@natlibfi/melinda-record-matching 3.0.1-alpha.1 → 3.0.1

package/.github/workflows/melinda-node-tests.yml

@@ -16,7 +16,7 @@ jobs:
 
     steps:
     - name: Checkout the code
-      uses: actions/checkout@v2
+      uses: actions/checkout@v3
     - name: Use Node.js ${{ matrix.node-version }}
       uses: actions/setup-node@v3
       with:
@@ -58,7 +58,7 @@ jobs:
     if: contains(github.ref, 'refs/tags/')
 
     steps:
-      - uses: actions/checkout@v2
+      - uses: actions/checkout@v3
       # Setup .npmrc file to publish to npm
       - uses: actions/setup-node@v3
         with:

package/dist/match-detection/features/bib/index.js

@@ -75,11 +75,18 @@ Object.defineProperty(exports, "title", {
     return _title.default;
   }
 });
+Object.defineProperty(exports, "titleVersionOriginal", {
+  enumerable: true,
+  get: function () {
+    return _titleVersionOriginal.default;
+  }
+});
 var _hostComponent = _interopRequireDefault(require("./host-component"));
 var _isbn = _interopRequireDefault(require("./isbn"));
 var _issn = _interopRequireDefault(require("./issn"));
 var _otherStandardIdentifier = _interopRequireDefault(require("./other-standard-identifier"));
 var _title = _interopRequireDefault(require("./title"));
+var _titleVersionOriginal = _interopRequireDefault(require("./title-version-original"));
 var _authors = _interopRequireDefault(require("./authors"));
 var _recordType = _interopRequireDefault(require("./record-type"));
 var _publicationTime = _interopRequireDefault(require("./publication-time"));

package/dist/match-detection/features/bib/index.js.map

@@ -1 +1 @@
-{"version":3,"file":"index.js","names":[],"sources":["../../../../src/match-detection/features/bib/index.js"],"sourcesContent":["/**\n*\n* @licstart  The following is the entire license notice for the JavaScript code in this file.\n*\n* Melinda record matching modules for Javascript\n*\n* Copyright (C) 2020 University Of Helsinki (The National Library Of Finland)\n*\n* This file is part of melinda-record-matching-js\n*\n* melinda-record-matching-js program is free software: you can redistribute it and/or modify\n* it under the terms of the GNU Lesser General Public License as\n* published by the Free Software Foundation, either version 3 of the\n* License, or (at your option) any later version.\n*\n* melinda-record-matching-js is distributed in the hope that it will be useful,\n* but WITHOUT ANY WARRANTY; without even the implied warranty of\n* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the\n* GNU Lesser General Public License for more details.\n*\n* You should have received a copy of the GNU Affero General Public License\n* along with this program.  If not, see <http://www.gnu.org/licenses/>.\n*\n* @licend  The above is the entire license notice\n* for the JavaScript code in this file.\n*\n*/\n\nexport {default as hostComponent} from './host-component';\nexport {default as isbn} from './isbn';\nexport {default as issn} from './issn';\nexport {default as otherStandardIdentifier} from './other-standard-identifier';\nexport {default as title} from './title';\nexport {default as authors} from './authors';\nexport {default as recordType} from './record-type';\nexport {default as publicationTime} from './publication-time';\nexport {default as language} from './language';\nexport {default as bibliographicLevel} from './bibliographic-level';\nexport {default as melindaId} from './melinda-id';\nexport {default as allSourceIds} from './all-source-ids';\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AA4BA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AAAyD"}
+{"version":3,"file":"index.js","names":[],"sources":["../../../../src/match-detection/features/bib/index.js"],"sourcesContent":["/**\n*\n* @licstart  The following is the entire license notice for the JavaScript code in this file.\n*\n* Melinda record matching modules for Javascript\n*\n* Copyright (C) 2020 University Of Helsinki (The National Library Of Finland)\n*\n* This file is part of melinda-record-matching-js\n*\n* melinda-record-matching-js program is free software: you can redistribute it and/or modify\n* it under the terms of the GNU Lesser General Public License as\n* published by the Free Software Foundation, either version 3 of the\n* License, or (at your option) any later version.\n*\n* melinda-record-matching-js is distributed in the hope that it will be useful,\n* but WITHOUT ANY WARRANTY; without even the implied warranty of\n* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the\n* GNU Lesser General Public License for more details.\n*\n* You should have received a copy of the GNU Affero General Public License\n* along with this program.  If not, see <http://www.gnu.org/licenses/>.\n*\n* @licend  The above is the entire license notice\n* for the JavaScript code in this file.\n*\n*/\n\nexport {default as hostComponent} from './host-component';\nexport {default as isbn} from './isbn';\nexport {default as issn} from './issn';\nexport {default as otherStandardIdentifier} from './other-standard-identifier';\nexport {default as title} from './title';\nexport {default as titleVersionOriginal} from './title-version-original';\nexport {default as authors} from './authors';\nexport {default as recordType} from './record-type';\nexport {default as publicationTime} from './publication-time';\nexport {default as language} from './language';\nexport {default as bibliographicLevel} from './bibliographic-level';\nexport {default as melindaId} from './melinda-id';\nexport {default as allSourceIds} from './all-source-ids';\n"],"mappings":";;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;;AA4BA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AAAyD"}

