Repository navigation
M2.17-G: exact DataSet MX schema fallback and truthful q/t evidence status - #612
Draft
masarray wants to merge 5 commits into
Conversation
…SCL when live schema is absent
…and no MMS read promotion
…from exposure gaps
This was referenced Oct 16, 2026
This branch has not been deployed
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
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
M2.17-G — DataSet MX schema completeness and truthful q/t evidence status
Stacked base: draft #611 at exact SHA
80bc2dab2d807c5636f86ffe244337daf97b93d9. Retain the prior PR chain and pinned engines; no main merge.Audited correctness gaps
IedBrowserStructuredDetails.Membercould use selected-IED SCL as a schema-only fallback for some compound ST members, but not for MX DataSet members when live discovery provided the DataSet reference without an exact typed DA hierarchy. This left declared phasor magnitudes, angles and q/t expansion unavailable even when a verified selected-IED SCL described them. The consumer was not missing a second decoder or cache: it lacked a consistent schema projection.Not exposed, conflating a verified received leaf with an unavailable typed expansion.Changes
IedBrowserEngineeringObjects.GroupExactand phase/vector presentation with a stateless exact-reference, exact-FC presentation owner resolver. Live schema always takes priority when it declares the requested member FC. Only in the absence of live FC declarations, the selected-IED SCL can provide the shape of MX member and lazy child rows. Different IED, FC, owner prefix or ambiguous owners fail closed. No SCL scalar is published and no MMS Read target is derived from SCL.MemberandDataSetMemberChildrenshare that one resolver. Consistent triplet/quality/time child structure can now be drawn without a shadow Live Model, while the actual value is still fetched from the original existing per-IED Read/Report ledger. UI provenance explicitly says "SCL structure only" when appropriate.Not decodedand gives the precise reason instead of claiming the IEC attribute was not exposed. This does not generate bit flags or change the existing engine decoder.Explicit boundaries / later field acceptance
This is a schema/presentation correctness slice for DataSet evidence reliability, not a claim that all BCUGE
Not readandNot exposedcells have vanished. If a live DataSet has no value in Read/Report ledger or a selected model cannot provide exact FCDA identity, the UI correctly remains unavailable. The next physical evidence gate must inspect actual MMS read failure/reference and semantic Report mapping; the purpose is to diagnose protocol absence rather than synthesize defaults.No new value cache, consumer decoder, background scan, GOOSE subscriber or engine repin. Keep PR draft until exact-head Windows compilation and regression suite pass.