Download file from inside resource's DwC-A

src/main/java/org/gbif/ipt/action/BaseAction.java

@@ -16,6 +16,7 @@
 import java.util.ResourceBundle;
 import javax.annotation.Nullable;
 import javax.servlet.http.HttpServletRequest;
+import javax.servlet.http.HttpServletResponse;
 
 import com.google.common.base.Strings;
 import com.google.inject.Inject;
@@ -25,13 +26,14 @@
 import org.apache.commons.lang3.StringUtils;
 import org.apache.log4j.Logger;
 import org.apache.struts2.interceptor.ServletRequestAware;
+import org.apache.struts2.interceptor.ServletResponseAware;
 import org.apache.struts2.interceptor.SessionAware;
 
 /**
  * The base of all IPT actions. This handles conditions such as menu items, a custom text provider, sessions, currently
  * logged in user, and hosting organization information.
  */
-public class BaseAction extends ActionSupport implements SessionAware, Preparable, ServletRequestAware {
+public class BaseAction extends ActionSupport implements SessionAware, Preparable, ServletRequestAware , ServletResponseAware {
 
   // logging
   private static final Logger LOG = Logger.getLogger(BaseAction.class);
@@ -48,6 +50,7 @@ public class BaseAction extends ActionSupport implements SessionAware, Preparabl
   protected List<String> warnings = new ArrayList<String>();
   protected Map<String, Object> session;
   protected HttpServletRequest req;
+  protected HttpServletResponse res;
   // a generic identifier for loading an object BEFORE the param interceptor sets values
   protected String id;
 
@@ -264,6 +267,10 @@ public void setServletRequest(HttpServletRequest req) {
     this.req = req;
   }
 
+  public void setServletResponse(HttpServletResponse res) {
+    this.res = res;
+  }
+
   public void setSession(Map<String, Object> session) {
     this.session = session;
     // always keep sth in the session otherwise the session is not maintained and e.g. the message redirect interceptor

src/main/java/org/gbif/ipt/action/portal/ResourceFileAction.java

@@ -1,5 +1,14 @@
 package org.gbif.ipt.action.portal;
 
+import com.google.common.base.Strings;
+import com.google.inject.Inject;
+import org.apache.commons.compress.archivers.zip.ZipArchiveEntry;
+import org.apache.commons.compress.archivers.zip.ZipFile;
+import org.apache.commons.compress.compressors.deflate.DeflateCompressorInputStream;
+import org.apache.commons.compress.compressors.gzip.GzipCompressorInputStream;
+import org.apache.commons.compress.utils.IOUtils;
+import org.apache.commons.lang3.StringUtils;
+import org.apache.log4j.Logger;
 import org.gbif.ipt.config.AppConfig;
 import org.gbif.ipt.config.Constants;
 import org.gbif.ipt.config.DataDir;
@@ -13,11 +22,10 @@
 import java.io.FileNotFoundException;
 import java.io.InputStream;
 import java.math.BigDecimal;
+import java.util.Enumeration;
+import java.util.zip.GZIPOutputStream;
 
-import com.google.common.base.Strings;
-import com.google.inject.Inject;
-import org.apache.commons.lang3.StringUtils;
-import org.apache.log4j.Logger;
+import javax.servlet.ServletOutputStream;
 
 /**
  * The Action responsible for serving datadir resource files.
@@ -42,6 +50,91 @@ public ResourceFileAction(SimpleTextProvider textProvider, AppConfig cfg, Regist
     this.dataDir = dataDir;
   }
 
+  /**
+   * Handles DwC-A internal file download request.
+   *
+   * @return Struts2 result string
+   */
+  public String dwcaIn() {
+    if (resource == null) {
+      return NOT_FOUND;
+    }
+    String internalFilename = StringUtils.trimToNull(req.getParameter(Constants.REQ_PARAM_FILE));
+
+    if(internalFilename == null || internalFilename.trim().isEmpty()) {
+      return NOT_FOUND;
+    }
+
+    // if no specific version is requested, use the latest published version
+    if (version == null) {
+      BigDecimal latestVersion = resource.getLastPublishedVersionsVersion();
+      if (latestVersion == null) {
+        return NOT_FOUND;
+      } else {
+        version = latestVersion;
+      }
+    }
+
+    boolean gz = internalFilename.endsWith(".gz");
+    if(gz) {
+      internalFilename = internalFilename.replace(".gz","");
+    }
+    boolean compress = gz || (req.getHeader("Accept-Encoding") != null && req.getHeader("Accept-Encoding").contains("gzip"));
+
+    // serve file
+    File dwcaFile = dataDir.resourceDwcaFile(resource.getShortname(), version);
+    try {
+      ZipFile dwcaZip = new ZipFile(dwcaFile);
+      ZipArchiveEntry entry = dwcaZip.getEntry(internalFilename);
+      if(entry != null) {
+          inputStream = dwcaZip.getInputStream(entry);
+      } else {
+        return NOT_FOUND;
+      }
+    }catch(Exception e) {
+      LOG.warn("failed to get internal file", e);
+      return ERROR;
+    }
+
+    // construct download filename
+    StringBuilder sb = new StringBuilder();
+    sb.append("dwca-" + resource.getShortname());
+    if (version != null) {
+      sb.append("-v" + version.toPlainString());
+    }
+    sb.append("-"+internalFilename);
+    if(gz) sb.append(".gz");
+    filename = sb.toString();
+
+    if(compress){
+      try {
+        if(gz) {
+          res.setContentType("application/x-gzip");
+        } else {
+          res.setContentType("application/octet-stream");
+          res.addHeader("Content-Encoding", "gzip");
+
+        }
+        res.addHeader("Content-Disposition", "attachment; filename=\""+filename+"\"");
+
+        ServletOutputStream sos = res.getOutputStream();
+        GZIPOutputStream zos = new GZIPOutputStream(sos);
+        IOUtils.copy(inputStream, zos);
+        zos.flush();
+        inputStream.close();
+        zos.close();
+        sos.close();
+      } catch (Exception e){
+        LOG.warn("error sending gzip",e);
+      }
+      return NONE;
+    } else {
+      mimeType = "application/octet-stream";
+      return SUCCESS;
+
+    }
+  }
+
   /**
    * Handles DwC-A file download request. The method checks if the request is a conditional get with If-Modified-Since
    * header. If the If-Modified-Since date is greater than the last published date, the NOT_MODIFIED string is returned.

src/main/java/org/gbif/ipt/config/Constants.java

@@ -19,6 +19,7 @@ public final class Constants {
   public static final String REQ_PATH_DWCA = "archive.do";
   public static final String REQ_PATH_LOGO = "logo.do";
   public static final String REQ_PARAM_RESOURCE = "r";
+  public static final String REQ_PARAM_FILE = "f";
   public static final String REQ_PARAM_ID = "id";
   public static final String REQ_PARAM_SOURCE = "s";
   public static final String REQ_PARAM_VERSION = "v";

src/main/resources/struts-portal.xml

@@ -88,6 +88,14 @@
         <param name="bufferSize">1024</param>
       </result>
     </action>
+    <action name="archiveFile" class="org.gbif.ipt.action.portal.ResourceFileAction" method="dwcaIn">
+      <result name="success" type="stream">
+        <param name="contentType">${mimeType}</param>
+        <param name="inputName">inputStream</param>
+        <param name="contentDisposition">filename="${filename}"</param>
+        <param name="bufferSize">1024</param>
+      </result>
+    </action>
     <action name="dcat" class="org.gbif.ipt.action.portal.DCATAction">
       <result type="stream">
         <param name="noCache">true</param>