Skip to content

Commit

Permalink
Merge pull request #5 from codeforkjeff/fix-bnf-ids
Browse files Browse the repository at this point in the history 
tweak matching logic in associateSourceIds() to account for missing I…
  • Loading branch information
@codeforkjeff
codeforkjeff authored Jun 17, 2016
2 parents 4921ed4 + f620b2a commit e29918d
Show file tree
Hide file tree
Showing 3 changed files with 259 additions and 0 deletions.
21 changes: 21 additions & 0 deletions src/main/java/com/codefork/refine/viaf/VIAFParser.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -278,6 +278,27 @@ private void associateSourceIds() {
for(NameEntry nameEntry : result.getNameEntries()) {
for(NameSource nameSource : nameEntry.getNameSources()) {
String sourceId = sourceIdMappings.get(nameSource.getViafSourceId());
if(sourceId == null) {
// sometimes ../mainHeadings/data/sources will list a source
// without an ID, but the XML will contain an ID under
// ../VIAFCluster/sources. This is the case for record 76304784,
// as of 6/17/2016.
//
// This screws up the association here, so to deal,
// we look for just the ORG prefix in sourceIdMappings keys.
for(String k : sourceIdMappings.keySet()) {
if(k.contains("|")) {
String[] pieces = k.split("\\|");
if(pieces.length == 2) {
String orgCode = pieces[0];
String id = pieces[1];
if (orgCode.equals(nameSource.getCode())) {
sourceId = id;
}
}
}
}
}
if(sourceId != null) {
nameSource.setSourceId(sourceId);
}
Expand Down
31 changes: 31 additions & 0 deletions src/test/java/com/codefork/refine/viaf/VIAFParserTest.java
View file
Original file line number Diff line number Diff line change
Expand Up @@ -132,6 +132,37 @@ public void testParseNames() throws Exception {

}

/**
* Test case for XML missing an ID in ../mainHeadings/data/sources
* but having an ID under ../VIAFCluster/sources.
* @throws Exception
*/
@Test
public void testParseXMLWithoutSourceID() throws Exception {
SAXParserFactory spf = SAXParserFactory.newInstance();
SAXParser parser = spf.newSAXParser();
VIAFParser viafParser = new VIAFParser();

InputStream is = getClass().getResourceAsStream("/alexandre.xml");
parser.parse(is, viafParser);

List<VIAFResult> results = viafParser.getResults();

assertEquals(1, results.size());

VIAFResult firstResult = results.get(0);

assertEquals(1, firstResult.getNameEntries().size());

assertEquals("Alexandre, Jean-François 1804-1874",
firstResult.getNameEntries().get(0).getName());
assertEquals("BNF",
joinSources(firstResult.getNameEntries().get(0).getNameSources(), ","));

// test that our source ID mappings work
assertEquals("10341017", firstResult.getSourceId("BNF"));
}

