@primmel/primmel 1.3.2 → 1.4.0

package/dist/src/ser-des/config/subject.js

@@ -0,0 +1,2093 @@
+"use strict";
+// ─────────────────────────────────────────────────────────────────────
+// Subject-chain constructs (Primmel v2, gap G1):
+//   instrument, attribute_definition, capability, behavior, condition_set
+//
+// Grammar sketch:
+//   instrument LoadCell {
+//     extends MeasuringInstrumentModel
+//     definition "..."
+//     variant AnaloguePassiveLoadCell { definition "..." }
+//     dimension accuracy_class {
+//       label "Accuracy class"  scope group  description "..."
+//       reference { doc "urn:..." clause "5.1.1" }
+//       values { A { description "..." payload { n_lc_limits: "..." } } B C D }
+//     }
+//     family_criteria { "same material..." "same design..." }
+//     family_defaults { dimensions { construction technology } parameters { rated_output } }
+//     model_group {
+//       definition "..."
+//       identical_characteristics { metrological_class n_lc y z }
+//       identical_attributes { accuracy_class n_lc y z t_min t_max }
+//     }
+//   }
+//
+//   attribute_definition e_max {
+//     symbol "E_max"  name "Maximum capacity"  definition "..."
+//     source { doc "urn:..." clause "3.5.5" }
+//     quantity_kind mass  unit kg  value_type QuantityValue
+//     origin design-fixed  scope model  category metrological
+//     is_dimension false  enum humidity_class  irdi "..."  derived "ocl{...}"
+//   }
+//
+//   capability digital {
+//     label "Digital"  description "..."  abstract false
+//     extends analogue-active  requires strain-gauge
+//     has_parameters { output_signal software_identification }
+//     satisfies_requirements { /req/electronic/software }
+//     verified_by_tests { /conf/electronic-tests/software }
+//   }
+//
+//   behavior creep {
+//     kind temporal  stimulus force  response "..."
+//     source { doc "..." clause "3.7.1" }  verified_by { /conf/... }
+//   }
+//
+//   condition_set load-cell-reference {
+//     role reference
+//     entries { temperature { value 20 unit degC tolerance 1 } }
+//   }
+//
+// Subject anatomy (Primmel v3, TODO.roadmap/01) — the `subject` construct
+// organizes the aspect catalog by the three families is / has / does:
+//
+//   subject LoadCell {
+//     extends MeasuringInstrumentModel
+//     is {
+//       metadata { name "Load cell" source "urn:oiml:pub:r:60-1:2021#clause-3.1.3" }
+//       provenance { manufacturer ACME }
+//       structure { }
+//       design_parameters { e_max : mass by design }
+//       designed_conditions { reference ref-conds rated rated-conds }
+//       promises {
+//         c6-envelope {
+//           target accuracy-class
+//           level symbolic C6
+//           conditions ocl{self.temperature >= rated.t_min and self.temperature <= rated.t_max}
+//           statement "Holds accuracy class C6 across the rated range −10…+40 °C."
+//           verified_by { /req/metrological/mpe /conf/metrological-tests/mpe-test }
+//           source { doc "urn:oiml:pub:r:60-1:2021" clause "5.1" }
+//         }
+//         "a statement-only claim (shorthand — linter C43 flags it unverifiable)"
+//       }
+//       artifacts { }
+//     }
+//     has {
+//       attributes { d_min : mass test_dependent }
+//       dimensions { accuracy_class in {A,B,C,D} }
+//       state OperationalStates
+//       characteristics { creep c_c = ocl{self.indication.delta / self.time.delta} }
+//       (block form, TODO.roadmap/10: creep { symbol "c_c" derivation ocl{…}
+//        behavior creep quantity_kind verification_interval unit "v"
+//        source { doc "…" clause "…" } })
+//       environmental_context { "logged 23.4 degC during run 7" }
+//       artifact_instances { }
+//     }
+//     does {
+//       behavior measure
+//       behavior creep
+//     }
+//   }
+//
+// Aspect blocks under the wrong family (or undeclared aspect kinds) are
+// recorded on Subject.misplacedAspects for the linter (C6) — the parse
+// itself stays lenient.
+//
+// `extends` merges at resolve time (resolveSubject): each subject merges
+// with its parent per aspect-kind rules — maps merge by key (child
+// wins), lists append (parent entries first), scalars override (child's
+// non-empty value wins). Missing parents and cycle links stay unmerged
+// (lint rule subject-extends-resolves reports the missing ones).
+// ─────────────────────────────────────────────────────────────────────
+Object.defineProperty(exports, "__esModule", { value: true });
+exports.subjectConstruct = exports.conditionSetConstruct = exports.behaviorConstruct = exports.capabilityConstruct = exports.attributeDefinitionConstruct = exports.instrumentConstruct = void 0;
+exports.resolveSubject = resolveSubject;
+const tokenize_1 = require("../tokenize");
+const tokenize_2 = require("../tokenize");
+const field_parser_1 = require("./field-parser");
+const field_parser_2 = require("./field-parser");
+const twin_1 = require("./twin");
+const composition_1 = require("./composition");
+const quantity_1 = require("./quantity");
+// ── shared little readers ────────────────────────────────────────────
+function dumpSource(keyword, src, indent) {
+    if (!src || (!src.doc && !src.clause)) {
+        return '';
+    }
+    const frag = src.fragment ? ` fragment "${(0, tokenize_2.escapeString)(src.fragment)}"` : '';
+    return `${indent}${keyword} { doc "${(0, tokenize_2.escapeString)(src.doc)}" clause "${(0, tokenize_2.escapeString)(src.clause)}"${frag} }\n`;
+}
+function readIdList(block) {
+    // Wrap-tolerant: `{ a b }` block OR bare scalar — stripWrapping only strips
+    // when actually wrapped (unwrapBlock would mangle bare ids).
