Merge pull request #3774 from inception-project/feature/3773-Support-…

…loading-resources-from-documents-that-are-actually-archives #3773 - Support loading resources from documents that are actually archives
inception-project · Feb 3, 2023 · 9e3131c · 9e3131c
2 parents deade4c + 852c53b
commit 9e3131c
Show file tree

Hide file tree

Showing 4 changed files with 95 additions and 1 deletion.
diff --git a/...api-formats/src/main/java/de/tudarmstadt/ukp/clarin/webanno/api/format/FormatSupport.java b/...api-formats/src/main/java/de/tudarmstadt/ukp/clarin/webanno/api/format/FormatSupport.java
@@ -19,7 +19,11 @@
 
 import static de.tudarmstadt.ukp.clarin.webanno.support.ZipUtils.zipFolder;
 import static java.io.File.createTempFile;
+import static java.util.Collections.unmodifiableSet;
 import static org.apache.commons.io.FileUtils.copyFile;
+import static org.apache.commons.io.FilenameUtils.getExtension;
+import static org.apache.commons.io.FilenameUtils.normalize;
+import static org.apache.commons.lang3.StringUtils.prependIfMissing;
 import static org.apache.uima.fit.factory.AnalysisEngineFactory.createEngine;
 import static org.apache.uima.fit.factory.CollectionReaderFactory.createReader;
 import static org.apache.uima.fit.factory.ConfigurationParameterFactory.addConfigurationParameters;
@@ -29,9 +33,12 @@
 import java.io.File;
 import java.io.FileNotFoundException;
 import java.io.IOException;
+import java.io.InputStream;
+import java.net.URL;
 import java.util.Collections;
 import java.util.List;
 import java.util.Optional;
+import java.util.Set;
 
 import org.apache.commons.io.FilenameUtils;
 import org.apache.commons.lang3.StringUtils;
@@ -55,6 +62,10 @@
 
 public interface FormatSupport
 {
+    Set<String> DEFAULT_PERMITTED_RESOURCE_EXTENSIONS = unmodifiableSet(
+            Set.of("apng", "avif", "gif", "jpg", "jpeg", "jfif", "pjpeg", "pjp", "png", "svg",
+                    "webp", "bmp", "tif", "tiff"));
+
     /**
      * Returns the format identifier which is stored in in {@link SourceDocument#setFormat(String)}.
      * 
@@ -95,6 +106,26 @@ default boolean isProneToInconsistencies()
         return true;
     }
 
+    default boolean hasResources()
+    {
+        return false;
+    }
+
+    default boolean isAccessibelResource(File aDocFile, String aResourcePath)
+    {
+        return DEFAULT_PERMITTED_RESOURCE_EXTENSIONS.contains(getExtension(aResourcePath));
+    }
+
+    default InputStream openResourceStream(File aDocFile, String aResourcePath) throws IOException
+    {
+        if (!hasResources() || !isAccessibelResource(aDocFile, aResourcePath)) {
+            throw new FileNotFoundException("Resource not found [" + aResourcePath + "]");
+        }
+
+        var path = prependIfMissing(normalize(aResourcePath, isProneToInconsistencies()), "/");
+        return new URL("jar:" + aDocFile.toURI().toURL() + "!" + path).openStream();
+    }
+
     /**
      * @return format-specific CSS style-sheets that styleable editors should load.
      */

diff --git a/inception/inception-external-editor/pom.xml b/inception/inception-external-editor/pom.xml
@@ -45,6 +45,10 @@
       <groupId>de.tudarmstadt.ukp.inception.app</groupId>
       <artifactId>inception-api-schema</artifactId>
     </dependency>
+    <dependency>
+      <groupId>de.tudarmstadt.ukp.inception.app</groupId>
+      <artifactId>inception-api-formats</artifactId>
+    </dependency>
     <dependency>
       <groupId>de.tudarmstadt.ukp.inception.app</groupId>
       <artifactId>inception-model</artifactId>

diff --git a/...ava/de/tudarmstadt/ukp/inception/externaleditor/xhtml/XHtmlXmlDocumentViewController.java b/...ava/de/tudarmstadt/ukp/inception/externaleditor/xhtml/XHtmlXmlDocumentViewController.java
@@ -22,6 +22,9 @@
 import java.security.Principal;
 import java.util.Optional;
 
