AIEP — Canonical Pointer Resolution and Deterministic Structural Diff Classification Engine
Field of The Invention
[0001] The present invention relates to deterministic structural comparison in computing systems.
[0002] More particularly, the invention relates to a canonical pointer resolution and deterministic structural diff classification engine for use within distributed arbitration systems operating under an Architected Instruction & Evidence Protocol (AIEP) system as defined in United Kingdom patent application GB2519711.2.
Background
[0003] Distributed deterministic systems frequently require comparison between structured state artefacts to verify equivalence, detect divergence, or enforce invariant compliance.
[0004] Conventional diff mechanisms operate on textual or unordered structural representations, leading to non-deterministic difference enumeration.
[0005] Many structural diff implementations depend upon runtime ordering of keys, memory layout, or interpreter behaviour.
[0006] Pointer-based diff systems are known; however, such systems commonly rely on implementation-dependent traversal order.
[0007] Non-deterministic diff classification prevents reliable replay certification and cross-node equivalence validation.
[0008] There exists a need for a deterministic structural diff engine that:
(a) resolves canonical pointers in a reproducible manner;
(b) enumerates structural differences in canonical order;
(c) classifies differences according to invariant-bound rules; and
(d) produces reproducible diff artefacts across distributed nodes.
Summary Of The Invention
[0009] The invention provides a computer-implemented method for canonical pointer resolution and deterministic structural diff classification.
[0010] Structured artefacts are serialised in canonical form.
[0011] A deterministic traversal algorithm enumerates structural elements according to a canonical ordering rule.
[0012] Canonical pointers are generated for each structural path.
[0013] Differences between two artefacts are identified by comparing corresponding canonical pointer paths.
[0014] Each identified difference is classified into one or more categories comprising:
(a) value change;
(b) structural addition;
(c) structural removal;
(d) type mismatch;
(e) invariant violation.
[0015] Enumeration of differences occurs in canonical pointer order.
[0016] A deterministic DiffHash is computed over the ordered difference set.
[0017] The method ensures identical diff outputs across nodes operating over identical artefacts.
[0018] The technical effect is reproducible structural comparison enabling deterministic arbitration enforcement and replay certification.
Definitions
[0019] CanonicalPointer: A deterministic structural path identifier derived from canonical serialisation of a structured artefact.
[0020] DiffSet: The ordered collection of classified structural differences between two artefacts.
[0021] DiffHash: A cryptographic hash computed over the canonical ordered DiffSet.
[0022] StructuralTraversalRule: A deterministic rule defining traversal order of structured elements.
Brief Description Of The Drawings
[0023] Figure 1 illustrates canonical serialisation and pointer generation.
[0024] Figure 2 illustrates deterministic structural traversal.
[0025] Figure 3 illustrates difference classification categories.
[0026] Figure 4 illustrates computation of DiffHash.
Detailed Description Of Preferred Embodiments
Canonical Serialisation
[0027] Structured artefacts are serialised using deterministic key ordering and stable structural representation.
[0028] Serialisation prohibits non-deterministic ordering behaviour.
[0029] Canonical serialisation ensures identical byte representation for identical logical structures.
Deterministic Pointer Generation
[0030] For each structural element, a CanonicalPointer is generated.
[0031] CanonicalPointer construction follows StructuralTraversalRule.
[0032] StructuralTraversalRule comprises deterministic ordering of object keys, array indices, and nested elements.
[0033] CanonicalPointer format is consistent across distributed nodes.
Structural Comparison
[0034] Two artefacts are compared by iterating through CanonicalPointer sets in canonical order.
[0035] Where a pointer exists in both artefacts, values are compared.
[0036] Where a pointer exists in only one artefact, a structural addition or removal is recorded.
[0037] Type mismatches are detected where corresponding values differ in structural type.
[0038] Invariant violations are identified where schema-bound constraints are breached.
Difference Classification
[0039] Each detected difference is classified deterministically.
[0040] Classification rules are schema-bound and version-pinned.
[0041] Classification output comprises pointer identifier, classification type, and affected values.
[0042] Differences are appended to DiffSet in canonical pointer order.
Diff Hash Computation
[0043] DiffSet is serialised in canonical order.
[0044] DiffHash is computed over the serialised DiffSet.
[0045] DiffHash provides reproducible verification of structural divergence.
Distributed Reproducibility
[0046] Nodes operating over identical artefacts and schema versions produce identical CanonicalPointer sets.
[0047] Nodes produce identical DiffSet enumeration and DiffHash values.
[0048] DiffHash comparison enables cross-node structural equivalence validation.
Fail-Closed Integration
[0049] Where DiffSet contains non-permitted classifications, execution enablement is suppressed.
[0050] Permitted classifications may be defined through schema-bound exclusion rules.
[0051] No structural difference bypasses classification evaluation.
Technical Effect
[0052] The invention enables deterministic structural comparison independent of runtime implementation differences.
[0053] The invention provides reproducible pointer-based diff enumeration.
[0054] The invention enables reliable cross-node replay certification.
[0055] The invention improves invariant enforcement in distributed arbitration systems.
[0056] The invention eliminates non-deterministic diff ordering errors.
CLAIMS
1. A computer-implemented method for deterministic structural comparison within a distributed arbitration system operating under an Architected Instruction & Evidence Protocol (AIEP), the method comprising:
(a) serialising structured artefacts in canonical form;
(b) generating CanonicalPointers according to a deterministic StructuralTraversalRule;
(c) comparing artefacts by iterating through CanonicalPointers in canonical order;
(d) classifying detected differences into one or more categories comprising value change, structural addition, structural removal, type mismatch, or invariant violation;
(e) constructing an ordered DiffSet; and
(f) computing a DiffHash over the ordered DiffSet.
2. The method of claim 1 wherein StructuralTraversalRule enforces deterministic ordering of object keys and array indices.
3. The method of claim 1 wherein classification rules are schema-version bound.
4. The method of claim 1 wherein DiffHash enables cross-node equivalence verification.
5. The method of claim 1 wherein non-permitted classifications result in fail-closed execution suppression.
6. A distributed computing system configured to perform the method of any preceding claim.
7. A non-transitory computer-readable medium storing instructions which, when executed, perform the method of any of claims 1–5.
Abstract
A canonical pointer resolution and deterministic structural diff classification engine is disclosed for distributed arbitration systems operating under an Architected Instruction & Evidence Protocol (AIEP). Structured artefacts are serialised in canonical form and traversed according to deterministic ordering rules to generate CanonicalPointers. Differences are classified in canonical order and recorded in an ordered DiffSet. A DiffHash computed over the DiffSet enables reproducible cross-node structural comparison. Non-permitted differences cause fail-closed execution suppression. The invention improves deterministic equivalence validation and invariant enforcement within distributed computing frameworks.
Brief Description of the Drawing
FIG. 1 — Branch Structure and Pruning
┌──────────────┐
│ Root State │
│ (canonical) │
└──────┬───────┘
┌───────────┼───────────┐
┌──────▼──────┐ │ ┌───────▼─────┐
│ Branch B1 │ │ │ Branch B2 │
│ viable │ │ │ stale → │
└──────┬──────┘ │ │ PRUNE │
│ │ └─────────────┘
┌──────▼──────┐ │
│ Branch B1a │ │ ┌─────────────┐
│ (deepened) │ └───▶│ Branch B3 │
└─────────────┘ │ candidate │
└──────┬──────┘
│
┌────────▼────────┐
│ Selected Path │
│ hash-committed │
└─────────────────┘