+    return (0, tokenize_1.default)((0, tokenize_2.stripWrapping)(block))
+        .map(field_parser_1.stripColon)
+        .map(tokenize_2.stripWrapping)
+        .filter(s => s.length > 0);
+}
+function dumpIdList(keyword, ids, indent) {
+    if (ids.length === 0) {
+        return '';
+    }
+    return `${indent}${keyword} { ${ids.join(' ')} }\n`;
+}
+function readReference(block) {
+    return (0, tokenize_1.default)((0, tokenize_2.stripWrapping)(block))
+        .map(tokenize_2.stripWrapping)
+        .filter(s => s.length > 0);
+}
+// ── instrument ───────────────────────────────────────────────────────
+const parseInstrument = function (id, data) {
+    const result = {
+        id,
+        extends: '',
+        definition: '',
+        variants: [],
+        dimensions: [],
+        perChannel: '',
+        familyCriteria: [],
+        familyDefaultDimensions: [],
+        familyDefaultParameters: [],
+        modelGroup: null,
+        referenceIds: [],
+        measurand: null,
+        components: [],
+        structure: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'extends') {
+            result.extends = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'measurand_kind') {
+            result.measurandKind = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'measurand') {
+            result.measurand = parseMeasurand((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'component') {
+            result.components.push(parseComponent((0, tokenize_2.stripWrapping)(t[i++]), (0, tokenize_2.unwrapBlock)(t[i++])));
+        }
+        else if (cmd === 'structure') {
+            result.structure.push(parseStructure((0, tokenize_2.stripWrapping)(t[i++]), (0, tokenize_2.unwrapBlock)(t[i++])));
+        }
+        else if (cmd === 'definition') {
+            result.definition = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'note') {
+            result.note = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source') {
+            result.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'variant') {
+            const variantId = (0, tokenize_2.stripWrapping)(t[i++]);
+            const vblock = i < t.length ? (0, tokenize_2.unwrapBlock)(t[i++]) : '';
+            const v = { id: variantId, definition: '' };
+            const vt = (0, tokenize_1.default)(vblock);
+            let j = 0;
+            while (j < vt.length) {
+                const vc = vt[j++];
+                if (j >= vt.length) {
+                    break;
+                }
+                if (vc === 'name') {
+                    v.name = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (vc === 'definition') {
+                    v.definition = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (vc === 'note') {
+                    v.note = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (vc === 'source') {
+                    v.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(vt[j++]));
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(vt[j++]);
+                }
+            }
+            result.variants.push(v);
+        }
+        else if (cmd === 'dimension') {
+            result.dimensions.push(parseDimension((0, tokenize_2.stripWrapping)(t[i++]), (0, tokenize_2.unwrapBlock)(t[i++])));
+        }
+        else if (cmd === 'per_channel') {
+            result.perChannel = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'family_criteria') {
+            result.familyCriteria = readReference(t[i++]);
+        }
+        else if (cmd === 'family') {
+            // family { metamodel_class X definition "..." note "..." source { … } }
+            const fblk = (0, tokenize_2.unwrapBlock)(t[i++]);
+            const ft = (0, tokenize_1.default)(fblk);
+            let j = 0;
+            while (j < ft.length) {
+                const fc = ft[j++];
+                if (j >= ft.length) {
+                    break;
+                }
+                if (fc === 'metamodel_class') {
+                    result.familyMetamodelClass = (0, tokenize_2.stripWrapping)(ft[j++]);
+                }
+                else if (fc === 'definition') {
+                    result.familyDefinition = (0, tokenize_2.stripWrapping)(ft[j++]);
+                }
+                else if (fc === 'note') {
+                    result.familyNote = (0, tokenize_2.stripWrapping)(ft[j++]);
+                }
+                else if (fc === 'source') {
+                    result.familySource = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(ft[j++]));
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(ft[j++]);
+                }
+            }
+        }
+        else if (cmd === 'family_defaults') {
+            const fblock = (0, tokenize_2.unwrapBlock)(t[i++]);
+            const ft = (0, tokenize_1.default)(fblock);
+            let j = 0;
+            while (j < ft.length) {
+                const fc = ft[j++];
+                if (j >= ft.length) {
+                    break;
+                }
+                if (fc === 'dimensions') {
+                    result.familyDefaultDimensions = readIdList(ft[j++]);
+                }
+                else if (fc === 'parameters') {
+                    result.familyDefaultParameters = readIdList(ft[j++]);
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(ft[j++]);
+                }
+            }
+        }
+        else if (cmd === 'model_group') {
+            result.modelGroup = parseModelGroup((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'reference') {
+            result.referenceIds = readReference(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.instruments[id] = result;
+        return ctx;
+    };
+};
+/** measurand { kind X description "..." context { … } source { … } } (TODO.roadmap/19, G8). */
+function parseMeasurand(block) {
+    const m = {
+        kind: '',
+        description: '',
+        context: null,
+        source: null,
+    };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'kind') {
+            m.kind = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'description') {
+            m.description = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'context') {
+            const ctx = { targetObject: '', identificationMethod: '' };
+            const ct = (0, tokenize_1.default)((0, tokenize_2.unwrapBlock)(t[i++]));
+            let j = 0;
+            while (j < ct.length) {
+                const cc = ct[j++];
+                if (j >= ct.length) {
+                    break;
+                }
+                if (cc === 'target_object') {
+                    ctx.targetObject = (0, tokenize_2.stripWrapping)(ct[j++]);
+                }
+                else if (cc === 'identification_method') {
+                    ctx.identificationMethod = (0, tokenize_2.stripWrapping)(ct[j++]);
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(ct[j++]);
+                }
+            }
+            m.context = ctx;
+        }
+        else if (cmd === 'source') {
+            m.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return m;
+}
+/** component <id> { class X definition "..." source { … } } (TODO.roadmap/19, G3). */
+function parseComponent(id, block) {
+    const c = {
+        id,
+        classId: '',
+        definition: '',
+        source: null,
+    };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'class') {
+            c.classId = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'definition') {
+            c.definition = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source') {
+            c.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return c;
+}
+/**
+ * structure <id> { predicate X subject Y target Z applicability { … }
+ * propagation { prop dir … } note "..." source { … } } (TODO.roadmap/19, G3).
+ */
+function parseStructure(id, block) {
+    const s = {
+        id,
+        predicate: '',
+        subject: '',
+        target: '',
+        applicability: [],
+        propagation: [],
+        note: '',
+        source: null,
+    };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'predicate') {
+            s.predicate = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'subject') {
+            s.subject = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'target') {
+            s.target = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'applicability') {
+            s.applicability = (0, field_parser_2.parseApplicability)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'propagation') {
+            const pt = (0, tokenize_1.default)((0, tokenize_2.unwrapBlock)(t[i++]));
+            for (let j = 0; j + 1 < pt.length; j += 2) {
+                s.propagation.push({
+                    property: (0, tokenize_2.stripWrapping)(pt[j]),
+                    direction: (0, tokenize_2.stripWrapping)(pt[j + 1]),
+                });
+            }
+        }
+        else if (cmd === 'note') {
+            s.note = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source') {
+            s.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return s;
+}
+function parseDimension(id, block) {
+    const dim = {
+        id,
+        label: '',
+        scope: '',
+        cardinality: '',
+        labelSeparator: '',
+        description: '',
+        source: null,
+        values: [],
+    };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'label') {
+            dim.label = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'scope') {
+            dim.scope = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'cardinality') {
+            const c = (0, tokenize_2.stripWrapping)(t[i++]);
+            if (c !== 'single' && c !== 'set') {
+                throw new Error(`Parsing error: dimension. ID ${id}: Unknown cardinality ${c} (valid: single, set)`);
+            }
+            dim.cardinality = c;
+        }
+        else if (cmd === 'label_separator') {
+            dim.labelSeparator = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'description') {
+            dim.description = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'reference' || cmd === 'source') {
+            dim.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'values') {
+            dim.values = parseDimensionValues((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return dim;
+}
+function parseDimensionValues(block) {
+    const out = [];
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const id = (0, field_parser_1.stripColon)(t[i++]);
+        if (!id) {
+            break;
+        }
+        const value = {
+            id,
+            label: '',
+            description: '',
+            payload: {},
+            implies: [],
+        };
