Merge pull request #1939 from atmire/DS-3817

DS-3817: Support the DSpace Cover page functionality in the bitstream endpoint

dspace-api/src/main/java/org/dspace/curate/CitationPage.java

@@ -7,6 +7,7 @@
  */
 package org.dspace.curate;
 
+import org.apache.commons.lang3.tuple.Pair;
 import org.apache.log4j.Logger;
 import org.dspace.authorize.AuthorizeException;
 import org.dspace.content.*;
@@ -150,9 +151,9 @@ protected void performItem(Item item) throws SQLException {
                             + bitstream.getName() + " is citable.");
                     try {
                         //Create the cited document
-                        File citedDocument = citationDocument.makeCitedDocument(Curator.curationContext(), bitstream);
+                        Pair<InputStream, Long> citedDocument = citationDocument.makeCitedDocument(Curator.curationContext(), bitstream);
                         //Add the cited document to the approiate bundle
-                        this.addCitedPageToItem(citedDocument, bundle, pBundle,
+                        this.addCitedPageToItem(citedDocument.getLeft(), bundle, pBundle,
                                 dBundle, displayMap, item, bitstream);
                     } catch (Exception e) {
                         //Could be many things, but nothing that should be
@@ -187,7 +188,7 @@ protected void performItem(Item item) throws SQLException {
      * A helper function for {@link CitationPage#performItem(Item)}. This function takes in the
      * cited document as a File and adds it to DSpace properly.
      *
-     * @param citedTemp The temporary File that is the cited document.
+     * @param citedDoc The inputstream that is the cited document.
      * @param bundle The bundle the cited file is from.
      * @param pBundle The preservation bundle. The original document should be
      * put in here if it is not already.
@@ -200,7 +201,7 @@ protected void performItem(Item item) throws SQLException {
      * @throws AuthorizeException if authorization error
      * @throws IOException if IO error
      */
-    protected void addCitedPageToItem(File citedTemp, Bundle bundle, Bundle pBundle,
+    protected void addCitedPageToItem(InputStream citedDoc, Bundle bundle, Bundle pBundle,
                                     Bundle dBundle, Map<String,Bitstream> displayMap, Item item,
                                     Bitstream bitstream) throws SQLException, AuthorizeException, IOException {
         //If we are modifying a file that is not in the
@@ -218,12 +219,11 @@ protected void addCitedPageToItem(File citedTemp, Bundle bundle, Bundle pBundle,
         //Create an input stream form the temporary file
         //that is the cited document and create a
         //bitstream from it.
-        InputStream inp = new FileInputStream(citedTemp);
         if (displayMap.containsKey(bitstream.getName())) {
             bundleService.removeBitstream(context, dBundle, displayMap.get(bitstream.getName()));
         }
-        Bitstream citedBitstream = bitstreamService.create(context, dBundle, inp);
-        inp.close(); //Close up the temporary InputStream
+        Bitstream citedBitstream = bitstreamService.create(context, dBundle, citedDoc);
+        citedDoc.close(); //Close up the temporary InputStream
 
         //Setup a good name for our bitstream and make
         //it the same format as the source document.

dspace-api/src/main/java/org/dspace/disseminate/CitationDocumentServiceImpl.java

@@ -8,6 +8,7 @@
 package org.dspace.disseminate;
 
 import org.apache.commons.lang.StringUtils;
+import org.apache.commons.lang3.tuple.Pair;
 import org.apache.log4j.Logger;
 import org.apache.pdfbox.pdmodel.PDDocument;
 import org.apache.pdfbox.pdmodel.PDPage;
@@ -118,6 +119,8 @@ public void afterPropertiesSet() throws Exception {
         //Populate VALID_TYPES
         VALID_TYPES.addAll(PDF_MIMES);
 
+        // Global enabled?
+        citationEnabledGlobally = configurationService.getBooleanProperty("citation-page.enable_globally", false);
 
         //Load enabled collections
         String[] citationEnabledCollections = configurationService.getArrayProperty("citation-page.enabled_collections");
@@ -206,10 +209,6 @@ protected CitationDocumentServiceImpl() {}
     protected Boolean citationEnabledGlobally = null;
 
     protected boolean isCitationEnabledGlobally() {
-        if(citationEnabledGlobally == null) {
-            citationEnabledGlobally = configurationService.getBooleanProperty("citation-page.enable_globally", false);
-        }
-
         return citationEnabledGlobally;
     }
 
