WebAPI: Upload picture & get byte array - asp.net-mvc

I want the user to be able to upload a file via my application. I don't have DB access, all my data calls get completed via a web-service that another person is writing. I needed to secure the web service, so I've consumed it & exposed it via WebAPI, & added OAuth security.
Now to my problem.
I've written the following.
public Task<FileResult> Post()
{
if (Request.Content.IsMimeMultipartContent())
{
var task = Request.Content.ReadAsByteArrayAsync().ContinueWith(
o =>
{
var result = this.Client.UploadPicture(this.UserId, o.Result);
if (result.ResultCode == 0)
{
return new FileResult()
{
Message = "Success",
FileId = result.ServerId
};
}
throw new HttpResponseException(...);
});
return task;
}
...
}
I'm pretty much a noob when it comes to WebAPI & multithreading (I'm not sure why this needs to be handled async? I'm sure there is a reason, but for now I'd just like a working example and get to the why later..).
My code is loosely based on some R&D & samples I've found on the net, but i haven't come across a scenario like I'm needing to complete... Yet it doesn't seem like I'm doing something out of the ordinary...
Upload a file to the server, and pass the image byte[] object to either sql or another service?
In this line
var result = this.Client.UploadPicture(this.UserId, o.Result);
I'm uploading a byte[] array of something....
Then later (the retrieval method works, I've managed to retrieve & view a test image)
When retrieving the byte array of the "image" i uploaded i get an array of idk what.. EG, i get a valid result of something, but it ain't no picture. Which leads me to believe that the uploaded data is bogus :|
O_o
How to get the image byte[]?

Mime Multipart is more than just your array of bytes. It also has metadata and boundary stuff. You need to treat it as MultiPartContent and then extract the image byte array out of that.
Filip has a blog post on the subject here.

Related

Data Gets Converted when passed to CREATE_STREAM

I have created an UI5 Application to read a file and send it to a custom OData Service in the Backend.
onUploadFile: function() {
var oFileUpload =
this.getView().byId("fileUploaderFS");
var domRef = oFileUpload.getFocusDomRef();
var file = domRef.files[0];
var that = this;
var reader = new FileReader();
var ftype = file.type;
reader.readAsArrayBuffer(file);
reader.onload = function(evt) {
var vContent = evt.currentTarget.result
console.log(vContent);
var hex = that.buf2hex(vContent);
that.updateFile(hex, ftype);
}
},
buf2hex: function(buffer) {
return [...new Uint8Array(buffer)]
.map(x => x.toString(16).padStart(2, '0'))
.join('');
}
When I print the content of hex on the console before sending it to the backend, the data starts with 89504e470d0a1a0a0000000d49484 ....
Even before sending the data in the payload to Odata Service it shows the correct data
Here is the Odata Service
Inside the Create Stream the data when received, is getting converted into something else. As a result the image that has been saved is not opening.
I tried to change the Data Type of Content in SEGW to Binary and it did not work. I also tried to convert the data in the create_stream but in vain. At last I tried reading the data in UI5 in different formats but of no use.
This whole Odata service works perfectly fine when I load the data through Postman Application.
Please help me resolve this Issue. Thanks In Advance.
The sap.ui.unified.FileUploader has everything built in. No need for conversions from Buffer to hex.
Make sure that your FileUploader knows where to upload the file
<unified:FileUploader xmlns:unified="sap.ui.unified"
id="fileUploaderFS"
uploadUrl="/sap/opu/odata/sap/Z_TEST_SRV/FileSet"
/>
The attribute uploadUrl points to the media entity for which you implemented the create_stream method.
Then when the upload is triggered via button press, simply get the FileUploader, set the token (for security reasons when doing a POST request), and fire the upload method.
onUploadFile: function () {
const oFileUpload = this.getView().byId("fileUploaderFS");
const sToken = this.getModel("nameOfTheModel").getSecurityToken();
const oTokenParam = new FileUploaderParameter({
name: "x-csrf-token",
value: sToken
});
oFileUpload.removeAllHeaderParameters()
oFileUpload.addHeaderParameter(oTokenParam);
oFileUpload.upload();
}
To use FileUploaderParameter, make sure to import it at the beginning:
sap.ui.define([
// ...,
"sap/ui/unified/FileUploaderParameter"
], function (/*..., */FileUploaderParameter) {
// ...
Now about your File entity. When working with it via create_stream or read_stream, you don't use the entity structure but is_media_resource. This means your entity doesn't need a property content. Or most of the other properties (except a unique id and the mime type). All other properties would only be used if you want to do one of the CRUD methods (which happens almost never when dealing with streams).

