@natlibfi/melinda-record-matching 1.0.9 → 2.0.1-alpha.1

package/src/match-detection/index.js

@@ -1,3 +1,4 @@
+/* eslint-disable no-console */
 /**
 *
 * @licstart  The following is the entire license notice for the JavaScript code in this file.
@@ -31,15 +32,17 @@ import * as features from './features';
 
 export {features};
 
+// eslint-disable-next-line max-statements
 export default ({strategy, treshold = 0.9}) => (recordA, recordB) => {
   const minProbabilityQuantifier = 0.5;
 
   const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection');
+  const debugData = debug.extend('data');
   const featuresA = extractFeatures(recordA);
   const featuresB = extractFeatures(recordB);
 
-  debug(`Features (a): ${JSON.stringify(featuresA)}`);
-  debug(`Features (b): ${JSON.stringify(featuresB)}`);
+  debugData(`Features (a): ${JSON.stringify(featuresA)}`);
+  debugData(`Features (b): ${JSON.stringify(featuresB)}`);
 
   const featurePairs = generateFeaturePairs();
   const similarityVector = generateSimilarityVector();
@@ -50,7 +53,7 @@ export default ({strategy, treshold = 0.9}) => (recordA, recordB) => {
     return {match: probability >= treshold, probability};
   }
 
-  debug(`No feature yielded minimum probability amount of points (${minProbabilityQuantifier})`);
+  debugData(`No feature yielded minimum probability amount of points (${minProbabilityQuantifier})`);
   return {match: false, probability: 0.0};
 
   function calculateprobability() {
@@ -69,7 +72,7 @@ export default ({strategy, treshold = 0.9}) => (recordA, recordB) => {
       return {name, points};
     });
 
-    debug(`Points: ${JSON.stringify(compared)}`);
+    debugData(`Points: ${JSON.stringify(compared)}`);
     return compared.map(({points}) => points);
   }
 

package/src/match-detection/index.spec.js

@@ -33,28 +33,35 @@ import {MarcRecord} from '@natlibfi/marc-record';
 import * as features from './features';
 import createDetectionInterface from '.';
 
-generateTests({
-  path: [__dirname, '..', '..', 'test-fixtures', 'match-detection', 'index'],
-  useMetadataFile: true,
-  recurse: false,
-  fixura: {
-    reader: READERS.JSON
-  },
-  callback: ({getFixture, options, expectedResults}) => {
-    const detect = createDetectionInterface(formatOptions());
-    const recordA = new MarcRecord(getFixture('recordA.json'));
-    const recordB = new MarcRecord(getFixture('recordB.json'));
-    const results = detect(recordA, recordB);
+describe('match-detection', () => {
+  generateTests({
+    path: [__dirname, '..', '..', 'test-fixtures', 'match-detection', 'index'],
+    useMetadataFile: true,
+    recurse: false,
+    fixura: {
+      reader: READERS.JSON
+    },
+    callback: ({getFixture, options, expectedResults, enabled = true}) => {
 
-    expect(results).to.eql(expectedResults);
+      if (!enabled) {
+        return;
+      }
 
-    function formatOptions() {
-      const contextFeatures = features[options.strategy.type];
+      const detect = createDetectionInterface(formatOptions());
+      const recordA = new MarcRecord(getFixture('recordA.json'));
+      const recordB = new MarcRecord(getFixture('recordB.json'));
+      const results = detect(recordA, recordB);
 
-      return {
-        ...options,
-        strategy: options.strategy.features.map(v => contextFeatures[v]())
-      };
+      expect(results).to.eql(expectedResults);
+
+      function formatOptions() {
+        const contextFeatures = features[options.strategy.type];
+
+        return {
+          ...options,
+          strategy: options.strategy.features.map(v => contextFeatures[v]())
+        };
+      }
     }
-  }
+  });
 });

package/src/matching-utils.js

@@ -0,0 +1,86 @@
+/**
+*
+* @licstart  The following is the entire license notice for the JavaScript code in this file.
+*
+* Melinda record matching modules for Javascript
+*
+* Copyright (C) 2020 University Of Helsinki (The National Library Of Finland)
+*
+* This file is part of melinda-record-matching-js
+*
+* melinda-record-matching-js program is free software: you can redistribute it and/or modify
+* it under the terms of the GNU Lesser General Public License as
+* published by the Free Software Foundation, either version 3 of the
+* License, or (at your option) any later version.
+*
+* melinda-record-matching-js is distributed in the hope that it will be useful,
+* but WITHOUT ANY WARRANTY; without even the implied warranty of
+* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+* GNU Lesser General Public License for more details.
+*
+* You should have received a copy of the GNU Affero General Public License
+* along with this program.  If not, see <http://www.gnu.org/licenses/>.
+*
+* @licend  The above is the entire license notice
+* for the JavaScript code in this file.
+*
+*/
+
+import createDebugLogger from 'debug';
+
+const debug = createDebugLogger('@natlibfi/melinda-record-matching:utils');
+const debugData = debug.extend('data');
+
+// eslint-disable-next-line max-statements
+export function getMelindaIdsF035(record) {
+
+  const debug = createDebugLogger('@natlibfi/melinda-record-matching:melinda-id');
+  const debugData = debug.extend('data');
+
+  const f035s = record.getFields('035');
+
+  if (f035s.length < 1) {
+    debug(`No f035s found.`);
+    return [];
+  }
+
+  const allF035MelindaIds = [].concat(...f035s.map(toMelindaIds));
+  const f035MelindaIds = [...new Set(allF035MelindaIds)];
+
+  debugData(`Fields (${f035s.length}): ${JSON.stringify(f035s)}`);
+  debugData(`Ids (${allF035MelindaIds.length}): ${JSON.stringify(allF035MelindaIds)}`);
+  debugData(`Unique ids (${f035MelindaIds.length}): ${JSON.stringify(f035MelindaIds)}`);
+
+  return f035MelindaIds;
+
+  function toMelindaIds({subfields}) {
+    const melindaIdRegExp = /^(?<prefix>\(FI-MELINDA\)|FCC)(?<id>\d{9})$/u;
+
+    return subfields
+      .filter(sub => ['a', 'z'].includes(sub.code))
+      .filter(sub => melindaIdRegExp.test(sub.value))
+      .map(({value}) => value.replace(melindaIdRegExp, '$<id>'));
+
+  }
+}
+
+export function validateSidFieldSubfieldCounts(field) {
+  // Valid SID-fields have just one $c and one $b
+  debugData(`Validating SID field ${JSON.stringify(field)}`);
+  const countC = countSubfields(field, 'c');
+  const countB = countSubfields(field, 'b');
+  debug(`Found ${countC} sf $cs and ${countB} sf $bs. IsValid: ${countC === 1 && countB === 1}`);
+
+  return countC === 1 && countB === 1;
+}
+
+function countSubfields(field, subfieldCode) {
+  // debug(`Counting subfields ${subfieldCode}`);
+  return field.subfields.filter(({code}) => code === subfieldCode).length;
+}
+
+export function getSubfieldValues(field, subfieldCode) {
+  debugData(`Get subfield(s) $${subfieldCode} from ${JSON.stringify(field)}`);
+  return field.subfields.filter(({code}) => code === subfieldCode).map(({value}) => value);
+}
+