+        if (i < t.length && t[i].startsWith('{')) {
+            const vblock = (0, tokenize_2.unwrapBlock)(t[i++]);
+            const vt = (0, tokenize_1.default)(vblock);
+            let j = 0;
+            while (j < vt.length) {
+                const cmd = vt[j++];
+                if (j >= vt.length) {
+                    break;
+                }
+                if (cmd === 'label') {
+                    value.label = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (cmd === 'description') {
+                    value.description = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (cmd === 'implies') {
+                    value.implies = readIdList(vt[j++]);
+                }
+                else if (cmd === 'term_ref') {
+                    value.termRef = (0, tokenize_2.stripWrapping)(vt[j++]);
+                }
+                else if (cmd === 'payload') {
+                    const pblock = (0, tokenize_2.unwrapBlock)(vt[j++]);
+                    const pt = (0, tokenize_1.default)(pblock);
+                    let k = 0;
+                    while (k < pt.length) {
+                        const key = (0, field_parser_1.stripColon)(pt[k++]);
+                        if (k >= pt.length) {
+                            break;
+                        }
+                        if (pt[k] === ':') {
+                            k++;
+                        }
+                        if (k < pt.length && pt[k].startsWith('{')) {
+                            // Nested payload block: n_lc_limits { lower: 50000 upper: unlimited }
+                            const nt = (0, tokenize_1.default)((0, tokenize_2.unwrapBlock)(pt[k++]));
+                            const nested = {};
+                            let n = 0;
+                            while (n < nt.length) {
+                                const nkey = (0, field_parser_1.stripColon)(nt[n++]);
+                                if (n >= nt.length) {
+                                    break;
+                                }
+                                if (nt[n] === ':') {
+                                    n++;
+                                }
+                                if (n < nt.length) {
+                                    nested[nkey] = (0, tokenize_2.stripWrapping)(nt[n++]);
+                                }
+                            }
+                            value.payload[key] = nested;
+                        }
+                        else if (k < pt.length) {
+                            value.payload[key] = (0, tokenize_2.stripWrapping)(pt[k++]);
+                        }
+                    }
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(vt[j++]);
+                }
+            }
+        }
+        out.push(value);
+    }
+    return out;
+}
+function parseModelGroup(block) {
+    const mg = {
+        definition: '',
+        identicalCharacteristics: [],
+        identicalAttributes: [],
+        sources: [],
+        sampleSelection: [],
+    };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'definition') {
+            mg.definition = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'identical_characteristics') {
+            mg.identicalCharacteristics = readIdList(t[i++]);
+        }
+        else if (cmd === 'identical_attributes') {
+            mg.identicalAttributes = readIdList(t[i++]);
+        }
+        else if (cmd === 'group_by') {
+            mg.groupBy = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'note') {
+            mg.note = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source') {
+            mg.sources.push((0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++])));
+        }
+        else if (cmd === 'sample_selection') {
+            mg.sampleSelection.push((0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++])));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return mg;
+}
+const dumpInstrument = function (inst) {
+    let out = 'instrument ' + inst.id + ' {\n';
+    if (inst.extends) {
+        out += '  extends ' + inst.extends + '\n';
+    }
+    if (inst.perChannel) {
+        out += '  per_channel ' + inst.perChannel + '\n';
+    }
+    if (inst.measurandKind) {
+        out += '  measurand_kind ' + inst.measurandKind + '\n';
+    }
+    if (inst.measurand) {
+        const m = inst.measurand;
+        out += '  measurand {\n';
+        if (m.kind) {
+            out += '    kind ' + m.kind + '\n';
+        }
+        if (m.description) {
+            out += '    description "' + (0, tokenize_2.escapeString)(m.description) + '"\n';
+        }
+        if (m.context) {
+            out += '    context {';
+            if (m.context.targetObject) {
+                out += ' target_object "' + (0, tokenize_2.escapeString)(m.context.targetObject) + '"';
+            }
+            if (m.context.identificationMethod) {
+                out +=
+                    ' identification_method "' +
+                        (0, tokenize_2.escapeString)(m.context.identificationMethod) +
+                        '"';
+            }
+            out += ' }\n';
+        }
+        out += dumpSource('source', m.source, '    ');
+        out += '  }\n';
+    }
+    for (const c of inst.components ?? []) {
+        let line = '  component ' + c.id + ' { ';
+        if (c.classId) {
+            line += 'class ' + c.classId + ' ';
+        }
+        if (c.definition) {
+            line += 'definition "' + (0, tokenize_2.escapeString)(c.definition) + '" ';
+        }
+        if (c.source && (c.source.doc || c.source.clause)) {
+            line +=
+                'source { doc "' +
+                    (0, tokenize_2.escapeString)(c.source.doc) +
+                    '" clause "' +
+                    (0, tokenize_2.escapeString)(c.source.clause) +
+                    '" } ';
+        }
+        out += line + '}\n';
+    }
+    for (const s of inst.structure ?? []) {
+        out += '  structure ' + s.id + ' {\n';
+        if (s.predicate) {
+            out += '    predicate ' + s.predicate + '\n';
+        }
+        if (s.subject) {
+            out += '    subject ' + s.subject + '\n';
+        }
+        if (s.target) {
+            out += '    target ' + s.target + '\n';
+        }
+        if (s.applicability.length > 0) {
+            out +=
+                '    applicability { ' +
+                    (0, field_parser_2.dumpApplicabilityEntries)(s.applicability).trim() +
+                    ' }\n';
+        }
+        if (s.propagation.length > 0) {
+            out += '    propagation { ';
+            for (const p of s.propagation) {
+                out += p.property + ' ' + p.direction + ' ';
+            }
+            out += '}\n';
+        }
+        if (s.note) {
+            out += '    note "' + (0, tokenize_2.escapeString)(s.note) + '"\n';
+        }
+        out += dumpSource('source', s.source, '    ');
+        out += '  }\n';
+    }
+    if (inst.definition) {
+        out += '  definition "' + (0, tokenize_2.escapeString)(inst.definition) + '"\n';
+    }
+    if (inst.note) {
+        out += '  note "' + (0, tokenize_2.escapeString)(inst.note) + '"\n';
+    }
+    for (const v of inst.variants) {
+        let line = '  variant ' + v.id + ' { ';
+        if (v.name) {
+            line += 'name "' + (0, tokenize_2.escapeString)(v.name) + '" ';
+        }
+        line += 'definition "' + (0, tokenize_2.escapeString)(v.definition) + '" ';
+        if (v.note) {
+            line += 'note "' + (0, tokenize_2.escapeString)(v.note) + '" ';
+        }
+        if (v.source && (v.source.doc || v.source.clause)) {
+            line +=
+                'source { doc "' +
+                    (0, tokenize_2.escapeString)(v.source.doc) +
+                    '" clause "' +
+                    (0, tokenize_2.escapeString)(v.source.clause) +
+                    '" } ';
+        }
+        out += line + '}\n';
+    }
+    for (const d of inst.dimensions) {
+        out += '  dimension ' + d.id + ' {\n';
+        if (d.label) {
+            out += '    label "' + (0, tokenize_2.escapeString)(d.label) + '"\n';
+        }
+        if (d.scope) {
+            out += '    scope ' + d.scope + '\n';
+        }
+        if (d.cardinality) {
+            out += '    cardinality ' + d.cardinality + '\n';
+        }
+        if (d.labelSeparator) {
+            out += '    label_separator "' + (0, tokenize_2.escapeString)(d.labelSeparator) + '"\n';
+        }
+        if (d.description) {
+            out += '    description "' + (0, tokenize_2.escapeString)(d.description) + '"\n';
+        }
+        out += dumpSource('reference', d.source, '    ');
+        if (d.values.length > 0) {
+            out += '    values {\n';
+            for (const v of d.values) {
+                let line = '      ' + v.id;
+                const hasProps = v.label ||
+                    v.description ||
+                    v.termRef ||
+                    Object.keys(v.payload).length > 0 ||
+                    v.implies.length > 0;
+                if (hasProps) {
+                    line += ' { ';
+                    if (v.label) {
+                        line += 'label "' + (0, tokenize_2.escapeString)(v.label) + '" ';
+                    }
+                    if (v.description) {
+                        line += 'description "' + (0, tokenize_2.escapeString)(v.description) + '" ';
+                    }
+                    if (v.termRef) {
+                        line += 'term_ref ' + v.termRef + ' ';
+                    }
+                    if (v.implies.length > 0) {
+                        line += 'implies { ' + v.implies.join(' ') + ' } ';
+                    }
+                    if (Object.keys(v.payload).length > 0) {
+                        line += 'payload { ';
+                        for (const [k, val] of Object.entries(v.payload)) {
+                            if (typeof val === 'object' && val !== null) {
+                                line += k + ' { ';
+                                for (const [nk, nv] of Object.entries(val)) {
+                                    line += nk + ': "' + (0, tokenize_2.escapeString)(String(nv)) + '" ';
+                                }
+                                line += '} ';
+                            }
+                            else {
+                                line += k + ': "' + (0, tokenize_2.escapeString)(String(val)) + '" ';
+                            }
+                        }
+                        line += '} ';
+                    }
+                    line += '}';
+                }
+                out += line + '\n';
+            }
+            out += '    }\n';
+        }
+        out += '  }\n';
