.py— impact, full-run-safetytests/test_binop_operand_order.py— impact, full-run-safetytests/test_block_shadow_primitive_types.py— impact, full-run-safetytests/test_block_var_plain_siblings.py— impact, full-run-safetytests/test_bool_na_refusal.py— impact, full-run-safetytests/test_bool_na_refusals_2.py— impact, full-run-safetytests/test_bool_na_values.py— impact, full-run-safetytests/test_bool_string_inference.py— impact, full-run-safetytests/test_builtin_name_shadow.py— impact, full-run-safetytests/test_builtin_named_locals_compile.py— impact, full-run-safetytests/test_builtin_named_locals_price_refusal.py— impact, full-run-safetytests/test_builtin_named_locals_receipt.py— impact, full-run-safetytests/test_builtin_named_locals_scope.py— impact, full-run-safetytests/test_builtin_named_parameters.py— impact, full-run-safetytests/test_builtin_named_parameters_reference.py— full-run-safetytests/test_calc_on_order_fills_codegen.py— impact, full-run-safetytests/test_calendar_lambda_locals.py— impact, full-run-safetytests/test_call_argument_order.py— impact, full-run-safetytests/test_callable_compound_type_flow.py— impact, changed-test, full-run-safetytests/test_callable_local_histories.py— impact, full-run-safetytests/test_callable_series_param_types.py— impact, full-run-safetytests/test_callable_string_series_type.py— impact, full-run-safetytests/test_callable_ta_forwarded_length.py— impact, full-run-safetytests/test_callable_var_first_reach.py— impact, full-run-safetytests/test_checked_settings.py— impact, full-run-safetytests/test_ci_engine_main_touched.py— full-run-safetytests/test_clone_name_allocation.py— impact, changed-test, full-run-safetytests/test_closed_trade_fold.py— full-run-safety- … and 486 more
non-code file(s) changed (
CHANGELOG.md,docs/reference/support-contracts.md,pineforge_codegen/diagnostics_catalog.json,tests/fixtures/diagnostic_codes_pin.json,tests/fixtures/diagnostic_notes/delta.json…) → running the full suite for safety (a code graph can't see config/fixture/data deps)
changed code file(s) with no mapped test (
CHANGELOG.md,docs/reference/support-contracts.md,pineforge_codegen/analyzer/call_handlers.py,pineforge_codegen/codegen/drawing.py,pineforge_codegen/codegen/emit_top.py…) — a coverage gap or a missing link — running the full suite rather than only the selected tests
Selection is safe under the controlled-regression assumption; always-run tests + a periodic full run are the backstops. Advisory — it never changes the check verdict.
Docs that may be stale (advisory)
AGENTS.md§ Engine-enabled tests (selected cases now run in native PR CI) (lines 100-124): references changed symbolssignaturesAGENTS.md§ Parity campaign gate (applies on EVERY harness) (lines 256-277): references changed symbolsrecordCLAUDE.md§ Engine-enabled tests (selected cases now run in native PR CI) (lines 100-124): references changed symbolssignaturesdocs/reference/adding-a-pine-function.md§ How to add a new Pine v6 function (lines 1-40): references changed symbolssignaturesdocs/reference/file-map.md§ File map (lines 1-189): references changed symbolssignaturesdocs/reference/known-codegen-quirks.md§ Known codegen quirks (read before changing) (lines 1-493): references changed symbols_type_spec_from_expr,signatures
Formal verification
Could not verify: Could not verify \_\_init\_\_.
The verifier did not have enough to check \_\_init\_\_, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_analyze.
The verifier did not have enough to check \_analyze, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_operator\_operand\_type.
The verifier did not have enough to check \_operator\_operand\_type, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_propagate\_call\_site\_counts.
The verifier did not have enough to check \_propagate\_call\_site\_counts, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_untyped\_param\_family\_conflicts.
The verifier did not have enough to check \_untyped\_param\_family\_conflicts, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_visit\_FuncCall.
The verifier did not have enough to check \_visit\_FuncCall, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_visit\_FuncDef.
The verifier did not have enough to check \_visit\_FuncDef, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_visit\_MemberAccess.
The verifier did not have enough to check \_visit\_MemberAccess, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_visit\_TupleAssign.
The verifier did not have enough to check \_visit\_TupleAssign, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_visit\_VarDecl.
The verifier did not have enough to check \_visit\_VarDecl, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_handle\_user\_func\_call.
The verifier did not have enough to check \_handle\_user\_func\_call, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_\_init\_\_.
