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

[Cleaning] added couple of tests for DOIBOOST publications

This commit is contained in:
Miriam Baglioni 2021-11-21 16:35:22 +01:00
parent 6110a2b984
commit fdb75b180e
3 changed files with 29 additions and 0 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

27
dhp-workflows/dhp-graph-mapper/src/test/java/eu/dnetlib/dhp/oa/graph/clean/GraphCleaningFunctionsTest.java
View File

@ -12,6 +12,8 @@ import java.util.stream.Collectors;
import java.util.stream.Stream; import java.util.stream.Stream;
import org.apache.commons.io.IOUtils; import org.apache.commons.io.IOUtils;
import org.apache.hadoop.hdfs.server.datanode.fsdataset.impl.MappableBlock;
import org.junit.jupiter.api.Assertions;
import org.junit.jupiter.api.BeforeEach; import org.junit.jupiter.api.BeforeEach;
import org.junit.jupiter.api.Test; import org.junit.jupiter.api.Test;
import org.junit.jupiter.api.extension.ExtendWith; import org.junit.jupiter.api.extension.ExtendWith;
@ -222,4 +224,29 @@ public class GraphCleaningFunctionsTest {
.readLines( .readLines(
GraphCleaningFunctionsTest.class.getResourceAsStream("/eu/dnetlib/dhp/oa/graph/clean/synonyms.txt")); GraphCleaningFunctionsTest.class.getResourceAsStream("/eu/dnetlib/dhp/oa/graph/clean/synonyms.txt"));
} }
@Test
public void testCleanDoiBoost() throws IOException {
String json = IOUtils.toString(getClass().getResourceAsStream("/eu/dnetlib/dhp/oa/graph/clean/doiboostpub.json"));
Publication p_in = MAPPER.readValue(json, Publication.class);
Publication p_out = OafCleaner.apply(GraphCleaningFunctions.fixVocabularyNames(p_in), mapping);
Publication cleaned = GraphCleaningFunctions.cleanup(p_out);
Assertions.assertEquals(true,GraphCleaningFunctions.filter(cleaned) );
}
@Test
public void testCleanDoiBoost2() throws IOException {
String json = IOUtils.toString(getClass().getResourceAsStream("/eu/dnetlib/dhp/oa/graph/clean/doiboostpub2.json"));
Publication p_in = MAPPER.readValue(json, Publication.class);
Publication p_out = OafCleaner.apply(GraphCleaningFunctions.fixVocabularyNames(p_in), mapping);
Publication cleaned = GraphCleaningFunctions.cleanup(p_out);
Assertions.assertEquals(true,GraphCleaningFunctions.filter(cleaned) );
}
} }

1
dhp-workflows/dhp-graph-mapper/src/test/resources/eu/dnetlib/dhp/oa/graph/clean/doiboostpub.json Normal file
View File

@ -0,0 +1 @@
{"context": [], "dataInfo": {"invisible": false, "trust": "0.9", "provenanceaction": {"classid": "sysimport:actionset", "classname": "sysimport:actionset", "schemeid": "dnet:provenanceActions", "schemename": "dnet:provenanceActions"}, "inferred": false, "deletedbyinference": false}, "resourcetype": {"classid": "0001", "classname": "Article", "schemeid": "dnet:publication_resource", "schemename": "dnet:publication_resource"}, "pid": [{"qualifier": {"classid": "doi", "classname": "doi", "schemeid": "dnet:pid_types", "schemename": "dnet:pid_types"}, "value": "10.1097/00132586-197308000-00003"}], "contributor": [], "bestaccessright": {"classid": "UNKNOWN", "classname": "not available", "schemeid": "dnet:access_modes", "schemename": "dnet:access_modes"}, "relevantdate": [{"qualifier": {"classid": "created", "classname": "created", "schemeid": "dnet:dataCite_date", "schemename": "dnet:dataCite_date"}, "value": "2006-11-06T11:36:37Z"}], "collectedfrom": [{"key": "10|openaire____::081b82f96300b6a6e3d282bad31cb6e2", "value": "Crossref"}], "id": "50|doi_________::b0baa0eb88a5788f0b8815560d2a32f2", "subject": [], "lastupdatetimestamp": 1620353302565, "author": [{"fullname": "N. S. AGRUSS", "surname": "AGRUSS", "name": "N. S.", "rank": 1}, {"fullname": "E. Y. ROSIN", "surname": "ROSIN", "name": "E. Y.", "rank": 2}, {"fullname": "R. J. ADOLPH", "surname": "ADOLPH", "name": "R. J.", "rank": 3}, {"fullname": "N. O. FOWLER", "surname": "FOWLER", "name": "N. O.", "rank": 4}], "instance": [{"hostedby": {"key": "10|issn___print::b8cee613d4f898f8c03956d57ea69be2", "value": "Survey of Anesthesiology"}, "url": ["https://doi.org/10.1097/00132586-197308000-00003"], "pid": [{"qualifier": {"classid": "doi", "classname": "doi", "schemeid": "dnet:pid_types", "schemename": "dnet:pid_types"}, "value": "10.1097/00132586-197308000-00003"}], "dateofacceptance": {"value": "2006-11-06T11:36:37Z"}, "collectedfrom": {"key": "10|openaire____::081b82f96300b6a6e3d282bad31cb6e2", "value": "Crossref"}, "accessright": {"classid": "UNKNOWN", "classname": "not available", "schemeid": "dnet:access_modes", "schemename": "dnet:access_modes"}, "instancetype": {"classid": "0001", "classname": "Article", "schemeid": "dnet:publication_resource", "schemename": "dnet:publication_resource"}}], "dateofcollection": "2021-05-07T02:08:22Z", "fulltext": [], "description": [], "format": [], "journal": {"issnPrinted": "0039-6206", "vol": "17", "sp": "304", "name": "Survey of Anesthesiology"}, "measures": [], "coverage": [], "externalReference": [], "publisher": {"value": "Ovid Technologies (Wolters Kluwer Health)"}, "resulttype": {"classid": "publication", "classname": "publication", "schemeid": "dnet:result_typologies", "schemename": "dnet:result_typologies"}, "country": [], "extraInfo": [], "originalId": ["10.1097/00132586-197308000-00003", "50|doiboost____::b0baa0eb88a5788f0b8815560d2a32f2"], "source": [{"value": "Crossref"}], "dateofacceptance": {"value": "2006-11-06T11:36:37Z"}, "title": [{"qualifier": {"classid": "main title", "classname": "main title", "schemeid": "dnet:dataCite_title", "schemename": "dnet:dataCite_title"}, "value": "SIGNIFICANCE OF CHRONIC SINUS BRADYCARDIA IN ELDERLY PEOPLE"}]}

