@larsgw/formica 0.10.2 → 0.10.4

package/README.md

@@ -20,6 +20,8 @@ Run locally:
   - `loir-validate-resources [./B1234.txt]`:
     Validate the `.txt` files containing information on the taxonomic scope of keys
     ([documentation](https://github.com/identification-resources/catalog/blob/main/docs/resources-txt.md)).
+  - `loir-validate-integrity .`:
+    Check if all values are mapped as expected.
   - `loir-resources-process [./resources]`:
     Interactive tool to convert the aforementioned `.txt` files to Darwin Core archives
     ([documentation](https://github.com/identification-resources/catalog/blob/main/docs/resources-dwc.md)),

package/lib/bin/generate-linked-data.js

@@ -69,6 +69,9 @@ const SCOPES = {
     'nypmhs': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
     'nymphs (instar V)': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
     'nymphs (instar IV)': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
+    'nymphs (instar III)': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
+    'nymphs (instar II)': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
+    'nymphs (instar I)': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/larva'],
     'eggs': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/embryo'],
     // plant life stage
     'flowering plants': ['dwciri:lifeStage', 'http://rs.gbif.org/vocabulary/gbif/life_stage/adult'],
@@ -87,6 +90,9 @@ const SCOPES = {
     'alatae': ['dwc:caste', 'alate'],
     'apterae': ['dwc:caste', 'aptera'],
     'viviparae': ['dwc:caste', 'vivipara'],
+    'brachypterae': ['dwc:caste', 'brachypterae'],
+    'macropterae': ['dwc:caste', 'macropterae'],
+    'micropterae': ['dwc:caste', 'micropterae'],
     // evidence
     'nests': ['ac:subjectPartLiteral', 'nest'],
     'galls': ['ac:subjectPartLiteral', 'gall'],
@@ -96,6 +102,7 @@ const SCOPES = {
     'bones (skulls)': ['ac:subjectPart', 'http://rs.tdwg.org/acpart/values/p0027'],
     'bones (upper jaws)': ['ac:subjectPart', 'http://rs.tdwg.org/acpart/values/p0028'],
     'bones (lower jaws)': ['ac:subjectPart', 'http://rs.tdwg.org/acpart/values/p0029'],
+    'spermathecae': ['ac:subjectPartLiteral', 'spermathecae'],
 };
 const PREFIXES = {
     'ac': 'http://rs.tdwg.org/ac/terms/',
@@ -129,15 +136,18 @@ const STATUSES = {
     'proparte synonym': 'http://rs.gbif.org/vocabulary/gbif/taxonomicStatus/proParteSynonym',
     'synonym': 'http://rs.gbif.org/vocabulary/gbif/taxonomicStatus/synonym',
 };
-function getCoveringTaxon(taxa) {
-    if (taxa.length === 1 && !taxa[0].has('gbif')) {
+function getCoveringTaxon(taxa, allTaxa) {
+    if (taxa.length === 1 && !taxa[0].has('col')) {
         return `${PREFIX}taxon/${taxa[0].get('id')}`;
     }
     function getAncestors(taxon) {
         var _a;
-        const [...parents] = ((_a = taxon.get('parent_taxa')) !== null && _a !== void 0 ? _a : []);
-        if (taxon.has('gbif')) {
-            parents.push(taxon.get('gbif'));
+        const [...parents] = ((_a = taxon.get('ancestors_col')) !== null && _a !== void 0 ? _a : []);
+        if (taxon.has('accepted_col')) {
+            parents.push(taxon.get('accepted_col'));
+        }
+        else if (taxon.has('col')) {
+            parents.push(taxon.get('col'));
         }
         return parents;
     }
@@ -150,7 +160,15 @@ function getCoveringTaxon(taxa) {
             }
         }
     }
-    return ancestors.length ? makeGbifUri(ancestors[ancestors.length - 1])['@id'] : null;
+    if (!ancestors.length) {
+        return null;
+    }
+    const ancestor = ancestors[ancestors.length - 1];
+    const ancestorEntity = allTaxa.find(taxon => taxon.get('col') === ancestor);
+    if (ancestorEntity) {
+        return makeTaxonUri(ancestorEntity.get('id'))['@id'];
+    }
