Develop Reference

Spring Boot File Download Progress

I have a small problem regarding a download controller I want to implement. I want to make files accessible through an url, therefore I am using a spring boot controller. Because there could be several data types, I use the Apache Tika lib to determine the correct media type. I am currently bound to use the JAX-RS Requests. To save the byte array to a File, I use guava.
#GET
#Path("/getMedia")
public Response downloadMedia(#QueryParam("company") final String company,
#QueryParam("username") final String username,
#QueryParam("password") final String password,
#QueryParam("messageId") final int messageId) throws IOException {
ApplicationContext ctx = CyMega.getInstance().getContext(company);
if (ctx == null) {
return Response.status(Response.Status.UNAUTHORIZED).entity("Company not found").build();
}
ExternalAuthenticationService extService = ctx.getBean(ExternalAuthenticationService.class);
ExternalAuthObj authObj = extService.validateLogin(company, username, password);
if (authObj.getCode() != 0) {
return Response.status(Response.Status.UNAUTHORIZED).entity(authObj)build();
}
ChatService chatService = ctx.getBean(ChatService.class);
ChatGroupMessage message = chatService.getSingleChatGroupMessage(messageId);
if (message != null) {
Byte[] blobs = chatService.getBlob(messageId);
byte[] blob = ArrayUtils.toPrimitive(blobs);
File file = new File(message.getFilename());
Files.write(blob, file);
TikaConfig config = TikaConfig.getDefaultConfig();
Detector detector = config.getDetector();
Metadata metadata = new Metadata();
ByteArrayInputStream inputStream = new ByteArrayInputStream(blob);
TikaInputStream tikaInputStream = TikaInputStream.get(inputStream);
metadata.add(TikaCoreProperties.RESOURCE_NAME_KEY, message.getFilename());
org.apache.tika.mime.MediaType mediaType = detector.detect(tikaInputStream, metadata);
return Response.status(Response.Status.OK)
.header("Content-Length", String.valueOf(blob.length))
.header(HttpHeaders.CONTENT_TYPE, mediaType)
.entity(file).build();
} else {
return Response.status(Response.Status.NOT_FOUND).build();
}
}
By using this code, the file successfully will be send to the client. The only problem is that I also want to track the progress of the download. I tried defining the Content-Length, but this causes my request to load way longer than necessary. Is there any way to achieve that? Should I use a ByteArrayStream as response entity? It would be awesome if someone could provide an example on how to do that properly.
Thanks in advance!

Reading resources from src/main/resources/.. in spring-MVC

I have a question that makes my head ache.
First, my project structure looks like below.
I made a controller, which returns image(*.png) file to the appropriate request.
The code of controller is written below.
#Controller
public class ImageController {
#GetMapping(value = "/ImageStore.do", produces = MediaType.IMAGE_PNG_VALUE)
public #ResponseBody byte[] getStoreImage(HttpServletRequest request) throws IOException {
String image_name = request.getParameter("image_name");
Resource resource = null;
try {
resource = new ClassPathResource("/images/stores/" + image_name);
if(resource == null) {
throw new NullPointerException();
}
} catch(NullPointerException e) {
resource = new ClassPathResource("/images/stores/noimage.png");
}
InputStream inputStream = resource.getInputStream();
return IOUtils.toByteArray(inputStream);
}
}
Q1. I added try-catch phrase to send noimage.png if the request parameter is wrong, or if the filename of request parameter image_name does not exist. But it doesn't seem to work, and it gives me log saying
class path resource [images/stores/noima.png] cannot be opened because it does not exist
(If you need to know the full stack trace, I will comment below.)
Q2. I have 2 image files, hello.png and noimage.png in the folder /resources/images/stores/. I can read noimage.png correctly, but if I make request localhost:8080/ImageStore.do?image_name=hello.png, then it makes an error, making the same log in Q1.

