forked from D-Net/dnet-hadoop
new implementation for country propagatio
This commit is contained in:
parent
beebbcf66b
commit
2afe971816
|
@ -53,14 +53,18 @@ public class SparkCountryPropagationJob {
|
|||
|
||||
List<String> whitelist = Arrays.asList(parser.get("whitelist").split(";"));
|
||||
List<String> allowedtypes = Arrays.asList(parser.get("allowedtypes").split(";"));
|
||||
boolean writeUpdates = TRUE.equals(parser.get("writeUpdate"));
|
||||
boolean saveGraph = TRUE.equals(parser.get("saveGraph"));
|
||||
|
||||
// datasource(spark, whitelist, outputPath, inputPath, "true".equals(parser.get("writeUpdate")),
|
||||
// "true".equals(parser.get("saveGraph")), allowedtypes);
|
||||
//
|
||||
// }
|
||||
//
|
||||
//
|
||||
// private static void datasource(SparkSession spark, List<String> whitelist, String outputPath, String inputPath,
|
||||
// boolean writeUpdates, boolean saveGraph, List<String> allowedtypes){
|
||||
|
||||
datasource(spark, whitelist, outputPath, inputPath);
|
||||
|
||||
}
|
||||
|
||||
|
||||
private static void datasource(SparkSession spark, List<String> whitelist, String outputPath, String inputPath){
|
||||
String whitelisted = "";
|
||||
for (String i : whitelist){
|
||||
whitelisted += " OR id = '" + i + "'";
|
||||
|
@ -96,26 +100,14 @@ public class SparkCountryPropagationJob {
|
|||
datasource.createOrReplaceTempView("datasource");
|
||||
relation.createOrReplaceTempView("relation");
|
||||
organization.createOrReplaceTempView("organization");
|
||||
// String query = "SELECT source ds, target org, country.classid country " +
|
||||
// "FROM ( SELECT id " +
|
||||
// "FROM openaire.datasource " +
|
||||
// "WHERE datasourcetype.classid = 'pubsrepository::institutional' " +
|
||||
// "AND (datainfo.deletedbyinference = false " + whitelisted + ") ) d " +
|
||||
// "JOIN ( SELECT source, target " +
|
||||
// "FROM openaire.relation " +
|
||||
// "WHERE relclass = 'provides' " +
|
||||
// "AND datainfo.deletedbyinference = false ) rel " +
|
||||
// "ON d.id = rel.source " +
|
||||
// "JOIN (SELECT id, country " +
|
||||
// "FROM openaire.organization " +
|
||||
// "WHERE datainfo.deletedbyinference = false ) o " +
|
||||
// "ON o.id = rel.target";
|
||||
|
||||
String query = "SELECT source ds, target org, country.classid country " +
|
||||
"FROM ( SELECT id " +
|
||||
"FROM datasource " +
|
||||
"WHERE datasourcetype.classid = 'pubsrepository::institutional' " +
|
||||
"AND (datainfo.deletedbyinference = false " + whitelisted + ") ) d " +
|
||||
"WHERE (datainfo.deletedbyinference = false " + whitelisted + ") " +
|
||||
getConstraintList("datasourcetype.classid = '", allowedtypes) +
|
||||
// "datasourcetype.classid = 'pubsrepository::institutional' " +
|
||||
// "AND (datainfo.deletedbyinference = false " + whitelisted + ") ) d " +
|
||||
"JOIN ( SELECT source, target " +
|
||||
"FROM relation " +
|
||||
"WHERE relclass = 'provides' " +
|
||||
|
@ -141,8 +133,11 @@ public class SparkCountryPropagationJob {
|
|||
publication.createOrReplaceTempView("publication");
|
||||
final JavaRDD<Row> toupdateresultpublication = propagateOnResult(spark, "publication");
|
||||
|
||||
if(writeUpdates){
|
||||
writeUpdates(toupdateresultsoftware, toupdateresultdataset, toupdateresultother, toupdateresultpublication, outputPath);
|
||||
}
|
||||
|
||||
if(saveGraph){
|
||||
createUpdateForSoftwareDataset(toupdateresultsoftware, inputPath, spark)
|
||||
.map(s -> new ObjectMapper().writeValueAsString(s))
|
||||
.saveAsTextFile(outputPath + "/software");
|
||||
|
@ -158,6 +153,7 @@ public class SparkCountryPropagationJob {
|
|||
createUpdateForPublicationDataset(toupdateresultpublication, inputPath, spark)
|
||||
.map(p -> new ObjectMapper().writeValueAsString(p))
|
||||
.saveAsTextFile(outputPath + "/publication");
|
||||
}
|
||||
|
||||
}
|
||||
|
||||
|
@ -334,4 +330,3 @@ public class SparkCountryPropagationJob {
|
|||
|
||||
|
||||
}
|
||||
|
||||
|
|
Loading…
Reference in New Issue