@natlibfi/melinda-record-matching 5.1.1-alpha.1 → 5.1.2-alpha.1

package/dist/cli.js

@@ -59,6 +59,7 @@ async function cli() {
         matchDetection.features.bib.hostComponent(),
         matchDetection.features.bib.isbn(),
         matchDetection.features.bib.issn(),
+        matchDetection.features.bib.publisherNumber(),
         // matchDetection.features.bib.sid(),
         matchDetection.features.bib.otherStandardIdentifier(),
         // Let's not use the same title matchDetection here
@@ -96,6 +97,7 @@ async function cli() {
         matchDetection.features.bib.hostComponent(),
         matchDetection.features.bib.isbn(),
         matchDetection.features.bib.issn(),
+        matchDetection.features.bib.publisherNumber(),
         // matchDetection.features.bib.sid(),
         matchDetection.features.bib.otherStandardIdentifier(),
         matchDetection.features.bib.title(),

package/dist/cli.js.map

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../src/cli.js"],
-  "sourcesContent": ["import fs from 'fs';\nimport yargs from 'yargs';\nimport createMatchOperator, {candidateSearch, matchDetection} from './index.js';\nimport createDebugLogger from 'debug';\nimport {MarcRecord} from '@natlibfi/marc-record';\n\ncli();\n\nasync function cli() {\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:cli');\n  const args = yargs(process.argv.slice(2))\n    .scriptName('melinda-record-matching-js')\n    .epilog('Copyright (C) 2022-2023 University Of Helsinki (The National Library Of Finland)')\n    .usage('$0 <file> [options] and env variable info in README')\n    .usage('Installed globally: $0 <file> [options] and env variable info in README')\n    .usage('Not installed: npx $0 <file> [options] and env variable info in README')\n    .usage('Build from source: node dist/index.js <file> [options] and env variable info in README')\n    .showHelpOnFail(true)\n    .example([['$ node /dist/cli.js record.json']])\n    .env('MELINDA_RECORD_MATCH')\n    .version()\n    .positional('file', {type: 'string', describe: 'Json file of records to match'})\n    .options({\n      t: {type: 'string', default: 'IDS', alias: 'searchType', describe: 'IDS, STANDARD_IDS, COMPONENT, CONTENT or CONTENTALT'},\n      m: {type: 'number', default: 1, alias: 'maxMatches', describe: ''},\n      c: {type: 'number', default: 1000, alias: 'maxCandidates', describe: ''},\n      s: {type: 'boolean', default: false, alias: 'returnStrategy', describe: ''},\n      q: {type: 'boolean', default: false, alias: 'returnQuery', describe: ''},\n      n: {type: 'boolean', default: false, alias: 'returnNonMatches', describe: ''}\n    })\n    .check((args) => {\n      const [file] = args._;\n      if (file === undefined) {\n        throw new Error('No file argument given');\n      }\n\n      if (!fs.existsSync(file)) {\n        throw new Error(`File ${file} does not exist`);\n      }\n\n      if (args.sruUrl === undefined) {\n        throw new Error('Setup sru url');\n      }\n\n      if (!['IDS', 'STANDARD_IDS', 'COMPONENT', 'CONTENT', 'CONTENTALT', 'YSO'].includes(args.searchType)) {\n        throw new Error('Invalid search type');\n      }\n\n      return true;\n    })\n    .parseSync();\n\n  const [file] = args._;\n  const {searchType} = args;\n  debug(JSON.stringify(args));\n\n  const detection = {\n    threshold: 0.8999,\n    strategy: generateStrategy(searchType)\n  };\n\n  const search = {\n    url: args.sruUrl, searchSpec: generateSearchSpec(searchType)\n  };\n\n  const matchOperator = await createMatchOperator({detection, search, ...args});\n\n  const fileRaw = fs.readFileSync(file, 'utf8');\n  const record = new MarcRecord(JSON.parse(fileRaw), {subfieldValues: false});\n\n  const result = await matchOperator({record});\n  debug(JSON.stringify(result));\n\n\n  function generateStrategy(searchType) {\n    if (['IDS'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.melindaId(),\n        matchDetection.features.bib.allSourceIds(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    // We could have differing strategy for STANDARD_IDS\n    // Let's not run title in strategy when we found the candidates through standard_ids search\n\n    if (['STANDARD_IDS'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.isbn(),\n        matchDetection.features.bib.issn(),\n        // matchDetection.features.bib.sid(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        // Let's not use the same title matchDetection here\n        //matchDetection.features.bib.title(),\n        matchDetection.features.bib.authors(),\n        // We probably should have some leeway here for notated music as BK etc.\n        matchDetection.features.bib.recordType(),\n        // Use publicationTimeAllowConsYearsMulti to\n        //  - ignore one year differences in publicationTime\n        //  - extract publicationTimes from f008, f26x and reprint notes in f500\n        //  - do not substract points for mismatching (normal) publicationTime, if there's a match between\n        //       normal publicationTime and a reprintPublication time\n        matchDetection.features.bib.publicationTimeAllowConsYearsMulti(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['COMPONENT'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        matchDetection.features.bib.recordType(),\n        matchDetection.features.bib.title(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.authors(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['CONTENT', 'CONTENTALT'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.isbn(),\n        matchDetection.features.bib.issn(),\n        // matchDetection.features.bib.sid(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        matchDetection.features.bib.title(),\n        matchDetection.features.bib.authors(),\n        matchDetection.features.bib.recordType(),\n        matchDetection.features.bib.publicationTime(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['YSO']) {\n      return [\n        // matchDetection.features.bib.sid(), // SIDs might be relevant on the auth side as well?\n        matchDetection.features.auth.yso()\n      ];\n    }\n\n    throw new Error('Unsupported match validation package');\n  }\n\n\n  function generateSearchSpec(searchType) {\n    if (['IDS'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.melindaId,\n        candidateSearch.searchTypes.bib.sourceIds\n      ];\n    }\n\n    if (['STANDARD_IDS'].includes(searchType)) {\n      return [candidateSearch.searchTypes.bib.standardIdentifiers];\n    }\n\n    if (['COMPONENT'].includes(searchType)) {\n      return [\n        //candidateSearch.searchTypes.bib.sourceIds,\n        candidateSearch.searchTypes.component.hostIdMelinda,\n        candidateSearch.searchTypes.component.hostIdOtherSource\n      ];\n    }\n\n    if (['CONTENT'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.hostComponents,\n        //candidateSearch.searchTypes.bib.titleAuthor,\n        candidateSearch.searchTypes.bib.title\n      ];\n    }\n    if (['CONTENTALT'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.hostComponents,\n        // titleAuthorYearAlternates searches for matchCandidates\n        // with alternate queries, starting from more tight searches\n        candidateSearch.searchTypes.bib.titleAuthorYearAlternates\n      ];\n    }\n\n    if (searchType === 'YSO') {\n      return [\n        candidateSearch.searchTypes.auth.authURX\n      ];\n    }\n\n    throw new Error('Unsupported match validation package');\n  }\n}\n"],
-  "mappings": "AAAA,OAAO,QAAQ;AACf,OAAO,WAAW;AAClB,OAAO,uBAAsB,iBAAiB,sBAAqB;AACnE,OAAO,uBAAuB;AAC9B,SAAQ,kBAAiB;AAEzB,IAAI;AAEJ,eAAe,MAAM;AACnB,QAAM,QAAQ,kBAAkB,uCAAuC;AACvE,QAAM,OAAO,MAAM,QAAQ,KAAK,MAAM,CAAC,CAAC,EACrC,WAAW,4BAA4B,EACvC,OAAO,kFAAkF,EACzF,MAAM,qDAAqD,EAC3D,MAAM,yEAAyE,EAC/E,MAAM,wEAAwE,EAC9E,MAAM,wFAAwF,EAC9F,eAAe,IAAI,EACnB,QAAQ,CAAC,CAAC,iCAAiC,CAAC,CAAC,EAC7C,IAAI,sBAAsB,EAC1B,QAAQ,EACR,WAAW,QAAQ,EAAC,MAAM,UAAU,UAAU,gCAA+B,CAAC,EAC9E,QAAQ;AAAA,IACP,GAAG,EAAC,MAAM,UAAU,SAAS,OAAO,OAAO,cAAc,UAAU,sDAAqD;AAAA,IACxH,GAAG,EAAC,MAAM,UAAU,SAAS,GAAG,OAAO,cAAc,UAAU,GAAE;AAAA,IACjE,GAAG,EAAC,MAAM,UAAU,SAAS,KAAM,OAAO,iBAAiB,UAAU,GAAE;AAAA,IACvE,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,kBAAkB,UAAU,GAAE;AAAA,IAC1E,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,eAAe,UAAU,GAAE;AAAA,IACvE,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,oBAAoB,UAAU,GAAE;AAAA,EAC9E,CAAC,EACA,MAAM,CAACA,UAAS;AACf,UAAM,CAACC,KAAI,IAAID,MAAK;AACpB,QAAIC,UAAS,QAAW;AACtB,YAAM,IAAI,MAAM,wBAAwB;AAAA,IAC1C;AAEA,QAAI,CAAC,GAAG,WAAWA,KAAI,GAAG;AACxB,YAAM,IAAI,MAAM,QAAQA,KAAI,iBAAiB;AAAA,IAC/C;AAEA,QAAID,MAAK,WAAW,QAAW;AAC7B,YAAM,IAAI,MAAM,eAAe;AAAA,IACjC;AAEA,QAAI,CAAC,CAAC,OAAO,gBAAgB,aAAa,WAAW,cAAc,KAAK,EAAE,SAASA,MAAK,UAAU,GAAG;AACnG,YAAM,IAAI,MAAM,qBAAqB;AAAA,IACvC;AAEA,WAAO;AAAA,EACT,CAAC,EACA,UAAU;AAEb,QAAM,CAAC,IAAI,IAAI,KAAK;AACpB,QAAM,EAAC,WAAU,IAAI;AACrB,QAAM,KAAK,UAAU,IAAI,CAAC;AAE1B,QAAM,YAAY;AAAA,IAChB,WAAW;AAAA,IACX,UAAU,iBAAiB,UAAU;AAAA,EACvC;AAEA,QAAM,SAAS;AAAA,IACb,KAAK,KAAK;AAAA,IAAQ,YAAY,mBAAmB,UAAU;AAAA,EAC7D;AAEA,QAAM,gBAAgB,MAAM,oBAAoB,EAAC,WAAW,QAAQ,GAAG,KAAI,CAAC;AAE5E,QAAM,UAAU,GAAG,aAAa,MAAM,MAAM;AAC5C,QAAM,SAAS,IAAI,WAAW,KAAK,MAAM,OAAO,GAAG,EAAC,gBAAgB,MAAK,CAAC;AAE1E,QAAM,SAAS,MAAM,cAAc,EAAC,OAAM,CAAC;AAC3C,QAAM,KAAK,UAAU,MAAM,CAAC;AAG5B,WAAS,iBAAiBE,aAAY;AACpC,QAAI,CAAC,KAAK,EAAE,SAASA,WAAU,GAAG;AAChC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,UAAU;AAAA,QACtC,eAAe,SAAS,IAAI,aAAa;AAAA,QACzC,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAKA,QAAI,CAAC,cAAc,EAAE,SAASA,WAAU,GAAG;AACzC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,QAEjC,eAAe,SAAS,IAAI,wBAAwB;AAAA;AAAA;AAAA,QAGpD,eAAe,SAAS,IAAI,QAAQ;AAAA;AAAA,QAEpC,eAAe,SAAS,IAAI,WAAW;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,QAMvC,eAAe,SAAS,IAAI,mCAAmC;AAAA,QAC/D,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,WAAW,EAAE,SAASA,WAAU,GAAG;AACtC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,wBAAwB;AAAA,QACpD,eAAe,SAAS,IAAI,WAAW;AAAA,QACvC,eAAe,SAAS,IAAI,MAAM;AAAA,QAClC,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,QAAQ;AAAA,QACpC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,WAAW,YAAY,EAAE,SAASA,WAAU,GAAG;AAClD,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,QAEjC,eAAe,SAAS,IAAI,wBAAwB;AAAA,QACpD,eAAe,SAAS,IAAI,MAAM;AAAA,QAClC,eAAe,SAAS,IAAI,QAAQ;AAAA,QACpC,eAAe,SAAS,IAAI,WAAW;AAAA,QACvC,eAAe,SAAS,IAAI,gBAAgB;AAAA,QAC5C,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,KAAK,GAAG;AACX,aAAO;AAAA;AAAA,QAEL,eAAe,SAAS,KAAK,IAAI;AAAA,MACnC;AAAA,IACF;AAEA,UAAM,IAAI,MAAM,sCAAsC;AAAA,EACxD;AAGA,WAAS,mBAAmBA,aAAY;AACtC,QAAI,CAAC,KAAK,EAAE,SAASA,WAAU,GAAG;AAChC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA,QAChC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AAEA,QAAI,CAAC,cAAc,EAAE,SAASA,WAAU,GAAG;AACzC,aAAO,CAAC,gBAAgB,YAAY,IAAI,mBAAmB;AAAA,IAC7D;AAEA,QAAI,CAAC,WAAW,EAAE,SAASA,WAAU,GAAG;AACtC,aAAO;AAAA;AAAA,QAEL,gBAAgB,YAAY,UAAU;AAAA,QACtC,gBAAgB,YAAY,UAAU;AAAA,MACxC;AAAA,IACF;AAEA,QAAI,CAAC,SAAS,EAAE,SAASA,WAAU,GAAG;AACpC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA;AAAA,QAEhC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AACA,QAAI,CAAC,YAAY,EAAE,SAASA,WAAU,GAAG;AACvC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA;AAAA;AAAA,QAGhC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AAEA,QAAIA,gBAAe,OAAO;AACxB,aAAO;AAAA,QACL,gBAAgB,YAAY,KAAK;AAAA,MACnC;AAAA,IACF;AAEA,UAAM,IAAI,MAAM,sCAAsC;AAAA,EACxD;AACF;",
+  "sourcesContent": ["import fs from 'fs';\nimport yargs from 'yargs';\nimport createMatchOperator, {candidateSearch, matchDetection} from './index.js';\nimport createDebugLogger from 'debug';\nimport {MarcRecord} from '@natlibfi/marc-record';\n\ncli();\n\nasync function cli() {\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:cli');\n  const args = yargs(process.argv.slice(2))\n    .scriptName('melinda-record-matching-js')\n    .epilog('Copyright (C) 2022-2023 University Of Helsinki (The National Library Of Finland)')\n    .usage('$0 <file> [options] and env variable info in README')\n    .usage('Installed globally: $0 <file> [options] and env variable info in README')\n    .usage('Not installed: npx $0 <file> [options] and env variable info in README')\n    .usage('Build from source: node dist/index.js <file> [options] and env variable info in README')\n    .showHelpOnFail(true)\n    .example([['$ node /dist/cli.js record.json']])\n    .env('MELINDA_RECORD_MATCH')\n    .version()\n    .positional('file', {type: 'string', describe: 'Json file of records to match'})\n    .options({\n      t: {type: 'string', default: 'IDS', alias: 'searchType', describe: 'IDS, STANDARD_IDS, COMPONENT, CONTENT or CONTENTALT'},\n      m: {type: 'number', default: 1, alias: 'maxMatches', describe: ''},\n      c: {type: 'number', default: 1000, alias: 'maxCandidates', describe: ''},\n      s: {type: 'boolean', default: false, alias: 'returnStrategy', describe: ''},\n      q: {type: 'boolean', default: false, alias: 'returnQuery', describe: ''},\n      n: {type: 'boolean', default: false, alias: 'returnNonMatches', describe: ''}\n    })\n    .check((args) => {\n      const [file] = args._;\n      if (file === undefined) {\n        throw new Error('No file argument given');\n      }\n\n      if (!fs.existsSync(file)) {\n        throw new Error(`File ${file} does not exist`);\n      }\n\n      if (args.sruUrl === undefined) {\n        throw new Error('Setup sru url');\n      }\n\n      if (!['IDS', 'STANDARD_IDS', 'COMPONENT', 'CONTENT', 'CONTENTALT', 'YSO'].includes(args.searchType)) {\n        throw new Error('Invalid search type');\n      }\n\n      return true;\n    })\n    .parseSync();\n\n  const [file] = args._;\n  const {searchType} = args;\n  debug(JSON.stringify(args));\n\n  const detection = {\n    threshold: 0.8999,\n    strategy: generateStrategy(searchType)\n  };\n\n  const search = {\n    url: args.sruUrl, searchSpec: generateSearchSpec(searchType)\n  };\n\n  const matchOperator = await createMatchOperator({detection, search, ...args});\n\n  const fileRaw = fs.readFileSync(file, 'utf8');\n  const record = new MarcRecord(JSON.parse(fileRaw), {subfieldValues: false});\n\n  const result = await matchOperator({record});\n  debug(JSON.stringify(result));\n\n\n  function generateStrategy(searchType) {\n    if (['IDS'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.melindaId(),\n        matchDetection.features.bib.allSourceIds(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    // We could have differing strategy for STANDARD_IDS\n    // Let's not run title in strategy when we found the candidates through standard_ids search\n\n    if (['STANDARD_IDS'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.isbn(),\n        matchDetection.features.bib.issn(),\n        matchDetection.features.bib.publisherNumber(),\n        // matchDetection.features.bib.sid(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        // Let's not use the same title matchDetection here\n        //matchDetection.features.bib.title(),\n        matchDetection.features.bib.authors(),\n        // We probably should have some leeway here for notated music as BK etc.\n        matchDetection.features.bib.recordType(),\n        // Use publicationTimeAllowConsYearsMulti to\n        //  - ignore one year differences in publicationTime\n        //  - extract publicationTimes from f008, f26x and reprint notes in f500\n        //  - do not substract points for mismatching (normal) publicationTime, if there's a match between\n        //       normal publicationTime and a reprintPublication time\n        matchDetection.features.bib.publicationTimeAllowConsYearsMulti(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['COMPONENT'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        matchDetection.features.bib.recordType(),\n        matchDetection.features.bib.title(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.authors(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['CONTENT', 'CONTENTALT'].includes(searchType)) {\n      return [\n        matchDetection.features.bib.hostComponent(),\n        matchDetection.features.bib.isbn(),\n        matchDetection.features.bib.issn(),\n        matchDetection.features.bib.publisherNumber(),\n        // matchDetection.features.bib.sid(),\n        matchDetection.features.bib.otherStandardIdentifier(),\n        matchDetection.features.bib.title(),\n        matchDetection.features.bib.authors(),\n        matchDetection.features.bib.recordType(),\n        matchDetection.features.bib.publicationTime(),\n        matchDetection.features.bib.language(),\n        matchDetection.features.bib.bibliographicLevel(),\n        matchDetection.features.bib.f773() // Host data\n      ];\n    }\n\n    if (['YSO']) {\n      return [\n        // matchDetection.features.bib.sid(), // SIDs might be relevant on the auth side as well?\n        matchDetection.features.auth.yso()\n      ];\n    }\n\n    throw new Error('Unsupported match validation package');\n  }\n\n\n  function generateSearchSpec(searchType) {\n    if (['IDS'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.melindaId,\n        candidateSearch.searchTypes.bib.sourceIds\n      ];\n    }\n\n    if (['STANDARD_IDS'].includes(searchType)) {\n      return [candidateSearch.searchTypes.bib.standardIdentifiers];\n    }\n\n    if (['COMPONENT'].includes(searchType)) {\n      return [\n        //candidateSearch.searchTypes.bib.sourceIds,\n        candidateSearch.searchTypes.component.hostIdMelinda,\n        candidateSearch.searchTypes.component.hostIdOtherSource\n      ];\n    }\n\n    if (['CONTENT'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.hostComponents,\n        //candidateSearch.searchTypes.bib.titleAuthor,\n        candidateSearch.searchTypes.bib.title\n      ];\n    }\n    if (['CONTENTALT'].includes(searchType)) {\n      return [\n        candidateSearch.searchTypes.bib.hostComponents,\n        // titleAuthorYearAlternates searches for matchCandidates\n        // with alternate queries, starting from more tight searches\n        candidateSearch.searchTypes.bib.titleAuthorYearAlternates\n      ];\n    }\n\n    if (searchType === 'YSO') {\n      return [\n        candidateSearch.searchTypes.auth.authURX\n      ];\n    }\n\n    throw new Error('Unsupported match validation package');\n  }\n}\n"],
+  "mappings": "AAAA,OAAO,QAAQ;AACf,OAAO,WAAW;AAClB,OAAO,uBAAsB,iBAAiB,sBAAqB;AACnE,OAAO,uBAAuB;AAC9B,SAAQ,kBAAiB;AAEzB,IAAI;AAEJ,eAAe,MAAM;AACnB,QAAM,QAAQ,kBAAkB,uCAAuC;AACvE,QAAM,OAAO,MAAM,QAAQ,KAAK,MAAM,CAAC,CAAC,EACrC,WAAW,4BAA4B,EACvC,OAAO,kFAAkF,EACzF,MAAM,qDAAqD,EAC3D,MAAM,yEAAyE,EAC/E,MAAM,wEAAwE,EAC9E,MAAM,wFAAwF,EAC9F,eAAe,IAAI,EACnB,QAAQ,CAAC,CAAC,iCAAiC,CAAC,CAAC,EAC7C,IAAI,sBAAsB,EAC1B,QAAQ,EACR,WAAW,QAAQ,EAAC,MAAM,UAAU,UAAU,gCAA+B,CAAC,EAC9E,QAAQ;AAAA,IACP,GAAG,EAAC,MAAM,UAAU,SAAS,OAAO,OAAO,cAAc,UAAU,sDAAqD;AAAA,IACxH,GAAG,EAAC,MAAM,UAAU,SAAS,GAAG,OAAO,cAAc,UAAU,GAAE;AAAA,IACjE,GAAG,EAAC,MAAM,UAAU,SAAS,KAAM,OAAO,iBAAiB,UAAU,GAAE;AAAA,IACvE,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,kBAAkB,UAAU,GAAE;AAAA,IAC1E,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,eAAe,UAAU,GAAE;AAAA,IACvE,GAAG,EAAC,MAAM,WAAW,SAAS,OAAO,OAAO,oBAAoB,UAAU,GAAE;AAAA,EAC9E,CAAC,EACA,MAAM,CAACA,UAAS;AACf,UAAM,CAACC,KAAI,IAAID,MAAK;AACpB,QAAIC,UAAS,QAAW;AACtB,YAAM,IAAI,MAAM,wBAAwB;AAAA,IAC1C;AAEA,QAAI,CAAC,GAAG,WAAWA,KAAI,GAAG;AACxB,YAAM,IAAI,MAAM,QAAQA,KAAI,iBAAiB;AAAA,IAC/C;AAEA,QAAID,MAAK,WAAW,QAAW;AAC7B,YAAM,IAAI,MAAM,eAAe;AAAA,IACjC;AAEA,QAAI,CAAC,CAAC,OAAO,gBAAgB,aAAa,WAAW,cAAc,KAAK,EAAE,SAASA,MAAK,UAAU,GAAG;AACnG,YAAM,IAAI,MAAM,qBAAqB;AAAA,IACvC;AAEA,WAAO;AAAA,EACT,CAAC,EACA,UAAU;AAEb,QAAM,CAAC,IAAI,IAAI,KAAK;AACpB,QAAM,EAAC,WAAU,IAAI;AACrB,QAAM,KAAK,UAAU,IAAI,CAAC;AAE1B,QAAM,YAAY;AAAA,IAChB,WAAW;AAAA,IACX,UAAU,iBAAiB,UAAU;AAAA,EACvC;AAEA,QAAM,SAAS;AAAA,IACb,KAAK,KAAK;AAAA,IAAQ,YAAY,mBAAmB,UAAU;AAAA,EAC7D;AAEA,QAAM,gBAAgB,MAAM,oBAAoB,EAAC,WAAW,QAAQ,GAAG,KAAI,CAAC;AAE5E,QAAM,UAAU,GAAG,aAAa,MAAM,MAAM;AAC5C,QAAM,SAAS,IAAI,WAAW,KAAK,MAAM,OAAO,GAAG,EAAC,gBAAgB,MAAK,CAAC;AAE1E,QAAM,SAAS,MAAM,cAAc,EAAC,OAAM,CAAC;AAC3C,QAAM,KAAK,UAAU,MAAM,CAAC;AAG5B,WAAS,iBAAiBE,aAAY;AACpC,QAAI,CAAC,KAAK,EAAE,SAASA,WAAU,GAAG;AAChC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,UAAU;AAAA,QACtC,eAAe,SAAS,IAAI,aAAa;AAAA,QACzC,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAKA,QAAI,CAAC,cAAc,EAAE,SAASA,WAAU,GAAG;AACzC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,gBAAgB;AAAA;AAAA,QAE5C,eAAe,SAAS,IAAI,wBAAwB;AAAA;AAAA;AAAA,QAGpD,eAAe,SAAS,IAAI,QAAQ;AAAA;AAAA,QAEpC,eAAe,SAAS,IAAI,WAAW;AAAA;AAAA;AAAA;AAAA;AAAA;AAAA,QAMvC,eAAe,SAAS,IAAI,mCAAmC;AAAA,QAC/D,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,WAAW,EAAE,SAASA,WAAU,GAAG;AACtC,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,wBAAwB;AAAA,QACpD,eAAe,SAAS,IAAI,WAAW;AAAA,QACvC,eAAe,SAAS,IAAI,MAAM;AAAA,QAClC,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,QAAQ;AAAA,QACpC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,WAAW,YAAY,EAAE,SAASA,WAAU,GAAG;AAClD,aAAO;AAAA,QACL,eAAe,SAAS,IAAI,cAAc;AAAA,QAC1C,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,KAAK;AAAA,QACjC,eAAe,SAAS,IAAI,gBAAgB;AAAA;AAAA,QAE5C,eAAe,SAAS,IAAI,wBAAwB;AAAA,QACpD,eAAe,SAAS,IAAI,MAAM;AAAA,QAClC,eAAe,SAAS,IAAI,QAAQ;AAAA,QACpC,eAAe,SAAS,IAAI,WAAW;AAAA,QACvC,eAAe,SAAS,IAAI,gBAAgB;AAAA,QAC5C,eAAe,SAAS,IAAI,SAAS;AAAA,QACrC,eAAe,SAAS,IAAI,mBAAmB;AAAA,QAC/C,eAAe,SAAS,IAAI,KAAK;AAAA;AAAA,MACnC;AAAA,IACF;AAEA,QAAI,CAAC,KAAK,GAAG;AACX,aAAO;AAAA;AAAA,QAEL,eAAe,SAAS,KAAK,IAAI;AAAA,MACnC;AAAA,IACF;AAEA,UAAM,IAAI,MAAM,sCAAsC;AAAA,EACxD;AAGA,WAAS,mBAAmBA,aAAY;AACtC,QAAI,CAAC,KAAK,EAAE,SAASA,WAAU,GAAG;AAChC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA,QAChC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AAEA,QAAI,CAAC,cAAc,EAAE,SAASA,WAAU,GAAG;AACzC,aAAO,CAAC,gBAAgB,YAAY,IAAI,mBAAmB;AAAA,IAC7D;AAEA,QAAI,CAAC,WAAW,EAAE,SAASA,WAAU,GAAG;AACtC,aAAO;AAAA;AAAA,QAEL,gBAAgB,YAAY,UAAU;AAAA,QACtC,gBAAgB,YAAY,UAAU;AAAA,MACxC;AAAA,IACF;AAEA,QAAI,CAAC,SAAS,EAAE,SAASA,WAAU,GAAG;AACpC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA;AAAA,QAEhC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AACA,QAAI,CAAC,YAAY,EAAE,SAASA,WAAU,GAAG;AACvC,aAAO;AAAA,QACL,gBAAgB,YAAY,IAAI;AAAA;AAAA;AAAA,QAGhC,gBAAgB,YAAY,IAAI;AAAA,MAClC;AAAA,IACF;AAEA,QAAIA,gBAAe,OAAO;AACxB,aAAO;AAAA,QACL,gBAAgB,YAAY,KAAK;AAAA,MACnC;AAAA,IACF;AAEA,UAAM,IAAI,MAAM,sCAAsC;AAAA,EACxD;AACF;",
   "names": ["args", "file", "searchType"]
 }