There's no reason to think that the constructor would result in a null value.
The exception you are getting is likely from the getInputStream method, which is documented to throw
FileNotFoundException - if the underlying resource doesn't exist
IOException - if the content stream could not be opened
A slight adjustment might help
#Controller
public class ImageController {
#GetMapping(value = "/ImageStore.do", produces = MediaType.IMAGE_PNG_VALUE)
public #ResponseBody byte[] getStoreImage(HttpServletRequest request) throws IOException {
InputStream is = null;
try {
String image_name = request.getParameter("image_name");
is = new ClassPathResource("/images/stores/" + image_name).getInputStream();
} catch(FileNotFoundException e) {
is = new ClassPathResource("/images/stores/noimage.png").getInputStream();
}
return IOUtils.toByteArray(is);
}
}
You should include the stack trace, and exception message, which might assist understanding your second query, but I would check that the file really does exist, with the exact name you're using.

Multipart File Upload via REST API corrupts file and increases file size

Am uploading a file (any format doc, docx, pdf, text, etc) as multipart form/data to a REST API from Postman or application UI. The text file uploads fine. All other non-text formats get corrupted. I cant open those files.
The size of the uploaded file increases drastically. Check the following server log:
File size just before call to request.getRequestDispatcher().forward(): 27583
Controller, first line in method:39439
The size of the uploaded file is 27.3Kb
I am guessing the files gets corrupted because of the other data appended to the file.
Controller method is
#RequestMapping(value="/entity/{entity}/{entityId}/type/{type}/{derive}",method = RequestMethod.POST)
#ResponseBody
public String uploadFile(#RequestParam("file") MultipartFile multipartFile,#PathVariable("entity")String entity,#PathVariable("type")String type,#PathVariable("entityId")Long entityId,#PathVariable("derive") boolean derive) throws Exception
Since text file is saving correctly and other files also get written correctly, don't think the code to write the file is incorrect.
Code to get inputStream
public String storeFile(MultipartFile multipartFile, String entity, Long id, String uploadType, boolean isDerive,String customName)
throws Exception
{
try
{
System.out.println(multipartFile.getSize());
String fileName = "";
String contentType = "";
if (multipartFile != null)
{
fileName = multipartFile.getOriginalFilename();
contentType = multipartFile.getContentType();
if (contentType == null)
{
contentType = "application/msword";
}
}
InputStream is = multipartFile.getInputStream();
String filePath = getFileName(entity, uploadType, id, fileName, isDerive,customName);
Helper.storeFile(is, filePath);
precedingPath = precedingPath.length() > 0 ? precedingPath + "/":"";
return precedingPath + filePath;
}
catch (WebException e)
{
e.printStackTrace();
throw e;
}
catch (Exception e)
{
e.printStackTrace();
throw new WebException(e.getMessage(), IHttpConstants.INTERNAL_SERVER_ERROR, e);
}
}
Helper.storeFile
public static File storeFile(InputStream is, String filePath) throws IOException {
try {
String staticRepoPath = null;
if (MasterData.getInstance().getSettingsMap().containsKey(Settings.REPO_LOCATION.toString())) {
staticRepoPath = MasterData.getInstance().getSettingsMap().get(Settings.REPO_LOCATION.toString());
} else {
throw new WebException("Invalid Settings");
}
byte[] buffer = new byte[is.available()];
is.read(buffer);
File targetFile = new File(staticRepoPath + File.separator + filePath);
#SuppressWarnings("resource")
OutputStream outStream = new FileOutputStream(targetFile);
outStream.write(buffer);
return targetFile;
} catch (Exception e) {
e.printStackTrace();
}
return null;
}
My Ajax request is as follows
var fd = new FormData();
//Take the first selected file
fd.append("file", document.actualFile);
//Generic AJAX call
CandidateService.ajax_uploadDocumentWithDocType($scope.candidate.id, fd, document.docType, function (error, json)
Content type while uploading:
var config = {headers:{'X-Auth-Token':authToken, 'Content-Type': undefined}, transformRequest: angular.identity};
Would anyone know how I can fix this and upload the file successfully?
Q1) Why does the file size change between the request dispatcher and the controller that handles the file data.
Q2) Could this change of file size be the cause of file corruption? Libre Office cause General Input/Output Error.

I figured the problem with the file upload. I had a spring filter in between that was changing the request to a wrappedRequest. This was adding additional data to the multipart data and causing the file to be corrupted.

Well in my case I had this exact same problem when accesing the API through Amazon API Gateway. Turned out I forgot to allow multipart ContentType on API Gateway.
Kind of weird the requests still made it to my server and text files worked fine.