package/dist/match-detection/features/bib/title-version-original.js

@@ -0,0 +1,79 @@
+"use strict";
+
+Object.defineProperty(exports, "__esModule", {
+  value: true
+});
+exports.default = void 0;
+var _debug = _interopRequireDefault(require("debug"));
+var _natural = require("natural");
+function _interopRequireDefault(obj) { return obj && obj.__esModule ? obj : { default: obj }; }
+/**
+*
+* @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.
+*
+*/
+var _default = ({
+  treshold = 10
+} = {}) => ({
+  name: 'titleVersionOriginal',
+  extract: ({
+    record
+  }) => {
+    const title = getTitle();
+    if (title) {
+      return [title.replace(/[^\p{Letter}\p{Number}]/gu, '').toLowerCase()];
+    }
+    return [];
+    function getTitle() {
+      const [field] = record.get(/^245$/u);
+      if (field) {
+        return field.subfields.filter(({
+          code
+        }) => ['a', 'b'].includes(code)).map(({
+          value
+        }) => value).join('');
+      }
+    }
+  },
+  compare: (a, b) => {
+    const debug = (0, _debug.default)('@natlibfi/melinda-record-matching:match-detection:features/bib/title-version-original');
+    const distance = (0, _natural.LevenshteinDistance)(a[0], b[0]);
+    if (distance === 0) {
+      return 0.5;
+    }
+    const maxLength = getMaxLength();
+    const percentage = distance / maxLength * 100;
+    debug(`'${a}' vs '${b}': Max length = ${maxLength}, distance = ${distance}, percentage = ${percentage}`);
+    if (percentage <= treshold) {
+      return 0.3;
+    }
+    return -0.5;
+    function getMaxLength() {
+      return a[0].length > b[0].length ? a[0].length : b[0].length;
+    }
+  }
+});
+exports.default = _default;
+//# sourceMappingURL=title-version-original.js.map