package/dist/index.js

@@ -146,7 +146,7 @@ export default ({ detection: detectionOptions, search: searchOptions, maxMatches
           debug(`${detectionResult.match ? `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is a match!` : `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is NOT a match!`}`);
           debugData(`Strategy: ${JSON.stringify(detectionResult.strategy)}, Threshold: ${JSON.stringify(detectionResult.threshold)}`);
           const matchResult = {
-            probability: detectionResult.probability,
+            probability: Math.round(detectionResult.probability * 100) / 100,
             candidate: {
               id: candidateId,
               record: candidateRecord

package/dist/index.js.map

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../src/index.js"],
-  "sourcesContent": ["import createDebugLogger from 'debug';\nimport createSearchInterface, * as candidateSearch from './candidate-search/index.js';\nimport createDetectionInterface, * as matchDetection from './match-detection/index.js';\n//import inspect from 'util';\n\nexport {candidateSearch, matchDetection};\n\nexport default ({detection: detectionOptions, search: searchOptions, maxMatches = 1, maxCandidates = 25, returnStrategy = false, returnQuery = false, returnNonMatches = false}) => {\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:index');\n  const debugData = debug.extend('data');\n\n  debugData(`DetectionOptions: ${JSON.stringify(detectionOptions)}`);\n  debugData(`SearchOptions: ${JSON.stringify(searchOptions)}`);\n  debugData(`MaxMatches: ${JSON.stringify(maxMatches)}`);\n  debugData(`MaxCandidates: ${JSON.stringify(maxCandidates)}`);\n  debugData(`ReturnStrategy: ${JSON.stringify(returnStrategy)}`);\n  debugData(`ReturnQuery: ${JSON.stringify(returnQuery)}`);\n  debugData(`ReturnNonMatches: ${JSON.stringify(returnNonMatches)}`);\n\n\n  const detect = createDetectionInterface(detectionOptions, returnStrategy);\n\n  return prepareSearch;\n\n  async function prepareSearch({record, recordExternal = {recordSource: 'incomingRecord', label: 'ic'}}) {\n\n    const {search} = await createSearchInterface({...searchOptions, record, maxCandidates, recordExternal});\n    return iterate({});\n\n    // candidateCount : amount of candidate records retrived from SRU for matching, NOT including current record set\n    // matches : candidates that have been detected as matches by current matcher job\n    // nonMatches : candidates that have been detected as non-matches by current matcher job (only if returnNonMatches is 'true')\n    // duplicateCount : amount of candidate records that were retrieved from the SRU but not handled further because they were already found in the matches/nonMatches\n\n    // state.totalRecords : amount of candidate records available to the current query (undefined, if there was no queries left)\n    // state.query : current query (undefined if there was no queries left)\n    // state.searchCounter : sequence for current search for current query (undefined, if there we no queries left)\n    // state.queryCandidateCounter: amount of candidate records retrieved from SRU for matching for current query, including the current record set (undefined if there were no queries left)\n    // state.queriesLeft : amount of queries left\n    // state.queryCounter : sequence for current query\n    // state.maxedQueries : queries that resulted in more than serverMaxResults hits\n\n    async function iterate({initialState = {}, matches = [], candidateCount = 0, nonMatches = [], duplicateCount = 0, nonMatchCount = 0, matchErrors = []}) {\n      debugData(`Starting next matcher iteration.`);\n      const {records, ...state} = await search(initialState);\n\n      debugData(`Current state: ${JSON.stringify(state)}, matches: ${matches.length}, candidateCount: ${candidateCount}, nonMatches: ${nonMatches.length}, nonMatchCount: ${nonMatchCount}, matchErrors: ${matchErrors.length}`);\n      const recordSetSize = records.length;\n      const newCandidateCount = candidateCount + recordSetSize;\n\n\n      if (recordSetSize > 0) {\n        return handleRecordSet();\n      }\n\n      if (state.queriesLeft > 0) {\n        debug(`Empty record set ${state.searchCounter} for ${state.query}, but there are ${state.queriesLeft} queries left`);\n        return iterate({initialState: state, matches, candidateCount: newCandidateCount, nonMatches, nonMatchCount, duplicateCount, matchErrors});\n      }\n\n      debug(`No (more) candidate records to check, no more queries left, matches: ${matches.length}`);\n      return returnResult({matches, state, stopReason: '', nonMatches, nonMatchCount, candidateCount: newCandidateCount, duplicateCount, matchErrors});\n\n      function handleRecordSet() {\n        debug(`Checking record set of ${recordSetSize} candidate records for possible matches, found by ${state.searchCounter} search for ${state.query}`);\n\n        const matchResult = iterateRecords({records, recordSetSize, maxMatches, matches, nonMatches, nonMatchCount});\n\n        const newDuplicateCount = duplicateCount + matchResult.duplicateCount;\n        const newNonMatchCount = nonMatchCount + matchResult.nonMatchCount;\n        const {newMatches, newNonMatches, newMatchErrors} = handleMatchResult(matchResult, matches, nonMatches, matchErrors);\n\n        if (maxMatchesFound({matches: newMatches, maxMatches})) {\n          return returnResult({matches: newMatches, state, stopReason: 'maxMatches', nonMatches: newNonMatches, duplicateCount: newDuplicateCount, candidateCount: newCandidateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n        }\n\n        if (maxCandidatesRetrieved(newCandidateCount, maxCandidates)) {\n          return returnResult({matches: newMatches, state, stopReason: 'maxCandidates', nonMatches: newNonMatches, duplicateCount: newDuplicateCount, candidateCount: newCandidateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n        }\n\n        return iterate({initialState: state, matches: newMatches, candidateCount: newCandidateCount, nonMatches: newNonMatches, duplicateCount: newDuplicateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n      }\n\n      function handleMatchResult(matchResult, matches, nonMatches, matchErrors) {\n        debugData(`- Amount of new matches from record set: ${matchResult.matches.length}`);\n        if (returnNonMatches) {\n          debugData(`- Amount of new nonMatches from record set: ${matchResult.nonMatches.length}`);\n        }\n\n        const newMatches = matches.concat(returnQuery ? addQuery(matchResult.matches) : matchResult.matches);\n        const newNonMatches = returnNonMatches ? nonMatches.concat(returnQuery ? addQuery(matchResult.nonMatches) : matchResult.nonMatches) : [];\n        const newMatchErrors = matchErrors.concat(matchResult.matchErrors);\n\n        debugData(`- Total amount of matches: ${newMatches.length}`);\n        if (returnNonMatches) {\n          debugData(`- Total amount of nonMatches: ${newNonMatches.length}`);\n        }\n\n        debugData(`MatchResult: ${JSON.stringify(matchResult)}`);\n        debugData(`Old matchErrors: ${JSON.stringify(matchErrors)}, matchErrors from matchResult: ${JSON.stringify(matchResult.matchErrors)}, New matchErrors: ${JSON.stringify(newMatchErrors)}`);\n\n        debugData(`- Total amount of matchErrors: ${newMatchErrors.length}`);\n\n        return {newMatches, newNonMatches, newMatchErrors};\n      }\n\n      function addQuery(matches) {\n        debugData(`Adding query ${state.query} to matches`);\n        return matches.map((match) => ({...match, matchQuery: state.query}));\n      }\n\n      function maxCandidatesRetrieved(candidateCount, maxCandidates) {\n        debugData(`Total amount of candidate records retrieved: ${newCandidateCount} (max: ${maxCandidates})`);\n        if (maxCandidates && candidateCount >= maxCandidates) {\n          debug(`Stopped matching because maximum number of candidate records ${candidateCount} / ${maxCandidates} have been retrieved`);\n          return true;\n        }\n      }\n    }\n\n    // matches : array of matching candidate records\n    // nonMatches : array of nonMatching candidate records (if returnNonMatches option is true, otherwise empty array)\n    // - candidate.id\n    // - candidate.record\n    // - probability\n    // - strategy (if returnStrategy option is true)\n    // - threshold (if returnStrategy option is true)\n    // - matchQuery (if returnQuery option is true)\n    // failures: array of conversionFailures from candidate-search and matchErrors from matchDetection in error format {status, payload: {message, id}} if returnFailures is true\n\n    // we could have here also returnRecords/returnMatchRecords/returnNonMatchRecord options that could be turned false for not to return actual record data\n\n    // matchStatus.status: boolean, true if matcher retrieved and handled all found candidate records, false if it did not\n    // matchStatus.stopReason: string ('maxMatches','maxCandidates','maxedQueries','conversionFailures', empty string/undefined), reason for stopping retrieving or handling the candidate records\n    // - only one stopReason is returned (if there would be several possible stopReasons, stopReason is picked in the above order)\n    // - currently stopReason can be non-empty also in cases where status is true, if matcher hit the stop reason when handling the last available candidate record\n\n    function returnResult({matches, state, stopReason, nonMatches, duplicateCount, candidateCount, nonMatchCount, matchErrors}) {\n      const matchErrorCount = matchErrors.length;\n      checkCounts({matches, nonMatches, candidateCount, duplicateCount, nonMatchCount, matchErrorCount});\n      const matchStatus = getMatchState(state, stopReason, matchErrorCount);\n      // add nonMatches to result only if returnNonMatches is 'true', otherwise nonMatches have not been gathered\n      const matchesResult = returnNonMatches ? {matches, matchStatus, nonMatches, candidateCount} : {matches, matchStatus, candidateCount};\n      const result = matchesResult;\n      debugData(`${JSON.stringify(result)}`);\n      return result;\n\n      // note that in cases where the matching has been stopped because of maxMatches checkCounts won't (in most cases) match\n\n      function checkCounts({matches, nonMatches, candidateCount, duplicateCount, nonMatchCount, matchErrorCount}) {\n        const matchCount = matches.length;\n        debugData(`Return nonMatches: ${returnNonMatches}`);\n        const chosenNonMatchCount = returnNonMatches ? nonMatches.length : nonMatchCount;\n        const totalHandled = matchCount + chosenNonMatchCount + duplicateCount;\n        debug(`candidateCount: ${candidateCount}, matches: ${matchCount}, nonMatches: ${chosenNonMatchCount}, duplicateCount: ${duplicateCount}, matchErrorCount: ${matchErrorCount}`);\n        debug(`We got result for ${totalHandled} / ${candidateCount} retrieved candidates`);\n        if (totalHandled !== candidateCount) {\n          debug(`WARNING: Missing results for ${candidateCount - totalHandled} candidates`);\n          return;\n        }\n        return;\n      }\n\n      function getMatchState(state, stopReason, matchErrorCount) {\n        debugData(`${JSON.stringify(state)}`);\n        debug(`We had ${matchErrorCount} retrieved candidates that errored in matchDetection.`);\n        debug(`Queries left ${state.queriesLeft}, Searches for current query left: ${state.resultSetOffset && state.resultSetOffset <= state.totalRecords}, non-retrieved records: ${state.totalRecords - state.queryCandidateCounter}, maxedQueries (${state.maxedQueries.length}): ${state.maxedQueries}`);\n\n        debugData(`StopReason: <${stopReason}>`);\n\n        const searchesLeft = state.resultSetOffset && state.resultSetOffset <= state.totalRecords;\n        const nonRetrieved = searchesLeft ? state.totalRecords - state.queryCandidateCounter : 0;\n        debugData(`nonRetrieved: ${nonRetrieved}`);\n\n        // matchStatus.stopReason: string ('maxMatches','maxCandidates','maxedQueries', empty string/undefined), reason for stopping retrieving or handling the candidate records\n        // 'maxMatches' and 'maxCandidates' are in stopReason, 'maxedQueries', 'conversionFailures' and 'matchErrors' are created here\n\n        if (state.queriesLeft > 0 || nonRetrieved > 0 || state.maxedQueries.length > 0 || matchErrorCount > 0) {\n          const maxedQueriesStopReason = state.maxedQueries.length > 0 ? 'maxedQueries' : undefined;\n          const matchErrorsStopReason = matchErrorCount > 0 ? 'matchErrors' : undefined;\n          const newStopReason = stopReason === '' || stopReason === undefined ? maxedQueriesStopReason || matchErrorsStopReason : stopReason;\n          debugData(`MaxedQueriesStopReason: <${maxedQueriesStopReason}>`);\n          debugData(`MatchErrorsStopReason <${matchErrorsStopReason}>`);\n          debugData(`NewStopReason: <${newStopReason}>`);\n          debug(`Match status: false`);\n          return {status: false, stopReason: newStopReason};\n        }\n\n        debug(`Match status: true`);\n        return {status: true, stopReason};\n      }\n    }\n\n    // NOTES:\n    // - we could optimize by creating the featureSet for the incoming record once and using it for all database/candidateRecords\n    // - if creating the featureSet for the incoming record fails we have an unprocessable entity\n    // - if creating the featureSet for a candidate record fails we could skip that candidate - but list the case as a detectionFailure, same as conversionFailures\n\n    function iterateRecords({records, recordSetSize, maxMatches, matches = [], nonMatches = [], recordMatches = [], recordNonMatches = [], recordCount = 0, recordDuplicateCount = 0, recordNonMatchCount = 0, recordMatchErrors = []}) {\n\n      // recordSetSize : total amount of records in the current record set\n      // recordCount : amount of records from the current record set that have been handled\n      // maxMatches : setting for maximum amount found by current matcher job before the matcher job is stopped\n      // recordDuplicateCount : amount of records from the current record set that are already included in matches/nonMatches results\n      // recordNonMatchCount: amount of records from the current record set that are nonMatches (only is returnNonMatches setting is false)\n\n      // records : non-handled records in the current record set\n      // matches : found matches in the current matcher job\n      // recordMatches : found matches in the current record set\n      // recordNonMatches : found nonMatches in the current record set (only if returnNonMatches setting is true)\n      // recordMatchErrors: errored matchDetection in the current record set\n\n      const [candidate] = records;\n      const newRecordCount = candidate ? recordCount + 1 : recordCount;\n\n      // The matcher uses same matchDetection strategy for candidates from all candidate-searches -> matchDetection result for the same candidate is always same\n      // Exceptions would happen if the candidate would have been updated in the database between candidate searches\n      // Note that if returnNonMatches is false, matcher won't remember candidates that didn't match, so they will be matched again everytime they are retrieved by\n      // different candidate search queries. Same candidate search query won't have duplicate records.\n\n      /* We could optimize and detect all retrieved candidates at once\n      const candidateRecords = records.map(record => record.record);\n      const recordsIsArray = Array.isArray(candidateRecords);\n      debug(`records is an array: ${recordsIsArray}`);\n      const result = detect(record, candidateRecords);\n      debugData(`${JSON.stringify(result)}`);\n      */\n\n      if (candidate) {\n        if (candidateNotInMatches(matches.concat(nonMatches), candidate)) {\n          const {record: candidateRecord, id: candidateId} = candidate;\n          const recordBExternal = {id: candidate.id, recordSource: 'databaseRecord', label: `db-${candidate.id}`};\n          try {\n            debug(`Running matchDetection for record ${candidateId} (${newRecordCount}/${recordSetSize})`);\n            // we should handle errors from detection somehow - ie. cases where either record or candidateRecord errors\n            const detectionResult = detect({recordA: record, recordB: candidateRecord, recordAExternal: recordExternal, recordBExternal});\n\n            return handleDetectionResult(detectionResult, candidateId, candidateRecord);\n          } catch (error) {\n            debug(`MatchDetection errored: database record ${candidateId}: ${error}`);\n\n            const matchError = {status: 422, payload: {message: `Matching errored for database record ${candidateId}. ${error.message}.`, id: candidateId}};\n            const newRecordMatchErrors = recordMatchErrors.concat(matchError);\n            return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount, recordNonMatchCount, recordMatchErrors: newRecordMatchErrors});\n          }\n        }\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount: recordDuplicateCount + 1, recordNonMatchCount, recordMatchErrors});\n      }\n\n      debug(`No more candidates, record set (${recordCount}/${recordSetSize}) done, ${recordMatches.length} matches found, ${recordDuplicateCount} candidates already handled, ${returnNonMatches ? `${recordNonMatches.length}` : `${recordNonMatchCount}`} nonMatches found.`);\n      return {matches: recordMatches, nonMatches: returnNonMatches ? recordNonMatches : [], duplicateCount: recordDuplicateCount, nonMatchCount: recordNonMatchCount, matchErrors: recordMatchErrors};\n\n      function handleDetectionResult(detectionResult, candidateId, candidateRecord) {\n        debugData(`MatchDetection results for ${candidateId} (${newRecordCount}/${recordSetSize}): ${JSON.stringify(detectionResult)}`);\n\n        if (detectionResult.match || returnNonMatches) {\n          debug(`${detectionResult.match ? `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is a match!` : `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is NOT a match!`}`);\n          debugData(`Strategy: ${JSON.stringify(detectionResult.strategy)}, Threshold: ${JSON.stringify(detectionResult.threshold)}`);\n\n          const matchResult = {\n            probability: detectionResult.probability,\n            candidate: {\n              id: candidateId,\n              record: candidateRecord\n            }\n          };\n          const strategyResult = {\n            strategy: detectionResult.strategy,\n            threshold: detectionResult.threshold\n          };\n          const newMatch = returnStrategy ? {...matchResult, ...strategyResult} : {...matchResult};\n\n          debugData(`${JSON.stringify(newMatch)}`);\n\n          return handleRecordMatch(detectionResult.match, newMatch);\n        }\n\n        const newRecordNonMatchCount = recordNonMatchCount + 1;\n        debugData(`- Total nonMatches after this detection: ${newRecordNonMatchCount}`);\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount, recordNonMatchCount: newRecordNonMatchCount, recordMatchErrors});\n      }\n\n      function handleRecordMatch(isMatch, newMatch) {\n        const newRecordMatches = isMatch ? recordMatches.concat(newMatch) : recordMatches;\n        const newRecordNonMatches = isMatch ? recordNonMatches : recordNonMatches.concat(newMatch);\n        const newRecordNonMatchCount = isMatch ? recordNonMatchCount : recordNonMatchCount + 1;\n\n        debugData(`- Total matches after this detection: ${matches.concat(newRecordMatches).length} (max: ${maxMatches})`);\n\n        if (returnNonMatches) {\n          debugData(`- Total nonMatches after this detection: ${nonMatches.concat(newRecordNonMatches).length}`);\n        }\n        debugData(`- Total nonMatchCount after this detection: ${recordNonMatchCount}`);\n\n        if (maxMatchesFound({matches: matches.concat(newRecordMatches), maxMatches})) {\n          debug(`MaxMatches (${maxMatches}) reached, handled candidates in record set: ${newRecordCount} non-handled candidates in record set ${recordSetSize - newRecordCount}`);\n          return {matches: newRecordMatches, nonMatches: returnNonMatches ? newRecordNonMatches : [], duplicateCount: recordDuplicateCount, nonMatchCount: newRecordNonMatchCount, matchErrors: recordMatchErrors};\n        }\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches: newRecordMatches, recordCount: newRecordCount, recordNonMatches: returnNonMatches ? newRecordNonMatches : [], duplicateCount: recordDuplicateCount, recordNonMatchCount: newRecordNonMatchCount, matchErrors: recordMatchErrors});\n      }\n\n      function candidateNotInMatches(matches, candidate) {\n        debug(`Checking that record ${candidate.id} is not already included in ${matches.length} matches/nonMatches`);\n        const newCandidateId = candidate.id;\n        debugData(`newCandidateId: ${newCandidateId}`);\n        const result = matches.find(({candidate}) => candidate.id === newCandidateId);\n        debugData(`Result: ${result}`);\n        if (result) {\n          debug(`${candidate.id} was already handled.`);\n          return false;\n        }\n        debug(`${candidate.id} not found in matches/nonMatches`);\n        return true;\n      }\n    }\n\n    function maxMatchesFound({matches, maxMatches}) {\n      if (maxMatches && matches.length >= maxMatches) {\n        debug(`Stopping recordSet iteration: maxMatches (${maxMatches}) for matcher job found.`);\n        return true;\n      }\n    }\n  }\n};\n"],