@@ -273,7 +272,7 @@ public boolean canGenerateCitationVersion(Context context, Bitstream bitstream)
     }
 
     @Override
-    public File makeCitedDocument(Context context, Bitstream bitstream)
+    public Pair<InputStream, Long> makeCitedDocument(Context context, Bitstream bitstream)
             throws IOException, SQLException, AuthorizeException {
         PDDocument document = new PDDocument();
         PDDocument sourceDocument = new PDDocument();
@@ -284,8 +283,14 @@ public File makeCitedDocument(Context context, Bitstream bitstream)
             generateCoverPage(context, document, coverPage, item);
             addCoverPageToDocument(document, sourceDocument, coverPage);
 
-            document.save(tempDir.getAbsolutePath() + "/bitstream.cover.pdf");
-            return new File(tempDir.getAbsolutePath() + "/bitstream.cover.pdf");
+            //We already have the full PDF in memory, so keep it there
+            try(ByteArrayOutputStream out = new ByteArrayOutputStream()) {
+                document.save(out);
+
+                byte[] data = out.toByteArray();
+                return Pair.of((InputStream) new ByteArrayInputStream(data), new Long(data.length));
+            }
+
         } finally {
             sourceDocument.close();
             document.close();

dspace-api/src/main/java/org/dspace/disseminate/service/CitationDocumentService.java

@@ -7,6 +7,7 @@
  */
 package org.dspace.disseminate.service;
 
+import org.apache.commons.lang3.tuple.Pair;
 import org.apache.pdfbox.pdmodel.PDPage;
 import org.apache.pdfbox.pdmodel.PDPageContentStream;
 import org.apache.pdfbox.pdmodel.font.PDFont;
@@ -15,8 +16,8 @@
 import org.dspace.content.Item;
 import org.dspace.core.Context;
 
-import java.io.File;
 import java.io.IOException;
+import java.io.InputStream;
 import java.sql.SQLException;
 
 /**
@@ -81,7 +82,7 @@ public interface CitationDocumentService {
      * @throws SQLException if database error
      * @throws AuthorizeException if authorization error
      */
-    public File makeCitedDocument(Context context, Bitstream bitstream)
+    public Pair<InputStream, Long> makeCitedDocument(Context context, Bitstream bitstream)
             throws IOException, SQLException, AuthorizeException;
 
     /**

dspace-spring-rest/src/main/java/org/dspace/app/rest/BitstreamContentRestController.java

@@ -17,6 +17,7 @@
 import javax.ws.rs.core.Response;
 
 import org.apache.catalina.connector.ClientAbortException;
+import org.apache.commons.lang3.tuple.Pair;
 import org.apache.log4j.Logger;
 import org.dspace.app.rest.model.BitstreamRest;
 import org.dspace.app.rest.utils.ContextUtil;
@@ -28,6 +29,7 @@
 import org.dspace.content.service.BitstreamService;
 import org.dspace.core.Constants;
 import org.dspace.core.Context;
+import org.dspace.disseminate.service.CitationDocumentService;
 import org.dspace.services.EventService;
 import org.dspace.usage.UsageEvent;
 import org.springframework.beans.factory.annotation.Autowired;
@@ -71,6 +73,9 @@ public class BitstreamContentRestController {
 	@Autowired
     private AuthorizeService authorizeService;
 
+	@Autowired
+	private CitationDocumentService citationDocumentService;
+
 	@RequestMapping(method = {RequestMethod.GET, RequestMethod.HEAD})
 	public void retrieve(@PathVariable UUID uuid, HttpServletResponse response,
 											 HttpServletRequest request) throws IOException, SQLException, AuthorizeException {
@@ -86,23 +91,18 @@ public void retrieve(@PathVariable UUID uuid, HttpServletResponse response,
         Long lastModified = bitstreamService.getLastModified(bit);
         BitstreamFormat format = bit.getFormat(context);
         String mimetype = format.getMIMEType();
+        String name = getBitstreamName(bit, format);
+
+        Pair<InputStream, Long> bitstreamTuple = getBitstreamInputStreamAndSize(context, bit);
 
