hazza/DSpace
1
0
Fork 0
mirror of https://github.com/DSpace/DSpace.git synced 2025-10-15 22:13:08 +00:00
Code Issues Packages Projects Releases Wiki Activity

DS-2952: Small improvements to FullTextContentStreams and added a unit test for it

Browse Source
This commit is contained in:
Tom Desair
2016-12-30 00:16:25 +01:00
parent 3b2d8f3669
commit df1f81bf9d
6 changed files with 226 additions and 14 deletions
Download Patch File Download Diff File Expand all files Collapse all files

34
dspace-api/src/main/java/org/dspace/discovery/FullTextContentStreams.java
View File

@@ -10,6 +10,7 @@ package org.dspace.discovery;
import com.google.common.base.Function;
import com.google.common.collect.Iterables;
import org.apache.commons.collections.CollectionUtils;
import org.apache.commons.io.Charsets;
import org.apache.commons.lang3.StringUtils;
import org.apache.log4j.Logger;
import org.apache.solr.common.util.ContentStreamBase;
@@ -43,19 +44,25 @@ public class FullTextContentStreams extends ContentStreamBase
public static final String FULLTEXT_BUNDLE = "TEXT";
protected final Context context;
protected final List<FullTextBitstream> fullTextStreams;
protected List<FullTextBitstream> fullTextStreams;
protected BitstreamService bitstreamService;
public FullTextContentStreams(Context context, Item parentItem) throws SQLException {
this.context = context;
init(parentItem);
}
protected void init(Item parentItem) {
fullTextStreams = new LinkedList<>();
sourceInfo = parentItem.getHandle();
bitstreamService = ContentServiceFactory.getInstance().getBitstreamService();
//extracted full text is always extracted as plain text
contentType = "text/plain";
if(parentItem != null) {
sourceInfo = parentItem.getHandle();
buildFullTextList(parentItem);
//extracted full text is always extracted as plain text
contentType = "text/plain";
buildFullTextList(parentItem);
}
}
private void buildFullTextList(Item parentItem) {
@@ -123,7 +130,7 @@ public class FullTextContentStreams extends ContentStreamBase
return new SequenceInputStream(new FullTextEnumeration(fullTextStreams.iterator()));
} catch (Exception e) {
log.error("Unable to add full text bitstreams to SOLR for item " + sourceInfo + ": " + e.getMessage(), e);
return new ByteArrayInputStream(e.getMessage().getBytes(StandardCharsets.UTF_8));
return new ByteArrayInputStream((e.getClass() + ": " + e.getMessage()).getBytes(StandardCharsets.UTF_8));
}
}
@@ -131,6 +138,13 @@ public class FullTextContentStreams extends ContentStreamBase
return CollectionUtils.isEmpty(fullTextStreams);
}
private BitstreamService getBitstreamService() {
if(bitstreamService == null) {
bitstreamService = ContentServiceFactory.getInstance().getBitstreamService();
}
return bitstreamService;
}
private class FullTextBitstream {
private String itemHandle;
private Bitstream bitstream;
@@ -154,7 +168,7 @@ public class FullTextContentStreams extends ContentStreamBase
}
public InputStream getInputStream() throws SQLException, IOException, AuthorizeException {
return bitstreamService.retrieve(context, bitstream);
return getBitstreamService().retrieve(context, bitstream);
}
public String getItemHandle() {
@@ -186,10 +200,10 @@ public class FullTextContentStreams extends ContentStreamBase
bitstream.getFileName() + " for item " + bitstream.getItemHandle())
+ " to SOLR:" + e.getMessage(), e);
inputStream = new ByteArrayInputStream(e.getMessage().getBytes(StandardCharsets.UTF_8));
inputStream = new ByteArrayInputStream((e.getClass() + ": " + e.getMessage()).getBytes(StandardCharsets.UTF_8));
}
return inputStream;
return new SequenceInputStream(new ByteArrayInputStream("\n".getBytes(Charsets.UTF_8)), inputStream);
}
}