-  "mappings": "AAAA,OAAO,uBAAuB;AAC9B,OAAO,4BAA4B,qBAAqB;AACxD,OAAO,+BAA+B,oBAAoB;AAG1D,SAAQ,iBAAiB;AAEzB,eAAe,CAAC,EAAC,WAAW,kBAAkB,QAAQ,eAAe,aAAa,GAAG,gBAAgB,IAAI,iBAAiB,OAAO,cAAc,OAAO,mBAAmB,MAAK,MAAM;AAClL,QAAM,QAAQ,kBAAkB,yCAAyC;AACzE,QAAM,YAAY,MAAM,OAAO,MAAM;AAErC,YAAU,qBAAqB,KAAK,UAAU,gBAAgB,CAAC,EAAE;AACjE,YAAU,kBAAkB,KAAK,UAAU,aAAa,CAAC,EAAE;AAC3D,YAAU,eAAe,KAAK,UAAU,UAAU,CAAC,EAAE;AACrD,YAAU,kBAAkB,KAAK,UAAU,aAAa,CAAC,EAAE;AAC3D,YAAU,mBAAmB,KAAK,UAAU,cAAc,CAAC,EAAE;AAC7D,YAAU,gBAAgB,KAAK,UAAU,WAAW,CAAC,EAAE;AACvD,YAAU,qBAAqB,KAAK,UAAU,gBAAgB,CAAC,EAAE;AAGjE,QAAM,SAAS,yBAAyB,kBAAkB,cAAc;AAExE,SAAO;AAEP,iBAAe,cAAc,EAAC,QAAQ,iBAAiB,EAAC,cAAc,kBAAkB,OAAO,KAAI,EAAC,GAAG;AAErG,UAAM,EAAC,OAAM,IAAI,MAAM,sBAAsB,EAAC,GAAG,eAAe,QAAQ,eAAe,eAAc,CAAC;AACtG,WAAO,QAAQ,CAAC,CAAC;AAejB,mBAAe,QAAQ,EAAC,eAAe,CAAC,GAAG,UAAU,CAAC,GAAG,iBAAiB,GAAG,aAAa,CAAC,GAAG,iBAAiB,GAAG,gBAAgB,GAAG,cAAc,CAAC,EAAC,GAAG;AACtJ,gBAAU,kCAAkC;AAC5C,YAAM,EAAC,SAAS,GAAG,MAAK,IAAI,MAAM,OAAO,YAAY;AAErD,gBAAU,kBAAkB,KAAK,UAAU,KAAK,CAAC,cAAc,QAAQ,MAAM,qBAAqB,cAAc,iBAAiB,WAAW,MAAM,oBAAoB,aAAa,kBAAkB,YAAY,MAAM,EAAE;AACzN,YAAM,gBAAgB,QAAQ;AAC9B,YAAM,oBAAoB,iBAAiB;AAG3C,UAAI,gBAAgB,GAAG;AACrB,eAAO,gBAAgB;AAAA,MACzB;AAEA,UAAI,MAAM,cAAc,GAAG;AACzB,cAAM,oBAAoB,MAAM,aAAa,QAAQ,MAAM,KAAK,mBAAmB,MAAM,WAAW,eAAe;AACnH,eAAO,QAAQ,EAAC,cAAc,OAAO,SAAS,gBAAgB,mBAAmB,YAAY,eAAe,gBAAgB,YAAW,CAAC;AAAA,MAC1I;AAEA,YAAM,wEAAwE,QAAQ,MAAM,EAAE;AAC9F,aAAO,aAAa,EAAC,SAAS,OAAO,YAAY,IAAI,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,YAAW,CAAC;AAE/I,eAAS,kBAAkB;AACzB,cAAM,0BAA0B,aAAa,qDAAqD,MAAM,aAAa,eAAe,MAAM,KAAK,EAAE;AAEjJ,cAAM,cAAc,eAAe,EAAC,SAAS,eAAe,YAAY,SAAS,YAAY,cAAa,CAAC;AAE3G,cAAM,oBAAoB,iBAAiB,YAAY;AACvD,cAAM,mBAAmB,gBAAgB,YAAY;AACrD,cAAM,EAAC,YAAY,eAAe,eAAc,IAAI,kBAAkB,aAAa,SAAS,YAAY,WAAW;AAEnH,YAAI,gBAAgB,EAAC,SAAS,YAAY,WAAU,CAAC,GAAG;AACtD,iBAAO,aAAa,EAAC,SAAS,YAAY,OAAO,YAAY,cAAc,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,QAC3O;AAEA,YAAI,uBAAuB,mBAAmB,aAAa,GAAG;AAC5D,iBAAO,aAAa,EAAC,SAAS,YAAY,OAAO,YAAY,iBAAiB,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,QAC9O;AAEA,eAAO,QAAQ,EAAC,cAAc,OAAO,SAAS,YAAY,gBAAgB,mBAAmB,YAAY,eAAe,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,MAC1N;AAEA,eAAS,kBAAkB,aAAaA,UAASC,aAAYC,cAAa;AACxE,kBAAU,4CAA4C,YAAY,QAAQ,MAAM,EAAE;AAClF,YAAI,kBAAkB;AACpB,oBAAU,+CAA+C,YAAY,WAAW,MAAM,EAAE;AAAA,QAC1F;AAEA,cAAM,aAAaF,SAAQ,OAAO,cAAc,SAAS,YAAY,OAAO,IAAI,YAAY,OAAO;AACnG,cAAM,gBAAgB,mBAAmBC,YAAW,OAAO,cAAc,SAAS,YAAY,UAAU,IAAI,YAAY,UAAU,IAAI,CAAC;AACvI,cAAM,iBAAiBC,aAAY,OAAO,YAAY,WAAW;AAEjE,kBAAU,8BAA8B,WAAW,MAAM,EAAE;AAC3D,YAAI,kBAAkB;AACpB,oBAAU,iCAAiC,cAAc,MAAM,EAAE;AAAA,QACnE;AAEA,kBAAU,gBAAgB,KAAK,UAAU,WAAW,CAAC,EAAE;AACvD,kBAAU,oBAAoB,KAAK,UAAUA,YAAW,CAAC,mCAAmC,KAAK,UAAU,YAAY,WAAW,CAAC,sBAAsB,KAAK,UAAU,cAAc,CAAC,EAAE;AAEzL,kBAAU,kCAAkC,eAAe,MAAM,EAAE;AAEnE,eAAO,EAAC,YAAY,eAAe,eAAc;AAAA,MACnD;AAEA,eAAS,SAASF,UAAS;AACzB,kBAAU,gBAAgB,MAAM,KAAK,aAAa;AAClD,eAAOA,SAAQ,IAAI,CAAC,WAAW,EAAC,GAAG,OAAO,YAAY,MAAM,MAAK,EAAE;AAAA,MACrE;AAEA,eAAS,uBAAuBG,iBAAgBC,gBAAe;AAC7D,kBAAU,gDAAgD,iBAAiB,UAAUA,cAAa,GAAG;AACrG,YAAIA,kBAAiBD,mBAAkBC,gBAAe;AACpD,gBAAM,gEAAgED,eAAc,MAAMC,cAAa,sBAAsB;AAC7H,iBAAO;AAAA,QACT;AAAA,MACF;AAAA,IACF;AAmBA,aAAS,aAAa,EAAC,SAAS,OAAO,YAAY,YAAY,gBAAgB,gBAAgB,eAAe,YAAW,GAAG;AAC1H,YAAM,kBAAkB,YAAY;AACpC,kBAAY,EAAC,SAAS,YAAY,gBAAgB,gBAAgB,eAAe,gBAAe,CAAC;AACjG,YAAM,cAAc,cAAc,OAAO,YAAY,eAAe;AAEpE,YAAM,gBAAgB,mBAAmB,EAAC,SAAS,aAAa,YAAY,eAAc,IAAI,EAAC,SAAS,aAAa,eAAc;AACnI,YAAM,SAAS;AACf,gBAAU,GAAG,KAAK,UAAU,MAAM,CAAC,EAAE;AACrC,aAAO;AAIP,eAAS,YAAY,EAAC,SAAAJ,UAAS,YAAAC,aAAY,gBAAAE,iBAAgB,gBAAAE,iBAAgB,eAAAC,gBAAe,iBAAAC,iBAAe,GAAG;AAC1G,cAAM,aAAaP,SAAQ;AAC3B,kBAAU,sBAAsB,gBAAgB,EAAE;AAClD,cAAM,sBAAsB,mBAAmBC,YAAW,SAASK;AACnE,cAAM,eAAe,aAAa,sBAAsBD;AACxD,cAAM,mBAAmBF,eAAc,cAAc,UAAU,iBAAiB,mBAAmB,qBAAqBE,eAAc,sBAAsBE,gBAAe,EAAE;AAC7K,cAAM,qBAAqB,YAAY,MAAMJ,eAAc,uBAAuB;AAClF,YAAI,iBAAiBA,iBAAgB;AACnC,gBAAM,gCAAgCA,kBAAiB,YAAY,aAAa;AAChF;AAAA,QACF;AACA;AAAA,MACF;AAEA,eAAS,cAAcK,QAAOC,aAAYF,kBAAiB;AACzD,kBAAU,GAAG,KAAK,UAAUC,MAAK,CAAC,EAAE;AACpC,cAAM,UAAUD,gBAAe,uDAAuD;AACtF,cAAM,gBAAgBC,OAAM,WAAW,sCAAsCA,OAAM,mBAAmBA,OAAM,mBAAmBA,OAAM,YAAY,4BAA4BA,OAAM,eAAeA,OAAM,qBAAqB,mBAAmBA,OAAM,aAAa,MAAM,MAAMA,OAAM,YAAY,EAAE;AAEnS,kBAAU,gBAAgBC,WAAU,GAAG;AAEvC,cAAM,eAAeD,OAAM,mBAAmBA,OAAM,mBAAmBA,OAAM;AAC7E,cAAM,eAAe,eAAeA,OAAM,eAAeA,OAAM,wBAAwB;AACvF,kBAAU,iBAAiB,YAAY,EAAE;AAKzC,YAAIA,OAAM,cAAc,KAAK,eAAe,KAAKA,OAAM,aAAa,SAAS,KAAKD,mBAAkB,GAAG;AACrG,gBAAM,yBAAyBC,OAAM,aAAa,SAAS,IAAI,iBAAiB;AAChF,gBAAM,wBAAwBD,mBAAkB,IAAI,gBAAgB;AACpE,gBAAM,gBAAgBE,gBAAe,MAAMA,gBAAe,SAAY,0BAA0B,wBAAwBA;AACxH,oBAAU,4BAA4B,sBAAsB,GAAG;AAC/D,oBAAU,0BAA0B,qBAAqB,GAAG;AAC5D,oBAAU,mBAAmB,aAAa,GAAG;AAC7C,gBAAM,qBAAqB;AAC3B,iBAAO,EAAC,QAAQ,OAAO,YAAY,cAAa;AAAA,QAClD;AAEA,cAAM,oBAAoB;AAC1B,eAAO,EAAC,QAAQ,MAAM,YAAAA,YAAU;AAAA,MAClC;AAAA,IACF;AAOA,aAAS,eAAe,EAAC,SAAS,eAAe,YAAAC,aAAY,UAAU,CAAC,GAAG,aAAa,CAAC,GAAG,gBAAgB,CAAC,GAAG,mBAAmB,CAAC,GAAG,cAAc,GAAG,uBAAuB,GAAG,sBAAsB,GAAG,oBAAoB,CAAC,EAAC,GAAG;AAclO,YAAM,CAAC,SAAS,IAAI;AACpB,YAAM,iBAAiB,YAAY,cAAc,IAAI;AAerD,UAAI,WAAW;AACb,YAAI,sBAAsB,QAAQ,OAAO,UAAU,GAAG,SAAS,GAAG;AAChE,gBAAM,EAAC,QAAQ,iBAAiB,IAAI,YAAW,IAAI;AACnD,gBAAM,kBAAkB,EAAC,IAAI,UAAU,IAAI,cAAc,kBAAkB,OAAO,MAAM,UAAU,EAAE,GAAE;AACtG,cAAI;AACF,kBAAM,qCAAqC,WAAW,KAAK,cAAc,IAAI,aAAa,GAAG;AAE7F,kBAAM,kBAAkB,OAAO,EAAC,SAAS,QAAQ,SAAS,iBAAiB,iBAAiB,gBAAgB,gBAAe,CAAC;AAE5H,mBAAO,sBAAsB,iBAAiB,aAAa,eAAe;AAAA,UAC5E,SAAS,OAAO;AACd,kBAAM,2CAA2C,WAAW,KAAK,KAAK,EAAE;AAExE,kBAAM,aAAa,EAAC,QAAQ,KAAK,SAAS,EAAC,SAAS,wCAAwC,WAAW,KAAK,MAAM,OAAO,KAAK,IAAI,YAAW,EAAC;AAC9I,kBAAM,uBAAuB,kBAAkB,OAAO,UAAU;AAChE,mBAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,qBAAqB,mBAAmB,qBAAoB,CAAC;AAAA,UACzO;AAAA,QACF;AAEA,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,uBAAuB,GAAG,qBAAqB,kBAAiB,CAAC;AAAA,MAC7O;AAEA,YAAM,mCAAmC,WAAW,IAAI,aAAa,WAAW,cAAc,MAAM,mBAAmB,oBAAoB,gCAAgC,mBAAmB,GAAG,iBAAiB,MAAM,KAAK,GAAG,mBAAmB,EAAE,oBAAoB;AACzQ,aAAO,EAAC,SAAS,eAAe,YAAY,mBAAmB,mBAAmB,CAAC,GAAG,gBAAgB,sBAAsB,eAAe,qBAAqB,aAAa,kBAAiB;AAE9L,eAAS,sBAAsB,iBAAiB,aAAa,iBAAiB;AAC5E,kBAAU,8BAA8B,WAAW,KAAK,cAAc,IAAI,aAAa,MAAM,KAAK,UAAU,eAAe,CAAC,EAAE;AAE9H,YAAI,gBAAgB,SAAS,kBAAkB;AAC7C,gBAAM,GAAG,gBAAgB,QAAQ,UAAU,WAAW,KAAK,cAAc,IAAI,aAAa,kBAAkB,UAAU,WAAW,KAAK,cAAc,IAAI,aAAa,mBAAmB,EAAE;AAC1L,oBAAU,aAAa,KAAK,UAAU,gBAAgB,QAAQ,CAAC,gBAAgB,KAAK,UAAU,gBAAgB,SAAS,CAAC,EAAE;AAE1H,gBAAM,cAAc;AAAA,YAClB,aAAa,gBAAgB;AAAA,YAC7B,WAAW;AAAA,cACT,IAAI;AAAA,cACJ,QAAQ;AAAA,YACV;AAAA,UACF;AACA,gBAAM,iBAAiB;AAAA,YACrB,UAAU,gBAAgB;AAAA,YAC1B,WAAW,gBAAgB;AAAA,UAC7B;AACA,gBAAM,WAAW,iBAAiB,EAAC,GAAG,aAAa,GAAG,eAAc,IAAI,EAAC,GAAG,YAAW;AAEvF,oBAAU,GAAG,KAAK,UAAU,QAAQ,CAAC,EAAE;AAEvC,iBAAO,kBAAkB,gBAAgB,OAAO,QAAQ;AAAA,QAC1D;AAEA,cAAM,yBAAyB,sBAAsB;AACrD,kBAAU,4CAA4C,sBAAsB,EAAE;AAE9E,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,qBAAqB,wBAAwB,kBAAiB,CAAC;AAAA,MAC3O;AAEA,eAAS,kBAAkB,SAAS,UAAU;AAC5C,cAAM,mBAAmB,UAAU,cAAc,OAAO,QAAQ,IAAI;AACpE,cAAM,sBAAsB,UAAU,mBAAmB,iBAAiB,OAAO,QAAQ;AACzF,cAAM,yBAAyB,UAAU,sBAAsB,sBAAsB;AAErF,kBAAU,yCAAyC,QAAQ,OAAO,gBAAgB,EAAE,MAAM,UAAUA,WAAU,GAAG;AAEjH,YAAI,kBAAkB;AACpB,oBAAU,4CAA4C,WAAW,OAAO,mBAAmB,EAAE,MAAM,EAAE;AAAA,QACvG;AACA,kBAAU,+CAA+C,mBAAmB,EAAE;AAE9E,YAAI,gBAAgB,EAAC,SAAS,QAAQ,OAAO,gBAAgB,GAAG,YAAAA,YAAU,CAAC,GAAG;AAC5E,gBAAM,eAAeA,WAAU,gDAAgD,cAAc,yCAAyC,gBAAgB,cAAc,EAAE;AACtK,iBAAO,EAAC,SAAS,kBAAkB,YAAY,mBAAmB,sBAAsB,CAAC,GAAG,gBAAgB,sBAAsB,eAAe,wBAAwB,aAAa,kBAAiB;AAAA,QACzM;AAEA,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,kBAAkB,aAAa,gBAAgB,kBAAkB,mBAAmB,sBAAsB,CAAC,GAAG,gBAAgB,sBAAsB,qBAAqB,wBAAwB,aAAa,kBAAiB,CAAC;AAAA,MACvU;AAEA,eAAS,sBAAsBV,UAASW,YAAW;AACjD,cAAM,wBAAwBA,WAAU,EAAE,+BAA+BX,SAAQ,MAAM,qBAAqB;AAC5G,cAAM,iBAAiBW,WAAU;AACjC,kBAAU,mBAAmB,cAAc,EAAE;AAC7C,cAAM,SAASX,SAAQ,KAAK,CAAC,EAAC,WAAAW,WAAS,MAAMA,WAAU,OAAO,cAAc;AAC5E,kBAAU,WAAW,MAAM,EAAE;AAC7B,YAAI,QAAQ;AACV,gBAAM,GAAGA,WAAU,EAAE,uBAAuB;AAC5C,iBAAO;AAAA,QACT;AACA,cAAM,GAAGA,WAAU,EAAE,kCAAkC;AACvD,eAAO;AAAA,MACT;AAAA,IACF;AAEA,aAAS,gBAAgB,EAAC,SAAS,YAAAD,YAAU,GAAG;AAC9C,UAAIA,eAAc,QAAQ,UAAUA,aAAY;AAC9C,cAAM,6CAA6CA,WAAU,0BAA0B;AACvF,eAAO;AAAA,MACT;AAAA,IACF;AAAA,EACF;AACF;",
+  "sourcesContent": ["import createDebugLogger from 'debug';\nimport createSearchInterface, * as candidateSearch from './candidate-search/index.js';\nimport createDetectionInterface, * as matchDetection from './match-detection/index.js';\n//import inspect from 'util';\n\nexport {candidateSearch, matchDetection};\n\nexport default ({detection: detectionOptions, search: searchOptions, maxMatches = 1, maxCandidates = 25, returnStrategy = false, returnQuery = false, returnNonMatches = false}) => {\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:index');\n  const debugData = debug.extend('data');\n\n  debugData(`DetectionOptions: ${JSON.stringify(detectionOptions)}`);\n  debugData(`SearchOptions: ${JSON.stringify(searchOptions)}`);\n  debugData(`MaxMatches: ${JSON.stringify(maxMatches)}`);\n  debugData(`MaxCandidates: ${JSON.stringify(maxCandidates)}`);\n  debugData(`ReturnStrategy: ${JSON.stringify(returnStrategy)}`);\n  debugData(`ReturnQuery: ${JSON.stringify(returnQuery)}`);\n  debugData(`ReturnNonMatches: ${JSON.stringify(returnNonMatches)}`);\n\n\n  const detect = createDetectionInterface(detectionOptions, returnStrategy);\n\n  return prepareSearch;\n\n  async function prepareSearch({record, recordExternal = {recordSource: 'incomingRecord', label: 'ic'}}) {\n\n    const {search} = await createSearchInterface({...searchOptions, record, maxCandidates, recordExternal});\n    return iterate({});\n\n    // candidateCount : amount of candidate records retrived from SRU for matching, NOT including current record set\n    // matches : candidates that have been detected as matches by current matcher job\n    // nonMatches : candidates that have been detected as non-matches by current matcher job (only if returnNonMatches is 'true')\n    // duplicateCount : amount of candidate records that were retrieved from the SRU but not handled further because they were already found in the matches/nonMatches\n\n    // state.totalRecords : amount of candidate records available to the current query (undefined, if there was no queries left)\n    // state.query : current query (undefined if there was no queries left)\n    // state.searchCounter : sequence for current search for current query (undefined, if there we no queries left)\n    // state.queryCandidateCounter: amount of candidate records retrieved from SRU for matching for current query, including the current record set (undefined if there were no queries left)\n    // state.queriesLeft : amount of queries left\n    // state.queryCounter : sequence for current query\n    // state.maxedQueries : queries that resulted in more than serverMaxResults hits\n\n    async function iterate({initialState = {}, matches = [], candidateCount = 0, nonMatches = [], duplicateCount = 0, nonMatchCount = 0, matchErrors = []}) {\n      debugData(`Starting next matcher iteration.`);\n      const {records, ...state} = await search(initialState);\n\n      debugData(`Current state: ${JSON.stringify(state)}, matches: ${matches.length}, candidateCount: ${candidateCount}, nonMatches: ${nonMatches.length}, nonMatchCount: ${nonMatchCount}, matchErrors: ${matchErrors.length}`);\n      const recordSetSize = records.length;\n      const newCandidateCount = candidateCount + recordSetSize;\n\n\n      if (recordSetSize > 0) {\n        return handleRecordSet();\n      }\n\n      if (state.queriesLeft > 0) {\n        debug(`Empty record set ${state.searchCounter} for ${state.query}, but there are ${state.queriesLeft} queries left`);\n        return iterate({initialState: state, matches, candidateCount: newCandidateCount, nonMatches, nonMatchCount, duplicateCount, matchErrors});\n      }\n\n      debug(`No (more) candidate records to check, no more queries left, matches: ${matches.length}`);\n      return returnResult({matches, state, stopReason: '', nonMatches, nonMatchCount, candidateCount: newCandidateCount, duplicateCount, matchErrors});\n\n      function handleRecordSet() {\n        debug(`Checking record set of ${recordSetSize} candidate records for possible matches, found by ${state.searchCounter} search for ${state.query}`);\n\n        const matchResult = iterateRecords({records, recordSetSize, maxMatches, matches, nonMatches, nonMatchCount});\n\n        const newDuplicateCount = duplicateCount + matchResult.duplicateCount;\n        const newNonMatchCount = nonMatchCount + matchResult.nonMatchCount;\n        const {newMatches, newNonMatches, newMatchErrors} = handleMatchResult(matchResult, matches, nonMatches, matchErrors);\n\n        if (maxMatchesFound({matches: newMatches, maxMatches})) {\n          return returnResult({matches: newMatches, state, stopReason: 'maxMatches', nonMatches: newNonMatches, duplicateCount: newDuplicateCount, candidateCount: newCandidateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n        }\n\n        if (maxCandidatesRetrieved(newCandidateCount, maxCandidates)) {\n          return returnResult({matches: newMatches, state, stopReason: 'maxCandidates', nonMatches: newNonMatches, duplicateCount: newDuplicateCount, candidateCount: newCandidateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n        }\n\n        return iterate({initialState: state, matches: newMatches, candidateCount: newCandidateCount, nonMatches: newNonMatches, duplicateCount: newDuplicateCount, nonMatchCount: newNonMatchCount, matchErrors: newMatchErrors});\n      }\n\n      function handleMatchResult(matchResult, matches, nonMatches, matchErrors) {\n        debugData(`- Amount of new matches from record set: ${matchResult.matches.length}`);\n        if (returnNonMatches) {\n          debugData(`- Amount of new nonMatches from record set: ${matchResult.nonMatches.length}`);\n        }\n\n        const newMatches = matches.concat(returnQuery ? addQuery(matchResult.matches) : matchResult.matches);\n        const newNonMatches = returnNonMatches ? nonMatches.concat(returnQuery ? addQuery(matchResult.nonMatches) : matchResult.nonMatches) : [];\n        const newMatchErrors = matchErrors.concat(matchResult.matchErrors);\n\n        debugData(`- Total amount of matches: ${newMatches.length}`);\n        if (returnNonMatches) {\n          debugData(`- Total amount of nonMatches: ${newNonMatches.length}`);\n        }\n\n        debugData(`MatchResult: ${JSON.stringify(matchResult)}`);\n        debugData(`Old matchErrors: ${JSON.stringify(matchErrors)}, matchErrors from matchResult: ${JSON.stringify(matchResult.matchErrors)}, New matchErrors: ${JSON.stringify(newMatchErrors)}`);\n\n        debugData(`- Total amount of matchErrors: ${newMatchErrors.length}`);\n\n        return {newMatches, newNonMatches, newMatchErrors};\n      }\n\n      function addQuery(matches) {\n        debugData(`Adding query ${state.query} to matches`);\n        return matches.map((match) => ({...match, matchQuery: state.query}));\n      }\n\n      function maxCandidatesRetrieved(candidateCount, maxCandidates) {\n        debugData(`Total amount of candidate records retrieved: ${newCandidateCount} (max: ${maxCandidates})`);\n        if (maxCandidates && candidateCount >= maxCandidates) {\n          debug(`Stopped matching because maximum number of candidate records ${candidateCount} / ${maxCandidates} have been retrieved`);\n          return true;\n        }\n      }\n    }\n\n    // matches : array of matching candidate records\n    // nonMatches : array of nonMatching candidate records (if returnNonMatches option is true, otherwise empty array)\n    // - candidate.id\n    // - candidate.record\n    // - probability\n    // - strategy (if returnStrategy option is true)\n    // - threshold (if returnStrategy option is true)\n    // - matchQuery (if returnQuery option is true)\n    // failures: array of conversionFailures from candidate-search and matchErrors from matchDetection in error format {status, payload: {message, id}} if returnFailures is true\n\n    // we could have here also returnRecords/returnMatchRecords/returnNonMatchRecord options that could be turned false for not to return actual record data\n\n    // matchStatus.status: boolean, true if matcher retrieved and handled all found candidate records, false if it did not\n    // matchStatus.stopReason: string ('maxMatches','maxCandidates','maxedQueries','conversionFailures', empty string/undefined), reason for stopping retrieving or handling the candidate records\n    // - only one stopReason is returned (if there would be several possible stopReasons, stopReason is picked in the above order)\n    // - currently stopReason can be non-empty also in cases where status is true, if matcher hit the stop reason when handling the last available candidate record\n\n    function returnResult({matches, state, stopReason, nonMatches, duplicateCount, candidateCount, nonMatchCount, matchErrors}) {\n      const matchErrorCount = matchErrors.length;\n      checkCounts({matches, nonMatches, candidateCount, duplicateCount, nonMatchCount, matchErrorCount});\n      const matchStatus = getMatchState(state, stopReason, matchErrorCount);\n      // add nonMatches to result only if returnNonMatches is 'true', otherwise nonMatches have not been gathered\n      const matchesResult = returnNonMatches ? {matches, matchStatus, nonMatches, candidateCount} : {matches, matchStatus, candidateCount};\n      const result = matchesResult;\n      debugData(`${JSON.stringify(result)}`);\n      return result;\n\n      // note that in cases where the matching has been stopped because of maxMatches checkCounts won't (in most cases) match\n\n      function checkCounts({matches, nonMatches, candidateCount, duplicateCount, nonMatchCount, matchErrorCount}) {\n        const matchCount = matches.length;\n        debugData(`Return nonMatches: ${returnNonMatches}`);\n        const chosenNonMatchCount = returnNonMatches ? nonMatches.length : nonMatchCount;\n        const totalHandled = matchCount + chosenNonMatchCount + duplicateCount;\n        debug(`candidateCount: ${candidateCount}, matches: ${matchCount}, nonMatches: ${chosenNonMatchCount}, duplicateCount: ${duplicateCount}, matchErrorCount: ${matchErrorCount}`);\n        debug(`We got result for ${totalHandled} / ${candidateCount} retrieved candidates`);\n        if (totalHandled !