@Test
public void testParseTime() throws Exception {

Expand Down
207 changes: 207 additions & 0 deletions src/test/resources/alexandre.xml
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,207 @@
<?xml version="1.0" encoding="UTF-8"?>
<searchRetrieveResponse xmlns:xsd="http://www.w3.org/2001/XMLSchema" xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance" xmlns="http://www.loc.gov/zing/srw/" xsi:schemaLocation="http://www.loc.gov/zing/srw/ http://www.loc.gov/standards/sru/sru1-1archive/xml-files/srw-types.xsd">
<version xsi:type="xsd:string">1.1</version>
<numberOfRecords xsi:type="xsd:nonNegativeInteger">1</numberOfRecords>
<resultSetIdleTime xsi:type="xsd:positiveInteger">1</resultSetIdleTime>
<records xmlns:ns1="http://www.loc.gov/zing/srw/" xsi:type="ns1:recordsType">
<record xsi:type="ns1:recordType">
<recordSchema xsi:type="xsd:string">http://viaf.org/VIAFCluster</recordSchema>
<recordPacking xsi:type="xsd:string">xml</recordPacking>
<recordData xsi:type="ns1:stringOrXmlFragment">
<ns2:VIAFCluster xmlns="http://viaf.org/viaf/terms#" xmlns:foaf="http://xmlns.com/foaf/0.1/" xmlns:owl="http://www.w3.org/2002/07/owl#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:void="http://rdfs.org/ns/void#" xmlns:ns2="http://viaf.org/viaf/terms#">
<ns2:viafID>76304784</ns2:viafID>
<ns2:Document about="http://viaf.org/viaf/76304784/">
<ns2:inDataset resource="http://viaf.org/viaf/data"/>
<ns2:primaryTopic resource="http://viaf.org/viaf/76304784"/>
</ns2:Document>
<ns2:nameType>Personal</ns2:nameType>
<ns2:sources>
<ns2:source nsid="http://catalogue.bnf.fr/ark:/12148/cb10341017g">BNF|10341017</ns2:source>
</ns2:sources>
<ns2:length>19</ns2:length>
<ns2:mainHeadings>
<ns2:data>
<ns2:text>Alexandre, Jean-François 1804-1874</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
<ns2:mainHeadingEl>
<ns2:datafield dtype="UNIMARC" ind1=" " ind2="|" tag="200">
<ns2:subfield code="7">ba0yba0y</ns2:subfield>
<ns2:subfield code="8">fre</ns2:subfield>
<ns2:subfield code="9">0</ns2:subfield>
<ns2:subfield code="a">Alexandre</ns2:subfield>
<ns2:subfield code="b">Jean-François</ns2:subfield>
<ns2:subfield code="f">1804-1874</ns2:subfield>
</ns2:datafield>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:id>BNF|10341017</ns2:id>
</ns2:mainHeadingEl>
</ns2:mainHeadings>
<ns2:fixed>
<ns2:gender>b</ns2:gender>
</ns2:fixed>
<ns2:coauthors>
<ns2:data count="1" tag="950">
<ns2:text>Lescoët, Jonathas-Marie-Joseph Barbier, Mis de</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
<ns2:data count="1" tag="950">
<ns2:text>Lescoet, Cte de</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:coauthors>
<ns2:publishers>
<ns2:data count="2">
<ns2:text>L. Prud'homme</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
<ns2:data count="1">
<ns2:text>moulet gand E. Blot</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
<ns2:data count="1">
<ns2:text>impr. de J.-B. Defournier aîné</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:publishers>
<ns2:birthDate>1804-04-26</ns2:birthDate>
<ns2:deathDate>1874-03-19</ns2:deathDate>
<ns2:dateType>lived</ns2:dateType>
<ns2:RecFormats>
<ns2:data count="5">
<ns2:text>am</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:RecFormats>
<ns2:RelatorCodes>
<ns2:data count="3">
<ns2:text>070</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
<ns2:data count="2">
<ns2:text>730</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:RelatorCodes>
<ns2:countries>
<ns2:data count="5" scaled="3">
<ns2:text>FR</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:countries>
<ns2:languageOfEntity>
<ns2:data>
<ns2:text>fre</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:languageOfEntity>
<ns2:nationalityOfEntity>
<ns2:data>
<ns2:text>FR</ns2:text>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
</ns2:data>
</ns2:nationalityOfEntity>
<ns2:titles>
<ns2:work>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:title>Ar Virionez da Vreiz-Izel, scrivet gant an aotrou'n abad Alexandre,...</ns2:title>
</ns2:work>
<ns2:work>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:title>Calon an aotrou Paul de Parcevaux,... Leoric great e gallec gant ar c'hont de Lescoet ha lekeat e brezounec gant an abad Alexandre, 1860...</ns2:title>
</ns2:work>
<ns2:work>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:title>Calon an aotrou Paul de Parcevaux, oficer a regimant ar zouaved en arme hon tad Santel ar Pab Pius IX...</ns2:title>
</ns2:work>
<ns2:work>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:title>Cantique en l'honneur de l'Immaculée Conception de la très Sainte Vierge Marie, par M. Alexandre,...</ns2:title>
</ns2:work>
<ns2:work>
<ns2:sources>
<ns2:s>BNF</ns2:s>
</ns2:sources>
<ns2:title>Leoriou ar baradoz : approche bibliographique du livre religieux en langue bretonne</ns2:title>
</ns2:work>
</ns2:titles>
<ns2:history>
<ns2:ht recid="BNF|10341017" time="2009-03-03T12:03:17+00:00" type="add"/>
</ns2:history>
</ns2:VIAFCluster>
</recordData>
<recordPosition xsi:type="xsd:positiveInteger">1</recordPosition>
</record>
</records>
<echoedSearchRetrieveRequest xmlns:ns3="http://www.loc.gov/zing/srw/" xsi:type="ns3:echoedSearchRetrieveRequestType">
<version xsi:type="xsd:string">1.1</version>
<query xsi:type="xsd:string">local.mainHeadingEl all "Jean-François Alexandre 1804 1874" and local.sources = "bnf"</query>
<xQuery>
<ns4:triple xmlns:ns4="http://www.loc.gov/zing/cql/xcql/" xsi:type="ns4:tripleType">
<ns4:boolean xsi:type="ns4:booleanType">
<ns4:value xsi:type="xsd:string">and</ns4:value>
</ns4:boolean>
<ns4:leftOperand xsi:type="ns4:operandType">
<ns4:searchClause xsi:type="ns4:searchClauseType">
<ns4:index xsi:type="xsd:string">local.mainHeadingEl</ns4:index>
<ns4:relation xsi:type="ns4:relationType">
<ns4:value xsi:type="xsd:string">all</ns4:value>
</ns4:relation>
<ns4:term xsi:type="xsd:string">Jean-François Alexandre 1804 1874</ns4:term>
</ns4:searchClause>
</ns4:leftOperand>
<ns4:rightOperand xsi:type="ns4:operandType">
<ns4:searchClause xsi:type="ns4:searchClauseType">
<ns4:index xsi:type="xsd:string">local.sources</ns4:index>
<ns4:relation xsi:type="ns4:relationType">
<ns4:value xsi:type="xsd:string">=</ns4:value>
</ns4:relation>
<ns4:term xsi:type="xsd:string">bnf</ns4:term>
</ns4:searchClause>
</ns4:rightOperand>
</ns4:triple>
</xQuery>
<maximumRecords xsi:type="xsd:nonNegativeInteger">3</maximumRecords>
<recordSchema xsi:type="xsd:string">default</recordSchema>
<sortKeys xsi:type="xsd:string">holdingscount</sortKeys>
</echoedSearchRetrieveRequest>
<extraResponseData xmlns:ns5="http://www.loc.gov/zing/srw/" xsi:type="ns5:extraDataType">
<ns6:extraData xmlns="http://oclc.org/srw/extraData" xmlns:ns6="http://oclc.org/srw/extraData">
<ns6:databaseTitle>VIAF: The Virtual International Authority File</ns6:databaseTitle>
</ns6:extraData>
</extraResponseData>
</searchRetrieveResponse>

0 comments on commit e29918d

Please sign in to comment.