The verifier did not have enough to check \_\_init\_\_, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_check\_node.
The verifier did not have enough to check \_check\_node, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
No difference found (not proven): No behavior difference found in \_selection\_type\_spec (not a proof).
The verifier ran both versions of \_selection\_type\_spec on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: the synth_receiver tier. A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
No difference found (not proven): No behavior difference found in \_type\_spec\_from\_expr (not a proof).
The verifier ran both versions of \_type\_spec\_from\_expr on many inputs and saw identical behavior every time. Strong evidence the change is safe, but evidence, not a proof.
Guarantee: Empirical: the synth_receiver tier. A divergence on an untested input remains possible, so this is 'no counterexample found', not 'proven equivalent'.
Note: An input the sampler did not try could still differ.
Could not verify: Could not verify \_visit.
The verifier did not have enough to check \_visit, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_\_init\_\_.
The verifier did not have enough to check \_\_init\_\_, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_build\_func\_instances.
The verifier did not have enough to check \_build\_func\_instances, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_generate\_cpp.
The verifier did not have enough to check \_generate\_cpp, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_inline\_history\_contexts\_for\_owner.
The verifier did not have enough to check \_inline\_history\_contexts\_for\_owner, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_is\_default\_drawing\_initializer.
The verifier did not have enough to check \_is\_default\_drawing\_initializer, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_is\_omitted\_udt\_field.
The verifier did not have enough to check \_is\_omitted\_udt\_field, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_is\_skip\_expr.
The verifier did not have enough to check \_is\_skip\_expr, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_prepare\_inline\_history\_members.
The verifier did not have enough to check \_prepare\_inline\_history\_members, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
Could not verify: Could not verify \_prepare\_runtime\_scalar\_var\_initializers.
The verifier did not have enough to check \_prepare\_runtime\_scalar\_var\_initializers, so it is saying so rather than guessing. No false assurance is the whole point.
Guarantee: No guarantee either way, this is an honest abstention, not a pass.
Note: Reason: method — first parameter is `self`, which needs a constructed instance (not synthesizable)
· 6 grounded finding(s) anchored inline below; 3 more finding(s) on lines outside this diff (see the check run).
| return STRATEGY_SIZING_CONSTANTS.get(f"{namespace}.{name}") | ||
|
|
||
|
|
||
| def get_variable_type_spec( |
There was a problem hiding this comment.
get_variable_type_spec()
high coupling complexity (Ca·Ce = 12).
Grounded coupling-delta finding (deterministic), not an LLM guess.
| @pytest.mark.parametrize("stem", [ | ||
| "_pf_collection_hist", "_sec7", "_precalc", "_strat", | ||
| ]) | ||
| def test_composed_reserved_storage_avoids_a_parameter_at_its_escape( |
There was a problem hiding this comment.
test_composed_reserved_storage_avoids_a_parameter_at_its_escape()
fans out to 6 callees (efferent coupling).
Grounded coupling-delta finding (deterministic), not an LLM guess.
|
|
||
|
|
||
| @pytest.fixture(scope="module") | ||
| def argument_runs(tmp_path_factory): |
There was a problem hiding this comment.
argument_runs()
fans out to 8 callees (efferent coupling).
Grounded coupling-delta finding (deterministic), not an LLM guess.
| ("1.25", "float"), | ||
| ("-2.5", "float"), | ||
| ]) | ||
| def test_numeric_timeframe_is_refused_during_analysis(function, named, argument, pine_type): |
There was a problem hiding this comment.
test_numeric_timeframe_is_refused_during_analysis()
fans out to 7 callees (efferent coupling).
Grounded coupling-delta finding (deterministic), not an LLM guess.
|
|
||
|
|
||
| @pytest.mark.parametrize("function", ("change", "in_seconds")) | ||
| def test_open_numeric_domain_is_not_reported_as_a_guessed_float(function): |
There was a problem hiding this comment.
test_open_numeric_domain_is_not_reported_as_a_guessed_float()
fans out to 7 callees (efferent coupling).
Grounded coupling-delta finding (deterministic), not an LLM guess.
| ]) | ||
| @pytest.mark.parametrize("conversion", [False, True], ids=["alone", "bool-to-text"]) | ||
| @pytest.mark.parametrize("declarations,argument,pine_type", NON_STRINGS) | ||
| def test_non_string_timeframes_are_refused_by_semantic_type( |
There was a problem hiding this comment.
test_non_string_timeframes_are_refused_by_semantic_type()
fans out to 7 callees (efferent coupling).
Grounded coupling-delta finding (deterministic), not an LLM guess.