== candidateCount) {\n          debug(`WARNING: Missing results for ${candidateCount - totalHandled} candidates`);\n          return;\n        }\n        return;\n      }\n\n      function getMatchState(state, stopReason, matchErrorCount) {\n        debugData(`${JSON.stringify(state)}`);\n        debug(`We had ${matchErrorCount} retrieved candidates that errored in matchDetection.`);\n        debug(`Queries left ${state.queriesLeft}, Searches for current query left: ${state.resultSetOffset && state.resultSetOffset <= state.totalRecords}, non-retrieved records: ${state.totalRecords - state.queryCandidateCounter}, maxedQueries (${state.maxedQueries.length}): ${state.maxedQueries}`);\n\n        debugData(`StopReason: <${stopReason}>`);\n\n        const searchesLeft = state.resultSetOffset && state.resultSetOffset <= state.totalRecords;\n        const nonRetrieved = searchesLeft ? state.totalRecords - state.queryCandidateCounter : 0;\n        debugData(`nonRetrieved: ${nonRetrieved}`);\n\n        // matchStatus.stopReason: string ('maxMatches','maxCandidates','maxedQueries', empty string/undefined), reason for stopping retrieving or handling the candidate records\n        // 'maxMatches' and 'maxCandidates' are in stopReason, 'maxedQueries', 'conversionFailures' and 'matchErrors' are created here\n\n        if (state.queriesLeft > 0 || nonRetrieved > 0 || state.maxedQueries.length > 0 || matchErrorCount > 0) {\n          const maxedQueriesStopReason = state.maxedQueries.length > 0 ? 'maxedQueries' : undefined;\n          const matchErrorsStopReason = matchErrorCount > 0 ? 'matchErrors' : undefined;\n          const newStopReason = stopReason === '' || stopReason === undefined ? maxedQueriesStopReason || matchErrorsStopReason : stopReason;\n          debugData(`MaxedQueriesStopReason: <${maxedQueriesStopReason}>`);\n          debugData(`MatchErrorsStopReason <${matchErrorsStopReason}>`);\n          debugData(`NewStopReason: <${newStopReason}>`);\n          debug(`Match status: false`);\n          return {status: false, stopReason: newStopReason};\n        }\n\n        debug(`Match status: true`);\n        return {status: true, stopReason};\n      }\n    }\n\n    // NOTES:\n    // - we could optimize by creating the featureSet for the incoming record once and using it for all database/candidateRecords\n    // - if creating the featureSet for the incoming record fails we have an unprocessable entity\n    // - if creating the featureSet for a candidate record fails we could skip that candidate - but list the case as a detectionFailure, same as conversionFailures\n\n    function iterateRecords({records, recordSetSize, maxMatches, matches = [], nonMatches = [], recordMatches = [], recordNonMatches = [], recordCount = 0, recordDuplicateCount = 0, recordNonMatchCount = 0, recordMatchErrors = []}) {\n\n      // recordSetSize : total amount of records in the current record set\n      // recordCount : amount of records from the current record set that have been handled\n      // maxMatches : setting for maximum amount found by current matcher job before the matcher job is stopped\n      // recordDuplicateCount : amount of records from the current record set that are already included in matches/nonMatches results\n      // recordNonMatchCount: amount of records from the current record set that are nonMatches (only is returnNonMatches setting is false)\n\n      // records : non-handled records in the current record set\n      // matches : found matches in the current matcher job\n      // recordMatches : found matches in the current record set\n      // recordNonMatches : found nonMatches in the current record set (only if returnNonMatches setting is true)\n      // recordMatchErrors: errored matchDetection in the current record set\n\n      const [candidate] = records;\n      const newRecordCount = candidate ? recordCount + 1 : recordCount;\n\n      // The matcher uses same matchDetection strategy for candidates from all candidate-searches -> matchDetection result for the same candidate is always same\n      // Exceptions would happen if the candidate would have been updated in the database between candidate searches\n      // Note that if returnNonMatches is false, matcher won't remember candidates that didn't match, so they will be matched again everytime they are retrieved by\n      // different candidate search queries. Same candidate search query won't have duplicate records.\n\n      /* We could optimize and detect all retrieved candidates at once\n      const candidateRecords = records.map(record => record.record);\n      const recordsIsArray = Array.isArray(candidateRecords);\n      debug(`records is an array: ${recordsIsArray}`);\n      const result = detect(record, candidateRecords);\n      debugData(`${JSON.stringify(result)}`);\n      */\n\n      if (candidate) {\n        if (candidateNotInMatches(matches.concat(nonMatches), candidate)) {\n          const {record: candidateRecord, id: candidateId} = candidate;\n          const recordBExternal = {id: candidate.id, recordSource: 'databaseRecord', label: `db-${candidate.id}`};\n          try {\n            debug(`Running matchDetection for record ${candidateId} (${newRecordCount}/${recordSetSize})`);\n            // we should handle errors from detection somehow - ie. cases where either record or candidateRecord errors\n            const detectionResult = detect({recordA: record, recordB: candidateRecord, recordAExternal: recordExternal, recordBExternal});\n\n            return handleDetectionResult(detectionResult, candidateId, candidateRecord);\n          } catch (error) {\n            debug(`MatchDetection errored: database record ${candidateId}: ${error}`);\n\n            const matchError = {status: 422, payload: {message: `Matching errored for database record ${candidateId}. ${error.message}.`, id: candidateId}};\n            const newRecordMatchErrors = recordMatchErrors.concat(matchError);\n            return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount, recordNonMatchCount, recordMatchErrors: newRecordMatchErrors});\n          }\n        }\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount: recordDuplicateCount + 1, recordNonMatchCount, recordMatchErrors});\n      }\n\n      debug(`No more candidates, record set (${recordCount}/${recordSetSize}) done, ${recordMatches.length} matches found, ${recordDuplicateCount} candidates already handled, ${returnNonMatches ? `${recordNonMatches.length}` : `${recordNonMatchCount}`} nonMatches found.`);\n      return {matches: recordMatches, nonMatches: returnNonMatches ? recordNonMatches : [], duplicateCount: recordDuplicateCount, nonMatchCount: recordNonMatchCount, matchErrors: recordMatchErrors};\n\n      function handleDetectionResult(detectionResult, candidateId, candidateRecord) {\n        debugData(`MatchDetection results for ${candidateId} (${newRecordCount}/${recordSetSize}): ${JSON.stringify(detectionResult)}`);\n\n        if (detectionResult.match || returnNonMatches) {\n          debug(`${detectionResult.match ? `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is a match!` : `Record ${candidateId} (${newRecordCount}/${recordSetSize}) is NOT a match!`}`);\n          debugData(`Strategy: ${JSON.stringify(detectionResult.strategy)}, Threshold: ${JSON.stringify(detectionResult.threshold)}`);\n\n          const matchResult = {\n            probability: Math.round(detectionResult.probability * 100)/100,\n            candidate: {\n              id: candidateId,\n              record: candidateRecord\n            }\n          };\n          const strategyResult = {\n            strategy: detectionResult.strategy,\n            threshold: detectionResult.threshold\n          };\n          const newMatch = returnStrategy ? {...matchResult, ...strategyResult} : {...matchResult};\n\n          debugData(`${JSON.stringify(newMatch)}`);\n\n          return handleRecordMatch(detectionResult.match, newMatch);\n        }\n\n        const newRecordNonMatchCount = recordNonMatchCount + 1;\n        debugData(`- Total nonMatches after this detection: ${newRecordNonMatchCount}`);\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches, recordCount: newRecordCount, recordNonMatches, recordDuplicateCount, recordNonMatchCount: newRecordNonMatchCount, recordMatchErrors});\n      }\n\n      function handleRecordMatch(isMatch, newMatch) {\n        const newRecordMatches = isMatch ? recordMatches.concat(newMatch) : recordMatches;\n        const newRecordNonMatches = isMatch ? recordNonMatches : recordNonMatches.concat(newMatch);\n        const newRecordNonMatchCount = isMatch ? recordNonMatchCount : recordNonMatchCount + 1;\n\n        debugData(`- Total matches after this detection: ${matches.concat(newRecordMatches).length} (max: ${maxMatches})`);\n\n        if (returnNonMatches) {\n          debugData(`- Total nonMatches after this detection: ${nonMatches.concat(newRecordNonMatches).length}`);\n        }\n        debugData(`- Total nonMatchCount after this detection: ${recordNonMatchCount}`);\n\n        if (maxMatchesFound({matches: matches.concat(newRecordMatches), maxMatches})) {\n          debug(`MaxMatches (${maxMatches}) reached, handled candidates in record set: ${newRecordCount} non-handled candidates in record set ${recordSetSize - newRecordCount}`);\n          return {matches: newRecordMatches, nonMatches: returnNonMatches ? newRecordNonMatches : [], duplicateCount: recordDuplicateCount, nonMatchCount: newRecordNonMatchCount, matchErrors: recordMatchErrors};\n        }\n\n        return iterateRecords({records: records.slice(1), recordSetSize, maxMatches, matches, recordMatches: newRecordMatches, recordCount: newRecordCount, recordNonMatches: returnNonMatches ? newRecordNonMatches : [], duplicateCount: recordDuplicateCount, recordNonMatchCount: newRecordNonMatchCount, matchErrors: recordMatchErrors});\n      }\n\n      function candidateNotInMatches(matches, candidate) {\n        debug(`Checking that record ${candidate.id} is not already included in ${matches.length} matches/nonMatches`);\n        const newCandidateId = candidate.id;\n        debugData(`newCandidateId: ${newCandidateId}`);\n        const result = matches.find(({candidate}) => candidate.id === newCandidateId);\n        debugData(`Result: ${result}`);\n        if (result) {\n          debug(`${candidate.id} was already handled.`);\n          return false;\n        }\n        debug(`${candidate.id} not found in matches/nonMatches`);\n        return true;\n      }\n    }\n\n    function maxMatchesFound({matches, maxMatches}) {\n      if (maxMatches && matches.length >= maxMatches) {\n        debug(`Stopping recordSet iteration: maxMatches (${maxMatches}) for matcher job found.`);\n        return true;\n      }\n    }\n  }\n};\n"],
+  "mappings": "AAAA,OAAO,uBAAuB;AAC9B,OAAO,4BAA4B,qBAAqB;AACxD,OAAO,+BAA+B,oBAAoB;AAG1D,SAAQ,iBAAiB;AAEzB,eAAe,CAAC,EAAC,WAAW,kBAAkB,QAAQ,eAAe,aAAa,GAAG,gBAAgB,IAAI,iBAAiB,OAAO,cAAc,OAAO,mBAAmB,MAAK,MAAM;AAClL,QAAM,QAAQ,kBAAkB,yCAAyC;AACzE,QAAM,YAAY,MAAM,OAAO,MAAM;AAErC,YAAU,qBAAqB,KAAK,UAAU,gBAAgB,CAAC,EAAE;AACjE,YAAU,kBAAkB,KAAK,UAAU,aAAa,CAAC,EAAE;AAC3D,YAAU,eAAe,KAAK,UAAU,UAAU,CAAC,EAAE;AACrD,YAAU,kBAAkB,KAAK,UAAU,aAAa,CAAC,EAAE;AAC3D,YAAU,mBAAmB,KAAK,UAAU,cAAc,CAAC,EAAE;AAC7D,YAAU,gBAAgB,KAAK,UAAU,WAAW,CAAC,EAAE;AACvD,YAAU,qBAAqB,KAAK,UAAU,gBAAgB,CAAC,EAAE;AAGjE,QAAM,SAAS,yBAAyB,kBAAkB,cAAc;AAExE,SAAO;AAEP,iBAAe,cAAc,EAAC,QAAQ,iBAAiB,EAAC,cAAc,kBAAkB,OAAO,KAAI,EAAC,GAAG;AAErG,UAAM,EAAC,OAAM,IAAI,MAAM,sBAAsB,EAAC,GAAG,eAAe,QAAQ,eAAe,eAAc,CAAC;AACtG,WAAO,QAAQ,CAAC,CAAC;AAejB,mBAAe,QAAQ,EAAC,eAAe,CAAC,GAAG,UAAU,CAAC,GAAG,iBAAiB,GAAG,aAAa,CAAC,GAAG,iBAAiB,GAAG,gBAAgB,GAAG,cAAc,CAAC,EAAC,GAAG;AACtJ,gBAAU,kCAAkC;AAC5C,YAAM,EAAC,SAAS,GAAG,MAAK,IAAI,MAAM,OAAO,YAAY;AAErD,gBAAU,kBAAkB,KAAK,UAAU,KAAK,CAAC,cAAc,QAAQ,MAAM,qBAAqB,cAAc,iBAAiB,WAAW,MAAM,oBAAoB,aAAa,kBAAkB,YAAY,MAAM,EAAE;AACzN,YAAM,gBAAgB,QAAQ;AAC9B,YAAM,oBAAoB,iBAAiB;AAG3C,UAAI,gBAAgB,GAAG;AACrB,eAAO,gBAAgB;AAAA,MACzB;AAEA,UAAI,MAAM,cAAc,GAAG;AACzB,cAAM,oBAAoB,MAAM,aAAa,QAAQ,MAAM,KAAK,mBAAmB,MAAM,WAAW,eAAe;AACnH,eAAO,QAAQ,EAAC,cAAc,OAAO,SAAS,gBAAgB,mBAAmB,YAAY,eAAe,gBAAgB,YAAW,CAAC;AAAA,MAC1I;AAEA,YAAM,wEAAwE,QAAQ,MAAM,EAAE;AAC9F,aAAO,aAAa,EAAC,SAAS,OAAO,YAAY,IAAI,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,YAAW,CAAC;AAE/I,eAAS,kBAAkB;AACzB,cAAM,0BAA0B,aAAa,qDAAqD,MAAM,aAAa,eAAe,MAAM,KAAK,EAAE;AAEjJ,cAAM,cAAc,eAAe,EAAC,SAAS,eAAe,YAAY,SAAS,YAAY,cAAa,CAAC;AAE3G,cAAM,oBAAoB,iBAAiB,YAAY;AACvD,cAAM,mBAAmB,gBAAgB,YAAY;AACrD,cAAM,EAAC,YAAY,eAAe,eAAc,IAAI,kBAAkB,aAAa,SAAS,YAAY,WAAW;AAEnH,YAAI,gBAAgB,EAAC,SAAS,YAAY,WAAU,CAAC,GAAG;AACtD,iBAAO,aAAa,EAAC,SAAS,YAAY,OAAO,YAAY,cAAc,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,QAC3O;AAEA,YAAI,uBAAuB,mBAAmB,aAAa,GAAG;AAC5D,iBAAO,aAAa,EAAC,SAAS,YAAY,OAAO,YAAY,iBAAiB,YAAY,eAAe,gBAAgB,mBAAmB,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,QAC9O;AAEA,eAAO,QAAQ,EAAC,cAAc,OAAO,SAAS,YAAY,gBAAgB,mBAAmB,YAAY,eAAe,gBAAgB,mBAAmB,eAAe,kBAAkB,aAAa,eAAc,CAAC;AAAA,MAC1N;AAEA,eAAS,kBAAkB,aAAaA,UAASC,aAAYC,cAAa;AACxE,kBAAU,4CAA4C,YAAY,QAAQ,MAAM,EAAE;AAClF,YAAI,kBAAkB;AACpB,oBAAU,+CAA+C,YAAY,WAAW,MAAM,EAAE;AAAA,QAC1F;AAEA,cAAM,aAAaF,SAAQ,OAAO,cAAc,SAAS,YAAY,OAAO,IAAI,YAAY,OAAO;AACnG,cAAM,gBAAgB,mBAAmBC,YAAW,OAAO,cAAc,SAAS,YAAY,UAAU,IAAI,YAAY,UAAU,IAAI,CAAC;AACvI,cAAM,iBAAiBC,aAAY,OAAO,YAAY,WAAW;AAEjE,kBAAU,8BAA8B,WAAW,MAAM,EAAE;AAC3D,YAAI,kBAAkB;AACpB,oBAAU,iCAAiC,cAAc,MAAM,EAAE;AAAA,QACnE;AAEA,kBAAU,gBAAgB,KAAK,UAAU,WAAW,CAAC,EAAE;AACvD,kBAAU,oBAAoB,KAAK,UAAUA,YAAW,CAAC,mCAAmC,KAAK,UAAU,YAAY,WAAW,CAAC,sBAAsB,KAAK,UAAU,cAAc,CAAC,EAAE;AAEzL,kBAAU,kCAAkC,eAAe,MAAM,EAAE;AAEnE,eAAO,EAAC,YAAY,eAAe,eAAc;AAAA,MACnD;AAEA,eAAS,SAASF,UAAS;AACzB,kBAAU,gBAAgB,MAAM,KAAK,aAAa;AAClD,eAAOA,SAAQ,IAAI,CAAC,WAAW,EAAC,GAAG,OAAO,YAAY,MAAM,MAAK,EAAE;AAAA,MACrE;AAEA,eAAS,uBAAuBG,iBAAgBC,gBAAe;AAC7D,kBAAU,gDAAgD,iBAAiB,UAAUA,cAAa,GAAG;AACrG,YAAIA,kBAAiBD,mBAAkBC,gBAAe;AACpD,gBAAM,gEAAgED,eAAc,MAAMC,cAAa,sBAAsB;AAC7H,iBAAO;AAAA,QACT;AAAA,MACF;AAAA,IACF;AAmBA,aAAS,aAAa,EAAC,SAAS,OAAO,YAAY,YAAY,gBAAgB,gBAAgB,eAAe,YAAW,GAAG;AAC1H,YAAM,kBAAkB,YAAY;AACpC,kBAAY,EAAC,SAAS,YAAY,gBAAgB,gBAAgB,eAAe,gBAAe,CAAC;AACjG,YAAM,cAAc,cAAc,OAAO,YAAY,eAAe;AAEpE,YAAM,gBAAgB,mBAAmB,EAAC,SAAS,aAAa,YAAY,eAAc,IAAI,EAAC,SAAS,aAAa,eAAc;AACnI,YAAM,SAAS;AACf,gBAAU,GAAG,KAAK,UAAU,MAAM,CAAC,EAAE;AACrC,aAAO;AAIP,eAAS,YAAY,EAAC,SAAAJ,UAAS,YAAAC,aAAY,gBAAAE,iBAAgB,gBAAAE,iBAAgB,eAAAC,gBAAe,iBAAAC,iBAAe,GAAG;AAC1G,cAAM,aAAaP,SAAQ;AAC3B,kBAAU,sBAAsB,gBAAgB,EAAE;AAClD,cAAM,sBAAsB,mBAAmBC,YAAW,SAASK;AACnE,cAAM,eAAe,aAAa,sBAAsBD;AACxD,cAAM,mBAAmBF,eAAc,cAAc,UAAU,iBAAiB,mBAAmB,qBAAqBE,eAAc,sBAAsBE,gBAAe,EAAE;AAC7K,cAAM,qBAAqB,YAAY,MAAMJ,eAAc,uBAAuB;AAClF,YAAI,iBAAiBA,iBAAgB;AACnC,gBAAM,gCAAgCA,kBAAiB,YAAY,aAAa;AAChF;AAAA,QACF;AACA;AAAA,MACF;AAEA,eAAS,cAAcK,QAAOC,aAAYF,kBAAiB;AACzD,kBAAU,GAAG,KAAK,UAAUC,MAAK,CAAC,EAAE;AACpC,cAAM,UAAUD,gBAAe,uDAAuD;AACtF,cAAM,gBAAgBC,OAAM,WAAW,sCAAsCA,OAAM,mBAAmBA,OAAM,mBAAmBA,OAAM,YAAY,4BAA4BA,OAAM,eAAeA,OAAM,qBAAqB,mBAAmBA,OAAM,aAAa,MAAM,MAAMA,OAAM,YAAY,EAAE;AAEnS,kBAAU,gBAAgBC,WAAU,GAAG;AAEvC,cAAM,eAAeD,OAAM,mBAAmBA,OAAM,mBAAmBA,OAAM;AAC7E,cAAM,eAAe,eAAeA,OAAM,eAAeA,OAAM,wBAAwB;AACvF,kBAAU,iBAAiB,YAAY,EAAE;AAKzC,YAAIA,OAAM,cAAc,KAAK,eAAe,KAAKA,OAAM,aAAa,SAAS,KAAKD,mBAAkB,GAAG;AACrG,gBAAM,yBAAyBC,OAAM,aAAa,SAAS,IAAI,iBAAiB;AAChF,gBAAM,wBAAwBD,mBAAkB,IAAI,gBAAgB;AACpE,gBAAM,gBAAgBE,gBAAe,MAAMA,gBAAe,SAAY,0BAA0B,wBAAwBA;AACxH,oBAAU,4BAA4B,sBAAsB,GAAG;AAC/D,oBAAU,0BAA0B,qBAAqB,GAAG;AAC5D,oBAAU,mBAAmB,aAAa,GAAG;AAC7C,gBAAM,qBAAqB;AAC3B,iBAAO,EAAC,QAAQ,OAAO,YAAY,cAAa;AAAA,QAClD;AAEA,cAAM,oBAAoB;AAC1B,eAAO,EAAC,QAAQ,MAAM,YAAAA,YAAU;AAAA,MAClC;AAAA,IACF;AAOA,aAAS,eAAe,EAAC,SAAS,eAAe,YAAAC,aAAY,UAAU,CAAC,GAAG,aAAa,CAAC,GAAG,gBAAgB,CAAC,GAAG,mBAAmB,CAAC,GAAG,cAAc,GAAG,uBAAuB,GAAG,sBAAsB,GAAG,oBAAoB,CAAC,EAAC,GAAG;AAclO,YAAM,CAAC,SAAS,IAAI;AACpB,YAAM,iBAAiB,YAAY,cAAc,IAAI;AAerD,UAAI,WAAW;AACb,YAAI,sBAAsB,QAAQ,OAAO,UAAU,GAAG,SAAS,GAAG;AAChE,gBAAM,EAAC,QAAQ,iBAAiB,IAAI,YAAW,IAAI;AACnD,gBAAM,kBAAkB,EAAC,IAAI,UAAU,IAAI,cAAc,kBAAkB,OAAO,MAAM,UAAU,EAAE,GAAE;AACtG,cAAI;AACF,kBAAM,qCAAqC,WAAW,KAAK,cAAc,IAAI,aAAa,GAAG;AAE7F,kBAAM,kBAAkB,OAAO,EAAC,SAAS,QAAQ,SAAS,iBAAiB,iBAAiB,gBAAgB,gBAAe,CAAC;AAE5H,mBAAO,sBAAsB,iBAAiB,aAAa,eAAe;AAAA,UAC5E,SAAS,OAAO;AACd,kBAAM,2CAA2C,WAAW,KAAK,KAAK,EAAE;AAExE,kBAAM,aAAa,EAAC,QAAQ,KAAK,SAAS,EAAC,SAAS,wCAAwC,WAAW,KAAK,MAAM,OAAO,KAAK,IAAI,YAAW,EAAC;AAC9I,kBAAM,uBAAuB,kBAAkB,OAAO,UAAU;AAChE,mBAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,qBAAqB,mBAAmB,qBAAoB,CAAC;AAAA,UACzO;AAAA,QACF;AAEA,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,uBAAuB,GAAG,qBAAqB,kBAAiB,CAAC;AAAA,MAC7O;AAEA,YAAM,mCAAmC,WAAW,IAAI,aAAa,WAAW,cAAc,MAAM,mBAAmB,oBAAoB,gCAAgC,mBAAmB,GAAG,iBAAiB,MAAM,KAAK,GAAG,mBAAmB,EAAE,oBAAoB;AACzQ,aAAO,EAAC,SAAS,eAAe,YAAY,mBAAmB,mBAAmB,CAAC,GAAG,gBAAgB,sBAAsB,eAAe,qBAAqB,aAAa,kBAAiB;AAE9L,eAAS,sBAAsB,iBAAiB,aAAa,iBAAiB;AAC5E,kBAAU,8BAA8B,WAAW,KAAK,cAAc,IAAI,aAAa,MAAM,KAAK,UAAU,eAAe,CAAC,EAAE;AAE9H,YAAI,gBAAgB,SAAS,kBAAkB;AAC7C,gBAAM,GAAG,gBAAgB,QAAQ,UAAU,WAAW,KAAK,cAAc,IAAI,aAAa,kBAAkB,UAAU,WAAW,KAAK,cAAc,IAAI,aAAa,mBAAmB,EAAE;AAC1L,oBAAU,aAAa,KAAK,UAAU,gBAAgB,QAAQ,CAAC,gBAAgB,KAAK,UAAU,gBAAgB,SAAS,CAAC,EAAE;AAE1H,gBAAM,cAAc;AAAA,YAClB,aAAa,KAAK,MAAM,gBAAgB,cAAc,GAAG,IAAE;AAAA,YAC3D,WAAW;AAAA,cACT,IAAI;AAAA,cACJ,QAAQ;AAAA,YACV;AAAA,UACF;AACA,gBAAM,iBAAiB;AAAA,YACrB,UAAU,gBAAgB;AAAA,YAC1B,WAAW,gBAAgB;AAAA,UAC7B;AACA,gBAAM,WAAW,iBAAiB,EAAC,GAAG,aAAa,GAAG,eAAc,IAAI,EAAC,GAAG,YAAW;AAEvF,oBAAU,GAAG,KAAK,UAAU,QAAQ,CAAC,EAAE;AAEvC,iBAAO,kBAAkB,gBAAgB,OAAO,QAAQ;AAAA,QAC1D;AAEA,cAAM,yBAAyB,sBAAsB;AACrD,kBAAU,4CAA4C,sBAAsB,EAAE;AAE9E,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,aAAa,gBAAgB,kBAAkB,sBAAsB,qBAAqB,wBAAwB,kBAAiB,CAAC;AAAA,MAC3O;AAEA,eAAS,kBAAkB,SAAS,UAAU;AAC5C,cAAM,mBAAmB,UAAU,cAAc,OAAO,QAAQ,IAAI;AACpE,cAAM,sBAAsB,UAAU,mBAAmB,iBAAiB,OAAO,QAAQ;AACzF,cAAM,yBAAyB,UAAU,sBAAsB,sBAAsB;AAErF,kBAAU,yCAAyC,QAAQ,OAAO,gBAAgB,EAAE,MAAM,UAAUA,WAAU,GAAG;AAEjH,YAAI,kBAAkB;AACpB,oBAAU,4CAA4C,WAAW,OAAO,mBAAmB,EAAE,MAAM,EAAE;AAAA,QACvG;AACA,kBAAU,+CAA+C,mBAAmB,EAAE;AAE9E,YAAI,gBAAgB,EAAC,SAAS,QAAQ,OAAO,gBAAgB,GAAG,YAAAA,YAAU,CAAC,GAAG;AAC5E,gBAAM,eAAeA,WAAU,gDAAgD,cAAc,yCAAyC,gBAAgB,cAAc,EAAE;AACtK,iBAAO,EAAC,SAAS,kBAAkB,YAAY,mBAAmB,sBAAsB,CAAC,GAAG,gBAAgB,sBAAsB,eAAe,wBAAwB,aAAa,kBAAiB;AAAA,QACzM;AAEA,eAAO,eAAe,EAAC,SAAS,QAAQ,MAAM,CAAC,GAAG,eAAe,YAAAA,aAAY,SAAS,eAAe,kBAAkB,aAAa,gBAAgB,kBAAkB,mBAAmB,sBAAsB,CAAC,GAAG,gBAAgB,sBAAsB,qBAAqB,wBAAwB,aAAa,kBAAiB,CAAC;AAAA,MACvU;AAEA,eAAS,sBAAsBV,UAASW,YAAW;AACjD,cAAM,wBAAwBA,WAAU,EAAE,+BAA+BX,SAAQ,MAAM,qBAAqB;AAC5G,cAAM,iBAAiBW,WAAU;AACjC,kBAAU,mBAAmB,cAAc,EAAE;AAC7C,cAAM,SAASX,SAAQ,KAAK,CAAC,EAAC,WAAAW,WAAS,MAAMA,WAAU,OAAO,cAAc;AAC5E,kBAAU,WAAW,MAAM,EAAE;AAC7B,YAAI,QAAQ;AACV,gBAAM,GAAGA,WAAU,EAAE,uBAAuB;AAC5C,iBAAO;AAAA,QACT;AACA,cAAM,GAAGA,WAAU,EAAE,kCAAkC;AACvD,eAAO;AAAA,MACT;AAAA,IACF;AAEA,aAAS,gBAAgB,EAAC,SAAS,YAAAD,YAAU,GAAG;AAC9C,UAAIA,eAAc,QAAQ,UAAUA,aAAY;AAC9C,cAAM,6CAA6CA,WAAU,0BAA0B;AACvF,eAAO;AAAA,MACT;AAAA,IACF;AAAA,EACF;AACF;",
   "names": ["matches", "nonMatches", "matchErrors", "candidateCount", "maxCandidates", "duplicateCount", "nonMatchCount", "matchErrorCount", "state", "stopReason", "maxMatches", "candidate"]
 }