package/dist/match-detection/features/bib/title-version-original.js.map

@@ -0,0 +1 @@
+{"version":3,"file":"title-version-original.js","names":["treshold","name","extract","record","title","getTitle","replace","toLowerCase","field","get","subfields","filter","code","includes","map","value","join","compare","a","b","debug","createDebugLogger","distance","leven","maxLength","getMaxLength","percentage","length"],"sources":["../../../../src/match-detection/features/bib/title-version-original.js"],"sourcesContent":["/**\n*\n* @licstart  The following is the entire license notice for the JavaScript code in this file.\n*\n* Melinda record matching modules for Javascript\n*\n* Copyright (C) 2020 University Of Helsinki (The National Library Of Finland)\n*\n* This file is part of melinda-record-matching-js\n*\n* melinda-record-matching-js program is free software: you can redistribute it and/or modify\n* it under the terms of the GNU Lesser General Public License as\n* published by the Free Software Foundation, either version 3 of the\n* License, or (at your option) any later version.\n*\n* melinda-record-matching-js is distributed in the hope that it will be useful,\n* but WITHOUT ANY WARRANTY; without even the implied warranty of\n* MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the\n* GNU Lesser General Public License for more details.\n*\n* You should have received a copy of the GNU Affero General Public License\n* along with this program.  If not, see <http://www.gnu.org/licenses/>.\n*\n* @licend  The above is the entire license notice\n* for the JavaScript code in this file.\n*\n*/\n\nimport createDebugLogger from 'debug';\nimport {LevenshteinDistance as leven} from 'natural';\n\nexport default ({treshold = 10} = {}) => ({\n  name: 'titleVersionOriginal',\n  extract: ({record}) => {\n    const title = getTitle();\n\n    if (title) {\n      return [title.replace(/[^\\p{Letter}\\p{Number}]/gu, '').toLowerCase()];\n    }\n\n    return [];\n\n    function getTitle() {\n      const [field] = record.get(/^245$/u);\n\n      if (field) {\n        return field.subfields\n          .filter(({code}) => ['a', 'b'].includes(code))\n          .map(({value}) => value)\n          .join('');\n      }\n    }\n  },\n  compare: (a, b) => {\n    const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features/bib/title-version-original');\n    const distance = leven(a[0], b[0]);\n\n    if (distance === 0) {\n      return 0.5;\n    }\n\n    const maxLength = getMaxLength();\n    const percentage = distance / maxLength * 100;\n\n    debug(`'${a}' vs '${b}': Max length = ${maxLength}, distance = ${distance}, percentage = ${percentage}`);\n\n    if (percentage <= treshold) {\n      return 0.3;\n    }\n\n    return -0.5;\n\n    function getMaxLength() {\n      return a[0].length > b[0].length ? a[0].length : b[0].length;\n    }\n\n  }\n});\n"],"mappings":";;;;;;AA4BA;AACA;AAAqD;AA7BrD;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AACA;AA1BA,eA+Be,CAAC;EAACA,QAAQ,GAAG;AAAE,CAAC,GAAG,CAAC,CAAC,MAAM;EACxCC,IAAI,EAAE,sBAAsB;EAC5BC,OAAO,EAAE,CAAC;IAACC;EAAM,CAAC,KAAK;IACrB,MAAMC,KAAK,GAAGC,QAAQ,EAAE;IAExB,IAAID,KAAK,EAAE;MACT,OAAO,CAACA,KAAK,CAACE,OAAO,CAAC,2BAA2B,EAAE,EAAE,CAAC,CAACC,WAAW,EAAE,CAAC;IACvE;IAEA,OAAO,EAAE;IAET,SAASF,QAAQ,GAAG;MAClB,MAAM,CAACG,KAAK,CAAC,GAAGL,MAAM,CAACM,GAAG,CAAC,QAAQ,CAAC;MAEpC,IAAID,KAAK,EAAE;QACT,OAAOA,KAAK,CAACE,SAAS,CACnBC,MAAM,CAAC,CAAC;UAACC;QAAI,CAAC,KAAK,CAAC,GAAG,EAAE,GAAG,CAAC,CAACC,QAAQ,CAACD,IAAI,CAAC,CAAC,CAC7CE,GAAG,CAAC,CAAC;UAACC;QAAK,CAAC,KAAKA,KAAK,CAAC,CACvBC,IAAI,CAAC,EAAE,CAAC;MACb;IACF;EACF,CAAC;EACDC,OAAO,EAAE,CAACC,CAAC,EAAEC,CAAC,KAAK;IACjB,MAAMC,KAAK,GAAG,IAAAC,cAAiB,EAAC,uFAAuF,CAAC;IACxH,MAAMC,QAAQ,GAAG,IAAAC,4BAAK,EAACL,CAAC,CAAC,CAAC,CAAC,EAAEC,CAAC,CAAC,CAAC,CAAC,CAAC;IAElC,IAAIG,QAAQ,KAAK,CAAC,EAAE;MAClB,OAAO,GAAG;IACZ;IAEA,MAAME,SAAS,GAAGC,YAAY,EAAE;IAChC,MAAMC,UAAU,GAAGJ,QAAQ,GAAGE,SAAS,GAAG,GAAG;IAE7CJ,KAAK,CAAE,IAAGF,CAAE,SAAQC,CAAE,mBAAkBK,SAAU,gBAAeF,QAAS,kBAAiBI,UAAW,EAAC,CAAC;IAExG,IAAIA,UAAU,IAAI1B,QAAQ,EAAE;MAC1B,OAAO,GAAG;IACZ;IAEA,OAAO,CAAC,GAAG;IAEX,SAASyB,YAAY,GAAG;MACtB,OAAOP,CAAC,CAAC,CAAC,CAAC,CAACS,MAAM,GAAGR,CAAC,CAAC,CAAC,CAAC,CAACQ,MAAM,GAAGT,CAAC,CAAC,CAAC,CAAC,CAACS,MAAM,GAAGR,CAAC,CAAC,CAAC,CAAC,CAACQ,MAAM;IAC9D;EAEF;AACF,CAAC,CAAC;AAAA"}