+    return makeColUri(ancestor)['@id'];
 }
 function mapEntities(names, entities) {
     const result = [];
@@ -171,11 +189,17 @@ function makeWikidataUri(qid) {
     return { '@id': `http://www.wikidata.org/entity/${qid}` };
 }
 function makeGbifUri(id) {
-    return { '@id': `https://gbif.org/species/${id}` };
+    return { '@id': `https://www.gbif.org/species/${id}` };
+}
+function makeColUri(id) {
+    return { '@id': `https://www.catalogueoflife.org/data/taxon/${id}` };
 }
 function makeWorkUri(id) {
     return { '@id': `${PREFIX}catalog/${id}` };
 }
+function makeTaxonUri(id) {
+    return { '@id': `${PREFIX}taxon/${id}` };
+}
 function makeScientificNameUri(id) {
     const resource = id.split(':').slice(0, -1).join(':');
     return { '@id': `${PREFIX}resource/${resource}#${id}` };
@@ -226,7 +250,7 @@ function makeLinkedDataForPublisher(publisher) {
 }
 function makeLinkedDataForTaxon(taxon) {
     const node = {
-        '@id': `${PREFIX}taxon/${taxon.get('id')}`,
+        '@id': makeTaxonUri(taxon.get('id'))['@id'],
         '@type': 'dwc:Taxon',
         'dwc:scientificName': taxon.get('name'),
     };
@@ -237,6 +261,9 @@ function makeLinkedDataForTaxon(taxon) {
     if (taxon.has('qid')) {
         ids.push(makeWikidataUri(taxon.get('qid')));
     }
+    if (taxon.has('col')) {
+        ids.push(makeColUri(taxon.get('col')));
+    }
     if (taxon.has('gbif')) {
         ids.push(makeGbifUri(taxon.get('gbif')));
     }
@@ -247,32 +274,89 @@ function makeLinkedDataForTaxon(taxon) {
 }
 function makeLinkedDataForTaxa(files) {
     var _a;
-    const nodes = [];
-    const gbifTaxa = {};
+    const nodes = {};
+    const parents = {};
+    const colIndex = files.taxa.entities.reduce((index, taxon) => {
+        if (taxon.has('col')) {
+            index[taxon.get('col')] = taxon.get('id');
+        }
+        return index;
+    }, {});
+    const getUriFromCol = (col) => {
+        if (col == null) {
+            // No parent -> Biota
+            return makeTaxonUri('T141');
+        }
+        else if (col in colIndex) {
+            return makeTaxonUri(colIndex[col]);
+        }
+        else {
+            return makeColUri(col);
+        }
+    };
+    const getUriFromGbif = (gbif) => {
+        if (gbif == null) {
+            // No parent -> Biota
+            return makeTaxonUri('T141');
+        }
+        else {
+            return makeGbifUri(gbif);
+        }
+    };
+    const addParent = (child, parent) => {
+        if (!parents[child]) {
+            parents[child] = new Set();
+        }
+        parents[child].add(parent);
+    };
     for (const taxon of files.taxa.entities) {
         const node = makeLinkedDataForTaxon(taxon);
-        const ancestors = (_a = taxon.get('ancestors_gbif')) !== null && _a !== void 0 ? _a : [];
-        if (ancestors.length) {
-            node['dwc:parentNameUsageID'] = makeGbifUri(ancestors[ancestors.length - 1]);
+        const ancestors = (_a = taxon.get('ancestors_col')) !== null && _a !== void 0 ? _a : [];
+        if (taxon.get('id') !== 'T141') {
+            node['dwc:parentNameUsageID'] = getUriFromCol(ancestors.length ? ancestors[ancestors.length - 1] : null);
         }
-        else if (taxon.get('id') !== 'T141') {
-            node['dwc:parentNameUsageID'] = { '@id': `${PREFIX}taxon/T141` };
+        if (taxon.has('accepted_col')) {
+            node['dwc:acceptedNameUsageID'] = getUriFromCol(taxon.get('accepted_col'));
         }
         for (let i = 0; i < ancestors.length; i++) {
-            gbifTaxa[ancestors[i]] = Object.assign(Object.assign({}, makeGbifUri(ancestors[i])), { 'dwc:taxonRank': makeTaxonRankUri(GBIF_RANKS[i]), 'dwc:parentNameUsageID': i ? makeGbifUri(ancestors[i - 1]) : { '@id': `${PREFIX}taxon/T141` } });