HTTPClient GetAsync post object

We currently have a generic MVC method that GET's data from ASP.NET Web API
public static T Get<T>(string apiURI, object p)
{
using (HttpClient client = new HttpClient())
{
client.BaseAddress = new Uri(Config.API_BaseSite);
HttpResponseMessage response = client.GetAsync(apiURI).Result;
// Check that response was successful or throw exception
if (response.IsSuccessStatusCode == false)
{
string responseBody = response.Content.ReadAsStringAsync().Result;
throw new HttpException((int)response.StatusCode, responseBody);
}
T res = response.Content.ReadAsAsync<T>().Result;
return (T)res;
}
}
Our question is:- obviously, we can not send 'p' as you would with a post,
client.PostAsync(apiURI, new StringContent(p.ToString(), Encoding.UTF8, "application/json")
but how we go about sending this object / JSON with a get.
We have seen sending it as part of the URL, however, is there an alternative?
GET sends the values with the query string (end of url), in regards to "but how we go about sending this object / JSON with a get. We have seen sending it as part of the URL, however, is there an alternative?".
The alternative is POST or PUT.
PUT is best used when the user creates the key/url. You can look at examples such as cnn.com - where the URL's are just short versions of the article title. You want to PUT a page at that URL.
Example:
http://newday.blogs.cnn.com/2014/03/19/five-things-to-know-for-your-new-day-wednesday-march-19-2014/?hpt=hp_t2
has the url of "five-things-to-know-for-your-new-day-wednesday-march-19-2014", which was generated from the article title of "Five Things to Know for Your New Day – Wednesday, March 19, 2014"
In general, you should follow these guidelines:
Use GET when you want to fetch data from the server. Think of search engines. You can see your search query in the query string. You can also book mark it. It doesn't change anything on the server at all.
Use POST when you want to create a resource.
Use PUT when you want to create resources, but it also overwrites them. If you PUT an object twice, the servers state is only changed once. The opposite is true for POST
Use DELETE when you want to delete stuff
Neither POST nor PUT use the query string. GET does

Facebook Graph API object_id missing sometimes

I am using the Facebook Graph API to get the news feed of a user.
My request URL is:
xxxxxxxxxxxxxx/feed?fields=from,id,created_time,picture,link,object_id,message, likes.fields(id)
With the object_id, I want to get the big picture of the post, using the following url:
http://graph.facebook.com/OBJECT_ID/picture?type=normal
The picture return field is always filled, but the object_id is not being returned at some posts. Why is this? I really need the high res picture, and didn't find another way to acquire this..
The object_id is only returned if the attachment is a facebook object (e.g. an image uploaded by the user). Some stories in feed don't have a picture at all, and some pictures are not facebook objects (e.g. thumbnails for shared links).
Sometimes Facebook keeps a thumbnail of an image and stores an external link to the larger version of the image in the URL returned by the graph request. In order to access the images in either case, I used the code below, where smallURL is the URL returned by the graph request:
private String getRealURL(String smallURL){
if (smallURL.contains("url=http")){
String[] pieces = smallURL.split("url=");
pieces[1].replace("%2F", "//");
pieces[1].replace("%3A", ":");
return pieces[1];
}
else{
StringBuilder stringBuilder = new StringBuilder();
stringBuilder.setLength(0);
stringBuilder.append("http://graph.facebook.com/");
stringBuilder.append(item.getObjectID());
stringBuilder.append("/picture?type=large");
return stringBuilder.toString();
}
}
I also noticed that some FB posts didn't have {object_id} for large photos, but realized that the {picture} thumbnail URL contains the encoded URL of the original larger image:
https://external.xx.fbcdn.net/safe_image.php?d=AQBe9UvGd0vPbAHP&w=130&h=130&url=http%3A%2F%2Fskift.com%2Fwp-content%2Fuploads%2F2015%2F12%2Fpollution.jpg&cfs=1
--> contains -->
http://skift.com/wp-content/uploads/2015/12/pollution.jpg
So I check for {object_id}, and if not then try to extract the original URL from {picture}:
if(isset($post['object_id'])) {
echo "http://graph.facebook.com/".$post['object_id']."/picture";
}
elseif(isset($post['picture'])) {
echo urldecode(preg_replace('/&cfs.*/', '', preg_replace('/.*url=/', '', $post['picture'])));
}
else {
echo "no_large_image";
}

ASP.NET Web API, unexpected end of MIME multi-part stream when uploading from Flex FileReference