package/dist/match-detection/features/bib/f773.js

@@ -1,8 +1,8 @@
 import createDebugLogger from "debug";
 import { isComponentRecord } from "@natlibfi/melinda-commons";
-import { uniqArray } from "./issn.js";
 import { parse773g } from "../../../candidate-search/query-list/component.js";
-const debug = createDebugLogger("@natlibfi/melinda-record-matching:match-detection:features:issn");
+import { uniqArray } from "@natlibfi/marc-record-validators-melinda/dist/utils.js";
+const debug = createDebugLogger("@natlibfi/melinda-record-matching:match-detection:features:f773");
 const debugData = debug.extend("data");
 const MAX_IDENTIFIER = 0.1;
 const MAX_G = 0.5;

package/dist/match-detection/features/bib/f773.js.map

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../../../../src/match-detection/features/bib/f773.js"],
-  "sourcesContent": ["// NB! This checks identifiers ($w=ID, $x=ISSN, $z=ISBN and $o=others) to check whether the host is the same.\n// BB! $g is not checked *yet*. It could very strongly indicate that the records are the same... (I doubt $q could be useful here)\n// Rationale:\n// 773$w indicates sameness. However, non-Melinda records probably refer to non-melinda hosts.\n// Thus check other identifiers subfields ($x, $z and $o as well)\n//\n\n\nimport createDebugLogger from 'debug';\n\nimport {isComponentRecord} from '@natlibfi/melinda-commons';\nimport {uniqArray} from './issn.js';\nimport {parse773g} from '../../../candidate-search/query-list/component.js';\n\nconst debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:issn');\nconst debugData = debug.extend('data');\n\nconst MAX_IDENTIFIER = 0.1; // This should be pretty low: it only says something about the host\nconst MAX_G = 0.5; // $g is about the comp itself, so score can be high here\n\nexport default () => ({\n  name: 'f773 ',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    if (!isComponentRecord(record, false, [])) {\n      return [];\n    }\n    const f773s = record.get(/^773$/u); // Identifier extract handles multiple 773...\n\n    // I think it's ok, if, say, $o and $x values match, so I'm not keeping the subfield codes\n    const values = f773s.map(f => f.subfields)\n      .flat()\n      .filter(sf => ['w', 'o', 'x', 'z'].includes(sf.code))\n      .map(sf =>  normalizeValue(sf.value));\n\n    // $g: only on $g is supported\n    const gData = parse773g(f773s[0]);\n\n    return [uniqArray(values), gData];\n\n    function normalizeValue(value) {\n      return value.replace(/(?:\\. -|\\.)$/u, '');\n    }\n\n\n  },\n\n  compare: (aa, bb) => {\n    const [aIdentifiers, ag] = aa;\n    const [bIdentifiers, bg] = bb;\n\n    const identifierScore = scoreIdentifiers();\n    const gScore = scoreG();\n\n    if (identifierScore === MAX_IDENTIFIER && gScore === MAX_G) { // Pretty impressive hit, even if title matches not\n      return 1.0;\n    }\n\n    return identifierScore + gScore;\n\n    function scoreG() {\n      // NB! $g contents are very noise, so wrong values may be extracted. Thus do not overpenalize.\n\n      // All exist match: things must be pretty good:\n      if (ag.number && ag.number === bg.number && ag.pages && ag.pages === bg.pages && ag.year && ag.year === bg.year) {\n        return MAX_G;\n      }\n      // Not comparing volume. It correlates with year.\n      return scoreYear() + scoreNumber() + scorePages();\n    }\n\n    function scoreNumber() {\n      if (!ag.number || !bg.number) {\n        return 0.0;\n      }\n      if (ag.number === bg.number) {\n        return 0.05;\n      }\n      return -0.02;\n    }\n\n    function scorePages() {\n      if (!ag.pages || !bg.pages) {\n        return 0.0;\n      }\n      if (ag.pages === bg.pages) { // If pages match, things must be pretty good\n        return 0.1;\n      }\n      return -0.05;\n    }\n\n    function scoreYear() { // publication-time.js also uses this, so don't score heavily here\n      if (!ag.year || !bg.year) {\n        return 0.0;\n      }\n      if (ag.year === bg.year) {\n        return 0.02;\n      }\n      return -0.02;\n    }\n\n    function scoreIdentifiers() {\n      debugData(`Comparing ISSN sets ${JSON.stringify(aIdentifiers)} and ${JSON.stringify(bIdentifiers)}`);\n      if (aIdentifiers.length === 0 || bIdentifiers.length === 0) {\n        // No data for decision\n        return 0;\n      }\n      const firstSharedIdentifier = aIdentifiers.find(val => bIdentifiers.includes(val));\n      if (firstSharedIdentifier) {\n        debug(`\\t Shared identifier found: '${firstSharedIdentifier}'`);\n        return MAX_IDENTIFIER;\n      }\n      return -0.5;\n    }\n  }\n});\n\n"],
+  "sourcesContent": ["// NB! This checks identifiers ($w=ID, $x=ISSN, $z=ISBN and $o=others) to check whether the host is the same.\n// BB! $g is not checked *yet*. It could very strongly indicate that the records are the same... (I doubt $q could be useful here)\n// Rationale:\n// 773$w indicates sameness. However, non-Melinda records probably refer to non-melinda hosts.\n// Thus check other identifiers subfields ($x, $z and $o as well)\n//\n\n\nimport createDebugLogger from 'debug';\n\nimport {isComponentRecord} from '@natlibfi/melinda-commons';\nimport {parse773g} from '../../../candidate-search/query-list/component.js';\nimport {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';\n\nconst debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:f773');\nconst debugData = debug.extend('data');\n\nconst MAX_IDENTIFIER = 0.1; // This should be pretty low: it only says something about the host\nconst MAX_G = 0.5; // $g is about the comp itself, so score can be high here\n\nexport default () => ({\n  name: 'f773 ',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    if (!isComponentRecord(record, false, [])) {\n      return [];\n    }\n    const f773s = record.get(/^773$/u); // Identifier extract handles multiple 773...\n\n    // I think it's ok, if, say, $o and $x values match, so I'm not keeping the subfield codes\n    const values = f773s.map(f => f.subfields)\n      .flat()\n      .filter(sf => ['w', 'o', 'x', 'z'].includes(sf.code))\n      .map(sf =>  normalizeValue(sf.value));\n\n    // $g: only on $g is supported\n    const gData = parse773g(f773s[0]);\n\n    return [uniqArray(values), gData];\n\n    function normalizeValue(value) {\n      return value.replace(/(?:\\. -|\\.)$/u, '');\n    }\n\n\n  },\n\n  compare: (aa, bb) => {\n    const [aIdentifiers, ag] = aa;\n    const [bIdentifiers, bg] = bb;\n\n    const identifierScore = scoreIdentifiers();\n    const gScore = scoreG();\n\n    if (identifierScore === MAX_IDENTIFIER && gScore === MAX_G) { // Pretty impressive hit, even if title matches not\n      return 1.0;\n    }\n\n    return identifierScore + gScore;\n\n    function scoreG() {\n      // NB! $g contents are very noise, so wrong values may be extracted. Thus do not overpenalize.\n\n      // All exist match: things must be pretty good:\n      if (ag.number && ag.number === bg.number && ag.pages && ag.pages === bg.pages && ag.year && ag.year === bg.year) {\n        return MAX_G;\n      }\n      // Not comparing volume. It correlates with year.\n      return scoreYear() + scoreNumber() + scorePages();\n    }\n\n    function scoreNumber() {\n      if (!ag.number || !bg.number) {\n        return 0.0;\n      }\n      if (ag.number === bg.number) {\n        return 0.05;\n      }\n      return -0.02;\n    }\n\n    function scorePages() {\n      if (!ag.pages || !bg.pages) {\n        return 0.0;\n      }\n      if (ag.pages === bg.pages) { // If pages match, things must be pretty good\n        return 0.1;\n      }\n      return -0.05;\n    }\n\n    function scoreYear() { // publication-time.js also uses this, so don't score heavily here\n      if (!ag.year || !bg.year) {\n        return 0.0;\n      }\n      if (ag.year === bg.year) {\n        return 0.02;\n      }\n      return -0.02;\n    }\n\n    function scoreIdentifiers() {\n      debugData(`Comparing ISSN sets ${JSON.stringify(aIdentifiers)} and ${JSON.stringify(bIdentifiers)}`);\n      if (aIdentifiers.length === 0 || bIdentifiers.length === 0) {\n        // No data for decision\n        return 0;\n      }\n      const firstSharedIdentifier = aIdentifiers.find(val => bIdentifiers.includes(val));\n      if (firstSharedIdentifier) {\n        debug(`\\t Shared identifier found: '${firstSharedIdentifier}'`);\n        return MAX_IDENTIFIER;\n      }\n      return -0.5;\n    }\n  }\n});\n\n"],
   "mappings": "AAQA,OAAO,uBAAuB;AAE9B,SAAQ,yBAAwB;AAChC,SAAQ,iBAAgB;AACxB,SAAQ,iBAAgB;AAExB,MAAM,QAAQ,kBAAkB,iEAAiE;AACjG,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,MAAM,iBAAiB;AACvB,MAAM,QAAQ;AAEd,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,QAAI,CAAC,kBAAkB,QAAQ,OAAO,CAAC,CAAC,GAAG;AACzC,aAAO,CAAC;AAAA,IACV;AACA,UAAM,QAAQ,OAAO,IAAI,QAAQ;AAGjC,UAAM,SAAS,MAAM,IAAI,OAAK,EAAE,SAAS,EACtC,KAAK,EACL,OAAO,QAAM,CAAC,KAAK,KAAK,KAAK,GAAG,EAAE,SAAS,GAAG,IAAI,CAAC,EACnD,IAAI,QAAO,eAAe,GAAG,KAAK,CAAC;AAGtC,UAAM,QAAQ,UAAU,MAAM,CAAC,CAAC;AAEhC,WAAO,CAAC,UAAU,MAAM,GAAG,KAAK;AAEhC,aAAS,eAAe,OAAO;AAC7B,aAAO,MAAM,QAAQ,iBAAiB,EAAE;AAAA,IAC1C;AAAA,EAGF;AAAA,EAEA,SAAS,CAAC,IAAI,OAAO;AACnB,UAAM,CAAC,cAAc,EAAE,IAAI;AAC3B,UAAM,CAAC,cAAc,EAAE,IAAI;AAE3B,UAAM,kBAAkB,iBAAiB;AACzC,UAAM,SAAS,OAAO;AAEtB,QAAI,oBAAoB,kBAAkB,WAAW,OAAO;AAC1D,aAAO;AAAA,IACT;AAEA,WAAO,kBAAkB;AAEzB,aAAS,SAAS;AAIhB,UAAI,GAAG,UAAU,GAAG,WAAW,GAAG,UAAU,GAAG,SAAS,GAAG,UAAU,GAAG,SAAS,GAAG,QAAQ,GAAG,SAAS,GAAG,MAAM;AAC/G,eAAO;AAAA,MACT;AAEA,aAAO,UAAU,IAAI,YAAY,IAAI,WAAW;AAAA,IAClD;AAEA,aAAS,cAAc;AACrB,UAAI,CAAC,GAAG,UAAU,CAAC,GAAG,QAAQ;AAC5B,eAAO;AAAA,MACT;AACA,UAAI,GAAG,WAAW,GAAG,QAAQ;AAC3B,eAAO;AAAA,MACT;AACA,aAAO;AAAA,IACT;AAEA,aAAS,aAAa;AACpB,UAAI,CAAC,GAAG,SAAS,CAAC,GAAG,OAAO;AAC1B,eAAO;AAAA,MACT;AACA,UAAI,GAAG,UAAU,GAAG,OAAO;AACzB,eAAO;AAAA,MACT;AACA,aAAO;AAAA,IACT;AAEA,aAAS,YAAY;AACnB,UAAI,CAAC,GAAG,QAAQ,CAAC,GAAG,MAAM;AACxB,eAAO;AAAA,MACT;AACA,UAAI,GAAG,SAAS,GAAG,MAAM;AACvB,eAAO;AAAA,MACT;AACA,aAAO;AAAA,IACT;AAEA,aAAS,mBAAmB;AAC1B,gBAAU,uBAAuB,KAAK,UAAU,YAAY,CAAC,QAAQ,KAAK,UAAU,YAAY,CAAC,EAAE;AACnG,UAAI,aAAa,WAAW,KAAK,aAAa,WAAW,GAAG;AAE1D,eAAO;AAAA,MACT;AACA,YAAM,wBAAwB,aAAa,KAAK,SAAO,aAAa,SAAS,GAAG,CAAC;AACjF,UAAI,uBAAuB;AACzB,cAAM,+BAAgC,qBAAqB,GAAG;AAC9D,eAAO;AAAA,MACT;AACA,aAAO;AAAA,IACT;AAAA,EACF;AACF;",
   "names": []
 }

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

@@ -1,6 +1,7 @@
 export { default as hostComponent } from "./host-component.js";
 export { default as isbn } from "./isbn.js";
 export { default as issn } from "./issn.js";
+export { default as publisherNumber } from "./publisherNumber.js";
 export { default as sid } from "./sid.js";
 export { default as otherStandardIdentifier } from "./other-standard-identifier.js";
 export { default as title } from "./title.js";

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

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../../../../src/match-detection/features/bib/index.js"],
-  "sourcesContent": ["\nexport {default as hostComponent} from './host-component.js';\nexport {default as isbn} from './isbn.js';\nexport {default as issn} from './issn.js';\nexport {default as sid} from './sid.js';\nexport {default as otherStandardIdentifier} from './other-standard-identifier.js';\nexport {default as title} from './title.js';\nexport {default as authors} from './authors.js';\nexport {default as recordType} from './record-type.js';\nexport {default as publicationTime} from './publication-time.js';\nexport {default as publicationTimeAllowConsYears} from './publication-time-allow-cons-years.js';\nexport {default as publicationTimeAllowConsYearsMulti} from './publication-time-allow-cons-years-multi.js';\nexport {default as language} from './language.js';\nexport {default as bibliographicLevel} from './bibliographic-level.js';\nexport {default as melindaId} from './melinda-id.js';\nexport {default as allSourceIds} from './all-source-ids.js';\nexport {default as mediaType} from './media-type.js';\nexport {default as f773} from './f773.js';\n"],
-  "mappings": "AACA,SAAQ,WAAW,qBAAoB;AACvC,SAAQ,WAAW,YAAW;AAC9B,SAAQ,WAAW,YAAW;AAC9B,SAAQ,WAAW,WAAU;AAC7B,SAAQ,WAAW,+BAA8B;AACjD,SAAQ,WAAW,aAAY;AAC/B,SAAQ,WAAW,eAAc;AACjC,SAAQ,WAAW,kBAAiB;AACpC,SAAQ,WAAW,uBAAsB;AACzC,SAAQ,WAAW,qCAAoC;AACvD,SAAQ,WAAW,0CAAyC;AAC5D,SAAQ,WAAW,gBAAe;AAClC,SAAQ,WAAW,0BAAyB;AAC5C,SAAQ,WAAW,iBAAgB;AACnC,SAAQ,WAAW,oBAAmB;AACtC,SAAQ,WAAW,iBAAgB;AACnC,SAAQ,WAAW,YAAW;",
+  "sourcesContent": ["\nexport {default as hostComponent} from './host-component.js';\nexport {default as isbn} from './isbn.js';\nexport {default as issn} from './issn.js';\nexport {default as publisherNumber} from './publisherNumber.js'; // 028$b$a\nexport {default as sid} from './sid.js';\nexport {default as otherStandardIdentifier} from './other-standard-identifier.js';\nexport {default as title} from './title.js';\nexport {default as authors} from './authors.js';\nexport {default as recordType} from './record-type.js';\nexport {default as publicationTime} from './publication-time.js';\nexport {default as publicationTimeAllowConsYears} from './publication-time-allow-cons-years.js';\nexport {default as publicationTimeAllowConsYearsMulti} from './publication-time-allow-cons-years-multi.js';\nexport {default as language} from './language.js';\nexport {default as bibliographicLevel} from './bibliographic-level.js';\nexport {default as melindaId} from './melinda-id.js';\nexport {default as allSourceIds} from './all-source-ids.js';\nexport {default as mediaType} from './media-type.js';\nexport {default as f773} from './f773.js';\n"],
+  "mappings": "AACA,SAAQ,WAAW,qBAAoB;AACvC,SAAQ,WAAW,YAAW;AAC9B,SAAQ,WAAW,YAAW;AAC9B,SAAQ,WAAW,uBAAsB;AACzC,SAAQ,WAAW,WAAU;AAC7B,SAAQ,WAAW,+BAA8B;AACjD,SAAQ,WAAW,aAAY;AAC/B,SAAQ,WAAW,eAAc;AACjC,SAAQ,WAAW,kBAAiB;AACpC,SAAQ,WAAW,uBAAsB;AACzC,SAAQ,WAAW,qCAAoC;AACvD,SAAQ,WAAW,0CAAyC;AAC5D,SAAQ,WAAW,gBAAe;AAClC,SAAQ,WAAW,0BAAyB;AAC5C,SAAQ,WAAW,iBAAgB;AACnC,SAAQ,WAAW,oBAAmB;AACtC,SAAQ,WAAW,iBAAgB;AACnC,SAAQ,WAAW,YAAW;",
   "names": []
 }

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

@@ -1,6 +1,6 @@
 import createDebugLogger from "debug";
 import { parse as isbnParse } from "isbn3";
-import { uniqArray } from "./issn.js";
+import { uniqArray } from "@natlibfi/marc-record-validators-melinda/dist/utils.js";
 const debug = createDebugLogger(`@natlibfi/melinda-record-matching:match-detection:features:standard-identifiers:ISBN`);
 const debugData = debug.extend("data");
 const MAX_SCORE = 0.75;