+            const parent = getUriFromCol(i ? ancestors[i - 1] : null);
+            addParent(getUriFromCol(ancestors[i])['@id'], parent['@id']);
+        }
+        if (taxon.has('children_col')) {
+            const children = taxon.get('children_col');
+            for (const child of children) {
+                addParent(getUriFromCol(child)['@id'], node['@id']);
+            }
+        }
+        // GBIF
+        if (taxon.has('ancestors_gbif')) {
+            const ancestors = taxon.get('ancestors_gbif');
+            for (let i = 0; i < ancestors.length; i++) {
+                const ancestor = makeGbifUri(ancestors[i])['@id'];
+                addParent(ancestor, getUriFromGbif(ancestors[i - 1])['@id']);
+                nodes[ancestor] = { '@id': ancestor, 'dwc:taxonRank': makeTaxonRankUri(GBIF_RANKS[i]) };
+            }
         }
         if (taxon.has('children_gbif')) {
             const children = taxon.get('children_gbif');
-            node['@reverse'] = { 'dwc:parentNameUsageID': children.map(makeGbifUri) };
             const childRank = GBIF_RANKS[ancestors.length];
             for (const child of children) {
-                gbifTaxa[child] = Object.assign(Object.assign({}, makeGbifUri(child)), { 'dwc:taxonRank': makeTaxonRankUri(childRank) });
+                const childUri = makeGbifUri(child)['@id'];
+                addParent(childUri, node['@id']);
+                nodes[childUri] = { '@id': childUri, 'dwc:taxonRank': makeTaxonRankUri(childRank) };
             }
         }
-        nodes.push(node);
+        nodes[node['@id']] = node;
+    }
+    for (const child in parents) {
+        if (!nodes[child]) {
+            nodes[child] = { '@id': child };
+        }
+        for (const parent of parents[child]) {
+            nodes[child]['dwc:parentNameUsageID'] = { '@id': parent };
+        }
     }
-    nodes.push(...Object.values(gbifTaxa));
-    return nodes;
+    return Object.values(nodes);
 }
 function makeLinkedDataForScientificName(name) {
     const node = Object.assign(Object.assign({}, makeScientificNameUri(name.scientificNameID)), { '@type': 'dwc:Taxon' });
@@ -298,12 +382,12 @@ function makeLinkedDataForScientificName(name) {
         node['dwc:parentNameUsageID'] = makeScientificNameUri(name.parentNameUsageID);
     }
     const identifiers = [];
+    if (name.colTaxonID) {
+        identifiers.push(makeColUri(name.colTaxonID));
+    }
     if (name.gbifTaxonID) {
         identifiers.push(makeGbifUri(name.gbifTaxonID));
     }
-    if (name.colTaxonID) {
-        identifiers.push({ '@id': `https://www.checklistbank.org/dataset/309120/taxon/${name.colTaxonID}` });
-    }
     if (identifiers.length) {
         node['dwc:taxonID'] = identifiers;
     }
@@ -341,7 +425,7 @@ function makeLinkedDataForResource(work, files, resourceId) {
     const taxonNames = (_c = resource.get('taxon')) !== null && _c !== void 0 ? _c : work.get('taxon');
     if (taxonNames) {
         const taxa = mapEntities(taxonNames, files.taxa.entities);
-        const coveringTaxon = taxa.length ? getCoveringTaxon(taxa) : null;
+        const coveringTaxon = taxa.length ? getCoveringTaxon(taxa, files.taxa.entities) : null;
         if (coveringTaxon !== null) {
             node['ac:taxonCoverage'] = { '@id': coveringTaxon };
         }
@@ -408,6 +492,7 @@ function makeLinkedDataForResource(work, files, resourceId) {
 function makeLinkedDataForWork(work, files) {
     const id = work.get('id');
     const node = makeWorkUri(id);
+    node['@reverse'] = {};
     const languages = work.get('language');
     node['dcterms:language'] = languages.map(language => ({ '@id': `http://id.loc.gov/vocabulary/iso639-1/${language}` }));
     if (work.has('title')) {
@@ -536,46 +621,60 @@ function makeLinkedDataForWork(work, files) {
             }