1
dhp-workflows/dhp-graph-mapper/src/test/resources/eu/dnetlib/dhp/oa/graph/clean/doiboostpub2.json Normal file
View File

@ -0,0 +1 @@
{"context": [], "dataInfo": {"invisible": false, "trust": "0.9", "provenanceaction": {"classid": "sysimport:actionset", "classname": "sysimport:actionset", "schemeid": "dnet:provenanceActions", "schemename": "dnet:provenanceActions"}, "inferred": false, "deletedbyinference": false}, "resourcetype": {"classid": "0001", "classname": "Article", "schemeid": "dnet:publication_resource", "schemename": "dnet:publication_resource"}, "pid": [{"qualifier": {"classid": "doi", "classname": "doi", "schemeid": "dnet:pid_types", "schemename": "dnet:pid_types"}, "value": "10.2143/tvg.62.1.5002364"}], "contributor": [], "bestaccessright": {"classid": "UNKNOWN", "classname": "not available", "schemeid": "dnet:access_modes", "schemename": "dnet:access_modes"}, "relevantdate": [{"qualifier": {"classid": "created", "classname": "created", "schemeid": "dnet:dataCite_date", "schemename": "dnet:dataCite_date"}, "value": "2007-08-20T08:35:04Z"}, {"qualifier": {"classid": "published-online", "classname": "published-online", "schemeid": "dnet:dataCite_date", "schemename": "dnet:dataCite_date"}, "value": "2006-01-01"}], "collectedfrom": [{"key": "10|openaire____::081b82f96300b6a6e3d282bad31cb6e2", "value": "Crossref"}], "id": "50|doi_________::4972b0ca81b96b225aed8038bb965656", "subject": [{"qualifier": {"classid": "keywords", "classname": "keywords", "schemeid": "dnet:subject_classification_typologies", "schemename": "dnet:subject_classification_typologies"}, "value": "General Medicine"}], "lastupdatetimestamp": 1620381522840, "author": [{"fullname": "null VERHAMME P", "surname": "VERHAMME P", "rank": 1}], "instance": [{"hostedby": {"key": "10|issn__online::7ec728ad1ac65c60cd563a5137111125", "value": "Tijdschrift voor Geneeskunde"}, "url": ["https://doi.org/10.2143/tvg.62.1.5002364"], "pid": [{"qualifier": {"classid": "doi", "classname": "doi", "schemeid": "dnet:pid_types", "schemename": "dnet:pid_types"}, "value": "10.2143/tvg.62.1.5002364"}], "dateofacceptance": {"value": "2006-01-01"}, "collectedfrom": {"key": "10|openaire____::081b82f96300b6a6e3d282bad31cb6e2", "value": "Crossref"}, "accessright": {"classid": "UNKNOWN", "classname": "not available", "schemeid": "dnet:access_modes", "schemename": "dnet:access_modes"}, "instancetype": {"classid": "0001", "classname": "Article", "schemeid": "dnet:publication_resource", "schemename": "dnet:publication_resource"}}], "dateofcollection": "2021-05-07T09:58:42Z", "fulltext": [], "description": [], "format": [], "journal": {"vol": "62", "sp": "55", "issnOnline": "0371-683X", "ep": "61", "name": "Tijdschrift voor Geneeskunde"}, "measures": [], "coverage": [], "externalReference": [], "publisher": {"value": "Peeters Publishers"}, "resulttype": {"classid": "publication", "classname": "publication", "schemeid": "dnet:result_typologies", "schemename": "dnet:result_typologies"}, "country": [], "extraInfo": [], "originalId": ["10.2143/tvg.62.1.5002364", "50|doiboost____::4972b0ca81b96b225aed8038bb965656"], "source": [{"value": "Crossref"}], "dateofacceptance": {"value": "2006-01-01"}, "title": [{"qualifier": {"classid": "main title", "classname": "main title", "schemeid": "dnet:dataCite_title", "schemename": "dnet:dataCite_title"}, "value": "Antitrombotica: nieuwe moleculen"}]}