Release v0.4.19#131
Merged
Merged
Conversation
…allable from generated code (#129) User-reported bug on openplc-web: an ST program using `ADD_TIME` on TIME variables failed at the arduino-cli stage with `'ADD_TIME' was not declared in this scope` (the AVR cross compiler couldn't see the symbol). Same project compiled fine on the desktop editor — but only by accident; it would have failed there too if the program had ever made it to the arduino-cli step on that path. Root cause is a two-layer dead-code problem in the runtime headers: 1. `iec_std_lib.hpp` was the only header `codegen.ts` emitted into every `generated.hpp`, but it didn't transitively `#include "iec_time.hpp"` / `iec_date.hpp` / `iec_dt.hpp` / `iec_tod.hpp`. A generated POU calling `ADD_TIME` saw no declaration — hence the "not declared in this scope" error. 2. Even with the includes restored, the functions in those four headers were templated on `TimeValue<T>` / `DateValue<T>` / `DateTimeValue<T>` / `TimeOfDayValue<T>` — a parallel value-class family the codegen never adopted. TIME variables are emitted as `IEC_TIME` (= `IECVar<TIME_t>`); the templates couldn't deduce against an IECVar so the overload didn't bind either. Strategy C — collapse the dead path: - The four headers are rewritten to operate on the canonical `IECVar<T>` aliases (`IEC_TIME`, `IEC_DATE`, `IEC_DT`, `IEC_TOD`) that codegen actually emits. Functions take/return IECVars, extract the underlying integer via `iec_unwrap()`, and re-wrap the result — matching the existing IECVar-shaped helpers in `iec_std_lib.hpp` (`TO_TIME`, `TIME_FROM_MS`, `TIME_TO_MS`, …). - `iec_std_lib.hpp` now transitively `#include`s all four time- family headers so the single `#include "iec_std_lib.hpp"` that `generated.hpp` emits is enough for every standard function. - Deleted: `TimeValue<T>`, `IECTimeVar<T>`, `DateValue<T>`, `IECDateVar<T>`, `DateTimeValue<T>`, `IECDtVar<T>`, `TimeOfDayValue<T>`, `IECTodVar<T>`, plus the `IEC_*_Value` / `IEC_*_Var` aliases and the `MAKE_TIME_*` / `MAKE_LTIME_*` constructors that targeted them. No call site in `src/`, in the libs, or in either downstream repo (openplc- editor, openplc-web) referenced any of these — `git grep` showed only the four headers themselves, this test file, and documentation comments. - Deleted: calendar component accessors (`YEAR`, `MONTH`, `DAY`, `DAY_OF_WEEK`, `DAY_OF_YEAR`, `HOUR`, `MINUTE`, `SECOND`, `MILLISECOND`, plus the `DT_*` variants). These collided with OSCAT-imported user-side `FUNCTION DAY_OF_YEAR : INT` declarations in the OSCAT codesys-v23 stdlib (overload-by-return-type errors). They were never IEC 61131-3 standard free functions anyway — IEC component access is via OSCAT-style library imports. - Also `#undef`s AVR `<time.h>`'s `ONE_HOUR` / `ONE_DEGREE` / `ONE_DAY` / `UNIX_OFFSET` / `NTP_OFFSET` macros that `<chrono>` transitively pulls in. Same class of macro-pollution bug as the existing `#undef OVERFLOW` codegen emits — without these undefs, an IEC variable named `one_hour` (upper-cased to `ONE_HOUR` by codegen) silently becomes the integer literal `3600` before the C++ parser sees the declaration, producing `expected unqualified-id before numeric constant`. - The dead `IEC_STRUCT_VALUE_FIELD` macro in `iec_struct.hpp` (referenced `IEC_TIME_Value` which is gone) is removed. No callers; `IEC_STRUCT_FIELD` was already not in use either — codegen emits `IECVar<T>` struct fields directly without the macro. Tests: - `runtime/tests/test_std_lib.cpp`: rewritten to exercise the IECVar surface (the only one codegen emits). Old tests using `IEC_TIME_Var` / `IEC_TIME_Value` / accessor functions removed. Coverage of the surface that ships to users is now what's tested. - `tests/integration/cpp-compile.test.ts`: new regression case `compiles IEC TIME / DATE / DT / TOD standard arithmetic` that drives an FB using `ADD_TIME` / `ADD_DATE` / `ADD_DT` / `ADD_TOD` / `TIME_TO_S` through `compile(...)` + `compileWithGpp(...)`. Would have caught the original bug — the resulting C++ doesn't compile if `iec_time.hpp` isn't reachable from `generated.hpp` or if the function signatures don't match `IECVar` arguments. End-to-end verification done locally: - Full suite: `npm run test` → 70/70 files, 1975 tests passing - Arduino-cli end-to-end: strucpp-emitted POU using ADD_TIME + ADD_DATE + ADD_DT + ADD_TOD + TIME_TO_S, compiled for arduino:avr:mega with the editor's bundled toolchain, produced a 4136-byte .hex. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Patch release rolling up the IEC TIME/DATE/DT/TOD runtime-header fix (#129): - `iec_std_lib.hpp` transitively pulls in `iec_time.hpp` / `iec_date. hpp` / `iec_dt.hpp` / `iec_tod.hpp` so generated POU code calling `ADD_TIME` et al. links cleanly. - The four time-family headers are rewritten to operate on `IECVar<T>` (the wrapper codegen actually emits) instead of the parallel `TimeValue<T>` / `IECTimeVar<T>` value-class design that was dead from generated code's perspective. - `TimeValue` / `IECTimeVar` / `DateValue` / `IECDateVar` / `DateTimeValue` / `IECDtVar` / `TimeOfDayValue` / `IECTodVar` deleted along with their aliases. Calendar component accessors (YEAR / MONTH / DAY / DAY_OF_WEEK / HOUR / …) removed too — those are OSCAT-side responsibilities and were colliding with user- imported function bodies. - AVR `<time.h>` macro pollution (`ONE_HOUR`, `ONE_DAY`, etc.) now `#undef`d in `iec_std_lib.hpp` alongside the existing `#undef OVERFLOW` codegen. End-to-end verification: strucpp + arduino-cli produced a working .hex for an ATmega2560 sketch using `ADD_TIME` / `ADD_DATE` / `ADD_DT` / `ADD_TOD` / `TIME_TO_S`. Co-authored-by: Claude Opus 4.7 (1M context) <noreply@anthropic.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
1 participant
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.
Promotes
development→mainfor the v0.4.19 release.Includes:
fix(runtime): IEC TIME/DATE/DT/TOD standard functions actually callable from generated code (the runtime-header dead-code cleanup)chore(release): v0.4.19version bumpAfter merge: tag
v0.4.19on main to kick the release workflow (which builds the strucpp binary + bundles runtime headers + libs and publishes the GitHub release assets that openplc-editor/web pin viabinary-versions.json).🤖 Generated with Claude Code