         });
     }
+    if (work.has('part_of')) {
+        const containers = work.get('part_of');
+        if (work.get('key_type').includes('supplement')) {
+            node['bibo:annotates'] = containers.map(makeWorkUri);
+        }
+        else {
+            node['dcterms:isPartOf'] = containers.map(makeWorkUri);
+            for (const container of containers) {
+                if (!Array.isArray(node['@reverse']['dcterms:hasPart'])) {
+                    node['@reverse']['dcterms:hasPart'] = [];
+                }
+                node['@reverse']['dcterms:hasPart'].push(makeWorkUri(container));
+            }
+        }
+    }
+    if (work.has('listed_in')) {
+        const referers = work.get('listed_in');
+        node['bibo:citedBy'] = node['dcterms:isReferencedBy'] = referers.map(makeWorkUri);
+        for (const referer of referers) {
+            if (!Array.isArray(node['@reverse']['bibo:cites'])) {
+                node['@reverse']['bibo:cites'] = [];
+            }
+            node['@reverse']['bibo:cites'].push(makeWorkUri(referer));
+            if (!Array.isArray(node['@reverse']['dcterms:references'])) {
+                node['@reverse']['dcterms:references'] = [];
+            }
+            node['@reverse']['dcterms:references'].push(makeWorkUri(referer));
+        }
+    }
+    if (work.has('version_of')) {
+        const originals = work.get('version_of');
+        const originalLanguages = originals.map(id => {
+            var _a;
+            const [workId, resourceId] = id.split(':');
+            const original = files.catalog.get(workId);
+            const language = original.get('language').join();
+            if (resourceId) {
+                const resourceLanguage = (_a = files.resources[id].metadata.catalog) === null || _a === void 0 ? void 0 : _a.language;
+                if (resourceLanguage) {
+                    return resourceLanguage.join(',');
+                }
+            }
+            return language;
+        });
+        node['bibo:translationOf'] = originals.filter((_, i) => originalLanguages[i] !== languages.join()).map(makeWorkUri);
+        node['dcterms:isVersionOf'] = originals.filter(id => work.get('id') !== id).map(makeWorkUri);
+    }
     return node;
 }
-function makeLinkedDataForWorks(files) {
-    const nodes = {};
+function* makeLinkedDataForWorks(files) {
     const works = files.catalog;
-    for (const work of works.entities) {
-        nodes[work.get('id')] = makeLinkedDataForWork(work, files);
-    }
     for (const work of works.entities) {
         const id = work.get('id');
-        const node = nodes[id];
-        const language = work.get('language');
-        if (work.has('part_of')) {
-            const containers = work.get('part_of');
-            if (work.get('key_type').includes('supplement')) {
-                node['bibo:annotates'] = containers.map(makeWorkUri);
-            }
-            else {
-                node['dcterms:isPartOf'] = containers.map(makeWorkUri);
-                for (const container of containers) {
-                    if (!Array.isArray(nodes[container]['dcterms:hasPart'])) {
-                        nodes[container]['dcterms:hasPart'] = [];
-                    }
-                    nodes[container]['dcterms:hasPart'].push(makeWorkUri(id));
-                }
-            }
-        }
-        if (work.has('listed_in')) {
-            const referers = work.get('listed_in');
-            node['bibo:citedBy'] = node['dcterms:isReferencedBy'] = referers.map(makeWorkUri);
-            for (const referer of referers) {
-                nodes[referer]['bibo:cites'] = nodes[referer]['dcterms:references'] = makeWorkUri(id);
-            }
-        }
-        if (work.has('version_of')) {