 		// Pipe the bits
-		try(InputStream is = bitstreamService.retrieve(context, bit)) {
-
-            String name = bit.getName();
-            if (name == null) {
-                // give a default name to the file based on the UUID and the primary extension of the format
-                name = bit.getID().toString();
-                if (format != null && format.getExtensions() != null && format.getExtensions().size() > 0) {
-                    name += "." + format.getExtensions().get(0);
-                }
-            }
+		try (InputStream is = bitstreamTuple.getLeft())
+		{
             MultipartFileSender sender = MultipartFileSender
 					.fromInputStream(is)
 					.withBufferSize(BUFFER_SIZE)
 					.withFileName(name)
-					.withLength(bit.getSize())
+					.withLength(bitstreamTuple.getRight())
 					.withChecksum(bit.getChecksum())
 					.withMimetype(mimetype)
 					.withLastModified(lastModified)
@@ -134,6 +134,40 @@ public void retrieve(@PathVariable UUID uuid, HttpServletResponse response,
 		}
 	}
 
+	private Pair<InputStream, Long> getBitstreamInputStreamAndSize(Context context, Bitstream bit) throws SQLException, IOException, AuthorizeException {
+
+		if (citationDocumentService.isCitationEnabledForBitstream(bit, context)) {
+			return generateBitstreamWithCitation(context, bit);
+		} else {
+			return Pair.of(bitstreamService.retrieve(context, bit),bit.getSize());
+		}
+	}
+
+	private Pair<InputStream, Long> generateBitstreamWithCitation(Context context, Bitstream bitstream) throws SQLException, IOException, AuthorizeException {
+		//Create the cited document
+		Pair<InputStream, Long> citationDocument = citationDocumentService.makeCitedDocument(context, bitstream);
+		if (citationDocument.getLeft() == null) {
+			log.error("CitedDocument was null");
+		} else {
+			if (log.isDebugEnabled()) {
+				log.debug("CitedDocument was ok, has size " + citationDocument.getRight());
+			}
+		}
+		return citationDocument;
+	}
+
+	private String getBitstreamName(Bitstream bit, BitstreamFormat format) {
+		String name = bit.getName();
+		if (name == null) {
+			// give a default name to the file based on the UUID and the primary extension of the format
+			name = bit.getID().toString();
+			if (format != null && format.getExtensions() != null && format.getExtensions().size() > 0) {
+				name += "." + format.getExtensions().get(0);
+			}
+		}
+		return name;
+	}
+
     private Bitstream getBitstream(Context context, @PathVariable UUID uuid, HttpServletResponse response) throws SQLException, IOException, AuthorizeException {
         Bitstream bit = bitstreamService.find(context, uuid);
         if (bit == null) {

dspace-spring-rest/src/test/java/org/dspace/app/rest/BitstreamContentRestControllerIT.java

@@ -7,17 +7,28 @@
  */
 package org.dspace.app.rest;
 
+import static org.hamcrest.CoreMatchers.not;
+import static org.hamcrest.CoreMatchers.nullValue;
 import static org.junit.Assert.assertEquals;
+import static org.junit.Assert.assertTrue;
 import static org.springframework.test.web.servlet.request.MockMvcRequestBuilders.get;
 import static org.springframework.test.web.servlet.request.MockMvcRequestBuilders.head;
 import static org.springframework.test.web.servlet.result.MockMvcResultMatchers.*;
 
+import java.io.ByteArrayInputStream;
+import java.io.File;
+import java.io.FileInputStream;
 import java.io.IOException;
 import java.io.InputStream;
+import java.io.StringWriter;
+import java.io.Writer;
 import java.util.UUID;
 
 import org.apache.commons.io.IOUtils;
 import org.apache.commons.lang3.CharEncoding;
+import org.apache.commons.lang3.StringUtils;
+import org.apache.pdfbox.pdmodel.PDDocument;
+import org.apache.pdfbox.text.PDFTextStripper;
 import org.apache.solr.client.solrj.SolrQuery;
 import org.apache.solr.client.solrj.SolrServerException;
 import org.apache.solr.client.solrj.response.QueryResponse;
@@ -30,11 +41,14 @@
 import org.dspace.content.Bitstream;
 import org.dspace.content.Collection;
 import org.dspace.content.Item;
+import org.dspace.disseminate.CitationDocumentServiceImpl;
 import org.dspace.eperson.Group;
+import org.dspace.services.ConfigurationService;
 import org.dspace.solr.MockSolrServer;
 import org.junit.After;
 import org.junit.Before;
 import org.junit.Test;
+import org.springframework.beans.factory.annotation.Autowired;
 
 /**
  * Integration test to test the /api/core/bitstreams/[id]/content endpoint
@@ -46,12 +60,19 @@ public class BitstreamContentRestControllerIT extends AbstractControllerIntegrat
 
     private MockSolrServer mockSolrServer;
 