+    }
+    if (inst.familyMetamodelClass ||
+        inst.familyDefinition ||
+        inst.familyNote ||
+        (inst.familySource && (inst.familySource.doc || inst.familySource.clause))) {
+        out += '  family {\n';
+        if (inst.familyMetamodelClass) {
+            out += '    metamodel_class ' + inst.familyMetamodelClass + '\n';
+        }
+        if (inst.familyDefinition) {
+            out += '    definition "' + (0, tokenize_2.escapeString)(inst.familyDefinition) + '"\n';
+        }
+        if (inst.familyNote) {
+            out += '    note "' + (0, tokenize_2.escapeString)(inst.familyNote) + '"\n';
+        }
+        out += dumpSource('source', inst.familySource ?? null, '    ');
+        out += '  }\n';
+    }
+    if (inst.familyCriteria.length > 0) {
+        out += '  family_criteria {\n';
+        for (const c of inst.familyCriteria) {
+            out += '    "' + (0, tokenize_2.escapeString)(c) + '"\n';
+        }
+        out += '  }\n';
+    }
+    if (inst.familyDefaultDimensions.length > 0 ||
+        inst.familyDefaultParameters.length > 0) {
+        out += '  family_defaults {\n';
+        out += dumpIdList('dimensions', inst.familyDefaultDimensions, '    ');
+        out += dumpIdList('parameters', inst.familyDefaultParameters, '    ');
+        out += '  }\n';
+    }
+    if (inst.modelGroup) {
+        out += '  model_group {\n';
+        if (inst.modelGroup.definition) {
+            out +=
+                '    definition "' + (0, tokenize_2.escapeString)(inst.modelGroup.definition) + '"\n';
+        }
+        if (inst.modelGroup.groupBy) {
+            out += '    group_by ' + inst.modelGroup.groupBy + '\n';
+        }
+        if (inst.modelGroup.note) {
+            out += '    note "' + (0, tokenize_2.escapeString)(inst.modelGroup.note) + '"\n';
+        }
+        out += dumpIdList('identical_characteristics', inst.modelGroup.identicalCharacteristics, '    ');
+        out += dumpIdList('identical_attributes', inst.modelGroup.identicalAttributes, '    ');
+        for (const s of inst.modelGroup.sources ?? []) {
+            out += dumpSource('source', s, '    ');
+        }
+        for (const s of inst.modelGroup.sampleSelection ?? []) {
+            out += dumpSource('sample_selection', s, '    ');
+        }
+        out += '  }\n';
+    }
+    out += dumpSource('source', inst.source ?? null, '  ');
+    out += dumpIdList('reference', inst.referenceIds, '  ');
+    out += '}\n';
+    return out;
+};
+// ── attribute_definition ─────────────────────────────────────────────
+const parseAttributeDefinition = function (id, data) {
+    const result = {
+        id,
+        symbol: '',
+        name: '',
+        definition: '',
+        source: null,
+        quantityKind: '',
+        unit: '',
+        valueType: '',
+        origin: '',
+        scope: '',
+        category: '',
+        isDimension: null,
+        enumRef: '',
+        irdi: '',
+        derived: '',
+        referenceIds: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'symbol') {
+            result.symbol = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'name') {
+            result.name = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'definition') {
+            result.definition = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source' || cmd === 'reference') {
+            result.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'quantity_kind') {
+            result.quantityKind = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'unit') {
+            result.unit = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'value_type') {
+            result.valueType = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'origin') {
+            result.origin = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'scope') {
+            result.scope = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'category') {
+            result.category = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'is_dimension') {
+            result.isDimension = (0, tokenize_2.stripWrapping)(t[i++]) === 'true';
+        }
+        else if (cmd === 'enum') {
+            result.enumRef = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'enum_values') {
+            result.enumValues = readIdList(t[i++]);
+        }
+        else if (cmd === 'note') {
+            result.note = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'irdi') {
+            result.irdi = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'derived') {
+            result.derived = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.attributeDefinitions[id] = result;
+        return ctx;
+    };
+};
+const dumpAttributeDefinition = function (a) {
+    let out = 'attribute_definition ' + a.id + ' {\n';
+    if (a.symbol) {
+        out += '  symbol "' + (0, tokenize_2.escapeString)(a.symbol) + '"\n';
+    }
+    if (a.name) {
+        out += '  name "' + (0, tokenize_2.escapeString)(a.name) + '"\n';
+    }
+    if (a.definition) {
+        out += '  definition "' + (0, tokenize_2.escapeString)(a.definition) + '"\n';
+    }
+    out += dumpSource('source', a.source, '  ');
+    if (a.quantityKind) {
+        out += '  quantity_kind ' + a.quantityKind + '\n';
+    }
+    if (a.unit) {
+        out += '  unit "' + (0, tokenize_2.escapeString)(a.unit) + '"\n';
+    }
+    if (a.valueType) {
+        out += '  value_type ' + a.valueType + '\n';
+    }
+    if (a.origin) {
+        out += '  origin ' + a.origin + '\n';
+    }
+    if (a.scope) {
+        out += '  scope ' + a.scope + '\n';
+    }
+    if (a.category) {
+        out += '  category ' + a.category + '\n';
+    }
+    if (a.isDimension !== null && a.isDimension !== undefined) {
+        out += '  is_dimension ' + (a.isDimension ? 'true' : 'false') + '\n';
+    }
+    if (a.enumRef) {
+        out += '  enum ' + a.enumRef + '\n';
+    }
+    if (a.irdi) {
+        out += '  irdi "' + (0, tokenize_2.escapeString)(a.irdi) + '"\n';
+    }
+    if (a.derived) {
+        out += '  derived "' + (0, tokenize_2.escapeString)(a.derived) + '"\n';
+    }
+    out += '}\n';
+    return out;
+};
+// ── capability ───────────────────────────────────────────────────────
+const parseCapability = function (id, data) {
+    const result = {
+        id,
+        label: '',
+        description: '',
+        abstract: false,
+        extends: [],
+        requires: [],
+        hasParameters: [],
+        satisfiesRequirements: [],
+        verifiedByTests: [],
+        referenceIds: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'label') {
+            result.label = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'description') {
+            result.description = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'abstract') {
+            result.abstract = (0, tokenize_2.stripWrapping)(t[i++]) === 'true';
+        }
+        else if (cmd === 'extends') {
+            result.extends = readIdList(t[i++]);
+        }
+        else if (cmd === 'requires') {
+            result.requires = readIdList(t[i++]);
+        }
+        else if (cmd === 'has_parameters') {
+            result.hasParameters = readIdList(t[i++]);
+        }
+        else if (cmd === 'satisfies_requirements') {
+            result.satisfiesRequirements = readReference(t[i++]);
+        }
+        else if (cmd === 'verified_by_tests' || cmd === 'verified_by') {
+            result.verifiedByTests = readReference(t[i++]);
+        }
+        else if (cmd === 'reference') {
+            result.referenceIds = readReference(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.capabilities[id] = result;
+        return ctx;
+    };
+};
+const dumpCapability = function (c) {
+    let out = 'capability ' + c.id + ' {\n';
+    if (c.label) {
+        out += '  label "' + (0, tokenize_2.escapeString)(c.label) + '"\n';
+    }
+    if (c.description) {
+        out += '  description "' + (0, tokenize_2.escapeString)(c.description) + '"\n';
+    }
+    if (c.abstract) {
+        out += '  abstract true\n';
+    }
+    out += dumpIdList('extends', c.extends, '  ');
+    out += dumpIdList('requires', c.requires, '  ');
+    out += dumpIdList('has_parameters', c.hasParameters, '  ');
+    out += dumpIdList('satisfies_requirements', c.satisfiesRequirements, '  ');
+    out += dumpIdList('verified_by_tests', c.verifiedByTests, '  ');
+    out += dumpIdList('reference', c.referenceIds, '  ');
+    out += '}\n';
+    return out;
+};
+// ── behavior ─────────────────────────────────────────────────────────
+const parseBehavior = function (id, data) {
+    const result = {
+        id,
+        kind: '',
+        stimulus: '',
+        response: '',
+        source: null,
+        verifiedBy: [],
+        referenceIds: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'kind') {
+            result.kind = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'stimulus') {
+            result.stimulus = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'response') {
+            result.response = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source' || cmd === 'reference') {
+            result.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'verified_by' || cmd === 'verified_by_tests') {
+            result.verifiedBy = readReference(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.behaviors[id] = result;
+        return ctx;
+    };
+};
+const dumpBehavior = function (b) {
+    let out = 'behavior ' + b.id + ' {\n';
+    if (b.kind) {
+        out += '  kind ' + b.kind + '\n';
+    }
+    if (b.stimulus) {
+        out += '  stimulus ' + b.stimulus + '\n';
+    }
+    if (b.response) {
+        out += '  response "' + (0, tokenize_2.escapeString)(b.response) + '"\n';
+    }
+    out += dumpSource('source', b.source, '  ');
+    out += dumpIdList('verified_by', b.verifiedBy, '  ');
+    out += '}\n';
+    return out;
+};
+// ── condition_set ────────────────────────────────────────────────────
+const parseConditionSet = function (id, data) {
+    const result = {
+        id,
+        role: '',
+        entries: [],
+        source: null,
+        sources: [],
+        referenceIds: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'role') {
+            result.role = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'subject') {
+            result.subject = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'description') {
+            result.description = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'entries') {
+            result.entries = parseConditionEntries((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'source') {
+            // Repeated source blocks accumulate; `source` stays the first entry.