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

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../../../../src/match-detection/features/bib/isbn.js"],
-  "sourcesContent": ["\nimport createDebugLogger from 'debug';\nimport {parse as isbnParse} from 'isbn3';\n\nimport {uniqArray} from './issn.js';\n\nconst debug = createDebugLogger(`@natlibfi/melinda-record-matching:match-detection:features:standard-identifiers:ISBN`);\nconst debugData = debug.extend('data');\n\nconst MAX_SCORE = 0.75;\n\nconst setRegexp = /(?:complete set|hela verket|helhet|koko kartasto|koko paketti|koko sarja|koko setti|koko teoksen isbn|koko teos|kokonaisuus|^set\\b|\\bset$|\\bsetti\\b|vol.*set)/ui;\n\nexport default () => ({\n  name: 'ISBN',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    const all = record.get('020').filter(f => f.subfields?.some(sf => ['a', 'z'].includes(sf.code) && sf.value));\n    const sets = all.filter(f => isSetIsbnField(f));\n    const noSets = all.filter(f => !isSetIsbnField(f));\n    if (sets.length > 0 && noSets.length > 0) {\n      return noSets;\n    }\n    return all;\n\n    function isSetIsbnField(field) {\n      return field.subfields.some(sf => sf.code === 'q' && setRegexp.test(sf.value));\n    }\n  },\n  compare: (aa, bb) => {\n    debugData(`Comparing ISBN sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);\n    if (aa.length === 0 || bb.length === 0) {\n      // No data for decision\n      return 0;\n    }\n\n    const subfieldCodeForGoodValues = 'a';\n    const subfieldCodeForBadValues = 'z';\n\n    const [aValidValuesA, aInvalidValuesA, zValidValuesA, zInvalidValuesA] = getValuesWrapper(aa, 'AA'); // initial 'a' and 'z' refer to 020 subfields codes\n    const [aValidValuesB, aInvalidValuesB, zValidValuesB, zInvalidValuesB] = getValuesWrapper(bb, 'BB');\n\n    function getValuesWrapper(data, prefix) {\n      const [aValidValues, aInvalidValues, zValidValues, zInvalidValues] = getValues(data);\n      if (aValidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForGoodValues} VALID: ${aValidValues.join(', ')}`);\n      }\n      if (aInvalidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForGoodValues} INVALID: ${aInvalidValues.join(', ')}`);\n      }\n      if (zValidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForBadValues} VALID: ${zValidValues.join(', ')}`);\n      }\n      if (zInvalidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForBadValues} INVALID: ${zInvalidValues.join(', ')}`);\n      }\n      return [aValidValues, aInvalidValues, zValidValues, zInvalidValues];\n    }\n\n    const [sharedGoodValues, goodValuesAOnly, goodValuesBOnly] = getUnionData(aValidValuesA, aValidValuesB);\n\n    debug(`GOOD\\tBOTH: ${sharedGoodValues.length}, A only: ${goodValuesAOnly.length}, B only: ${goodValuesBOnly.length}`);\n\n    if (sharedGoodValues.length > 0) {\n      // Third argument (aka 'N times') is >= 0\n      return scoreData(MAX_SCORE, 0.8, goodValuesAOnly.length + goodValuesBOnly.length);\n    }\n\n    const hitScore = scoreSuboptimalHit();\n\n    function scoreSuboptimalHit() {\n      // One record consider ISBN good and the other record considered it's canceled:\n      if (aValidValuesA.some(valA => zValidValuesB.includes(valA)) || aValidValuesB.some(valB => zValidValuesA.includes(valB))) {\n        return MAX_SCORE;\n      }\n\n      // Subfield is for cancelled/whatever values, but the value is syntactically valid:\n      // Could happen for two canceled ISBNs for example. I'll give this two thirds of the full score\n      const zzValid = zValidValuesA.find(valA => zValidValuesB.includes(valA));\n      if (zzValid) {\n        debug(`Both contain a valid value in 020$z: ${zzValid}`);\n        return MAX_SCORE * 2 / 3;\n      }\n      // Shared invalid identifiers:\n      const aaInvalid = aInvalidValuesA.find(valA => aInvalidValuesB.includes(valA) || zInvalidValuesB.includes(valA)) || aInvalidValuesB.find(valB => zInvalidValuesA.includes(valB));\n      if (aaInvalid) {\n        debug(`Shared invalid value in 020$a and 020$a-or-$z subfields: ${aaInvalid}`);\n        return MAX_SCORE * 2 / 3;\n      }\n\n      /* // Currently I think that paired invalid idenfiers in 020$z are meaningless...\n      const zzInvalid = zInvalidValuesB.find(valB => zInvalidValuesA.includes(valB)) || zInvalidValuesA.find(valA => zInvalidValuesB.includes(valA));\n      if (zzInvalid) {\n        debug(`Shared invalid value in 020$z subfields: ${zzInvalid}`);\n        return MAX_SCORE / 3;\n      }\n      */\n\n      return 0;\n    }\n\n\n\n    if (hitScore === MAX_SCORE) {\n      // -1 is needed to make the third argument >= 0 (otherwise min val would be 0)\n      return scoreData(hitScore, 0.8, goodValuesAOnly.length + goodValuesBOnly.length - 1);\n    }\n\n    if (hitScore > 0) { // Canceled/invalid ISBNs match\n      // Note that this is not (currently) penalized for non-matching canceled/invalid ISBNs. Maybe it should be.\n      return scoreData(hitScore, 0.8, goodValuesAOnly.length + goodValuesBOnly.length);\n    }\n\n    // No match:\n\n    if (aValidValuesA.length + aInvalidValuesA.length === 0 || aValidValuesB.length + aInvalidValuesB.length === 0) { // At least one record did not have any good ISBNs, so not penalizing here! (Invalid 020$as are counted a bad.)\n      return 0.0;\n    }\n    // We have same matching invalid identifiers. Don't penalize:\n    const allA = [...aValidValuesA, ...aInvalidValuesA, ...zValidValuesA, ...zInvalidValuesA];\n    const allB = [...aValidValuesB, ...aInvalidValuesB, ...zValidValuesB, ...zInvalidValuesB];\n    const [sharedValues, tmp1, tmp2] = getUnionData(allA, allB);\n    debug(`WHATEVER\\tBOTH: ${sharedGoodValues.length}, A only: ${tmp1.length}, B only: ${tmp2.length}`);\n\n    if (sharedValues.length > 0) {\n      return 0;\n    }\n    // We have values but they disagree:\n    return -0.75; // Has good ISBNs on both records, but they did not match\n\n    function getValues(fields) {\n      // Valid values are normalized to their isbn-13 form. Invalid values get their '-'s removed.\n      const goodValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForGoodValues)).map(sf => validatorAndNormalizer(sf.value));\n      const trueGoodValues = goodValues.filter(val => val.valid).map(val => val.value);\n      const wannabeGoodValues = goodValues.filter(val => !val.valid).map(val => val.value);\n      if (!subfieldCodeForBadValues) { // 773\n        return [trueGoodValues, wannabeGoodValues, [], []];\n      }\n      const badValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForBadValues)).map(sf => validatorAndNormalizer(sf.value));\n      const validBadValues = badValues.filter(val => val.valid).map(val => val.value);\n      const invalidBadValues = badValues.filter(val => !val.valid).map(val => val.value);\n      //const badValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForBadValues)).map(sf => validatorAndNormalizer(sf.value).value);\n      return [uniqArray(trueGoodValues), uniqArray(wannabeGoodValues), uniqArray(validBadValues), uniqArray(invalidBadValues)];\n    }\n\n    function validatorAndNormalizer(string) {\n      const string2 = string.replace(/\\. -$/u, ''); // Remove punctuation (773$z)\n\n      // Hack: Historically we 020$a \"1234567890 sidottu\" etc. Try the LHS alone:\n      const string3 = string2.replace(/ .*$/u, '');\n      if (string2 !== string3) {\n        const altResult = validatorAndNormalizer(string3);\n        if (altResult.valid) {\n          return altResult;\n        }\n      }\n\n      const isbnParseResult = isbnParse(string2, '') || '';\n      debugData(`isbnParseResult: ${JSON.stringify(isbnParseResult)}`);\n      if (!isbnParseResult.isValid) {\n        debug(`Not parseable ISBN '${string2}', just removing hyphens`);\n        return {valid: false, value: string2.replace(/-/ug, '')};\n      }\n\n      debug(`Parseable ISBN '${string2}', normalizing to ISBN-13 '${isbnParseResult.isbn13}'`);\n      return {valid: true, value: isbnParseResult.isbn13};\n\n\n    }\n  }\n});\n\n// These are outside the default function as I'll probably want to export these later on\n\nfunction getUnionData(set1, set2) {\n  const shared = set1.filter(val => set2.includes(val));\n  const onlyInSet1 = set1.filter(val => !shared.includes(val));\n  const onlyInSet2 = set2.filter(val => !shared.includes(val));\n  return [shared, onlyInSet1, onlyInSet2];\n}\n\nfunction scoreData(score, factor, n) {\n  return innerScoreData(score, n);\n  function innerScoreData(currScore, remaining) {\n    if (remaining > 0) {\n      return innerScoreData(currScore * factor, remaining-1);\n    }\n    return Math.round(currScore * 100)/100; // 0.600000000001 => 0.6\n  }\n}"],
-  "mappings": "AACA,OAAO,uBAAuB;AAC9B,SAAQ,SAAS,iBAAgB;AAEjC,SAAQ,iBAAgB;AAExB,MAAM,QAAQ,kBAAkB,sFAAsF;AACtH,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,MAAM,YAAY;AAElB,MAAM,YAAY;AAElB,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,UAAM,MAAM,OAAO,IAAI,KAAK,EAAE,OAAO,OAAK,EAAE,WAAW,KAAK,QAAM,CAAC,KAAK,GAAG,EAAE,SAAS,GAAG,IAAI,KAAK,GAAG,KAAK,CAAC;AAC3G,UAAM,OAAO,IAAI,OAAO,OAAK,eAAe,CAAC,CAAC;AAC9C,UAAM,SAAS,IAAI,OAAO,OAAK,CAAC,eAAe,CAAC,CAAC;AACjD,QAAI,KAAK,SAAS,KAAK,OAAO,SAAS,GAAG;AACxC,aAAO;AAAA,IACT;AACA,WAAO;AAEP,aAAS,eAAe,OAAO;AAC7B,aAAO,MAAM,UAAU,KAAK,QAAM,GAAG,SAAS,OAAO,UAAU,KAAK,GAAG,KAAK,CAAC;AAAA,IAC/E;AAAA,EACF;AAAA,EACA,SAAS,CAAC,IAAI,OAAO;AACnB,cAAU,uBAAuB,KAAK,UAAU,EAAE,CAAC,QAAQ,KAAK,UAAU,EAAE,CAAC,EAAE;AAC/E,QAAI,GAAG,WAAW,KAAK,GAAG,WAAW,GAAG;AAEtC,aAAO;AAAA,IACT;AAEA,UAAM,4BAA4B;AAClC,UAAM,2BAA2B;AAEjC,UAAM,CAAC,eAAe,iBAAiB,eAAe,eAAe,IAAI,iBAAiB,IAAI,IAAI;AAClG,UAAM,CAAC,eAAe,iBAAiB,eAAe,eAAe,IAAI,iBAAiB,IAAI,IAAI;AAElG,aAAS,iBAAiB,MAAM,QAAQ;AACtC,YAAM,CAAC,cAAc,gBAAgB,cAAc,cAAc,IAAI,UAAU,IAAI;AACnF,UAAI,aAAa,QAAQ;AACvB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,yBAAyB,WAAW,aAAa,KAAK,IAAI,CAAC,EAAE;AAAA,MAChG;AACA,UAAI,eAAe,QAAQ;AACzB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,yBAAyB,aAAa,eAAe,KAAK,IAAI,CAAC,EAAE;AAAA,MACpG;AACA,UAAI,aAAa,QAAQ;AACvB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,wBAAwB,WAAW,aAAa,KAAK,IAAI,CAAC,EAAE;AAAA,MAC/F;AACA,UAAI,eAAe,QAAQ;AACzB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,wBAAwB,aAAa,eAAe,KAAK,IAAI,CAAC,EAAE;AAAA,MACnG;AACA,aAAO,CAAC,cAAc,gBAAgB,cAAc,cAAc;AAAA,IACpE;AAEA,UAAM,CAAC,kBAAkB,iBAAiB,eAAe,IAAI,aAAa,eAAe,aAAa;AAEtG,UAAM,cAAe,iBAAiB,MAAM,aAAa,gBAAgB,MAAM,aAAa,gBAAgB,MAAM,EAAE;AAEpH,QAAI,iBAAiB,SAAS,GAAG;AAE/B,aAAO,UAAU,WAAW,KAAK,gBAAgB,SAAS,gBAAgB,MAAM;AAAA,IAClF;AAEA,UAAM,WAAW,mBAAmB;AAEpC,aAAS,qBAAqB;AAE5B,UAAI,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC,KAAK,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC,GAAG;AACxH,eAAO;AAAA,MACT;AAIA,YAAM,UAAU,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC;AACvE,UAAI,SAAS;AACX,cAAM,wCAAwC,OAAO,EAAE;AACvD,eAAO,YAAY,IAAI;AAAA,MACzB;AAEA,YAAM,YAAY,gBAAgB,KAAK,UAAQ,gBAAgB,SAAS,IAAI,KAAK,gBAAgB,SAAS,IAAI,CAAC,KAAK,gBAAgB,KAAK,UAAQ,gBAAgB,SAAS,IAAI,CAAC;AAC/K,UAAI,WAAW;AACb,cAAM,4DAA4D,SAAS,EAAE;AAC7E,eAAO,YAAY,IAAI;AAAA,MACzB;AAUA,aAAO;AAAA,IACT;AAIA,QAAI,aAAa,WAAW;AAE1B,aAAO,UAAU,UAAU,KAAK,gBAAgB,SAAS,gBAAgB,SAAS,CAAC;AAAA,IACrF;AAEA,QAAI,WAAW,GAAG;AAEhB,aAAO,UAAU,UAAU,KAAK,gBAAgB,SAAS,gBAAgB,MAAM;AAAA,IACjF;AAIA,QAAI,cAAc,SAAS,gBAAgB,WAAW,KAAK,cAAc,SAAS,gBAAgB,WAAW,GAAG;AAC9G,aAAO;AAAA,IACT;AAEA,UAAM,OAAO,CAAC,GAAG,eAAe,GAAG,iBAAiB,GAAG,eAAe,GAAG,eAAe;AACxF,UAAM,OAAO,CAAC,GAAG,eAAe,GAAG,iBAAiB,GAAG,eAAe,GAAG,eAAe;AACxF,UAAM,CAAC,cAAc,MAAM,IAAI,IAAI,aAAa,MAAM,IAAI;AAC1D,UAAM,kBAAmB,iBAAiB,MAAM,aAAa,KAAK,MAAM,aAAa,KAAK,MAAM,EAAE;AAElG,QAAI,aAAa,SAAS,GAAG;AAC3B,aAAO;AAAA,IACT;AAEA,WAAO;AAEP,aAAS,UAAU,QAAQ;AAEzB,YAAM,aAAa,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,GAAG,SAAS,yBAAyB,CAAC,EAAE,IAAI,QAAM,uBAAuB,GAAG,KAAK,CAAC;AAClJ,YAAM,iBAAiB,WAAW,OAAO,SAAO,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAC/E,YAAM,oBAAoB,WAAW,OAAO,SAAO,CAAC,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AACnF,UAAI,CAAC,0BAA0B;AAC7B,eAAO,CAAC,gBAAgB,mBAAmB,CAAC,GAAG,CAAC,CAAC;AAAA,MACnD;AACA,YAAM,YAAY,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,GAAG,SAAS,wBAAwB,CAAC,EAAE,IAAI,QAAM,uBAAuB,GAAG,KAAK,CAAC;AAChJ,YAAM,iBAAiB,UAAU,OAAO,SAAO,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAC9E,YAAM,mBAAmB,UAAU,OAAO,SAAO,CAAC,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAEjF,aAAO,CAAC,UAAU,cAAc,GAAG,UAAU,iBAAiB,GAAG,UAAU,cAAc,GAAG,UAAU,gBAAgB,CAAC;AAAA,IACzH;AAEA,aAAS,uBAAuB,QAAQ;AACtC,YAAM,UAAU,OAAO,QAAQ,UAAU,EAAE;AAG3C,YAAM,UAAU,QAAQ,QAAQ,SAAS,EAAE;AAC3C,UAAI,YAAY,SAAS;AACvB,cAAM,YAAY,uBAAuB,OAAO;AAChD,YAAI,UAAU,OAAO;AACnB,iBAAO;AAAA,QACT;AAAA,MACF;AAEA,YAAM,kBAAkB,UAAU,SAAS,EAAE,KAAK;AAClD,gBAAU,oBAAoB,KAAK,UAAU,eAAe,CAAC,EAAE;AAC/D,UAAI,CAAC,gBAAgB,SAAS;AAC5B,cAAM,uBAAuB,OAAO,0BAA0B;AAC9D,eAAO,EAAC,OAAO,OAAO,OAAO,QAAQ,QAAQ,OAAO,EAAE,EAAC;AAAA,MACzD;AAEA,YAAM,mBAAmB,OAAO,8BAA8B,gBAAgB,MAAM,GAAG;AACvF,aAAO,EAAC,OAAO,MAAM,OAAO,gBAAgB,OAAM;AAAA,IAGpD;AAAA,EACF;AACF;AAIA,SAAS,aAAa,MAAM,MAAM;AAChC,QAAM,SAAS,KAAK,OAAO,SAAO,KAAK,SAAS,GAAG,CAAC;AACpD,QAAM,aAAa,KAAK,OAAO,SAAO,CAAC,OAAO,SAAS,GAAG,CAAC;AAC3D,QAAM,aAAa,KAAK,OAAO,SAAO,CAAC,OAAO,SAAS,GAAG,CAAC;AAC3D,SAAO,CAAC,QAAQ,YAAY,UAAU;AACxC;AAEA,SAAS,UAAU,OAAO,QAAQ,GAAG;AACnC,SAAO,eAAe,OAAO,CAAC;AAC9B,WAAS,eAAe,WAAW,WAAW;AAC5C,QAAI,YAAY,GAAG;AACjB,aAAO,eAAe,YAAY,QAAQ,YAAU,CAAC;AAAA,IACvD;AACA,WAAO,KAAK,MAAM,YAAY,GAAG,IAAE;AAAA,EACrC;AACF;",
+  "sourcesContent": ["import createDebugLogger from 'debug';\nimport {parse as isbnParse} from 'isbn3';\nimport {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';\n\nconst debug = createDebugLogger(`@natlibfi/melinda-record-matching:match-detection:features:standard-identifiers:ISBN`);\nconst debugData = debug.extend('data');\n\nconst MAX_SCORE = 0.75;\n\nconst setRegexp = /(?:complete set|hela verket|helhet|koko kartasto|koko paketti|koko sarja|koko setti|koko teoksen isbn|koko teos|kokonaisuus|^set\\b|\\bset$|\\bsetti\\b|vol.*set)/ui;\n\nexport default () => ({\n  name: 'ISBN',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    const all = record.get('020').filter(f => f.subfields?.some(sf => ['a', 'z'].includes(sf.code) && sf.value));\n    const sets = all.filter(f => isSetIsbnField(f));\n    const noSets = all.filter(f => !isSetIsbnField(f));\n    if (sets.length > 0 && noSets.length > 0) {\n      return noSets;\n    }\n    return all;\n\n    function isSetIsbnField(field) {\n      return field.subfields.some(sf => sf.code === 'q' && setRegexp.test(sf.value));\n    }\n  },\n  compare: (aa, bb) => {\n    debugData(`Comparing ISBN sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);\n    if (aa.length === 0 || bb.length === 0) {\n      // No data for decision\n      return 0;\n    }\n\n    const subfieldCodeForGoodValues = 'a';\n    const subfieldCodeForBadValues = 'z';\n\n    const [aValidValuesA, aInvalidValuesA, zValidValuesA, zInvalidValuesA] = getValuesWrapper(aa, 'AA'); // initial 'a' and 'z' refer to 020 subfields codes\n    const [aValidValuesB, aInvalidValuesB, zValidValuesB, zInvalidValuesB] = getValuesWrapper(bb, 'BB');\n\n    function getValuesWrapper(data, prefix) {\n      const [aValidValues, aInvalidValues, zValidValues, zInvalidValues] = getValues(data);\n      if (aValidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForGoodValues} VALID: ${aValidValues.join(', ')}`);\n      }\n      if (aInvalidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForGoodValues} INVALID: ${aInvalidValues.join(', ')}`);\n      }\n      if (zValidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForBadValues} VALID: ${zValidValues.join(', ')}`);\n      }\n      if (zInvalidValues.length) {\n        debug(`${prefix}: ${aa[0].tag}$${subfieldCodeForBadValues} INVALID: ${zInvalidValues.join(', ')}`);\n      }\n      return [aValidValues, aInvalidValues, zValidValues, zInvalidValues];\n    }\n\n    const [sharedGoodValues, goodValuesAOnly, goodValuesBOnly] = getUnionData(aValidValuesA, aValidValuesB);\n\n    debug(`GOOD\\tBOTH: ${sharedGoodValues.length}, A only: ${goodValuesAOnly.length}, B only: ${goodValuesBOnly.length}`);\n\n    if (sharedGoodValues.length > 0) {\n      // Third argument (aka 'N times') is >= 0\n      return scoreData(MAX_SCORE, 0.8, goodValuesAOnly.length + goodValuesBOnly.length);\n    }\n\n    const hitScore = scoreSuboptimalHit();\n\n    function scoreSuboptimalHit() {\n      // One record consider ISBN good and the other record considered it's canceled:\n      if (aValidValuesA.some(valA => zValidValuesB.includes(valA)) || aValidValuesB.some(valB => zValidValuesA.includes(valB))) {\n        return MAX_SCORE;\n      }\n\n      // Subfield is for cancelled/whatever values, but the value is syntactically valid:\n      // Could happen for two canceled ISBNs for example. I'll give this two thirds of the full score\n      const zzValid = zValidValuesA.find(valA => zValidValuesB.includes(valA));\n      if (zzValid) {\n        debug(`Both contain a valid value in 020$z: ${zzValid}`);\n        return MAX_SCORE * 2 / 3;\n      }\n      // Shared invalid identifiers:\n      const aaInvalid = aInvalidValuesA.find(valA => aInvalidValuesB.includes(valA) || zInvalidValuesB.includes(valA)) || aInvalidValuesB.find(valB => zInvalidValuesA.includes(valB));\n      if (aaInvalid) {\n        debug(`Shared invalid value in 020$a and 020$a-or-$z subfields: ${aaInvalid}`);\n        return MAX_SCORE * 2 / 3;\n      }\n\n      /* // Currently I think that paired invalid idenfiers in 020$z are meaningless...\n      const zzInvalid = zInvalidValuesB.find(valB => zInvalidValuesA.includes(valB)) || zInvalidValuesA.find(valA => zInvalidValuesB.includes(valA));\n      if (zzInvalid) {\n        debug(`Shared invalid value in 020$z subfields: ${zzInvalid}`);\n        return MAX_SCORE / 3;\n      }\n      */\n\n      return 0;\n    }\n\n\n\n    if (hitScore === MAX_SCORE) {\n      // -1 is needed to make the third argument >= 0 (otherwise min val would be 0)\n      return scoreData(hitScore, 0.8, goodValuesAOnly.length + goodValuesBOnly.length - 1);\n    }\n\n    if (hitScore > 0) { // Canceled/invalid ISBNs match\n      // Note that this is not (currently) penalized for non-matching canceled/invalid ISBNs. Maybe it should be.\n      return scoreData(hitScore, 0.8, goodValuesAOnly.length + goodValuesBOnly.length);\n    }\n\n    // No match:\n\n    if (aValidValuesA.length + aInvalidValuesA.length === 0 || aValidValuesB.length + aInvalidValuesB.length === 0) { // At least one record did not have any good ISBNs, so not penalizing here! (Invalid 020$as are counted a bad.)\n      return 0.0;\n    }\n    // We have same matching invalid identifiers. Don't penalize:\n    const allA = [...aValidValuesA, ...aInvalidValuesA, ...zValidValuesA, ...zInvalidValuesA];\n    const allB = [...aValidValuesB, ...aInvalidValuesB, ...zValidValuesB, ...zInvalidValuesB];\n    const [sharedValues, tmp1, tmp2] = getUnionData(allA, allB);\n    debug(`WHATEVER\\tBOTH: ${sharedGoodValues.length}, A only: ${tmp1.length}, B only: ${tmp2.length}`);\n\n    if (sharedValues.length > 0) {\n      return 0;\n    }\n    // We have values but they disagree:\n    return -0.75; // Has good ISBNs on both records, but they did not match\n\n    function getValues(fields) {\n      // Valid values are normalized to their isbn-13 form. Invalid values get their '-'s removed.\n      const goodValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForGoodValues)).map(sf => validatorAndNormalizer(sf.value));\n      const trueGoodValues = goodValues.filter(val => val.valid).map(val => val.value);\n      const wannabeGoodValues = goodValues.filter(val => !val.valid).map(val => val.value);\n      if (!subfieldCodeForBadValues) { // 773\n        return [trueGoodValues, wannabeGoodValues, [], []];\n      }\n      const badValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForBadValues)).map(sf => validatorAndNormalizer(sf.value));\n      const validBadValues = badValues.filter(val => val.valid).map(val => val.value);\n      const invalidBadValues = badValues.filter(val => !val.valid).map(val => val.value);\n      //const badValues = fields.flatMap(f => f.subfields.filter(sf => sf.code === subfieldCodeForBadValues)).map(sf => validatorAndNormalizer(sf.value).value);\n      return [uniqArray(trueGoodValues), uniqArray(wannabeGoodValues), uniqArray(validBadValues), uniqArray(invalidBadValues)];\n    }\n\n    function validatorAndNormalizer(string) {\n      const string2 = string.replace(/\\. -$/u, ''); // Remove punctuation (773$z)\n\n      // Hack: Historically we 020$a \"1234567890 sidottu\" etc. Try the LHS alone:\n      const string3 = string2.replace(/ .*$/u, '');\n      if (string2 !== string3) {\n        const altResult = validatorAndNormalizer(string3);\n        if (altResult.valid) {\n          return altResult;\n        }\n      }\n\n      const isbnParseResult = isbnParse(string2, '') || '';\n      debugData(`isbnParseResult: ${JSON.stringify(isbnParseResult)}`);\n      if (!isbnParseResult.isValid) {\n        debug(`Not parseable ISBN '${string2}', just removing hyphens`);\n        return {valid: false, value: string2.replace(/-/ug, '')};\n      }\n\n      debug(`Parseable ISBN '${string2}', normalizing to ISBN-13 '${isbnParseResult.isbn13}'`);\n      return {valid: true, value: isbnParseResult.isbn13};\n\n\n    }\n  }\n});\n\n// These are outside the default function as I'll probably want to export these later on\n\nfunction getUnionData(set1, set2) {\n  const shared = set1.filter(val => set2.includes(val));\n  const onlyInSet1 = set1.filter(val => !shared.includes(val));\n  const onlyInSet2 = set2.filter(val => !shared.includes(val));\n  return [shared, onlyInSet1, onlyInSet2];\n}\n\nfunction scoreData(score, factor, n) {\n  return innerScoreData(score, n);\n  function innerScoreData(currScore, remaining) {\n    if (remaining > 0) {\n      return innerScoreData(currScore * factor, remaining-1);\n    }\n    return Math.round(currScore * 100)/100; // 0.600000000001 => 0.6\n  }\n}"],
+  "mappings": "AAAA,OAAO,uBAAuB;AAC9B,SAAQ,SAAS,iBAAgB;AACjC,SAAQ,iBAAgB;AAExB,MAAM,QAAQ,kBAAkB,sFAAsF;AACtH,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,MAAM,YAAY;AAElB,MAAM,YAAY;AAElB,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,UAAM,MAAM,OAAO,IAAI,KAAK,EAAE,OAAO,OAAK,EAAE,WAAW,KAAK,QAAM,CAAC,KAAK,GAAG,EAAE,SAAS,GAAG,IAAI,KAAK,GAAG,KAAK,CAAC;AAC3G,UAAM,OAAO,IAAI,OAAO,OAAK,eAAe,CAAC,CAAC;AAC9C,UAAM,SAAS,IAAI,OAAO,OAAK,CAAC,eAAe,CAAC,CAAC;AACjD,QAAI,KAAK,SAAS,KAAK,OAAO,SAAS,GAAG;AACxC,aAAO;AAAA,IACT;AACA,WAAO;AAEP,aAAS,eAAe,OAAO;AAC7B,aAAO,MAAM,UAAU,KAAK,QAAM,GAAG,SAAS,OAAO,UAAU,KAAK,GAAG,KAAK,CAAC;AAAA,IAC/E;AAAA,EACF;AAAA,EACA,SAAS,CAAC,IAAI,OAAO;AACnB,cAAU,uBAAuB,KAAK,UAAU,EAAE,CAAC,QAAQ,KAAK,UAAU,EAAE,CAAC,EAAE;AAC/E,QAAI,GAAG,WAAW,KAAK,GAAG,WAAW,GAAG;AAEtC,aAAO;AAAA,IACT;AAEA,UAAM,4BAA4B;AAClC,UAAM,2BAA2B;AAEjC,UAAM,CAAC,eAAe,iBAAiB,eAAe,eAAe,IAAI,iBAAiB,IAAI,IAAI;AAClG,UAAM,CAAC,eAAe,iBAAiB,eAAe,eAAe,IAAI,iBAAiB,IAAI,IAAI;AAElG,aAAS,iBAAiB,MAAM,QAAQ;AACtC,YAAM,CAAC,cAAc,gBAAgB,cAAc,cAAc,IAAI,UAAU,IAAI;AACnF,UAAI,aAAa,QAAQ;AACvB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,yBAAyB,WAAW,aAAa,KAAK,IAAI,CAAC,EAAE;AAAA,MAChG;AACA,UAAI,eAAe,QAAQ;AACzB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,yBAAyB,aAAa,eAAe,KAAK,IAAI,CAAC,EAAE;AAAA,MACpG;AACA,UAAI,aAAa,QAAQ;AACvB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,wBAAwB,WAAW,aAAa,KAAK,IAAI,CAAC,EAAE;AAAA,MAC/F;AACA,UAAI,eAAe,QAAQ;AACzB,cAAM,GAAG,MAAM,KAAK,GAAG,CAAC,EAAE,GAAG,IAAI,wBAAwB,aAAa,eAAe,KAAK,IAAI,CAAC,EAAE;AAAA,MACnG;AACA,aAAO,CAAC,cAAc,gBAAgB,cAAc,cAAc;AAAA,IACpE;AAEA,UAAM,CAAC,kBAAkB,iBAAiB,eAAe,IAAI,aAAa,eAAe,aAAa;AAEtG,UAAM,cAAe,iBAAiB,MAAM,aAAa,gBAAgB,MAAM,aAAa,gBAAgB,MAAM,EAAE;AAEpH,QAAI,iBAAiB,SAAS,GAAG;AAE/B,aAAO,UAAU,WAAW,KAAK,gBAAgB,SAAS,gBAAgB,MAAM;AAAA,IAClF;AAEA,UAAM,WAAW,mBAAmB;AAEpC,aAAS,qBAAqB;AAE5B,UAAI,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC,KAAK,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC,GAAG;AACxH,eAAO;AAAA,MACT;AAIA,YAAM,UAAU,cAAc,KAAK,UAAQ,cAAc,SAAS,IAAI,CAAC;AACvE,UAAI,SAAS;AACX,cAAM,wCAAwC,OAAO,EAAE;AACvD,eAAO,YAAY,IAAI;AAAA,MACzB;AAEA,YAAM,YAAY,gBAAgB,KAAK,UAAQ,gBAAgB,SAAS,IAAI,KAAK,gBAAgB,SAAS,IAAI,CAAC,KAAK,gBAAgB,KAAK,UAAQ,gBAAgB,SAAS,IAAI,CAAC;AAC/K,UAAI,WAAW;AACb,cAAM,4DAA4D,SAAS,EAAE;AAC7E,eAAO,YAAY,IAAI;AAAA,MACzB;AAUA,aAAO;AAAA,IACT;AAIA,QAAI,aAAa,WAAW;AAE1B,aAAO,UAAU,UAAU,KAAK,gBAAgB,SAAS,gBAAgB,SAAS,CAAC;AAAA,IACrF;AAEA,QAAI,WAAW,GAAG;AAEhB,aAAO,UAAU,UAAU,KAAK,gBAAgB,SAAS,gBAAgB,MAAM;AAAA,IACjF;AAIA,QAAI,cAAc,SAAS,gBAAgB,WAAW,KAAK,cAAc,SAAS,gBAAgB,WAAW,GAAG;AAC9G,aAAO;AAAA,IACT;AAEA,UAAM,OAAO,CAAC,GAAG,eAAe,GAAG,iBAAiB,GAAG,eAAe,GAAG,eAAe;AACxF,UAAM,OAAO,CAAC,GAAG,eAAe,GAAG,iBAAiB,GAAG,eAAe,GAAG,eAAe;AACxF,UAAM,CAAC,cAAc,MAAM,IAAI,IAAI,aAAa,MAAM,IAAI;AAC1D,UAAM,kBAAmB,iBAAiB,MAAM,aAAa,KAAK,MAAM,aAAa,KAAK,MAAM,EAAE;AAElG,QAAI,aAAa,SAAS,GAAG;AAC3B,aAAO;AAAA,IACT;AAEA,WAAO;AAEP,aAAS,UAAU,QAAQ;AAEzB,YAAM,aAAa,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,GAAG,SAAS,yBAAyB,CAAC,EAAE,IAAI,QAAM,uBAAuB,GAAG,KAAK,CAAC;AAClJ,YAAM,iBAAiB,WAAW,OAAO,SAAO,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAC/E,YAAM,oBAAoB,WAAW,OAAO,SAAO,CAAC,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AACnF,UAAI,CAAC,0BAA0B;AAC7B,eAAO,CAAC,gBAAgB,mBAAmB,CAAC,GAAG,CAAC,CAAC;AAAA,MACnD;AACA,YAAM,YAAY,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,GAAG,SAAS,wBAAwB,CAAC,EAAE,IAAI,QAAM,uBAAuB,GAAG,KAAK,CAAC;AAChJ,YAAM,iBAAiB,UAAU,OAAO,SAAO,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAC9E,YAAM,mBAAmB,UAAU,OAAO,SAAO,CAAC,IAAI,KAAK,EAAE,IAAI,SAAO,IAAI,KAAK;AAEjF,aAAO,CAAC,UAAU,cAAc,GAAG,UAAU,iBAAiB,GAAG,UAAU,cAAc,GAAG,UAAU,gBAAgB,CAAC;AAAA,IACzH;AAEA,aAAS,uBAAuB,QAAQ;AACtC,YAAM,UAAU,OAAO,QAAQ,UAAU,EAAE;AAG3C,YAAM,UAAU,QAAQ,QAAQ,SAAS,EAAE;AAC3C,UAAI,YAAY,SAAS;AACvB,cAAM,YAAY,uBAAuB,OAAO;AAChD,YAAI,UAAU,OAAO;AACnB,iBAAO;AAAA,QACT;AAAA,MACF;AAEA,YAAM,kBAAkB,UAAU,SAAS,EAAE,KAAK;AAClD,gBAAU,oBAAoB,KAAK,UAAU,eAAe,CAAC,EAAE;AAC/D,UAAI,CAAC,gBAAgB,SAAS;AAC5B,cAAM,uBAAuB,OAAO,0BAA0B;AAC9D,eAAO,EAAC,OAAO,OAAO,OAAO,QAAQ,QAAQ,OAAO,EAAE,EAAC;AAAA,MACzD;AAEA,YAAM,mBAAmB,OAAO,8BAA8B,gBAAgB,MAAM,GAAG;AACvF,aAAO,EAAC,OAAO,MAAM,OAAO,gBAAgB,OAAM;AAAA,IAGpD;AAAA,EACF;AACF;AAIA,SAAS,aAAa,MAAM,MAAM;AAChC,QAAM,SAAS,KAAK,OAAO,SAAO,KAAK,SAAS,GAAG,CAAC;AACpD,QAAM,aAAa,KAAK,OAAO,SAAO,CAAC,OAAO,SAAS,GAAG,CAAC;AAC3D,QAAM,aAAa,KAAK,OAAO,SAAO,CAAC,OAAO,SAAS,GAAG,CAAC;AAC3D,SAAO,CAAC,QAAQ,YAAY,UAAU;AACxC;AAEA,SAAS,UAAU,OAAO,QAAQ,GAAG;AACnC,SAAO,eAAe,OAAO,CAAC;AAC9B,WAAS,eAAe,WAAW,WAAW;AAC5C,QAAI,YAAY,GAAG;AACjB,aAAO,eAAe,YAAY,QAAQ,YAAU,CAAC;AAAA,IACvD;AACA,WAAO,KAAK,MAAM,YAAY,GAAG,IAAE;AAAA,EACrC;AACF;",
   "names": []
 }