+    @Autowired
+    private ConfigurationService configurationService;
+
+    @Autowired
+    private CitationDocumentServiceImpl citationDocumentService;
+
     @Before
     public void setup() throws Exception {
         super.setUp();
         mockSolrServer = new MockSolrServer("statistics");
         mockSolrServer.getSolrServer().deleteByQuery("*:*");
         mockSolrServer.getSolrServer().commit();
+        configurationService.setProperty("citation-page.enable_globally", false);
     }
 
     @After
@@ -301,4 +322,94 @@ private void checkNumberOfStatsRecords(Bitstream bitstream, int expectedNumberOf
         assertEquals(expectedNumberOfStatsRecords, queryResponse.getResults().getNumFound());
     }
 
+    @Test
+    public void retrieveCitationCoverpageOfBitstream() throws Exception {
+        configurationService.setProperty("citation-page.enable_globally", true);
+        citationDocumentService.afterPropertiesSet();
+        context.turnOffAuthorisationSystem();
+
+        //** GIVEN **
+        //1. A community-collection structure with one parent community and one collections.
+        parentCommunity = CommunityBuilder.createCommunity(context)
+                .withName("Parent Community")
+                .build();
+
+        Collection col1 = CollectionBuilder.createCollection(context, parentCommunity).withName("Collection 1").build();
+
+        //2. A public item with a bitstream
+        File originalPdf = new File(testProps.getProperty("test.bitstream"));
+
+
+        try(InputStream is = new FileInputStream(originalPdf)) {
+
+            Item publicItem1 = ItemBuilder.createItem(context, col1)
+                    .withTitle("Public item citation cover page test 1")
+                    .withIssueDate("2017-10-17")
+                    .withAuthor("Smith, Donald").withAuthor("Doe, John")
+                    .build();
+
+            Bitstream bitstream = BitstreamBuilder
+                    .createBitstream(context, publicItem1, is)
+                    .withName("Test bitstream")
+                    .withDescription("This is a bitstream to test the citation cover page.")
+                    .withMimeType("application/pdf")
+                    .build();
+
+            //** WHEN **
+            //We download the bitstream
+            byte[] content = getClient().perform(get("/api/core/bitstreams/" + bitstream.getID() + "/content"))
+
+                    //** THEN **
+                    .andExpect(status().isOk())
+
+                    //The Content Length must match the full length
+                    .andExpect(header().string("Content-Length", not(nullValue())))
+                    //The server should indicate we support Range requests
+                    .andExpect(header().string("Accept-Ranges", "bytes"))
+                    //The ETag has to be based on the checksum
+                    .andExpect(header().string("ETag", bitstream.getChecksum()))
+                    //We expect the content type to match the bitstream mime type
+                    .andExpect(content().contentType("application/pdf"))
+                    //THe bytes of the content must match the original content
+                    .andReturn().getResponse().getContentAsByteArray();
+
+            // The citation cover page contains the item title. We will now verify that the pdf text contains this title.
+            String pdfText = extractPDFText(content);
+            System.out.println(pdfText);
+            assertTrue(StringUtils.contains(pdfText,"Public item citation cover page test 1"));
+
+            // The dspace-api/src/test/data/dspaceFolder/assetstore/ConstitutionofIreland.pdf file contains 64 pages, manually counted + 1 citation cover page
+            assertEquals(65,getNumberOfPdfPages(content));
+
+            //A If-None-Match HEAD request on the ETag must tell is the bitstream is not modified
+            getClient().perform(head("/api/core/bitstreams/" + bitstream.getID() + "/content")
+                    .header("If-None-Match", bitstream.getChecksum()))
+                    .andExpect(status().isNotModified());
+
+            //The download and head request should also be logged as a statistics record
+            checkNumberOfStatsRecords(bitstream, 2);
+        }
+    }
+
+    private String extractPDFText(byte[] content) throws IOException {
+        PDFTextStripper pts = new PDFTextStripper();
+        pts.setSortByPosition(true);
+
+        try (ByteArrayInputStream source = new ByteArrayInputStream(content);
+            Writer writer = new StringWriter();
+            PDDocument pdfDoc = PDDocument.load(source)){
+
+            pts.writeText(pdfDoc, writer);
+            return writer.toString();
+        }
+    }
+
+    private int getNumberOfPdfPages(byte[] content) throws IOException {
+        try (ByteArrayInputStream source = new ByteArrayInputStream(content);
+             PDDocument pdfDoc = PDDocument.load(source)){
+            return pdfDoc.getNumberOfPages();
+        }
+    }
+
+
 }