+            const src = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+            result.sources.push(src);
+            if (!result.source) {
+                result.source = src;
+            }
+        }
+        else if (cmd === 'reference') {
+            result.referenceIds = readReference(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.conditionSets[id] = result;
+        return ctx;
+    };
+};
+function parseConditionEntries(block) {
+    const out = [];
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const qk = (0, field_parser_1.stripColon)(t[i++]);
+        if (!qk) {
+            break;
+        }
+        const entry = {
+            quantityKind: qk,
+            value: '',
+            unit: '',
+            tolerance: '',
+        };
+        if (i < t.length && t[i].startsWith('{')) {
+            const eblock = (0, tokenize_2.unwrapBlock)(t[i++]);
+            const et = (0, tokenize_1.default)(eblock);
+            let j = 0;
+            while (j < et.length) {
+                const cmd = et[j++];
+                if (j >= et.length) {
+                    break;
+                }
+                if (cmd === 'value') {
+                    entry.value = (0, tokenize_2.stripWrapping)(et[j++]);
+                }
+                else if (cmd === 'unit') {
+                    entry.unit = (0, tokenize_2.stripWrapping)(et[j++]);
+                }
+                else if (cmd === 'tolerance') {
+                    entry.tolerance = (0, tokenize_2.stripWrapping)(et[j++]);
+                }
+                else if (cmd === 'note') {
+                    entry.note = (0, tokenize_2.stripWrapping)(et[j++]);
+                }
+                else {
+                    (0, tokenize_2.unwrapBlock)(et[j++]);
+                }
+            }
+        }
+        out.push(entry);
+    }
+    return out;
+}
+const dumpConditionSet = function (cs) {
+    let out = 'condition_set ' + cs.id + ' {\n';
+    if (cs.role) {
+        out += '  role ' + cs.role + '\n';
+    }
+    if (cs.subject) {
+        out += '  subject ' + cs.subject + '\n';
+    }
+    if (cs.description) {
+        out += '  description "' + (0, tokenize_2.escapeString)(cs.description) + '"\n';
+    }
+    if (cs.entries.length > 0) {
+        out += '  entries {\n';
+        for (const e of cs.entries) {
+            // Entry values may be free text ("nominal, per installation
+            // conditions") — quote whenever not a single safe token.
+            let line = '    ' + e.quantityKind + ' { value ' + (0, field_parser_1.dumpBareSafe)(e.value);
+            if (e.unit) {
+                line += ' unit "' + (0, tokenize_2.escapeString)(e.unit) + '"';
+            }
+            if (e.tolerance) {
+                line += ' tolerance ' + (0, field_parser_1.dumpBareSafe)(e.tolerance);
+            }
+            if (e.note) {
+                line += ' note "' + (0, tokenize_2.escapeString)(e.note) + '"';
+            }
+            out += line + ' }\n';
+        }
+        out += '  }\n';
+    }
+    const sources = cs.sources && cs.sources.length > 0
+        ? cs.sources
+        : cs.source
+            ? [cs.source]
+            : [];
+    for (const s of sources) {
+        out += dumpSource('source', s, '  ');
+    }
+    out += dumpIdList('reference', cs.referenceIds, '  ');
+    out += '}\n';
+    return out;
+};
+// ── subject (v3 anatomy: is / has / does) ────────────────────────────
+/** Read `key value` / `key: value` pairs into a string map. */
+function readStringMap(block) {
+    const out = {};
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const key = (0, field_parser_1.stripColon)(t[i++]);
+        if (!key) {
+            break;
+        }
+        if (t[i] === ':') {
+            i++;
+        }
+        if (i < t.length) {
+            out[key] = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+    }
+    return out;
+}
+/**
+ * Read `<id> : <qualifier tokens…>` entries (design_parameters / attributes).
+ * The qualifier runs to the next `<id> :` head or the end of the block;
+ * the stored value is the tokens joined with single spaces. The dump
+ * re-emits the joined form, so the fixpoint holds.
+ */
+function readValueMap(block) {
+    const out = {};
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const key = (0, field_parser_1.stripColon)(t[i++]);
+        if (!key) {
+            break;
+        }
+        if (t[i] === ':') {
+            i++;
+        }
+        const parts = [];
+        while (i < t.length && t[i + 1] !== ':') {
+            parts.push((0, tokenize_2.stripWrapping)(t[i++]));
+        }
+        out[key] = parts.join(' ');
+    }
+    return out;
+}
+/** Read whitespace-separated entries (quoted phrases or bare ids). */
+function readStringEntries(block) {
+    return (0, tokenize_1.default)(block)
+        .map(tokenize_2.stripWrapping)
+        .filter(s => s.length > 0);
+}
+/** Read `<id> in { A, B }` dimension-membership entries. */
+function readDimensionMap(block) {
+    const out = {};
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const key = (0, field_parser_1.stripColon)(t[i++]);
+        if (!key) {
+            break;
+        }
+        if (t[i] === ':') {
+            i++;
+        }
+        if (t[i] === 'in') {
+            i++;
+        }
+        if (i < t.length) {
+            const raw = t[i++];
+            const inner = raw.startsWith('{') ? (0, tokenize_2.unwrapBlock)(raw) : (0, tokenize_2.stripWrapping)(raw);
+            out[key] = inner
+                .split(/[,\s]+/)
+                .map(s => s.trim())
+                .filter(s => s.length > 0);
+        }
+    }
+    return out;
+}
+/**
+ * Read `name [symbol] = derivation` characteristic entries. The
+ * derivation is an ocl{…} expression (reassembled across whitespace by
+ * readValueToken) or a quoted phrase; an absent derivation parses as ''
+ * and is flagged by the linter (C7).
+ *
+ * The `=` separator may arrive FUSED to the name or symbol token
+ * (`creep c_c= ocl{…}`, `drift= ocl{…}`) — the tokenizer splits on
+ * whitespace only, so a trailing `=` on either token is treated as the
+ * separator and the derivation is read from the next token.
+ *
+ * Block form (TODO.roadmap/10): an entry whose name is followed by a
+ * `{ … }` block carries the full register entry — symbol, derivation,
+ * behavior (the behavior it quantifies), quantity_kind, unit, source:
+ *   mdlo_normalized {
+ *     symbol "C_M"
+ *     derivation ocl{abs(c_m * t_f / delta_t * (d_max - d_min) / (n * v_min))}
+ *     behavior temp-effect-min-dead-load
+ *     quantity_kind dimensionless
+ *     source { doc "urn:oiml:pub:r:60-3:2021" clause "2.1.4" }
+ *   }
+ */
+function readCharacteristics(block) {
+    const out = {};
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        let name = (0, field_parser_1.stripColon)(t[i++]);
+        if (!name) {
+            break;
+        }
+        if (i < t.length && t[i].startsWith('{')) {
+            out[name] = readCharacteristicBlock((0, tokenize_2.unwrapBlock)(t[i++]));
+            continue;
+        }
+        let symbol = '';
+        let derivation = '';
+        let fused = false;
+        if (name.length > 1 && name.endsWith('=')) {
+            name = name.slice(0, -1);
+            fused = true;
+        }
+        if (!fused && i < t.length && t[i] !== '=') {
+            // t[i] !== '=' here, so a trailing '=' on the symbol token is fused.
+            fused = t[i].endsWith('=');
+            symbol = fused ? t[i].slice(0, -1) : t[i];
+            i++;
+        }
+        if (fused || t[i] === '=') {
+            if (!fused) {
+                i++;
+            }
+            const read = (0, field_parser_1.readValueToken)(t, i);
+            derivation = (0, tokenize_2.stripWrapping)(read.text);
+            i = read.next;
+        }
+        out[name] = { symbol, derivation };
+    }
+    return out;
+}
+/** Read one block-form characteristic entry (TODO.roadmap/10). */
+function readCharacteristicBlock(block) {
+    const c = { symbol: '', derivation: '' };
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'symbol') {
+            c.symbol = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'derivation' || cmd === 'derive') {
+            // ocl{…} expressions reassemble across whitespace (readValueToken).
+            const read = (0, field_parser_1.readValueToken)(t, i);
+            c.derivation = (0, tokenize_2.stripWrapping)(read.text);
+            i = read.next;
+        }
+        else if (cmd === 'behavior') {
+            c.behavior = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'quantity_kind') {
+            c.quantityKind = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'unit') {
+            c.unit = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'source' || cmd === 'reference') {
+            c.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return c;
+}
+/**
+ * Aspects whose payload is a scalar (not a `{ … }` block). When such an
+ * aspect is misplaced, its value token must be skipped together with the
+ * keyword — otherwise the value is re-read as another (phantom) aspect.