… and keep generated names distinct (#405) * Resolve empty timeframe arguments to the chart period * Preserve timeframe arguments and generated-name safety Resolve dotted builtin signatures consistently and preserve explicit strings, chart-period fallback and single evaluation. Reserve generated conversion names. * Update anchor test expectations for conditional timeframe calls The existing symbol-clock anchor tests now check both calls in the emitted empty-argument conditional. Add pure-Python checks of branch direction and single evaluation, and make the temporary-name sweep reach the argument lambda. Product code is unchanged by this test update. * Preserve reference types across fields and calls Lower missing reference values to their destination storage and retain drawing handle types through user functions and tuple returns. Support declared polyline types, fields and deletion arguments. * Handle emitters without function metadata Read optional function metadata safely when inferring reference casts, so bare emitters retain matrix element types without raising an AttributeError. * Validate timeframe argument types Use the chart period for a generic missing timeframe argument and refuse numeric values for string parameters during analysis. Cover named arguments, helper calls, source locations and generated translation units. * Register the timeframe argument diagnostic Include only the stock-generated PF-E2095 definition and pin. Preserve all prior catalog entry bytes and their existing order. * Disambiguate diagnostic templates by argument kinds Use recognized Pine types in declared type slots to resolve equally literal diagnostic templates. Keep stable code order for other ties and preserve qualifier identities, raw arguments and rendering. Pin the timeframe type diagnostic and its named arguments with direct regression controls. * Validate timeframe arguments using semantic types Reject non-string timeframe arguments through the intrinsic signature before C++ lowering, including helper parameters and boolean values. Preserve valid string calls and exercise complete native controls at distinct optimization levels. Declare the new diagnostic in the existing catalog delta fixture. * fix(codegen): allocate function clone names against authored members Keep generated state and callable clone identities distinct from authored members and use the same allocated identities at declarations and calls. Preserve noncolliding names and duplicate-state assertions. * Use the generated timeframe type diagnostic presentation Carry the generated argument-type entry and its pin into the member while preserving all other catalog bytes and order. Use the grammatical message and exact approved user presentation with the existing argument kinds. Retain the same numeric refusal cases and qualifier controls. * Locate helper argument expectations inside the builtin call Use the argument occurrence within the call instead of an identically named helper parameter. Keep the exact line and column assertions and all source inputs, diagnostic identities and test rows unchanged. * Resolve registered argument types and report their source starts Carry registered namespace types through the semantic query used for written helper calls while respecting lexical shadowing and version boundaries. Reuse the existing expression-start locator for timeframe type refusals. * fix(codegen): separate composed helper and lexical storage names Allocate composed callable and persistent-storage identities outside reserved helper families, and reserve all authored spellings before sibling-block qualification. Keep declarations, calls and history consumers on the same allocated identity. Exercise collision cases through the public pipeline. * Preserve sizing string values across expressions and host settings * Exercise v5 receiver shadowing through an imported library * Preserve integer risk type boundary codes * Strengthen sizing regression and reserve generated name * Lower computed risk type arms at numeric boundary * Convert sizing selection arms at host boundaries * fix(codegen): preserve established typed method names Keep raw typed-method names on the function channel when they collide with authored data members. Continue allocating encoded names and collisions with generated storage so method definitions, calls and state stay distinct. * Assert the risk type conversion for reassigned sizing variables (#394) Sizing constants keep their Pine string type for the string consumers this change fixes, so a reassigned sizing variable reaches the risk host boundary as a string and is converted to its integer code there. Keeping the integer comparison would need integer storage for these variables, which the string consumers cannot use. The test still requires the call to read t after its write, never its initializer, and now also checks the code chosen for the assigned value.
Bring the branch up to date with main after #405 merged. No conflicts.
This pull request adds the compound value type fix (#391) on top of #405, which it contains unchanged. Once #405 has merged, the remaining change here is #391's. It supersedes #404.
Compound value types across user function calls (#391)
Valid Pine v6 scripts could generate C++ that did not compile, because an array, matrix, UDT or string was inferred as
doubleat a function return, an untyped parameter or a reassigned local.The fix:
The admission of an unresolved script-level selection is unchanged, so its history output stays the same.
Reproduction tests:
tests/test_callable_compound_type_flow.py::test_callable_boundaries_keep_the_value_type[array-block-return]tests/test_callable_compound_type_flow.py::test_untyped_array_calls_keep_distinct_element_typestests/test_callable_compound_type_flow.py::test_reassigned_matrix_local_keeps_its_initializer_typeFixes #391