Trouble downloading binary file with angular $http or jquery.ajax

My problem is that I am getting the wrong sized file on the client side. Here is my #Controller ...
#RequestMapping(value = "/download/{id}", method = RequestMethod.GET)
public ResponseEntity<?> download(final HttpServletRequest request,
final HttpServletResponse response,
#PathVariable("id") final int id) throws IOException {
try {
// Pseudo-code for retrieving file from ID.
Path zippath = getZipFile(id);
if (!Files.exists(zippath)) {
throw new IOException("File not found.");
}
ResponseEntity<InputStreamResource> result;
return ResponseEntity.ok()
.contentLength(Files.size(zippath))
.contentType(MediaType.APPLICATION_OCTET_STREAM)
.body(new InputStreamResource(new FileInputStream(zippath.toFile())));
} catch (Exception ex) {
// ErrorInfo is another class, unimportant
return ResponseEntity.status(HttpStatus.INTERNAL_SERVER_ERROR).body(new ErrorInfo(ex));
}
}
... and here is my client-side code using angular-file-saver ...
$http({url: "export/download/" + exportitem.exportId, withCredentials: true})
.then(function(response) {
function str2bytes(str) {
var bytes = new Uint8Array(str.length);
for (var i=0; i<str.length; i++) {
bytes[i] = str.charCodeAt(i);
}
return bytes;
}
var blob = new Blob([str2bytes(response.data)], {type: 'application/octet-stream'});
FileSaver.saveAs(blob, "download.zip");
}, $exceptionHandler);
The original file is 935673 bytes but response.data is 900728 and passing it through the transformation to Uint8Array results in a Blob that is 900728 in size as well. Either way, the resulting saved file is 900728 bytes (34945 bytes shy). Also it is not quite the same in what gets written. It seems to slightly get bloated but then the last part just seems to be truncated. Any ideas what I might be doing wrong?
UPDATE
I just updated my controller method to be the following and got the exact same result. Grrr.
#RequestMapping(value = "/download/{id}", method = RequestMethod.GET)
public void download(final HttpServletRequest request,
final HttpServletResponse response,
#PathVariable("id") final int id) throws IOException {
// Pseudo-code for retrieving file from ID.
Path zippath = getZipFile(id);
if (!Files.exists(zippath)) {
throw new IOException("File not found.");
}
response.setContentType("application/zip");
response.setHeader("Content-Disposition",
"attachment; filename=download.zip");
InputStream inputStream = new FileInputStream(zippath.toFile());
org.apache.commons.io.IOUtils.copy(inputStream, response.getOutputStream());
response.flushBuffer();
inputStream.close();
}

So the problem turned out to be angular's $http service. I also tried jQuery's ajax method. Both gave the same result. If I instead use the native XMLHttpRequest it works correctly. So the Java code was sound. I first verified this by exposing the file directly to the internet and then both using curl and directly accessing in the browser I managed to download the file of the correct size. Then I found this solution so that I could also download the file via javascript.
var xhr = new XMLHttpRequest();
xhr.open('GET', url, true);
xhr.responseType = "blob";
xhr.withCredentials = true;
xhr.onreadystatechange = function (){
if (xhr.readyState === 4) {
var blob = xhr.response;
FileSaver.saveAs(blob, filename);
}
};
xhr.send();
Why does angular or jQuery give the wrong result? I still don't know but if anyone wishes to give an answer that uses those it would be appreciated.

responseType: blob
did the trick for a zip file
Angular 2 +
this.http.get('http://localhost:8080/export', { responseType: ResponseContentType.Blob })
.subscribe((res: any) => {
const blob = new Blob([res._body], { type: 'application/zip' });
saveAs(blob, "fileName.zip");

i just stumbled over the 'responseType' in $http requests, you are probably looking for 'blob': https://docs.angularjs.org/api/ng/service/$http#usage

