sabeel
/
dnet-hadoop
forked from D-Net/dnet-hadoop
1
0
Fork 0
Code Pull Requests Activity

Merge pull request 'missing landingPage urls in instances' (#22) from instances-with-landing-page into master 

Looks good, thanks!
This commit is contained in:
Claudio Atzori 2020-06-16 15:32:44 +02:00
parent 89859111ee c2e1b66e83
commit 5441f01586
1 changed files with 11 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

13
dhp-workflows/dhp-graph-mapper/src/main/java/eu/dnetlib/dhp/oa/graph/raw/OdfToOafMapper.java
View File

@ -126,9 +126,16 @@ public class OdfToOafMapper extends AbstractMdRecordToOafMapper {
for (final Object o : doc.selectNodes("//datacite:alternateIdentifier[@alternateIdentifierType='URL']")) { for (final Object o : doc.selectNodes("//datacite:alternateIdentifier[@alternateIdentifierType='URL']")) {
url.add(((Node) o).getText().trim()); url.add(((Node) o).getText().trim());
} }
for (final Object o : doc
.selectNodes("//datacite:alternateIdentifier[@alternateIdentifierType='landingPage']")) {
url.add(((Node) o).getText().trim());
}
for (final Object o : doc.selectNodes("//datacite:identifier[@identifierType='URL']")) { for (final Object o : doc.selectNodes("//datacite:identifier[@identifierType='URL']")) {
url.add(((Node) o).getText().trim()); url.add(((Node) o).getText().trim());
} }
for (final Object o : doc.selectNodes("//datacite:identifier[@identifierType='landingPage']")) {
url.add(((Node) o).getText().trim());
}
for (final Object o : doc.selectNodes("//datacite:alternateIdentifier[@alternateIdentifierType='DOI']")) { for (final Object o : doc.selectNodes("//datacite:alternateIdentifier[@alternateIdentifierType='DOI']")) {
url.add(HTTP_DX_DOI_PREIFX + ((Node) o).getText().trim()); url.add(HTTP_DX_DOI_PREIFX + ((Node) o).getText().trim());
} }
@ -367,11 +374,13 @@ public class OdfToOafMapper extends AbstractMdRecordToOafMapper {
res res
.addAll( .addAll(
prepareListStructPropsWithValidQualifier( prepareListStructPropsWithValidQualifier(
doc, "//datacite:identifier[@identifierType != 'URL']", "@identifierType", DNET_PID_TYPES, info)); doc, "//datacite:identifier[@identifierType != 'URL' and @identifierType != 'landingPage']",
"@identifierType", DNET_PID_TYPES, info));
res res
.addAll( .addAll(
prepareListStructPropsWithValidQualifier( prepareListStructPropsWithValidQualifier(
doc, "//datacite:alternateIdentifier[@alternateIdentifierType != 'URL']", doc,
"//datacite:alternateIdentifier[@alternateIdentifierType != 'URL' and @alternateIdentifierType != 'landingPage']",
"@alternateIdentifierType", DNET_PID_TYPES, info)); "@alternateIdentifierType", DNET_PID_TYPES, info));
return res; return res;
} }