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

@@ -1,3 +1,4 @@
+import { uniqArray } from "@natlibfi/marc-record-validators-melinda/dist/utils.js";
 import createDebugLogger from "debug";
 const debug = createDebugLogger("@natlibfi/melinda-record-matching:match-detection:features:issn");
 const debugData = debug.extend("data");
@@ -42,7 +43,4 @@ export default () => ({
     return -0.2;
   }
 });
-export function uniqArray(arr) {
-  return arr.filter((val, i) => arr.indexOf(val) === i);
-}
 //# sourceMappingURL=issn.js.map

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

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../../../../src/match-detection/features/bib/issn.js"],
-  "sourcesContent": ["import createDebugLogger from 'debug';\n\nconst debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:issn');\nconst debugData = debug.extend('data');\n\nexport default () => ({\n  name: 'ISSN',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    return getIssns();\n\n    function getIssns() {\n      const fields = record.get(/^022$/u);\n      if (fields.length === 0) {\n        return [];\n      }\n\n      debug(`\\t ${fields.length} potential ISSN fields (${fields[0].tag})`);\n      const subfieldCodes = ['a', 'y', 'z']\n      //debug(`\\t subfield codes: '${subfieldCodes.join(\"', '\")}'`);\n      const subfieldValues = fields.flatMap(f => f.subfields.filter(sf => subfieldCodes.includes(sf.code)).map(sf => sf.value));\n      //debug(`\\t cand values: '${subfieldValues.join(\"', '\")}'`);\n      // Stripping punctuaction with substring here is pretty quick and dirty approach...\n      const validSubfieldValues = subfieldValues?.map(val => normalizeSubfieldValue(val)).filter(val => isValidIssn(val));\n      if (!validSubfieldValues) {\n        return [];\n      }\n      return uniqArray(validSubfieldValues);\n    }\n\n    function normalizeSubfieldValue(val) {\n      return val.replace(/[., :;].*$/u, '');\n    }\n\n    function isValidIssn(val) {\n      return /^[0-9][0-9][0-9][0-9]-[0-9][0-9][0-9][0-9X]$/u.test(val);\n    }\n  },\n  compare: (aa, bb) => {\n    debugData(`Comparing ISSN sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);\n    if (aa.length === 0 || bb.length === 0) {\n      // No data for decision\n      return 0;\n    }\n    const firstSharedIssn = aa.find(val => bb.includes(val));\n    if (firstSharedIssn) {\n      debug(`\\t Shared ISSN found: '${firstSharedIssn}'`);\n      // Maybe less for comps?\n      return 0.2;\n    }\n    return -0.2;\n\n  }\n});\n\nexport function uniqArray(arr) {\n  return arr.filter((val, i) => arr.indexOf(val) === i);\n}"],
-  "mappings": "AAAA,OAAO,uBAAuB;AAE9B,MAAM,QAAQ,kBAAkB,iEAAiE;AACjG,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,WAAO,SAAS;AAEhB,aAAS,WAAW;AAClB,YAAM,SAAS,OAAO,IAAI,QAAQ;AAClC,UAAI,OAAO,WAAW,GAAG;AACvB,eAAO,CAAC;AAAA,MACV;AAEA,YAAM,KAAM,OAAO,MAAM,2BAA2B,OAAO,CAAC,EAAE,GAAG,GAAG;AACpE,YAAM,gBAAgB,CAAC,KAAK,KAAK,GAAG;AAEpC,YAAM,iBAAiB,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,cAAc,SAAS,GAAG,IAAI,CAAC,EAAE,IAAI,QAAM,GAAG,KAAK,CAAC;AAGxH,YAAM,sBAAsB,gBAAgB,IAAI,SAAO,uBAAuB,GAAG,CAAC,EAAE,OAAO,SAAO,YAAY,GAAG,CAAC;AAClH,UAAI,CAAC,qBAAqB;AACxB,eAAO,CAAC;AAAA,MACV;AACA,aAAO,UAAU,mBAAmB;AAAA,IACtC;AAEA,aAAS,uBAAuB,KAAK;AACnC,aAAO,IAAI,QAAQ,eAAe,EAAE;AAAA,IACtC;AAEA,aAAS,YAAY,KAAK;AACxB,aAAO,gDAAgD,KAAK,GAAG;AAAA,IACjE;AAAA,EACF;AAAA,EACA,SAAS,CAAC,IAAI,OAAO;AACnB,cAAU,uBAAuB,KAAK,UAAU,EAAE,CAAC,QAAQ,KAAK,UAAU,EAAE,CAAC,EAAE;AAC/E,QAAI,GAAG,WAAW,KAAK,GAAG,WAAW,GAAG;AAEtC,aAAO;AAAA,IACT;AACA,UAAM,kBAAkB,GAAG,KAAK,SAAO,GAAG,SAAS,GAAG,CAAC;AACvD,QAAI,iBAAiB;AACnB,YAAM,yBAA0B,eAAe,GAAG;AAElD,aAAO;AAAA,IACT;AACA,WAAO;AAAA,EAET;AACF;AAEO,gBAAS,UAAU,KAAK;AAC7B,SAAO,IAAI,OAAO,CAAC,KAAK,MAAM,IAAI,QAAQ,GAAG,MAAM,CAAC;AACtD;",
+  "sourcesContent": ["import {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';\nimport createDebugLogger from 'debug';\n\nconst debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:issn');\nconst debugData = debug.extend('data');\n\nexport default () => ({\n  name: 'ISSN',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    return getIssns();\n\n    function getIssns() {\n      const fields = record.get(/^022$/u);\n      if (fields.length === 0) {\n        return [];\n      }\n\n      debug(`\\t ${fields.length} potential ISSN fields (${fields[0].tag})`);\n      const subfieldCodes = ['a', 'y', 'z']\n      //debug(`\\t subfield codes: '${subfieldCodes.join(\"', '\")}'`);\n      const subfieldValues = fields.flatMap(f => f.subfields.filter(sf => subfieldCodes.includes(sf.code)).map(sf => sf.value));\n      //debug(`\\t cand values: '${subfieldValues.join(\"', '\")}'`);\n      // Stripping punctuaction with substring here is pretty quick and dirty approach...\n      const validSubfieldValues = subfieldValues?.map(val => normalizeSubfieldValue(val)).filter(val => isValidIssn(val));\n      if (!validSubfieldValues) {\n        return [];\n      }\n      return uniqArray(validSubfieldValues);\n    }\n\n    function normalizeSubfieldValue(val) {\n      return val.replace(/[., :;].*$/u, '');\n    }\n\n    function isValidIssn(val) {\n      return /^[0-9][0-9][0-9][0-9]-[0-9][0-9][0-9][0-9X]$/u.test(val);\n    }\n  },\n  compare: (aa, bb) => {\n    debugData(`Comparing ISSN sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);\n    if (aa.length === 0 || bb.length === 0) {\n      // No data for decision\n      return 0;\n    }\n    const firstSharedIssn = aa.find(val => bb.includes(val));\n    if (firstSharedIssn) {\n      debug(`\\t Shared ISSN found: '${firstSharedIssn}'`);\n      // Maybe less for comps?\n      return 0.2;\n    }\n    return -0.2;\n\n  }\n});\n"],
+  "mappings": "AAAA,SAAQ,iBAAgB;AACxB,OAAO,uBAAuB;AAE9B,MAAM,QAAQ,kBAAkB,iEAAiE;AACjG,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,WAAO,SAAS;AAEhB,aAAS,WAAW;AAClB,YAAM,SAAS,OAAO,IAAI,QAAQ;AAClC,UAAI,OAAO,WAAW,GAAG;AACvB,eAAO,CAAC;AAAA,MACV;AAEA,YAAM,KAAM,OAAO,MAAM,2BAA2B,OAAO,CAAC,EAAE,GAAG,GAAG;AACpE,YAAM,gBAAgB,CAAC,KAAK,KAAK,GAAG;AAEpC,YAAM,iBAAiB,OAAO,QAAQ,OAAK,EAAE,UAAU,OAAO,QAAM,cAAc,SAAS,GAAG,IAAI,CAAC,EAAE,IAAI,QAAM,GAAG,KAAK,CAAC;AAGxH,YAAM,sBAAsB,gBAAgB,IAAI,SAAO,uBAAuB,GAAG,CAAC,EAAE,OAAO,SAAO,YAAY,GAAG,CAAC;AAClH,UAAI,CAAC,qBAAqB;AACxB,eAAO,CAAC;AAAA,MACV;AACA,aAAO,UAAU,mBAAmB;AAAA,IACtC;AAEA,aAAS,uBAAuB,KAAK;AACnC,aAAO,IAAI,QAAQ,eAAe,EAAE;AAAA,IACtC;AAEA,aAAS,YAAY,KAAK;AACxB,aAAO,gDAAgD,KAAK,GAAG;AAAA,IACjE;AAAA,EACF;AAAA,EACA,SAAS,CAAC,IAAI,OAAO;AACnB,cAAU,uBAAuB,KAAK,UAAU,EAAE,CAAC,QAAQ,KAAK,UAAU,EAAE,CAAC,EAAE;AAC/E,QAAI,GAAG,WAAW,KAAK,GAAG,WAAW,GAAG;AAEtC,aAAO;AAAA,IACT;AACA,UAAM,kBAAkB,GAAG,KAAK,SAAO,GAAG,SAAS,GAAG,CAAC;AACvD,QAAI,iBAAiB;AACnB,YAAM,yBAA0B,eAAe,GAAG;AAElD,aAAO;AAAA,IACT;AACA,WAAO;AAAA,EAET;AACF;",
   "names": []
 }

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

@@ -0,0 +1,61 @@
+import { uniqArray } from "@natlibfi/marc-record-validators-melinda/dist/utils.js";
+import createDebugLogger from "debug";
+const debug = createDebugLogger("@natlibfi/melinda-record-matching:match-detection:features:publisher-number");
+const debugData = debug.extend("data");
+export default () => ({
+  name: "Publisher or Distributor Number",
+  extract: ({
+    record
+    /*, recordExternal*/
+  }) => {
+    return getIdentifiers();
+    function getIdentifiers() {
+      const values = fieldsToPublisherNumber(record.get(/^028$/u));
+      debug(`	 ${values.length} potential publisher identifiers: ${values.join(", ")}`);
+      return uniqArray(values);
+      function fieldsToPublisherNumber(fields, result = []) {
+        const [field, ...remainingFields] = fields;
+        if (!field) {
+          return result;
+        }
+        const a = field.subfields.find((sf) => sf.code === "a");
+        if (!a || !a.value) {
+          return fieldsToPublisherNumber(remainingFields, result);
+        }
+        const aval = normalizePublisherNumber(`${a.value}`);
+        const b = field.subfields.find((sf) => sf.code === "b");
+        if (!b || !b.value) {
+          return fieldsToPublisherNumber(remainingFields, [...result, aval]);
+        }
+        const baval = normalizePublisherNumber(`${b.value}${a.value}`);
+        return fieldsToPublisherNumber(remainingFields, [...result, baval, aval]);
+      }
+      function normalizePublisherNumber(val) {
+        return val.replace(/[- .,:;]/ug, "");
+      }
+    }
+  },
+  compare: (aa, bb) => {
+    debugData(`Comparing identifier sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);
+    if (aa.length === 0 || bb.length === 0) {
+      return 0;
+    }
+    const sharedIdentifiers = aa.filter((val) => bb.includes(val));
+    if (sharedIdentifiers.length > 0) {
+      const goodMatch = sharedIdentifiers.find((identifier) => identifier.match(/[^0-9]/u));
+      if (goodMatch) {
+        debug(`	Shared identifier found: '${goodMatch}'`);
+        return 0.5;
+      }
+      debug(`	Shared identifier (numbers only) found: '${sharedIdentifiers[0]}'`);
+      return 0.2;
+    }
+    const aa2 = aa.map((a) => a.replace(/[^0-9]/ug, ""));
+    const bb2 = bb.map((b) => b.replace(/[^0-9]/ug, ""));
+    if (aa2.find((val) => val !== "" && bb2.includes(val))) {
+      return 0;
+    }
+    return -0.2;
+  }
+});
+//# sourceMappingURL=publisherNumber.js.map

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

@@ -0,0 +1,7 @@
+{
+  "version": 3,
+  "sources": ["../../../../src/match-detection/features/bib/publisherNumber.js"],
+  "sourcesContent": ["// Compare publusher number (028$b$a or 028$a)\nimport {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';\nimport createDebugLogger from 'debug';\n\nconst debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:publisher-number');\nconst debugData = debug.extend('data');\n\nexport default () => ({\n  name: 'Publisher or Distributor Number',\n  extract: ({record/*, recordExternal*/}) => {\n    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';\n\n    return getIdentifiers();\n\n    function getIdentifiers() {\n      const values = fieldsToPublisherNumber(record.get(/^028$/u));\n\n      debug(`\\t ${values.length} potential publisher identifiers: ${values.join(', ')}`);\n\n      return uniqArray(values);\n\n      function fieldsToPublisherNumber(fields, result = []) {\n        const [field, ...remainingFields] = fields;\n        if (!field) {\n          return result;\n        }\n        const a = field.subfields.find(sf => sf.code === 'a');\n        if (!a || !a.value) {\n          return fieldsToPublisherNumber(remainingFields, result);\n        }\n        const aval = normalizePublisherNumber(`${a.value}`);\n\n        const b = field.subfields.find(sf => sf.code === 'b');\n        if (!b || !b.value) {\n          return fieldsToPublisherNumber(remainingFields, [...result, aval]);\n        }\n        const baval = normalizePublisherNumber(`${b.value}${a.value}`);\n        return fieldsToPublisherNumber(remainingFields, [...result, baval, aval]);\n      }\n\n      function normalizePublisherNumber(val) {\n        return val.replace(/[- .,:;]/ug, '');\n      }\n    }\n  },\n  compare: (aa, bb) => {\n    debugData(`Comparing identifier sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);\n    if (aa.length === 0 || bb.length === 0) {\n      // No data for decision\n      return 0;\n    }\n    const sharedIdentifiers = aa.filter(val => bb.includes(val));\n    if (sharedIdentifiers.length > 0) {\n      const goodMatch = sharedIdentifiers.find(identifier => identifier.match(/[^0-9]/u));\n      if (goodMatch) {\n        debug(`\\tShared identifier found: '${goodMatch}'`);\n        return 0.5;\n      }\n      debug(`\\tShared identifier (numbers only) found: '${sharedIdentifiers[0]}'`);\n      return 0.2;\n    }\n\n    const aa2 = aa.map(a => a.replace(/[^0-9]/ug, ''));\n    const bb2 = bb.map(b => b.replace(/[^0-9]/ug, ''));\n\n    // Numbers matches (aka don't penalize \"Fazer F-123\" vs \"F-123\")\n    // Note that FOOLP-123 vs FOOCD-123 will now also return 0.0. Still nothing positive is returned, so no damage done.\n    if (aa2.find(val => val !== '' && bb2.includes(val))) {\n      return 0.0;\n    }\n\n    return -0.2;\n\n  }\n});\n\n"],
+  "mappings": "AACA,SAAQ,iBAAgB;AACxB,OAAO,uBAAuB;AAE9B,MAAM,QAAQ,kBAAkB,6EAA6E;AAC7G,MAAM,YAAY,MAAM,OAAO,MAAM;AAErC,eAAe,OAAO;AAAA,EACpB,MAAM;AAAA,EACN,SAAS,CAAC;AAAA,IAAC;AAAA;AAAA,EAA0B,MAAM;AAGzC,WAAO,eAAe;AAEtB,aAAS,iBAAiB;AACxB,YAAM,SAAS,wBAAwB,OAAO,IAAI,QAAQ,CAAC;AAE3D,YAAM,KAAM,OAAO,MAAM,qCAAqC,OAAO,KAAK,IAAI,CAAC,EAAE;AAEjF,aAAO,UAAU,MAAM;AAEvB,eAAS,wBAAwB,QAAQ,SAAS,CAAC,GAAG;AACpD,cAAM,CAAC,OAAO,GAAG,eAAe,IAAI;AACpC,YAAI,CAAC,OAAO;AACV,iBAAO;AAAA,QACT;AACA,cAAM,IAAI,MAAM,UAAU,KAAK,QAAM,GAAG,SAAS,GAAG;AACpD,YAAI,CAAC,KAAK,CAAC,EAAE,OAAO;AAClB,iBAAO,wBAAwB,iBAAiB,MAAM;AAAA,QACxD;AACA,cAAM,OAAO,yBAAyB,GAAG,EAAE,KAAK,EAAE;AAElD,cAAM,IAAI,MAAM,UAAU,KAAK,QAAM,GAAG,SAAS,GAAG;AACpD,YAAI,CAAC,KAAK,CAAC,EAAE,OAAO;AAClB,iBAAO,wBAAwB,iBAAiB,CAAC,GAAG,QAAQ,IAAI,CAAC;AAAA,QACnE;AACA,cAAM,QAAQ,yBAAyB,GAAG,EAAE,KAAK,GAAG,EAAE,KAAK,EAAE;AAC7D,eAAO,wBAAwB,iBAAiB,CAAC,GAAG,QAAQ,OAAO,IAAI,CAAC;AAAA,MAC1E;AAEA,eAAS,yBAAyB,KAAK;AACrC,eAAO,IAAI,QAAQ,cAAc,EAAE;AAAA,MACrC;AAAA,IACF;AAAA,EACF;AAAA,EACA,SAAS,CAAC,IAAI,OAAO;AACnB,cAAU,6BAA6B,KAAK,UAAU,EAAE,CAAC,QAAQ,KAAK,UAAU,EAAE,CAAC,EAAE;AACrF,QAAI,GAAG,WAAW,KAAK,GAAG,WAAW,GAAG;AAEtC,aAAO;AAAA,IACT;AACA,UAAM,oBAAoB,GAAG,OAAO,SAAO,GAAG,SAAS,GAAG,CAAC;AAC3D,QAAI,kBAAkB,SAAS,GAAG;AAChC,YAAM,YAAY,kBAAkB,KAAK,gBAAc,WAAW,MAAM,SAAS,CAAC;AAClF,UAAI,WAAW;AACb,cAAM,8BAA+B,SAAS,GAAG;AACjD,eAAO;AAAA,MACT;AACA,YAAM,6CAA8C,kBAAkB,CAAC,CAAC,GAAG;AAC3E,aAAO;AAAA,IACT;AAEA,UAAM,MAAM,GAAG,IAAI,OAAK,EAAE,QAAQ,YAAY,EAAE,CAAC;AACjD,UAAM,MAAM,GAAG,IAAI,OAAK,EAAE,QAAQ,YAAY,EAAE,CAAC;AAIjD,QAAI,IAAI,KAAK,SAAO,QAAQ,MAAM,IAAI,SAAS,GAAG,CAAC,GAAG;AACpD,aAAO;AAAA,IACT;AAEA,WAAO;AAAA,EAET;AACF;",
+  "names": []
+}

package/dist/match-detection/index.js

@@ -26,7 +26,7 @@ B: ${recordB}`);
     const featurePairs = generateFeaturePairs(featuresA2, featuresB);
     const similarityVector = generateSimilarityVector(featurePairs);
     if (similarityVector.some((v) => v >= MIN_PROPABILITY_QUALIFIER)) {
-      const probability = calculateProbability(similarityVector);
+      const probability = Math.round(calculateProbability(similarityVector) * 100) / 100;
       debug(`probability: ${probability} (Threshold: ${threshold})`);
       if (similarityVector.some((v) => v <= FORCE_FAIL)) {
         debug(`Some feature resulted in utter failure (${FORCE_FAIL} or less)`);