How to download file in Spring mvc and mongodb

Hi have written controller class like below. I am trying to get file from mongo db and try to download it.
organizationFileAttachmentService.setUser(getUser());
GridFSDBFile file = organizationFileAttachmentService.getGridFSDBFileById(new ObjectId(id), "File");
if (file != null) {
byte[] content = organizationFileAttachmentService.findByIdAndBucket(new ObjectId(id), "File");
try {
int size = content.length;
InputStream is = null;
byte[] b = new byte[size];
try {
is = new ByteArrayInputStream(content);
is.read(b);
} catch (IOException e) {
e.printStackTrace();
} finally {
try {
if (is != null)
is.close();
} catch (Exception ex) {
}
}
response.setContentType(file.getContentType());
// String attachment =
// "attachment; filename=\""+file.getFilename()+"\"";
String attachment = "attachment; filename=" + file.getFilename();
// response.setContentLength(new
// Long(file.getLength()).intValue());
response.setCharacterEncoding(file.getMD5());
response.setHeader("content-Disposition", attachment);// "attachment;filename=test.xls"
// copy it to response's OutputStream
// FileCopyUtils.copy(is, response.getOutputStream());
IOUtils.copy(is, response.getOutputStream());
response.flushBuffer();
is.close();
} catch (IOException ex) {
_logger.info("Error writing file to output stream. Filename was '" + id + "'");
throw new RuntimeException("IOError writing file to output stream");
}
but i am not able to down load file. can any one help me.

In case you missed it, Spring provides various built in resource handlers.
http://docs.spring.io/spring/docs/3.2.5.RELEASE/spring-framework-reference/html/resources.html#resources-implementations
If your method returns one of those (perhaps the ByteArrayResource in your case), then you just need a couple of annotations on the interface like so:
#RequestMapping(value = "/foo/bar/{fileId}",
method = RequestMethod.GET,
produces = { MediaType.APPLICATION_OCTET_STREAM_VALUE })
#ResponseBody FileSystemResource downloadFile(Long fileId);
No fiddling with encodings and headers for you that way. I'd recommend trying that before rolling your own.
Edit: The above worked fine in Spring 3.1.4. It no longer works for 3.2.x or 4.x. Whereas previously, the produces = { MediaType.APPLICATION_OCTET_STREAM_VALUE } would cause Spring to add the appropriate headers, it now treats that as a restriction. If accessing the URL with a standard web browser, an accept header of "application/octet-stream" will not be sent. Spring will therefore return a 406 error. To get it working again, such a method needs to be re-written without the "produces" attribute. Instead, add HttpServletResponse to the method arguments and add the header inside the method. i.e.:
#RequestMapping(value = "/foo/bar/{fileId}",
method = RequestMethod.GET)
#ResponseBody FileSystemResource downloadFile(
Long fileId, HttpServletResponse response) {
...
response.setHeader( "Content-Disposition", "attachment;filename=" + fileName );
...
}
Edit redux:
Now using Spring 4.0.7 via Spring Boot 1.1.8. It would appear that setting the produces = { MediaType.APPLICATION_OCTET_STREAM_VALUE } instruction is now working again. Just having that instruction seems to be enough for all the browsers I have tried. Note however, that I have also found that it does not set the Content-Disposition, which is left as application/json. Although this doesn't seem to be an issue for browsers, I have come across bugs in PHP client applications, which seem to behave only based on the Content-Disposition. So it seems that the current solution is to do both of the above!

I have changed my request as GET and added request in anchor tag in html. Aslo changed my code as
#RequestMapping(value = "/getFileById/{id}", method = RequestMethod.GET)
public #ResponseBody
void download(#PathVariable String id, HttpServletRequest request, HttpServletResponse response) throws IOException {
organizationFileAttachmentService.setUser(getUser());
GridFSDBFile file = organizationFileAttachmentService.getGridFSDBFileById(new ObjectId(id), "File");
if (file != null) {
try {
response.setContentType(file.getContentType());
response.setContentLength((new Long(file.getLength()).intValue()));
response.setHeader("content-Disposition", "attachment; filename=" + file.getFilename());// "attachment;filename=test.xls"
// copy it to response's OutputStream
IOUtils.copyLarge(file.getInputStream(), response.getOutputStream());
} catch (IOException ex) {
_logger.info("Error writing file to output stream. Filename was '" + id + "'");
throw new RuntimeException("IOError writing file to output stream");
}
}
}
Now it is working fine for me.