Conversation
Introduces an initial CFC transpiler implementation with some basic support for block execution and test coverage. However, the transpiler is not yet wired into the compiler pipeline and as such does not have any effects on it.
Introduces (un)conditional return statements for CFC.
Introduces a "resolver" to replace the placeholder types for the
temporary variables to their actually referenced types. For example
```
VAR_TEMP
temp_0: __return@foo;
temp_1: __output@bar@baz;
END_VAR
```
becomes
```
VAR_TEMP
// Assuming function foo returns a DINT
temp_0: DINT;
// Assuming the output variable of function block bar is a SINT
temp_1: SINT;
END_VAR
```
The synthetic `temp_N` variables introduced for block outputs are now named `__temp_N`, keeping them out of the user-facing identifier namespace. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
A CFC block whose type name is qualified (`function_block_0.myAction`) targets an action of the block's instance. It lowers to a member call on that instance — `myInstance.myAction()` — with no arguments, since actions have no inputs, outputs, in-outs, or return value. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Build Artifacts🐧 Linux
From workflow run 🪟 Windows
From workflow run |
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
volsa
commented
Jul 1, 2026
volsa
commented
Jul 6, 2026
This was referenced Jul 6, 2026
The `OutputVariable` field of `Object::BlockOutput` is only read in tests, making `cargo clippy --workspace -- -Dwarnings` (i.e. the CI style check, which excludes test code) fail. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
ghaith
previously approved these changes
Jul 9, 2026
ghaith
left a comment
ghaith
left a comment
Collaborator
There was a problem hiding this comment.
Review was offline, we'll get this in as is and followup with other PRs to avoid this getting stale
A negated output pin now inverts the value it feeds to each of its consumers. Also hardens two seams uncovered along the way: sinks and returns require an `EvaluationPriority` entry at deserialize time (a missing one silently sorted the object in front of the network, dead-coding the POU for an unconditional return), and the transpiled network and its temporaries are attached to the POU implementation by name instead of position (declarations containing methods pushed the body to the back of the implementation list). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A sink naming a compound target (`results[1]`, `myStruct.field`) now lowers through the expression parser instead of a flat identifier token, which previously aborted compilation with a bogus E048 and made the validator's deferred E050 check unreachable. Result temporaries of generic callees (e.g. the builtin `SEL`) can no longer be typed from the index alone -- their declared return type is a generic marker (`__SEL__U`) only a call site can specialize -- so the placeholder resolution moved from `post_index` to `post_annotate`, where the annotated call provides the concrete type. Non-CFC projects skip the pass through a cheap placeholder probe. Also renders inline `ARRAY[..] OF` declarations, range statements and VLA bounds properly in the AST serializer (previously `ARRAY[1..5] OF DINT` printed as `15DINT`). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
A method block arrives like an action (dotted `typeName` plus `instanceName`) but, unlike an action, carries pins. Its return pin is named after the method alone, so return-pin detection and the result temporary's name now use the callee's unqualified name (`Block::callee_name()`) -- previously the pin could never match the dotted `typeName` and was emitted as a bogus output argument with a dot inside the variable name. The IDE does not export method blocks yet; the fixture pins the shape anticipated by analogy with actions and functions. Also removes the invented `ConnectionPointOut` from the in-out pins of four hand-written fixtures: the IDE renders in-out pins on the left side of a block only and never exports an out-side connection point (verified against a real export). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Adds the missing coverage for direct self-feedback wires (design doc FB3): `function_block_feedback` proves cycle persistence end-to-end (y = 1, 2, 3 -- a VAR_TEMP regression would transpile identically and only fail there) and `feedback_in_function` pins the per-invocation counterpart (a function's feedback wire reads the type default on every call). Also refreshes the stale `temp_N` renderings in the fixture and lit READMEs to the actual `__<callee>_res_<n>` names, removes the dangling `OPEN.md` references, labels unconditional returns as "return" instead of "conditional return", and adds the crate's fixture-first workflow notes (AGENTS.md). Co-Authored-By: Claude Fable 5 <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
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.
Introduces the new CFC implementation based on the PLCOpen XML standard (IEC61131-10, https://www.plcopen.org/standards/logic/iec-61131-10/). Consequently the old implementation was removed altogether (
plc_xml) and a newplc_cfccrate has been introduced in its place. The changes are mostly self-isolated.Recommendation for reviewing this: Start with the transpiler.rs file and some fixtures and expand from there onwards. The changes are rather small and well documented (imo), most of the additions are also just test fixtures.