-            const originals = work.get('version_of');
-            const originalLanguages = originals.map(id => works.get(id).get('language').join());
-            node['bibo:translationOf'] = originals.filter((_, i) => originalLanguages[i] !== language.join()).map(makeWorkUri);
-            node['dcterms:isVersionOf'] = originals.filter(id => work.get('id') !== id).map(makeWorkUri);
-        }
+        const node = makeLinkedDataForWork(work, files);
         let resourceIndex = 1;
         let resourceId;
         while ((resourceId = `${id}:${resourceIndex++}`) in files.resources) {
@@ -589,26 +688,52 @@ function makeLinkedDataForWorks(files) {
             }
             resource['dcterms:isPartOf'].push({ '@id': node['@id'] });
         }
+        yield node;
     }
-    return Object.values(nodes);
 }
-function writeOutput(document_1) {
-    return __awaiter(this, arguments, void 0, function* (document, format = 'jsonld') {
-        if (format === 'jsonld') {
-            process.stdout.write(JSON.stringify(document, null, 2));
-            return;
-        }
-        const nquads = yield jsonld.toRDF(document, { format: 'application/n-quads' });
-        if (format === 'nquads') {
-            process.stdout.write(nquads);
-            return;
-        }
-        const parser = new N3.StreamParser();
-        const writer = new N3.StreamWriter({ prefixes: PREFIXES });
-        parser.write(nquads);
-        parser.pipe(writer);
-        writer.pipe(process.stdout);
-    });
+class Writer {
+    constructor(format) {
+        this.format = format;
+        this.issuer = new jsonld.util.IdentifierIssuer('_:b');
+        this.n3parser = new N3.StreamParser();
+        this.n3writer = new N3.StreamWriter({ prefixes: PREFIXES });
+        if (this.format === 'jsonld') {
+            const [start] = JSON.stringify({ '@context': PREFIXES, '@graph': [] }, null, 2).split(/(?<="@graph": \[)/);
+            process.stdout.write(start + '\n');
+        }
+        else if (this.format === 'nquads') {
+            // no preparation required
+        }
+        else {
+            this.n3parser.pipe(this.n3writer);
+            this.n3writer.pipe(process.stdout);
+        }
+    }
+    write(graph) {
+        return __awaiter(this, void 0, void 0, function* () {
+            if (this.format === 'jsonld') {
+                for (const entity of graph) {
+                    process.stdout.write(JSON.stringify(entity, null, 2).replace(/^/gm, '    ') + ',\n');
+                }
+                return;
+            }
+            const document = {
+                '@context': PREFIXES,
+                '@graph': graph
+            };
+            const nquads = (yield jsonld.toRDF(document, { format: 'application/n-quads', issuer: this.issuer }));
+            if (this.format === 'nquads') {
+                process.stdout.write(nquads);
+                return;
+            }
+            this.n3parser.write(nquads);
+        });
+    }
+    end() {
+        if (this.format === 'jsonld') {
+            process.stdout.write('    null\n  ]\n}\n');
+        }
+    }
 }
 function main() {
     return __awaiter(this, void 0, void 0, function* () {
@@ -644,8 +769,11 @@ function main() {
             }, {}));
             files.resources[id] = { metadata: resources[id], taxa };
         }
+        const writer = new Writer(args.values.format);
+        for (const entity of makeLinkedDataForWorks(files)) {
+            yield writer.write([entity]);
+        }
         const graph = [
-            ...makeLinkedDataForWorks(files),
             ...makeLinkedDataForTaxa(files),
         ];
         for (const entity of files.authors.entities) {
@@ -657,11 +785,8 @@ function main() {
         for (const entity of files.publishers.entities) {