+import javax.servlet.http.HttpServletRequest;
+
+import org.springframework.core.io.InputStreamResource;
 import org.springframework.http.ResponseEntity;
 
 import de.tudarmstadt.ukp.clarin.webanno.model.SourceDocument;
@@ -35,4 +38,8 @@ public interface XHtmlXmlDocumentViewController
     ResponseEntity<String> getDocument(long aProjectId, long aDocumentId, Optional<String> aEditor,
             Principal aPrincipal)
         throws Exception;
+
+    ResponseEntity<InputStreamResource> getResource(long aProjectId, long aDocumentId,
+            HttpServletRequest aRequest, Principal aPrincipal)
+        throws Exception;
 }
diff --git a/...de/tudarmstadt/ukp/inception/externaleditor/xhtml/XHtmlXmlDocumentViewControllerImpl.java b/...de/tudarmstadt/ukp/inception/externaleditor/xhtml/XHtmlXmlDocumentViewControllerImpl.java
@@ -17,19 +17,29 @@
  */
 package de.tudarmstadt.ukp.inception.externaleditor.xhtml;
 
+import static java.lang.invoke.MethodHandles.lookup;
+import static org.apache.commons.lang3.StringUtils.substringAfter;
+import static org.slf4j.LoggerFactory.getLogger;
+import static org.springframework.http.HttpStatus.OK;
+
+import java.io.FileNotFoundException;
 import java.io.StringWriter;
 import java.security.Principal;
 import java.util.Locale;
 import java.util.Optional;
 import java.util.stream.Collectors;
 
 import javax.servlet.ServletContext;
+import javax.servlet.http.HttpServletRequest;
 
 import org.apache.uima.cas.CAS;
 import org.dkpro.core.api.xml.type.XmlDocument;
 import org.dkpro.core.api.xml.type.XmlElement;
+import org.slf4j.Logger;
 import org.springframework.beans.factory.annotation.Autowired;
 import org.springframework.boot.autoconfigure.condition.ConditionalOnWebApplication;
+import org.springframework.core.io.InputStreamResource;
+import org.springframework.http.HttpHeaders;
 import org.springframework.http.MediaType;
 import org.springframework.http.ResponseEntity;
 import org.springframework.security.access.prepost.PreAuthorize;
@@ -60,7 +70,12 @@ public class XHtmlXmlDocumentViewControllerImpl
     extends XmlDocumentViewControllerImplBase
     implements XHtmlXmlDocumentViewController
 {
-    private static final String GET_DOCUMENT_PATH = "/p/{projectId}/xml/{documentId}";
+    private static final Logger LOG = getLogger(lookup().lookupClass());
+
+    private static final String RESOURCE_ELEMENT = "/res/";
+    private static final String GET_DOCUMENT_PATH = "/p/{projectId}/d/{documentId}/xml";
+    private static final String GET_RESOURCE_PATH = "/p/{projectId}/d/{documentId}"
+            + RESOURCE_ELEMENT + "**";
 
     private static final String XHTML_NS_URI = "http://www.w3.org/1999/xhtml";
     private static final String HTML = "html";
@@ -178,6 +193,43 @@ public ResponseEntity<String> getDocument(@PathVariable("projectId") long aProje
         }
     }
 
+    @PreAuthorize("@documentAccess.canViewAnnotationDocument(#aProjectId, #aDocumentId, #principal.name)")
+    @Override
+    @GetMapping(path = GET_RESOURCE_PATH)
+    public ResponseEntity<InputStreamResource> getResource(
+            @PathVariable("projectId") long aProjectId,
+            @PathVariable("documentId") long aDocumentId, HttpServletRequest aRequest,
+            Principal principal)
+        throws Exception
+    {
+        var srcDoc = documentService.getSourceDocument(aProjectId, aDocumentId);
+
+        var maybeFormatSupport = formatRegistry.getFormatById(srcDoc.getFormat());
+        if (!maybeFormatSupport.isPresent()) {
+            return ResponseEntity.notFound().build();
+        }
+
+        var srcDocFile = documentService.getSourceDocumentFile(srcDoc);
+        var path = aRequest.getServletPath();
+        var resource = substringAfter(path, RESOURCE_ELEMENT);
+
+        var formatSupport = maybeFormatSupport.get();
+
+        try {
+            var inputStream = formatSupport.openResourceStream(srcDocFile, resource);
+            HttpHeaders httpHeaders = new HttpHeaders();
+            return new ResponseEntity<>(new InputStreamResource(inputStream), httpHeaders, OK);
+        }
+        catch (FileNotFoundException e) {
+            LOG.error("Resource [{}] for document {} not found", resource, srcDoc);
+            return ResponseEntity.notFound().build();
+        }
+        catch (Exception e) {
+            LOG.error("Unable to load resource [{}] for document {}", resource, srcDoc, e);
+            return ResponseEntity.notFound().build();
+        }
+    }
+
     private ResponseEntity<String> toResponse(StringWriter aOut)
     {
         return ResponseEntity.ok() //