+ */
+const SCALAR_ASPECTS = new Set(['state', 'behavior']);
+/**
+ * Misplaced-payload skipping for the twin aspects (TODO.roadmap/32):
+ *   endpoint <id> { … }                      — skip the id AND the block;
+ *   serve <aspect> via <op> [{ … }]          — skip aspect/via/op + block.
+ * Without the skip, a misplaced endpoint/serve cascades into phantom
+ * aspects (the id, `via`, the operation name would each re-read as an
+ * aspect keyword).
+ */
+function skipMisplacedPayload(aspect, t, i) {
+    if (aspect === 'endpoint') {
+        if (i < t.length && !t[i].startsWith('{')) {
+            i++; // the endpoint id
+        }
+        if (i < t.length && t[i].startsWith('{')) {
+            i++; // the endpoint body
+        }
+        return i;
+    }
+    if (aspect === 'serve') {
+        if (i < t.length) {
+            i++; // the aspect path
+        }
+        if (t[i] === 'via') {
+            i += 2; // `via` + the operation name
+        }
+        if (i < t.length && t[i].startsWith('{')) {
+            i++; // the freshness block
+        }
+        return i;
+    }
+    return i;
+}
+/**
+ * Record an aspect key found under the wrong family (or undeclared) for
+ * the linter (C6), and skip its payload: one block token, or one scalar
+ * value token for the known-scalar aspects.
+ */
+function recordMisplaced(result, family, cmd, t, i) {
+    const aspect = (0, field_parser_1.stripColon)(cmd);
+    result.misplacedAspects.push({ family, aspect });
+    if (i < t.length && t[i].startsWith('{')) {
+        i++;
+    }
+    else if (SCALAR_ASPECTS.has(aspect)) {
+        i++;
+    }
+    else {
+        i = skipMisplacedPayload(aspect, t, i);
+    }
+    return i;
+}
+/**
+ * Read one promise `level` value (TODO.roadmap/08). Three shapes:
+ *   level symbolic C6                        → { kind: 'symbolic', symbolic }
+ *   level range { min -10 max 40 unit degC } → { kind: 'range', … }
+ *   level { value 0.7 unit v }               → { kind: 'quantity', quantity }
+ * Returns the level and the next unconsumed token index.
+ */
+function readPromiseLevel(t, i) {
+    const tok = t[i] ?? '';
+    if (tok === 'symbolic') {
+        return {
+            level: { kind: 'symbolic', symbolic: (0, tokenize_2.stripWrapping)(t[i + 1] ?? '') },
+            next: i + 2,
+        };
+    }
+    if (tok === 'range') {
+        const level = { kind: 'range' };
+        const inner = (0, tokenize_1.default)((0, tokenize_2.unwrapBlock)(t[i + 1] ?? ''));
+        let j = 0;
+        while (j < inner.length) {
+            const cmd = inner[j++];
+            if (j > inner.length) {
+                break;
+            }
+            if (cmd === 'min') {
+                level.min = (0, quantity_1.coerceValueToken)(inner[j++]);
+            }
+            else if (cmd === 'max') {
+                level.max = (0, quantity_1.coerceValueToken)(inner[j++]);
+            }
+            else if (cmd === 'unit') {
+                level.unit = (0, tokenize_2.stripWrapping)(inner[j++]);
+            }
+            else {
+                (0, tokenize_2.unwrapBlock)(inner[j++]);
+            }
+        }
+        return { level, next: i + 2 };
+    }
+    // QuantityValue block form (the full INV-1 contract).
+    return {
+        level: { kind: 'quantity', quantity: (0, quantity_1.readQuantityBlock)((0, tokenize_2.unwrapBlock)(tok)) },
+        next: i + 1,
+    };
+}
+/**
+ * Read is.promises entries (TODO.roadmap/08). Two entry forms:
+ *   - rich:  `<id> { target … level … conditions … statement …
+ *              verified_by { … } source { … } }`
+ *   - shorthand: a quoted phrase or a bare token NOT followed by a block —
+ *     the legacy string-list form, parsed as a statement-only promise
+ *     (empty id/target/conditions; the linter's C43 flags it as
+ *     unverifiable at authoring).
+ */
+function readPromises(block) {
+    const out = [];
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const tok = t[i++];
+        if (!tok) {
+            break;
+        }
+        if (tok.startsWith('"') || i >= t.length || !t[i].startsWith('{')) {
+            out.push({
+                id: '',
+                target: '',
+                level: null,
+                conditions: '',
+                statement: (0, tokenize_2.stripWrapping)(tok),
+                verifiedBy: [],
+                source: null,
+            });
+            continue;
+        }
+        const p = {
+            id: (0, field_parser_1.stripColon)(tok),
+            target: '',
+            level: null,
+            conditions: '',
+            statement: '',
+            verifiedBy: [],
+            source: null,
+        };
+        const inner = (0, tokenize_1.default)((0, tokenize_2.unwrapBlock)(t[i++]));
+        let j = 0;
+        while (j < inner.length) {
+            const cmd = inner[j++];
+            if (j >= inner.length) {
+                break;
+            }
+            if (cmd === 'target') {
+                p.target = (0, tokenize_2.stripWrapping)(inner[j++]);
+            }
+            else if (cmd === 'level') {
+                const read = readPromiseLevel(inner, j);
+                p.level = read.level;
+                j = read.next;
+            }
+            else if (cmd === 'conditions') {
+                // ocl{…} expressions reassemble across whitespace (readValueToken).
+                const read = (0, field_parser_1.readValueToken)(inner, j);
+                p.conditions = (0, tokenize_2.stripWrapping)(read.text);
+                j = read.next;
+            }
+            else if (cmd === 'statement') {
+                p.statement = (0, tokenize_2.stripWrapping)(inner[j++]);
+            }
+            else if (cmd === 'verified_by') {
+                p.verifiedBy = readIdList(inner[j++]);
+            }
+            else if (cmd === 'source') {
+                p.source = (0, field_parser_1.readSource)((0, tokenize_2.unwrapBlock)(inner[j++]));
+            }
+            else {
+                (0, tokenize_2.unwrapBlock)(inner[j++]);
+            }
+        }
+        out.push(p);
+    }
+    return out;
+}
+/** A promise carrying nothing but its prose statement (the shorthand form). */
+function isStatementOnlyPromise(p) {
+    return (p.id === '' &&
+        !p.target &&
+        !p.level &&
+        !p.conditions &&
+        p.verifiedBy.length === 0 &&
+        !p.source);
+}
+function parseSubjectIs(block, result) {
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'metadata') {
+            result.is.metadata = readStringMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'provenance') {
+            result.is.provenance = readStringMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'structure') {
+            result.is.structure = readStringEntries((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'design_parameters') {
+            result.is.designParameters = readValueMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'designed_conditions') {
+            result.is.designedConditions = readStringMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'promises') {
+            result.is.promises = readPromises((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'artifacts') {
+            result.is.artifacts = readStringEntries((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'endpoint') {
+            // endpoint <id> { … } (TODO.roadmap/32) — the declared API surface.
+            const endpointId = (0, field_parser_1.stripColon)(t[i++] ?? '');
+            const body = i < t.length ? (0, tokenize_2.unwrapBlock)(t[i++]) : '';
+            if (endpointId) {
+                result.is.endpoints.push((0, twin_1.parseEndpoint)(endpointId, body));
+            }
+        }
+        else if (cmd === 'composed_of') {
+            // composed_of { … } (TODO.integration/14) — the composition facet
+            // of a composite subject.