             graph.push(makeLinkedDataForPublisher(entity));
         }
-        const document = {
-            '@context': PREFIXES,
-            '@graph': graph
-        };
-        writeOutput(document, args.values.format);
+        yield writer.write(graph);
+        writer.end();
     });
 }
 main().catch((error) => {

package/lib/bin/validate-integrity.d.ts

@@ -0,0 +1,2 @@
+#!/usr/bin/env node
+export {};

package/lib/bin/validate-integrity.js

@@ -0,0 +1,135 @@
+#!/usr/bin/env node
+"use strict";
+var __createBinding = (this && this.__createBinding) || (Object.create ? (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    var desc = Object.getOwnPropertyDescriptor(m, k);
+    if (!desc || ("get" in desc ? !m.__esModule : desc.writable || desc.configurable)) {
+      desc = { enumerable: true, get: function() { return m[k]; } };
+    }
+    Object.defineProperty(o, k2, desc);
+}) : (function(o, m, k, k2) {
+    if (k2 === undefined) k2 = k;
+    o[k2] = m[k];
+}));
+var __setModuleDefault = (this && this.__setModuleDefault) || (Object.create ? (function(o, v) {
+    Object.defineProperty(o, "default", { enumerable: true, value: v });
+}) : function(o, v) {
+    o["default"] = v;
+});
+var __importStar = (this && this.__importStar) || (function () {
+    var ownKeys = function(o) {
+        ownKeys = Object.getOwnPropertyNames || function (o) {
+            var ar = [];
+            for (var k in o) if (Object.prototype.hasOwnProperty.call(o, k)) ar[ar.length] = k;
+            return ar;
+        };
+        return ownKeys(o);
+    };
+    return function (mod) {
+        if (mod && mod.__esModule) return mod;
+        var result = {};
+        if (mod != null) for (var k = ownKeys(mod), i = 0; i < k.length; i++) if (k[i] !== "default") __createBinding(result, mod, k[i]);
+        __setModuleDefault(result, mod);
+        return result;
+    };
+})();
+var __awaiter = (this && this.__awaiter) || function (thisArg, _arguments, P, generator) {
+    function adopt(value) { return value instanceof P ? value : new P(function (resolve) { resolve(value); }); }
+    return new (P || (P = Promise))(function (resolve, reject) {
+        function fulfilled(value) { try { step(generator.next(value)); } catch (e) { reject(e); } }
+        function rejected(value) { try { step(generator["throw"](value)); } catch (e) { reject(e); } }
+        function step(result) { result.done ? resolve(result.value) : adopt(result.value).then(fulfilled, rejected); }
+        step((generator = generator.apply(thisArg, _arguments || [])).next());
+    });
+};
+Object.defineProperty(exports, "__esModule", { value: true });
+const fs_1 = require("fs");
+const path = __importStar(require("path"));
+const index_1 = require("../index");
+function loadCatalogFile(file, base) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const contents = yield fs_1.promises.readFile(path.join(base, `${file}.csv`), 'utf8');
+        return index_1.catalog.loadData(contents, file);
+    });
+}
+function loadResourcesIndex(base) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const contents = yield fs_1.promises.readFile(path.join(base, 'resources', 'index.json'), 'utf8');
+        return JSON.parse(contents);
+    });
+}
+function getAll(entity, field) {
+    const value = entity.get(field);
+    if (Array.isArray(value)) {
+        return value;
+    }
+    else if (value) {
+        return [value];
+    }
+    else {
+        return [];
+    }
+}
+function checkIntegrity(catalog, resources, field, sheet) {
+    const errors = [];
+    const index = new Set();
+    for (const entity of sheet) {
+        for (const value of getAll(entity, 'name')) {
+            index.add(value);
+        }
+    }
+    for (const entity of catalog) {
+        const id = entity.get('id');