package/package.json

@@ -14,7 +14,7 @@
 		"url": "git@github.com:natlibfi/melinda-record-matching-js.git"
 	},
 	"license": "LGPL-3.0+",
-	"version": "3.0.1-alpha.1",
+	"version": "3.0.1",
 	"main": "./dist/index.js",
 	"engines": {
 		"node": ">=14"

package/src/match-detection/features/bib/index.js

@@ -31,6 +31,7 @@ export {default as isbn} from './isbn';
 export {default as issn} from './issn';
 export {default as otherStandardIdentifier} from './other-standard-identifier';
 export {default as title} from './title';
+export {default as titleVersionOriginal} from './title-version-original';
 export {default as authors} from './authors';
 export {default as recordType} from './record-type';
 export {default as publicationTime} from './publication-time';

package/src/match-detection/features/bib/title-version-original.js

@@ -0,0 +1,78 @@
+/**
+*
+* @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';
+import {LevenshteinDistance as leven} from 'natural';
+
+export default ({treshold = 10} = {}) => ({
+  name: 'titleVersionOriginal',
+  extract: ({record}) => {
+    const title = getTitle();
+
+    if (title) {
+      return [title.replace(/[^\p{Letter}\p{Number}]/gu, '').toLowerCase()];
+    }
+
+    return [];
+
+    function getTitle() {
+      const [field] = record.get(/^245$/u);
+
+      if (field) {
+        return field.subfields
+          .filter(({code}) => ['a', 'b'].includes(code))
+          .map(({value}) => value)
+          .join('');
+      }
+    }
+  },
+  compare: (a, b) => {
+    const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features/bib/title-version-original');
+    const distance = leven(a[0], b[0]);
+
+    if (distance === 0) {
+      return 0.5;
+    }
+
+    const maxLength = getMaxLength();
+    const percentage = distance / maxLength * 100;
+
+    debug(`'${a}' vs '${b}': Max length = ${maxLength}, distance = ${distance}, percentage = ${percentage}`);
+
+    if (percentage <= treshold) {
+      return 0.3;
+    }
+
+    return -0.5;
+
+    function getMaxLength() {
+      return a[0].length > b[0].length ? a[0].length : b[0].length;
+    }
+
+  }
+});