Following the tutorial found on ASP.NET, implemented a Web API controller method for doing asynchronous file uploads that looks like this:
public Task<HttpResponseMessage> PostFormData()
{
// Check if the request contains multipart/form-data.
if (!Request.Content.IsMimeMultipartContent())
{
throw new HttpResponseException(HttpStatusCode.UnsupportedMediaType);
}
string root = HttpContext.Current.Server.MapPath("~/App_Data");
var provider = new MultipartFormDataStreamProvider(root);
// Read the form data and return an async task.
var task = Request.Content.ReadAsMultipartAsync(provider).
ContinueWith<HttpResponseMessage>(t =>
{
if (t.IsFaulted || t.IsCanceled)
{
Request.CreateErrorResponse(HttpStatusCode.InternalServerError, t.Exception);
}
return Request.CreateResponse(HttpStatusCode.OK);
});
return task;
}
Uploading a file via a standard multipart HTML form works perfectly. However, when another developer attempts to upload a file via multipart form constructed by Flex's FileReference class, an error is thrown:
Unexpected end of MIME multipart stream. MIME multipart message is not complete.
I have no idea if the problem lies in Web API or Flex. I've found some sort of related fixes that had no affect (Multipart form POST using ASP.Net Web API), and more recently this one ("MIME multipart stream. MIME multipart message is not complete" error on webapi upload). If the second link holds true, does anyone know if it's out in the current release of Web API available via Nuget? The discussion was in May, the most recent release from Nuget was August, so I assume this fix was deployed already, and is not the root cause of my issue.
I had the same problem with MVC4, but Will is correct, add a name to your input.....
<input type="file" id="fileInput" name="fileInput"/>
and all the magic is back up and working!
I had the same problem with flex. And below is the code that solved it. Basically I used a custom stream to append the newline that asp.net web api is expecting.
Stream reqStream = Request.Content.ReadAsStreamAsync().Result;
MemoryStream tempStream = new MemoryStream();
reqStream.CopyTo(tempStream);
tempStream.Seek(0, SeekOrigin.End);
StreamWriter writer = new StreamWriter(tempStream);
writer.WriteLine();
writer.Flush();
tempStream.Position = 0;
StreamContent streamContent = new StreamContent(tempStream);
foreach(var header in Request.Content.Headers)
{
streamContent.Headers.Add(header.Key, header.Value);
}
// Read the form data and return an async task.
await streamContent.ReadAsMultipartAsync(provider);
Hope this helps.
Reading through your existing research and following through to the codeplex issue reported it looks like someone else confirmed this issue to still exist in September.
They believe that MVC 4 fails to parse uploads without a terminating "\r\n".
The issue is really simple but extremely hard to fix. The problem is that Uploadify does > not add an "\r\n" at the end of the MultiPartForm message
http://aspnetwebstack.codeplex.com/discussions/354215
It may be worth checking that the Flex upload adds the "\r\n"
For those landing here googling:
Unexpected end of MIME multipart stream. MIME multipart message is not complete.
Reading the request stream more than once will also cause this exception. I struggled with it for hours until I found a source explaining that the request stream only could be read once.
In my case, I combined trying to read the request stream using a MultipartMemoryStreamProvider and at the same time letting ASP.NET do some magic for me by specifying parameters (coming from the request body) for my api method.
Make sure the virtual directory ("~/App_Data" directory as below example) where the image files are first uploaded are physically existance. When you publish the project, it may not be in the output files.
string root = HttpContext.Current.Server.MapPath("~/App_Data");
var provider = new MultipartFormDataStreamProvider(root);
I just removed my headers I was setting on my post method which ended up solving this issue.
The problem is this line:
string root = HttpContext.Current.Server.MapPath("~/App_Data");
It will only work in localhost, you can use HostingEnvironment.MapPath instead in any context where System.Web objects like HttpContext.Current are not available (e.g also from a static method).
var mappedPath = System.Web.Hosting.HostingEnvironment.MapPath("~/SomePath");
See also What is the difference between Server.MapPath and HostingEnvironment.MapPath?
Reference to this answer How to do a Server Map Path.

IMAP - javax.mail. - Fetching Only Body Without Attachment

I am trying to develop an IMAP email client using the javax.mail API. I have been able to contact the servers, fetch emails, attachments and other operations without any problem.
However, we would not want to fetch the attachment until the user wants to view it explit in order to improve the performance. This would mean that we would need information which would indicate if the email has an attachment, if yes, then the filename(s) and the size(s)but would not send me the actual content of the email. However, I was not able to find a method in the API description which would return just the multi-part content.
Is there a way I could fetch just the body and the details of the attachment but not the actual attachment?
Thanks,
Aravind
The method getContent() of javax.Part returns
the content as a Java object. The type
of the returned object is of course
dependent on the content itself. For
example, the object returned for
"text/plain" content is usually a
String object. The object returned for
a "multipart" content is always a
Multipart subclass.
See http://download.oracle.com/javaee/6/api/javax/mail/Part.html#getContent%28%29
If a Message (which is a Part) contains attachments, the getContent method will return a Multipart object. In addition you can know the MIME type of the part with getContentType.
The information that is missing from the documentation is that this MultiPart object returned by getContent is just an empty representation of the Message's structure. The content of each Part of the Multipart will only be fetched from the server when you specifically ask for it. For example, to parse a multipart Message you would do:
if( p.isMimeType("multipart/*") ) {
Multipart mp = (Multipart)p.getContent();
// the content was not fetched from the server
// parse each Part
for (int i = 0; i < mp.getCount(); i++) {
Part inner_part = mp.getBodyPart(i)
if( inner_part.isMimeType("text/plain") ) {
String text = inner_part.getText();
// the content of this Part was fetched from the server
}
}
}
Something else to consider is that for optimal performance you should get the body structure information from the server in batch for the collection of messages you want to parse (see http://download.oracle.com/javaee/6/api/javax/mail/FetchProfile.Item.html#CONTENT_INFO). If not, each
Multipart mp = (Multipart)p.getContent();
will result in a request to the server to fetch the body structure. If the FetchProfile approach is used, the body structure for the collection of messages will be fetched with only one request. You can see the requests to the server in the log if you activate debug mode on the session:
session.setDebug(true);
This said, to get size and filenames of attachments just use Part.getSize() and Part.getFileName(), respectively.

Resources