+            result.is.composedOf = (0, composition_1.parseComposedOf)((0, tokenize_2.unwrapBlock)(t[i++] ?? ''));
+        }
+        else {
+            i = recordMisplaced(result, 'is', cmd, t, i);
+        }
+    }
+}
+function parseSubjectHas(block, result) {
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'attributes') {
+            result.has.attributes = readValueMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'dimensions') {
+            result.has.dimensions = readDimensionMap((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'state') {
+            result.has.state = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'characteristics') {
+            result.has.characteristics = readCharacteristics((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'environmental_context') {
+            result.has.environmentalContext = readStringEntries((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'artifact_instances') {
+            result.has.artifactInstances = readStringEntries((0, tokenize_2.unwrapBlock)(t[i++]));
+        }
+        else if (cmd === 'serve') {
+            // serve <aspect> via <operation> [{ fresh_within <duration> }]
+            // (TODO.roadmap/32) — the HAS-level binding with freshness semantics.
+            const read = (0, twin_1.parseServeEntry)(t, i);
+            result.has.serves.push(read.binding);
+            i = read.next;
+        }
+        else {
+            i = recordMisplaced(result, 'has', cmd, t, i);
+        }
+    }
+}
+function parseSubjectDoes(block, result) {
+    const t = (0, tokenize_1.default)(block);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'behavior') {
+            result.does.behaviors.push((0, tokenize_2.stripWrapping)(t[i++]));
+        }
+        else {
+            i = recordMisplaced(result, 'does', cmd, t, i);
+        }
+    }
+}
+const parseSubject = function (id, data) {
+    const result = {
+        id,
+        extends: '',
+        is: {
+            metadata: {},
+            provenance: {},
+            structure: [],
+            designParameters: {},
+            designedConditions: {},
+            promises: [],
+            artifacts: [],
+            endpoints: [],
+        },
+        has: {
+            attributes: {},
+            dimensions: {},
+            state: '',
+            characteristics: {},
+            environmentalContext: [],
+            artifactInstances: [],
+            serves: [],
+        },
+        does: { behaviors: [] },
+        referenceIds: [],
+        misplacedAspects: [],
+    };
+    const t = (0, tokenize_2.tokenizePackage)(data);
+    let i = 0;
+    while (i < t.length) {
+        const cmd = t[i++];
+        if (i >= t.length) {
+            break;
+        }
+        if (cmd === 'extends') {
+            result.extends = (0, tokenize_2.stripWrapping)(t[i++]);
+        }
+        else if (cmd === 'is') {
+            parseSubjectIs((0, tokenize_2.unwrapBlock)(t[i++]), result);
+        }
+        else if (cmd === 'has') {
+            parseSubjectHas((0, tokenize_2.unwrapBlock)(t[i++]), result);
+        }
+        else if (cmd === 'does') {
+            parseSubjectDoes((0, tokenize_2.unwrapBlock)(t[i++]), result);
+        }
+        else if (cmd === 'reference') {
+            result.referenceIds = readReference(t[i++]);
+        }
+        else {
+            (0, tokenize_2.unwrapBlock)(t[i++]);
+        }
+    }
+    return ctx => {
+        ctx.subjects[id] = result;
+        return ctx;
+    };
+};
+// ── subject extends resolution ───────────────────────────────────────
+/**
+ * Merge PARENT into CHILD per aspect-kind rules (TODO.roadmap/01): maps
+ * merge by key (child wins on conflict), lists append (parent entries
+ * first, then child's), scalars override (child's non-empty value wins).
+ * referenceIds and misplacedAspects are NOT merged: references stay the
+ * declaring subject's own, and misplaced aspects are parse-time lint
+ * capture the linter already sees on the parent itself.
+ */
+function mergeSubject(parent, child) {
+    return {
+        id: child.id,
+        // The extends link is consumed by the merge — the resolved subject is
+        // self-contained, so a dump→load cycle can never re-merge (list
+        // aspects would duplicate). Unmerged links keep `extends` (below).
+        extends: '',
+        is: {
+            metadata: { ...parent.is.metadata, ...child.is.metadata },
+            provenance: { ...parent.is.provenance, ...child.is.provenance },
+            structure: [...parent.is.structure, ...child.is.structure],
+            designParameters: {
+                ...parent.is.designParameters,
+                ...child.is.designParameters,
+            },
+            designedConditions: {
+                ...parent.is.designedConditions,
+                ...child.is.designedConditions,
+            },
+            promises: [...parent.is.promises, ...child.is.promises],
+            artifacts: [...parent.is.artifacts, ...child.is.artifacts],
+            endpoints: [...parent.is.endpoints, ...child.is.endpoints],
+            // The composition facet: the child's declaration wins when present
+            // (a composite refines its composition; the parent's carries
+            // through unchanged otherwise).
+            ...(child.is.composedOf
+                ? { composedOf: child.is.composedOf }
+                : parent.is.composedOf
+                    ? { composedOf: parent.is.composedOf }
+                    : {}),
+        },
+        has: {
+            attributes: { ...parent.has.attributes, ...child.has.attributes },
+            dimensions: { ...parent.has.dimensions, ...child.has.dimensions },
+            state: child.has.state || parent.has.state,
+            characteristics: {
+                ...parent.has.characteristics,
+                ...child.has.characteristics,
+            },
+            environmentalContext: [
+                ...parent.has.environmentalContext,
+                ...child.has.environmentalContext,
+            ],
+            artifactInstances: [
+                ...parent.has.artifactInstances,
+                ...child.has.artifactInstances,
+            ],
+            serves: [...parent.has.serves, ...child.has.serves],
+        },
+        does: { behaviors: [...parent.does.behaviors, ...child.does.behaviors] },
+        referenceIds: child.referenceIds,
+        misplacedAspects: child.misplacedAspects,
+    };
+}
+/**
+ * Resolve a subject's `extends` chain: merge each subject with its
+ * parent from ctx.subjects, recursively when the parent itself extends.
+ * Pure — parents are read from the raw parse table and never mutated
+ * (several subjects may extend the same parent). Lenient links:
+ *   - missing parent → the subject is returned unmerged (the lint rule
+ *     subject-extends-resolves reports it);
+ *   - cycle → the visited set breaks the cyclic LINK (treated like a
+ *     missing parent); acyclic prefixes still merge, so resolution
+ *     always terminates.
+ */
+function resolveSubject(ctx, item) {
+    const chain = (s, visited) => {
+        const parentId = s.extends;
+        if (!parentId || visited.has(parentId)) {
+            return s;
+        }
+        const parent = ctx.subjects[parentId];
+        if (!parent) {
+            return s;
+        }
+        visited.add(parentId);
+        return mergeSubject(chain(parent, visited), s);
+    };
+    return chain(item, new Set([item.id]));
+}
+// ── subject dumpers ──────────────────────────────────────────────────
+function dumpSubjectStringMap(keyword, map) {
+    const keys = Object.keys(map);
+    if (keys.length === 0) {
+        return '';
+    }
+    return ('    ' +
+        keyword +
+        ' { ' +
+        keys.map(k => k + ' ' + (0, field_parser_1.dumpBareSafe)(map[k])).join(' ') +
+        ' }\n');
+}
+function dumpSubjectValueMap(keyword, map) {
+    const keys = Object.keys(map);
+    if (keys.length === 0) {
+        return '';
+    }
+    return ('    ' +
+        keyword +
+        ' { ' +
+        // Values are free-form qualifiers — route through dumpBareSafe so
+        // quotes/braces/whitespace can't break re-parse (same as
+        // dumpSubjectStringMap).