+        for (const value of getAll(entity, field)) {
+            if (!index.has(value)) {
+                errors.push({
+                    entity: id,
+                    field: field,
+                    error: `Value "${value}" is not mapped`
+                });
+            }
+        }
+    }
+    for (const id in resources) {
+        const resource = resources[id];
+        if (!resource.catalog || !Object.prototype.hasOwnProperty.call(resource.catalog, field)) {
+            continue;
+        }
+        const value = resource.catalog[field];
+        const values = Array.isArray(value) ? value : [value];
+        for (const value of values) {
+            if (!index.has(value)) {
+                errors.push({
+                    entity: id,
+                    field: field,
+                    error: `Value "${value}" is not mapped`
+                });
+            }
+        }
+    }
+    return errors;
+}
+function main(args) {
+    return __awaiter(this, void 0, void 0, function* () {
+        const base = path.resolve(args[0]);
+        const [catalog, authors, places, publishers, taxa, resources] = yield Promise.all([
+            loadCatalogFile('catalog', base),
+            loadCatalogFile('authors', base),
+            loadCatalogFile('places', base),
+            loadCatalogFile('publishers', base),
+            loadCatalogFile('taxa', base),
+            loadResourcesIndex(base)
+        ]);
+        const errors = [
+            ...checkIntegrity(catalog, resources, 'author', authors),
+            ...checkIntegrity(catalog, resources, 'publisher', publishers),
+            ...checkIntegrity(catalog, resources, 'taxon', taxa),
+            ...checkIntegrity(catalog, resources, 'region', places)
+        ];
+        if (errors.length) {
+            console.table(errors);
+            process.exit(1);
+        }
+    });
+}
+main(process.argv.slice(2));

package/lib/resources/parse-name.js

@@ -233,7 +233,17 @@ function parseName(name, rank, parent) {
                 name = name.slice(species.length + 1);
             }
             // Remove rank abbreviations
-            name = name.replace(/^(st|r|ab|f|var|ssp|subsp)\. /, '');
+            for (const abbreviation in RANK_LABELS_REVERSE) {
+                const prefix = abbreviation + '. ';
+                if (name.startsWith(prefix)) {
+                    const abbreviationRank = RANK_LABELS_REVERSE[abbreviation];
+                    if (rank !== abbreviationRank) {
+                        throw new RecoverableSyntaxError(`Infraspecific rank label has an unexpected rank: "${abbreviation}." meaning ${abbreviationRank}, expected ${rank}`, item);
+                    }
+                    name = name.slice(prefix.length);
+                    break;
+                }
+            }
         }
     }
     else if (compareRanks('genus', rank) <= 0) {

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@larsgw/formica",
-  "version": "0.10.2",
+  "version": "0.10.4",
   "description": "SDK and tools for data from the Library of Identification Resources",
   "main": "lib/index.js",
   "types": "lib/index.d.ts",
@@ -10,6 +10,7 @@
   "bin": {
     "loir-validate-catalog": "lib/bin/validate-catalog.js",
     "loir-validate-resources": "lib/bin/validate-resources-text.js",
+    "loir-validate-integrity": "lib/bin/validate-integrity.js",
     "loir-resources-process": "lib/bin/process-resources.js",
     "loir-resources-index": "lib/bin/process-resources-index.js",
     "loir-resources-check-diff": "lib/bin/check-resources-diff.js",
@@ -52,6 +53,7 @@
     "@typescript-eslint/parser": "^8.32.0",
     "conventional-changelog-cli": "^5.0.0",
     "eslint": "^10.6.0",
+    "patch-package": "^8.0.1",
     "typescript": "^6.0.3"
   }
 }