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

@@ -1,7 +1,7 @@
 {
   "version": 3,
   "sources": ["../../src/match-detection/index.js"],
-  "sourcesContent": ["\nimport createDebugLogger from 'debug';\nimport * as features from './features/index.js';\n\nexport {features};\n\nexport default ({strategy, threshold = 0.8999}, returnStrategy = false, localSettings = {}) => ({recordA, recordB, recordAExternal = {}, recordBExternal = {}}) => {\n  const MIN_PROPABILITY_QUALIFIER = 0.4; // Title perfect match: 0.5\n  const FORCE_FAIL = -1.0;\n\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection');\n  const debugData = debug.extend('data');\n\n  debugData(`Strategy: ${JSON.stringify(strategy)}, Threshold: ${JSON.stringify(threshold)}, ReturnStrategy: ${JSON.stringify(returnStrategy)}`);\n  debugData(`Records: A: ${recordA}\\nB: ${recordB}`);\n  debug(`Externals: A: ${JSON.stringify(recordAExternal)}, B: ${JSON.stringify(recordBExternal)}`);\n  // We could add here labels for records if we didn't get external labels\n\n  const featuresA = extractFeatures({record: recordA, recordExternal: recordAExternal, localSettings});\n\n  debug(`We got an array of records: ${Array.isArray(recordB)}`);\n  const recordsB = Array.isArray(recordB) ? recordB : [recordB];\n  const recordsBExternal = Array.isArray(recordB) ? recordBExternal : [recordBExternal];\n\n  const detectionResults = recordsB.map((record, index) => actualDetection({featuresA, recordAExternal, record, recordExternal: recordsBExternal[index], index}));\n\n  // if we got array of records, we return an array of result\n  // if we got a singular record, we return a singular result\n  return Array.isArray(recordB) ? detectionResults : detectionResults[0];\n\n  function actualDetection({featuresA, record, recordExternal, index, localSettings}) {\n    const labelA = recordAExternal && recordAExternal.label ? recordAExternal.label : 'a';\n    const labelB = recordExternal && recordExternal.label ? recordExternal.label : 'b';\n\n\n    debug(`Actual detection for record ${index + 1} ${labelB}`);\n    const featuresB = extractFeatures({record, recordExternal, localSettings});\n\n    debugData(`Features (a: ${labelA}): ${JSON.stringify(featuresA)}`);\n    debugData(`Features (b: ${labelB}): ${JSON.stringify(featuresB)}`);\n\n    const featurePairs = generateFeaturePairs(featuresA, featuresB);\n    const similarityVector = generateSimilarityVector(featurePairs);\n\n    if (similarityVector.some(v => v >= MIN_PROPABILITY_QUALIFIER)) {\n      const probability = calculateProbability(similarityVector);\n      debug(`probability: ${probability} (Threshold: ${threshold})`);\n      if (similarityVector.some(v => v <= FORCE_FAIL)) {\n        debug(`Some feature resulted in utter failure (${FORCE_FAIL} or less)`);\n        return returnResult({match: false, probability: 0.0});\n      }\n\n      return returnResult({match: probability >= threshold, probability});\n    }\n\n    debugData(`No feature yielded minimum probability amount of points (${MIN_PROPABILITY_QUALIFIER})`);\n    return returnResult({match: false, probability: 0.0});\n  }\n\n  function extractFeatures({record, recordExternal}) {\n    return strategy.reduce((acc, {name, extract}) => acc.concat({name, value: extract({record, recordExternal})}), []);\n  }\n\n  function returnResult(result) {\n    if (returnStrategy) {\n      debug(`Returning detection strategy with the result`);\n      const resultWithStrategy = {match: result.match, probability: result.probability, strategy: formatStrategy(strategy), threshold};\n      debugData(`${JSON.stringify(resultWithStrategy)}`);\n      return resultWithStrategy;\n    }\n    return result;\n  }\n\n  function formatStrategy(strategy) {\n    const strategyNames = strategy.map(element => element.name);\n    return strategyNames || [];\n  }\n\n  function calculateProbability(similarityVector) {\n    const probability = similarityVector.reduce((acc, v) => acc + v, 0.0);\n    return probability > 1.0 ? 1.0 : probability;\n  }\n\n  function generateSimilarityVector(featurePairs) {\n    const compared = featurePairs.map(({name, a, b}) => {\n      const {compare} = strategy.find(({name: featureName}) => name === featureName);\n      const points = compare(a, b);\n      return {name, points};\n    });\n\n    debugData(`Points: ${JSON.stringify(compared)}`);\n    return compared.map(({points}) => points);\n  }\n\n  function generateFeaturePairs(featuresA, featuresB) {\n    const pairs = generatePairs();\n    const missingFeatures = findMissing();\n\n    debug(`Not comparing the following features because one, or both records are missing features: ${JSON.stringify(missingFeatures)}`);\n    return pairs;\n\n    function generatePairs() {\n      return featuresA\n        .reduce((acc, {name, value}, index) => acc.concat({\n          name,\n          a: value,\n          b: featuresB[index].value\n        }), [])\n        .filter(({a, b}) => {\n          if (a.length === 0 || b.length === 0) {\n            return false;\n          }\n\n          return true;\n        });\n    }\n\n    function findMissing() {\n      return featuresA\n        .map(({name}) => name)\n        .filter(v => pairs.some(({name}) => name === v) === false);\n    }\n  }\n\n};\n"],
-  "mappings": "AACA,OAAO,uBAAuB;AAC9B,YAAY,cAAc;AAE1B,SAAQ;AAER,eAAe,CAAC,EAAC,UAAU,YAAY,OAAM,GAAG,iBAAiB,OAAO,gBAAgB,CAAC,MAAM,CAAC,EAAC,SAAS,SAAS,kBAAkB,CAAC,GAAG,kBAAkB,CAAC,EAAC,MAAM;AACjK,QAAM,4BAA4B;AAClC,QAAM,aAAa;AAEnB,QAAM,QAAQ,kBAAkB,mDAAmD;AACnF,QAAM,YAAY,MAAM,OAAO,MAAM;AAErC,YAAU,aAAa,KAAK,UAAU,QAAQ,CAAC,gBAAgB,KAAK,UAAU,SAAS,CAAC,qBAAqB,KAAK,UAAU,cAAc,CAAC,EAAE;AAC7I,YAAU,eAAe,OAAO;AAAA,KAAQ,OAAO,EAAE;AACjD,QAAM,iBAAiB,KAAK,UAAU,eAAe,CAAC,QAAQ,KAAK,UAAU,eAAe,CAAC,EAAE;AAG/F,QAAM,YAAY,gBAAgB,EAAC,QAAQ,SAAS,gBAAgB,iBAAiB,cAAa,CAAC;AAEnG,QAAM,+BAA+B,MAAM,QAAQ,OAAO,CAAC,EAAE;AAC7D,QAAM,WAAW,MAAM,QAAQ,OAAO,IAAI,UAAU,CAAC,OAAO;AAC5D,QAAM,mBAAmB,MAAM,QAAQ,OAAO,IAAI,kBAAkB,CAAC,eAAe;AAEpF,QAAM,mBAAmB,SAAS,IAAI,CAAC,QAAQ,UAAU,gBAAgB,EAAC,WAAW,iBAAiB,QAAQ,gBAAgB,iBAAiB,KAAK,GAAG,MAAK,CAAC,CAAC;AAI9J,SAAO,MAAM,QAAQ,OAAO,IAAI,mBAAmB,iBAAiB,CAAC;AAErE,WAAS,gBAAgB,EAAC,WAAAA,YAAW,QAAQ,gBAAgB,OAAO,eAAAC,eAAa,GAAG;AAClF,UAAM,SAAS,mBAAmB,gBAAgB,QAAQ,gBAAgB,QAAQ;AAClF,UAAM,SAAS,kBAAkB,eAAe,QAAQ,eAAe,QAAQ;AAG/E,UAAM,+BAA+B,QAAQ,CAAC,IAAI,MAAM,EAAE;AAC1D,UAAM,YAAY,gBAAgB,EAAC,QAAQ,gBAAgB,eAAAA,eAAa,CAAC;AAEzE,cAAU,gBAAgB,MAAM,MAAM,KAAK,UAAUD,UAAS,CAAC,EAAE;AACjE,cAAU,gBAAgB,MAAM,MAAM,KAAK,UAAU,SAAS,CAAC,EAAE;AAEjE,UAAM,eAAe,qBAAqBA,YAAW,SAAS;AAC9D,UAAM,mBAAmB,yBAAyB,YAAY;AAE9D,QAAI,iBAAiB,KAAK,OAAK,KAAK,yBAAyB,GAAG;AAC9D,YAAM,cAAc,qBAAqB,gBAAgB;AACzD,YAAM,gBAAgB,WAAW,gBAAgB,SAAS,GAAG;AAC7D,UAAI,iBAAiB,KAAK,OAAK,KAAK,UAAU,GAAG;AAC/C,cAAM,2CAA2C,UAAU,WAAW;AACtE,eAAO,aAAa,EAAC,OAAO,OAAO,aAAa,EAAG,CAAC;AAAA,MACtD;AAEA,aAAO,aAAa,EAAC,OAAO,eAAe,WAAW,YAAW,CAAC;AAAA,IACpE;AAEA,cAAU,4DAA4D,yBAAyB,GAAG;AAClG,WAAO,aAAa,EAAC,OAAO,OAAO,aAAa,EAAG,CAAC;AAAA,EACtD;AAEA,WAAS,gBAAgB,EAAC,QAAQ,eAAc,GAAG;AACjD,WAAO,SAAS,OAAO,CAAC,KAAK,EAAC,MAAM,QAAO,MAAM,IAAI,OAAO,EAAC,MAAM,OAAO,QAAQ,EAAC,QAAQ,eAAc,CAAC,EAAC,CAAC,GAAG,CAAC,CAAC;AAAA,EACnH;AAEA,WAAS,aAAa,QAAQ;AAC5B,QAAI,gBAAgB;AAClB,YAAM,8CAA8C;AACpD,YAAM,qBAAqB,EAAC,OAAO,OAAO,OAAO,aAAa,OAAO,aAAa,UAAU,eAAe,QAAQ,GAAG,UAAS;AAC/H,gBAAU,GAAG,KAAK,UAAU,kBAAkB,CAAC,EAAE;AACjD,aAAO;AAAA,IACT;AACA,WAAO;AAAA,EACT;AAEA,WAAS,eAAeE,WAAU;AAChC,UAAM,gBAAgBA,UAAS,IAAI,aAAW,QAAQ,IAAI;AAC1D,WAAO,iBAAiB,CAAC;AAAA,EAC3B;AAEA,WAAS,qBAAqB,kBAAkB;AAC9C,UAAM,cAAc,iBAAiB,OAAO,CAAC,KAAK,MAAM,MAAM,GAAG,CAAG;AACpE,WAAO,cAAc,IAAM,IAAM;AAAA,EACnC;AAEA,WAAS,yBAAyB,cAAc;AAC9C,UAAM,WAAW,aAAa,IAAI,CAAC,EAAC,MAAM,GAAG,EAAC,MAAM;AAClD,YAAM,EAAC,QAAO,IAAI,SAAS,KAAK,CAAC,EAAC,MAAM,YAAW,MAAM,SAAS,WAAW;AAC7E,YAAM,SAAS,QAAQ,GAAG,CAAC;AAC3B,aAAO,EAAC,MAAM,OAAM;AAAA,IACtB,CAAC;AAED,cAAU,WAAW,KAAK,UAAU,QAAQ,CAAC,EAAE;AAC/C,WAAO,SAAS,IAAI,CAAC,EAAC,OAAM,MAAM,MAAM;AAAA,EAC1C;AAEA,WAAS,qBAAqBF,YAAW,WAAW;AAClD,UAAM,QAAQ,cAAc;AAC5B,UAAM,kBAAkB,YAAY;AAEpC,UAAM,2FAA2F,KAAK,UAAU,eAAe,CAAC,EAAE;AAClI,WAAO;AAEP,aAAS,gBAAgB;AACvB,aAAOA,WACJ,OAAO,CAAC,KAAK,EAAC,MAAM,MAAK,GAAG,UAAU,IAAI,OAAO;AAAA,QAChD;AAAA,QACA,GAAG;AAAA,QACH,GAAG,UAAU,KAAK,EAAE;AAAA,MACtB,CAAC,GAAG,CAAC,CAAC,EACL,OAAO,CAAC,EAAC,GAAG,EAAC,MAAM;AAClB,YAAI,EAAE,WAAW,KAAK,EAAE,WAAW,GAAG;AACpC,iBAAO;AAAA,QACT;AAEA,eAAO;AAAA,MACT,CAAC;AAAA,IACL;AAEA,aAAS,cAAc;AACrB,aAAOA,WACJ,IAAI,CAAC,EAAC,KAAI,MAAM,IAAI,EACpB,OAAO,OAAK,MAAM,KAAK,CAAC,EAAC,KAAI,MAAM,SAAS,CAAC,MAAM,KAAK;AAAA,IAC7D;AAAA,EACF;AAEF;",
+  "sourcesContent": ["\nimport createDebugLogger from 'debug';\nimport * as features from './features/index.js';\n\nexport {features};\n\nexport default ({strategy, threshold = 0.8999}, returnStrategy = false, localSettings = {}) => ({recordA, recordB, recordAExternal = {}, recordBExternal = {}}) => {\n  const MIN_PROPABILITY_QUALIFIER = 0.4; // Title perfect match: 0.5\n  const FORCE_FAIL = -1.0;\n\n  const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection');\n  const debugData = debug.extend('data');\n\n  debugData(`Strategy: ${JSON.stringify(strategy)}, Threshold: ${JSON.stringify(threshold)}, ReturnStrategy: ${JSON.stringify(returnStrategy)}`);\n  debugData(`Records: A: ${recordA}\\nB: ${recordB}`);\n  debug(`Externals: A: ${JSON.stringify(recordAExternal)}, B: ${JSON.stringify(recordBExternal)}`);\n  // We could add here labels for records if we didn't get external labels\n\n  const featuresA = extractFeatures({record: recordA, recordExternal: recordAExternal, localSettings});\n\n  debug(`We got an array of records: ${Array.isArray(recordB)}`);\n  const recordsB = Array.isArray(recordB) ? recordB : [recordB];\n  const recordsBExternal = Array.isArray(recordB) ? recordBExternal : [recordBExternal];\n\n  const detectionResults = recordsB.map((record, index) => actualDetection({featuresA, recordAExternal, record, recordExternal: recordsBExternal[index], index}));\n\n  // if we got array of records, we return an array of result\n  // if we got a singular record, we return a singular result\n  return Array.isArray(recordB) ? detectionResults : detectionResults[0];\n\n  function actualDetection({featuresA, record, recordExternal, index, localSettings}) {\n    const labelA = recordAExternal && recordAExternal.label ? recordAExternal.label : 'a';\n    const labelB = recordExternal && recordExternal.label ? recordExternal.label : 'b';\n\n\n    debug(`Actual detection for record ${index + 1} ${labelB}`);\n    const featuresB = extractFeatures({record, recordExternal, localSettings});\n\n    debugData(`Features (a: ${labelA}): ${JSON.stringify(featuresA)}`);\n    debugData(`Features (b: ${labelB}): ${JSON.stringify(featuresB)}`);\n\n    const featurePairs = generateFeaturePairs(featuresA, featuresB);\n    const similarityVector = generateSimilarityVector(featurePairs);\n\n    if (similarityVector.some(v => v >= MIN_PROPABILITY_QUALIFIER)) {\n      const probability = Math.round(calculateProbability(similarityVector) * 100)/100;\n      debug(`probability: ${probability} (Threshold: ${threshold})`);\n      if (similarityVector.some(v => v <= FORCE_FAIL)) {\n        debug(`Some feature resulted in utter failure (${FORCE_FAIL} or less)`);\n        return returnResult({match: false, probability: 0.0});\n      }\n\n      return returnResult({match: probability >= threshold, probability});\n    }\n\n    debugData(`No feature yielded minimum probability amount of points (${MIN_PROPABILITY_QUALIFIER})`);\n    return returnResult({match: false, probability: 0.0});\n  }\n\n  function extractFeatures({record, recordExternal}) {\n    return strategy.reduce((acc, {name, extract}) => acc.concat({name, value: extract({record, recordExternal})}), []);\n  }\n\n  function returnResult(result) {\n    if (returnStrategy) {\n      debug(`Returning detection strategy with the result`);\n      const resultWithStrategy = {match: result.match, probability: result.probability, strategy: formatStrategy(strategy), threshold};\n      debugData(`${JSON.stringify(resultWithStrategy)}`);\n      return resultWithStrategy;\n    }\n    return result;\n  }\n\n  function formatStrategy(strategy) {\n    const strategyNames = strategy.map(element => element.name);\n    return strategyNames || [];\n  }\n\n  function calculateProbability(similarityVector) {\n    const probability = similarityVector.reduce((acc, v) => acc + v, 0.0);\n    return probability > 1.0 ? 1.0 : probability;\n  }\n\n  function generateSimilarityVector(featurePairs) {\n    const compared = featurePairs.map(({name, a, b}) => {\n      const {compare} = strategy.find(({name: featureName}) => name === featureName);\n      const points = compare(a, b);\n      return {name, points};\n    });\n\n    debugData(`Points: ${JSON.stringify(compared)}`);\n    return compared.map(({points}) => points);\n  }\n\n  function generateFeaturePairs(featuresA, featuresB) {\n    const pairs = generatePairs();\n    const missingFeatures = findMissing();\n\n    debug(`Not comparing the following features because one, or both records are missing features: ${JSON.stringify(missingFeatures)}`);\n    return pairs;\n\n    function generatePairs() {\n      return featuresA\n        .reduce((acc, {name, value}, index) => acc.concat({\n          name,\n          a: value,\n          b: featuresB[index].value\n        }), [])\n        .filter(({a, b}) => {\n          if (a.length === 0 || b.length === 0) {\n            return false;\n          }\n\n          return true;\n        });\n    }\n\n    function findMissing() {\n      return featuresA\n        .map(({name}) => name)\n        .filter(v => pairs.some(({name}) => name === v) === false);\n    }\n  }\n\n};\n"],
+  "mappings": "AACA,OAAO,uBAAuB;AAC9B,YAAY,cAAc;AAE1B,SAAQ;AAER,eAAe,CAAC,EAAC,UAAU,YAAY,OAAM,GAAG,iBAAiB,OAAO,gBAAgB,CAAC,MAAM,CAAC,EAAC,SAAS,SAAS,kBAAkB,CAAC,GAAG,kBAAkB,CAAC,EAAC,MAAM;AACjK,QAAM,4BAA4B;AAClC,QAAM,aAAa;AAEnB,QAAM,QAAQ,kBAAkB,mDAAmD;AACnF,QAAM,YAAY,MAAM,OAAO,MAAM;AAErC,YAAU,aAAa,KAAK,UAAU,QAAQ,CAAC,gBAAgB,KAAK,UAAU,SAAS,CAAC,qBAAqB,KAAK,UAAU,cAAc,CAAC,EAAE;AAC7I,YAAU,eAAe,OAAO;AAAA,KAAQ,OAAO,EAAE;AACjD,QAAM,iBAAiB,KAAK,UAAU,eAAe,CAAC,QAAQ,KAAK,UAAU,eAAe,CAAC,EAAE;AAG/F,QAAM,YAAY,gBAAgB,EAAC,QAAQ,SAAS,gBAAgB,iBAAiB,cAAa,CAAC;AAEnG,QAAM,+BAA+B,MAAM,QAAQ,OAAO,CAAC,EAAE;AAC7D,QAAM,WAAW,MAAM,QAAQ,OAAO,IAAI,UAAU,CAAC,OAAO;AAC5D,QAAM,mBAAmB,MAAM,QAAQ,OAAO,IAAI,kBAAkB,CAAC,eAAe;AAEpF,QAAM,mBAAmB,SAAS,IAAI,CAAC,QAAQ,UAAU,gBAAgB,EAAC,WAAW,iBAAiB,QAAQ,gBAAgB,iBAAiB,KAAK,GAAG,MAAK,CAAC,CAAC;AAI9J,SAAO,MAAM,QAAQ,OAAO,IAAI,mBAAmB,iBAAiB,CAAC;AAErE,WAAS,gBAAgB,EAAC,WAAAA,YAAW,QAAQ,gBAAgB,OAAO,eAAAC,eAAa,GAAG;AAClF,UAAM,SAAS,mBAAmB,gBAAgB,QAAQ,gBAAgB,QAAQ;AAClF,UAAM,SAAS,kBAAkB,eAAe,QAAQ,eAAe,QAAQ;AAG/E,UAAM,+BAA+B,QAAQ,CAAC,IAAI,MAAM,EAAE;AAC1D,UAAM,YAAY,gBAAgB,EAAC,QAAQ,gBAAgB,eAAAA,eAAa,CAAC;AAEzE,cAAU,gBAAgB,MAAM,MAAM,KAAK,UAAUD,UAAS,CAAC,EAAE;AACjE,cAAU,gBAAgB,MAAM,MAAM,KAAK,UAAU,SAAS,CAAC,EAAE;AAEjE,UAAM,eAAe,qBAAqBA,YAAW,SAAS;AAC9D,UAAM,mBAAmB,yBAAyB,YAAY;AAE9D,QAAI,iBAAiB,KAAK,OAAK,KAAK,yBAAyB,GAAG;AAC9D,YAAM,cAAc,KAAK,MAAM,qBAAqB,gBAAgB,IAAI,GAAG,IAAE;AAC7E,YAAM,gBAAgB,WAAW,gBAAgB,SAAS,GAAG;AAC7D,UAAI,iBAAiB,KAAK,OAAK,KAAK,UAAU,GAAG;AAC/C,cAAM,2CAA2C,UAAU,WAAW;AACtE,eAAO,aAAa,EAAC,OAAO,OAAO,aAAa,EAAG,CAAC;AAAA,MACtD;AAEA,aAAO,aAAa,EAAC,OAAO,eAAe,WAAW,YAAW,CAAC;AAAA,IACpE;AAEA,cAAU,4DAA4D,yBAAyB,GAAG;AAClG,WAAO,aAAa,EAAC,OAAO,OAAO,aAAa,EAAG,CAAC;AAAA,EACtD;AAEA,WAAS,gBAAgB,EAAC,QAAQ,eAAc,GAAG;AACjD,WAAO,SAAS,OAAO,CAAC,KAAK,EAAC,MAAM,QAAO,MAAM,IAAI,OAAO,EAAC,MAAM,OAAO,QAAQ,EAAC,QAAQ,eAAc,CAAC,EAAC,CAAC,GAAG,CAAC,CAAC;AAAA,EACnH;AAEA,WAAS,aAAa,QAAQ;AAC5B,QAAI,gBAAgB;AAClB,YAAM,8CAA8C;AACpD,YAAM,qBAAqB,EAAC,OAAO,OAAO,OAAO,aAAa,OAAO,aAAa,UAAU,eAAe,QAAQ,GAAG,UAAS;AAC/H,gBAAU,GAAG,KAAK,UAAU,kBAAkB,CAAC,EAAE;AACjD,aAAO;AAAA,IACT;AACA,WAAO;AAAA,EACT;AAEA,WAAS,eAAeE,WAAU;AAChC,UAAM,gBAAgBA,UAAS,IAAI,aAAW,QAAQ,IAAI;AAC1D,WAAO,iBAAiB,CAAC;AAAA,EAC3B;AAEA,WAAS,qBAAqB,kBAAkB;AAC9C,UAAM,cAAc,iBAAiB,OAAO,CAAC,KAAK,MAAM,MAAM,GAAG,CAAG;AACpE,WAAO,cAAc,IAAM,IAAM;AAAA,EACnC;AAEA,WAAS,yBAAyB,cAAc;AAC9C,UAAM,WAAW,aAAa,IAAI,CAAC,EAAC,MAAM,GAAG,EAAC,MAAM;AAClD,YAAM,EAAC,QAAO,IAAI,SAAS,KAAK,CAAC,EAAC,MAAM,YAAW,MAAM,SAAS,WAAW;AAC7E,YAAM,SAAS,QAAQ,GAAG,CAAC;AAC3B,aAAO,EAAC,MAAM,OAAM;AAAA,IACtB,CAAC;AAED,cAAU,WAAW,KAAK,UAAU,QAAQ,CAAC,EAAE;AAC/C,WAAO,SAAS,IAAI,CAAC,EAAC,OAAM,MAAM,MAAM;AAAA,EAC1C;AAEA,WAAS,qBAAqBF,YAAW,WAAW;AAClD,UAAM,QAAQ,cAAc;AAC5B,UAAM,kBAAkB,YAAY;AAEpC,UAAM,2FAA2F,KAAK,UAAU,eAAe,CAAC,EAAE;AAClI,WAAO;AAEP,aAAS,gBAAgB;AACvB,aAAOA,WACJ,OAAO,CAAC,KAAK,EAAC,MAAM,MAAK,GAAG,UAAU,IAAI,OAAO;AAAA,QAChD;AAAA,QACA,GAAG;AAAA,QACH,GAAG,UAAU,KAAK,EAAE;AAAA,MACtB,CAAC,GAAG,CAAC,CAAC,EACL,OAAO,CAAC,EAAC,GAAG,EAAC,MAAM;AAClB,YAAI,EAAE,WAAW,KAAK,EAAE,WAAW,GAAG;AACpC,iBAAO;AAAA,QACT;AAEA,eAAO;AAAA,MACT,CAAC;AAAA,IACL;AAEA,aAAS,cAAc;AACrB,aAAOA,WACJ,IAAI,CAAC,EAAC,KAAI,MAAM,IAAI,EACpB,OAAO,OAAK,MAAM,KAAK,CAAC,EAAC,KAAI,MAAM,SAAS,CAAC,MAAM,KAAK;AAAA,IAC7D;AAAA,EACF;AAEF;",
   "names": ["featuresA", "localSettings", "strategy"]
 }

package/package.json

@@ -13,7 +13,7 @@
 		"url": "https://github.com/NatLibFi/melinda-record-matching-js"
 	},
 	"license": "MIT",
-	"version": "5.1.1-alpha.1",
+	"version": "5.1.2-alpha.1",
 	"type": "module",
 	"main": "./dist/index.js",
 	"bin": "./dist/cli.js",
@@ -57,7 +57,7 @@
 		"@natlibfi/fixura": "^4.0.1",
 		"cross-env": "^10.1.0",
 		"esbuild": "^0.28.0",
-		"eslint": "^10.3.0"
+		"eslint": "^10.4.0"
 	},
 	"overrides": {
 		"uuid": "^14.0.0"

package/src/cli.js

@@ -89,6 +89,7 @@ async function cli() {
         matchDetection.features.bib.hostComponent(),
         matchDetection.features.bib.isbn(),
         matchDetection.features.bib.issn(),
+        matchDetection.features.bib.publisherNumber(),
         // matchDetection.features.bib.sid(),
         matchDetection.features.bib.otherStandardIdentifier(),
         // Let's not use the same title matchDetection here
@@ -126,6 +127,7 @@ async function cli() {
         matchDetection.features.bib.hostComponent(),
         matchDetection.features.bib.isbn(),
         matchDetection.features.bib.issn(),
+        matchDetection.features.bib.publisherNumber(),
         // matchDetection.features.bib.sid(),
         matchDetection.features.bib.otherStandardIdentifier(),
         matchDetection.features.bib.title(),

package/src/index.js

@@ -259,7 +259,7 @@ export default ({detection: detectionOptions, search: searchOptions, maxMatches
           debugData(`Strategy: ${JSON.stringify(detectionResult.strategy)}, Threshold: ${JSON.stringify(detectionResult.threshold)}`);
 
           const matchResult = {
-            probability: detectionResult.probability,
+            probability: Math.round(detectionResult.probability * 100)/100,
             candidate: {
               id: candidateId,
               record: candidateRecord

package/src/match-detection/features/bib/f773.js

@@ -9,10 +9,10 @@
 import createDebugLogger from 'debug';
 
 import {isComponentRecord} from '@natlibfi/melinda-commons';
-import {uniqArray} from './issn.js';
 import {parse773g} from '../../../candidate-search/query-list/component.js';
+import {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';
 
-const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:issn');
+const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:f773');
 const debugData = debug.extend('data');
 
 const MAX_IDENTIFIER = 0.1; // This should be pretty low: it only says something about the host

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

@@ -2,6 +2,7 @@
 export {default as hostComponent} from './host-component.js';
 export {default as isbn} from './isbn.js';
 export {default as issn} from './issn.js';
+export {default as publisherNumber} from './publisherNumber.js'; // 028$b$a
 export {default as sid} from './sid.js';
 export {default as otherStandardIdentifier} from './other-standard-identifier.js';
 export {default as title} from './title.js';

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

@@ -1,8 +1,6 @@
-
 import createDebugLogger from 'debug';
 import {parse as isbnParse} from 'isbn3';
-
-import {uniqArray} from './issn.js';
+import {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';
 
 const debug = createDebugLogger(`@natlibfi/melinda-record-matching:match-detection:features:standard-identifiers:ISBN`);
 const debugData = debug.extend('data');

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

@@ -1,3 +1,4 @@
+import {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';
 import createDebugLogger from 'debug';
 
 const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:issn');
@@ -53,7 +54,3 @@ export default () => ({
 
   }
 });
-
-export function uniqArray(arr) {
-  return arr.filter((val, i) => arr.indexOf(val) === i);
-}

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

@@ -0,0 +1,76 @@
+// Compare publusher number (028$b$a or 028$a)
+import {uniqArray} from '@natlibfi/marc-record-validators-melinda/dist/utils.js';
+import createDebugLogger from 'debug';
+
+const debug = createDebugLogger('@natlibfi/melinda-record-matching:match-detection:features:publisher-number');
+const debugData = debug.extend('data');
+
+export default () => ({
+  name: 'Publisher or Distributor Number',
+  extract: ({record/*, recordExternal*/}) => {
+    //const label = recordExternal && recordExternal.label ? recordExternal.label : 'record';
+
+    return getIdentifiers();
+
+    function getIdentifiers() {
+      const values = fieldsToPublisherNumber(record.get(/^028$/u));
+
+      debug(`\t ${values.length} potential publisher identifiers: ${values.join(', ')}`);
+
+      return uniqArray(values);
+
+      function fieldsToPublisherNumber(fields, result = []) {
+        const [field, ...remainingFields] = fields;
+        if (!field) {
+          return result;
+        }
+        const a = field.subfields.find(sf => sf.code === 'a');
+        if (!a || !a.value) {
+          return fieldsToPublisherNumber(remainingFields, result);
+        }
+        const aval = normalizePublisherNumber(`${a.value}`);
+
+        const b = field.subfields.find(sf => sf.code === 'b');
+        if (!b || !b.value) {
+          return fieldsToPublisherNumber(remainingFields, [...result, aval]);
+        }
+        const baval = normalizePublisherNumber(`${b.value}${a.value}`);
+        return fieldsToPublisherNumber(remainingFields, [...result, baval, aval]);
+      }
+
+      function normalizePublisherNumber(val) {
+        return val.replace(/[- .,:;]/ug, '');
+      }
+    }
+  },
+  compare: (aa, bb) => {
+    debugData(`Comparing identifier sets ${JSON.stringify(aa)} and ${JSON.stringify(bb)}`);
+    if (aa.length === 0 || bb.length === 0) {
+      // No data for decision
+      return 0;
+    }
+    const sharedIdentifiers = aa.filter(val => bb.includes(val));
+    if (sharedIdentifiers.length > 0) {
+      const goodMatch = sharedIdentifiers.find(identifier => identifier.match(/[^0-9]/u));
+      if (goodMatch) {
+        debug(`\tShared identifier found: '${goodMatch}'`);
+        return 0.5;
+      }
+      debug(`\tShared identifier (numbers only) found: '${sharedIdentifiers[0]}'`);
+      return 0.2;
+    }
+
+    const aa2 = aa.map(a => a.replace(/[^0-9]/ug, ''));
+    const bb2 = bb.map(b => b.replace(/[^0-9]/ug, ''));
+
+    // Numbers matches (aka don't penalize "Fazer F-123" vs "F-123")
+    // Note that FOOLP-123 vs FOOCD-123 will now also return 0.0. Still nothing positive is returned, so no damage done.
+    if (aa2.find(val => val !== '' && bb2.includes(val))) {
+      return 0.0;
+    }
+
+    return -0.2;
+
+  }
+});
+

package/src/match-detection/index.js

@@ -43,7 +43,7 @@ export default ({strategy, threshold = 0.8999}, returnStrategy = false, localSet
     const similarityVector = generateSimilarityVector(featurePairs);
 
     if (similarityVector.some(v => v >= MIN_PROPABILITY_QUALIFIER)) {
-      const probability = calculateProbability(similarityVector);
+      const probability = Math.round(calculateProbability(similarityVector) * 100)/100;
       debug(`probability: ${probability} (Threshold: ${threshold})`);
       if (similarityVector.some(v => v <= FORCE_FAIL)) {
         debug(`Some feature resulted in utter failure (${FORCE_FAIL} or less)`);