+        keys
+            .map(k => k + ' :' + (map[k] ? ' ' + (0, field_parser_1.dumpBareSafe)(map[k]) : ''))
+            .join(' ') +
+        ' }\n');
+}
+function dumpSubjectEntries(keyword, entries) {
+    if (entries.length === 0) {
+        return '';
+    }
+    return ('    ' + keyword + ' { ' + entries.map(field_parser_1.dumpBareSafe).join(' ') + ' }\n');
+}
+function dumpSubjectDimensions(map) {
+    const keys = Object.keys(map);
+    if (keys.length === 0) {
+        return '';
+    }
+    return ('    dimensions { ' +
+        keys.map(k => k + ' in { ' + map[k].join(', ') + ' }').join(' ') +
+        ' }\n');
+}
+/** Derivations emit bare when ocl{…}/single-token, quoted otherwise. */
+function dumpSubjectDerivation(d) {
+    return /^ocl\{[\s\S]*\}$/.test(d) ? d : (0, field_parser_1.dumpBareSafe)(d);
+}
+/** Dump one promise level (TODO.roadmap/08) — the inverse of readPromiseLevel. */
+function dumpPromiseLevel(level) {
+    if (level.kind === 'symbolic') {
+        return 'symbolic ' + (0, field_parser_1.dumpBareSafe)(level.symbolic ?? '');
+    }
+    if (level.kind === 'range') {
+        let inner = '';
+        if (level.min !== undefined) {
+            inner += 'min ' + (0, quantity_1.dumpScalarToken)(level.min) + ' ';
+        }
+        if (level.max !== undefined) {
+            inner += 'max ' + (0, quantity_1.dumpScalarToken)(level.max) + ' ';
+        }
+        if (level.unit) {
+            inner += 'unit "' + (0, tokenize_2.escapeString)(level.unit) + '"';
+        }
+        return 'range { ' + inner.trimEnd() + ' }';
+    }
+    return (0, quantity_1.dumpQuantityBlock)(level.quantity ?? { value: '' });
+}
+/**
+ * Dump is.promises (TODO.roadmap/08). Byte-compat with the legacy string
+ * list: a block of statement-only entries keeps the single-line form
+ * (`promises { "a" "b" }`); any rich entry switches the block to the
+ * multi-line form, one entry per `<id> { … }` block (statement-only
+ * entries stay quoted strings on their own lines).
+ */
+function dumpSubjectPromises(promises) {
+    if (promises.length === 0) {
+        return '';
+    }
+    if (promises.every(isStatementOnlyPromise)) {
+        return ('    promises { ' +
+            promises.map(p => (0, field_parser_1.dumpBareSafe)(p.statement)).join(' ') +
+            ' }\n');
+    }
+    let out = '    promises {\n';
+    for (const p of promises) {
+        if (isStatementOnlyPromise(p)) {
+            out += '      ' + (0, field_parser_1.dumpBareSafe)(p.statement) + '\n';
+            continue;
+        }
+        out += '      ' + (0, field_parser_1.dumpBareSafe)(p.id) + ' {\n';
+        if (p.target) {
+            out += '        target ' + (0, field_parser_1.dumpBareSafe)(p.target) + '\n';
+        }
+        if (p.level) {
+            out += '        level ' + dumpPromiseLevel(p.level) + '\n';
+        }
+        if (p.conditions) {
+            out += '        conditions ' + dumpSubjectDerivation(p.conditions) + '\n';
+        }
+        if (p.statement) {
+            out += '        statement "' + (0, tokenize_2.escapeString)(p.statement) + '"\n';
+        }
+        out += dumpIdList('verified_by', p.verifiedBy, '        ');
+        out += dumpSource('source', p.source ?? null, '        ');
+        out += '      }\n';
+    }
+    out += '    }\n';
+    return out;
+}
+function dumpSubjectCharacteristics(map) {
+    const keys = Object.keys(map);
+    if (keys.length === 0) {
+        return '';
+    }
+    let out = '    characteristics {\n';
+    for (const k of keys) {
+        const c = map[k];
+        const rich = c.behavior ||
+            c.quantityKind ||
+            c.unit ||
+            (c.source && (c.source.doc || c.source.clause));
+        if (rich) {
+            // Block form (TODO.roadmap/10) — the full register entry.
+            out += '      ' + k + ' {\n';
+            if (c.symbol) {
+                out += '        symbol "' + (0, tokenize_2.escapeString)(c.symbol) + '"\n';
+            }
+            if (c.derivation) {
+                out +=
+                    '        derivation ' + dumpSubjectDerivation(c.derivation) + '\n';
+            }
+            if (c.behavior) {
+                out += '        behavior ' + (0, field_parser_1.dumpBareSafe)(c.behavior) + '\n';
+            }
+            if (c.quantityKind) {
+                out += '        quantity_kind ' + (0, field_parser_1.dumpBareSafe)(c.quantityKind) + '\n';
+            }
+            if (c.unit) {
+                out += '        unit "' + (0, tokenize_2.escapeString)(c.unit) + '"\n';
+            }
+            out += dumpSource('source', c.source ?? null, '        ');
+            out += '      }\n';
+            continue;
+        }
+        out += '      ' + k + (c.symbol ? ' ' + c.symbol : '');
+        if (c.derivation) {
+            out += ' = ' + dumpSubjectDerivation(c.derivation);
+        }
+        out += '\n';
+    }
+    out += '    }\n';
+    return out;
+}
+/** Dump is.endpoints (TODO.roadmap/32) — one `endpoint <id> { … }` block per
+ *  declared endpoint, after the other IS aspects. */
+function dumpSubjectEndpoints(endpoints) {
+    let out = '';
+    for (const e of endpoints) {
+        out += (0, twin_1.dumpEndpoint)(e, '    ');
+    }
+    return out;
+}
+/** Dump has.serves (TODO.roadmap/32) — one `serve … via …` line per binding. */
+function dumpSubjectServes(serves) {
+    let out = '';
+    for (const b of serves) {
+        out += (0, twin_1.dumpServe)(b, '    ');
+    }
+    return out;
+}
+const dumpSubject = function (s) {
+    let out = 'subject ' + s.id + ' {\n';
+    if (s.extends) {
+        out += '  extends ' + s.extends + '\n';
+    }
+    const isBody = dumpSubjectStringMap('metadata', s.is.metadata) +
+        dumpSubjectStringMap('provenance', s.is.provenance) +
+        dumpSubjectEntries('structure', s.is.structure) +
+        dumpSubjectValueMap('design_parameters', s.is.designParameters) +
+        dumpSubjectStringMap('designed_conditions', s.is.designedConditions) +
+        dumpSubjectPromises(s.is.promises) +
+        dumpSubjectEntries('artifacts', s.is.artifacts) +
+        dumpSubjectEndpoints(s.is.endpoints) +
+        (s.is.composedOf ? (0, composition_1.dumpComposedOf)(s.is.composedOf, '    ') : '');
+    if (isBody) {
+        out += '  is {\n' + isBody + '  }\n';
+    }
+    const hasBody = dumpSubjectValueMap('attributes', s.has.attributes) +
+        dumpSubjectDimensions(s.has.dimensions) +
+        (s.has.state ? '    state ' + (0, field_parser_1.dumpBareSafe)(s.has.state) + '\n' : '') +
+        dumpSubjectCharacteristics(s.has.characteristics) +
+        dumpSubjectEntries('environmental_context', s.has.environmentalContext) +
+        dumpSubjectEntries('artifact_instances', s.has.artifactInstances) +
+        dumpSubjectServes(s.has.serves);
+    if (hasBody) {
+        out += '  has {\n' + hasBody + '  }\n';
+    }
+    let doesBody = '';
+    for (const b of s.does.behaviors) {
+        doesBody += '    behavior ' + (0, field_parser_1.dumpBareSafe)(b) + '\n';
+    }
+    if (doesBody) {
+        out += '  does {\n' + doesBody + '  }\n';
+    }
+    out += dumpIdList('reference', s.referenceIds, '  ');
+    out += '}\n';
+    return out;
+};
+// ── construct registry entries ───────────────────────────────────────
+exports.instrumentConstruct = {
+    keyword: 'instrument',
+    field: 'instruments',
+    takesID: true,
+    parse: parseInstrument,
+    dump: dumpInstrument,
+};
+exports.attributeDefinitionConstruct = {
+    keyword: 'attribute_definition',
+    field: 'attributeDefinitions',
+    takesID: true,
+    parse: parseAttributeDefinition,
+    dump: dumpAttributeDefinition,
+};
+exports.capabilityConstruct = {
+    keyword: 'capability',
+    field: 'capabilities',
+    takesID: true,
+    parse: parseCapability,
+    dump: dumpCapability,
+};
+exports.behaviorConstruct = {
+    keyword: 'behavior',
+    field: 'behaviors',
+    takesID: true,
+    parse: parseBehavior,
+    dump: dumpBehavior,
+};
+exports.conditionSetConstruct = {
+    keyword: 'condition_set',
+    field: 'conditionSets',
+    takesID: true,
+    parse: parseConditionSet,
+    dump: dumpConditionSet,
+};
+exports.subjectConstruct = {
+    keyword: 'subject',
+    field: 'subjects',
+    takesID: true,
+    parse: parseSubject,
+    resolve: resolveSubject,
+    dump: dumpSubject,
+};
+//# sourceMappingURL=subject.js.map