I'm attempting to upload a file using Mediafire's API. Its not clear to me what they expect in the body of the message. I'm attempting to follow the API described in https://www.mediafire.com/developers/upload.php#upload
My understanding is that some parameters are passed in a query string as part of the URL. I'm passing the session_token on the URL.
I set an HTTP header for the file size, x-filesize.
I'm setting the method to POST and sending to base url (before the query string) http://www.mediafire.com/api/upload/upload.php
Its not clear to me what should be in the body. I tried including the pure data from the file being uploaded. I've also tried adding more data to make it look like multipart form.
In either case I'm getting no response at all from the server. I'm doing this in objective-c on a Mac. The NSURLConnection request comes back with nil response and nil error. I'm using
NSData *responseData = [NSURLConnection sendSynchronousRequest:request returningResponse:&response error:&error];
The Mediafire upload API is expecting parameters to be send in the URL as the query string, GET style.
The data from the file is included in the body as a multipart form. If the body is improperly constructed, Mediafire may not return a relevant error.
Headers specified in the API are sent as headers in the body.
The 'path' parameter only appears to be honored if a uploadkey is also provided. No error is provided indicating if the path parameter was honored.
Maximum file size appears to be around 4MB.
Related
I am trying to create the upload PUT request for the OneDrive API. It's the large file "resumable upload" version which requires the createUploadSession.
I have read the Microsoft docs here: As a warning the docs are VERY inaccurate and full of factual errors...
The docs simply say:
PUT
https://sn3302.up.1drv.com/up/fe6987415ace7X4e1eF866337Content-Length:
26Content-Range: bytes 0-25/128 <bytes 0-25 of the
file>
I am authenticated and have the upload session created, however when I pass the JSON body containing my binary file I receive this error:
{ "error": {
"code": "BadRequest",
"message": "Property file in payload has a value that does not match schema.", .....
Can anyone point me at the schema definition? Or explain how the JSON should be constructed?
As a side question, am I right in using "application/json" for this at all? What format should the request use?
Just to confirm, I am able to see the temp file created ready and waiting on OneDrive for the upload, so I know I'm close.
Thanks for any help!
If you're uploading the entire file in a single request then why do you use upload session when you can use the simple PUT request?
url = https://graph.microsoft.com/v1.0/{user_id}/items/{parent_folder_ref_id}:/{filename}:/content
and "Content-Type": "text/plain" header and in body simply put the file bytes.
If for some reason I don't understand you have to use single-chunk upload session then:
Create upload session (you didn't specified any problems here so i'm not elaborating)
Get uploadUrl from createUploadSession response and send PUT request with the following headers:
2.1 "Content-Length": str(file_size_in_bytes)
2.2 "Content-Range": "bytes 0-{file_size_in_bytes - 1}/{file_size_in_bytes}"
2.3 "Content-Type": "text/plain"
Pass the file bytes in body.
Note that in the PUT request the body is not json but simply bytes (as specified by the content-type header.
Also note that max chuck size is 4MB so if your file is larger than that, you will have to split into more than one chunks.
Goodlcuk
Let's say we gave a generic POST request with Python's requests.
req = requests.post('http://someapi.someservice.com', files=files)
req will be a Response object. In my case, the .content of the response can be very, very large so I do not wish to read it all into memory. Luckily, requests provides an iterator .iter_content that allows iteration over the contents. My question is, though, does req contain all contents of the response already (and as such everything is already read into memory), or does calling .content and as such .iter_content initiate a download which really fetches the content? This is important, because if assigning the POST request to a variable already reads the Response's content into memory, then of course using .iter_content makes no difference.
You will need to set the stream parameter to True in your request in order to avoid downloading the entire content of the response into the response object:
req = requests.post('http://someapi.someservice.com', files=files, stream=True)
Excerpt from the documentation of Body Content Workflow:
By default, when you make a request, the body of the response is
downloaded immediately. You can override this behaviour and defer
downloading the response body until you access the Response.content
attribute with the stream parameter... You can further control the workflow by use of the Response.iter_content() and Response.iter_lines() methods.
I'm using RESTKit to get data from a rest-api.
This is the URL i set for my request, here's the log just before the request goes off.
2014-04-03 15:51:10.186 xxx[35745:60b] Just sent URL: /api/dspObjGetNewsList?action=coverage&count=30&start=0&open=0&user=xxx&unique=36027&type=all&country=Sweden,global&division=Strategic Industries,Regional Sales and Service,Automotive
Then i log the reponse URL.
- (void)request:(RKRequest*)request didLoadResponse:(RKResponse*)response {
NSLog(#"xxx: %d, url: %#", [response statusCode], response.URL);
And i get this?
xxx: response code: 200, url: url/api/dspObjGetNewsList?unique=26791&type=all&division=Strategic%20Industries%2CRegional%20Sales%20and%20Service%2CAutomotive&user=xxx&action=coverage&open=0&country=Sweden%2Cglobal&count=30&start=0
Why am i getting a different URL in my response? Does RESTKit modify my url?
Have you configured HTTPClient properly? Use - (id)initWithHTTPClient:(AFHTTPClient *)client method to configure HTTPClient. For instance:
AFHTTPClient *HTTPClient = [AFHTTPClient clientWithBaseURL:[NSURL URLWithString:#"www.url.com"]];
Your first log appears to be a raw string of the URL. The second log appears to be the % escaped URL version of that string. This isn't a RestKit thing, it's a URL loading system thing. Certain characters need to be escaped so that they are valid for use in a URL.
For example, your original string has a number of spaces in it. This isn't allowed in a URL and each must be changed to %20.
Why the parameters change order isn't clear - it depends on how you created the string and supplied the parameters to RestKit. But, the order doesn't matter to the processing so you shouldn't need to worry about it.
I have an iOS app that is using RestKit 0.20.1 to pull data from a server. The server at this time can send JSON formatted data but is not able to receive JSON formatted data.
This is where my problem is. POST requests require HTTP Body and the server is set up to receive XML. I have implemented the RKXMLReaderSerialization add on so I can receive XML but I can't find any current way of sending an XML formatted HTTP Body with RestKit.
This question "Send post request in XML format using RestKit " is what I was looking for but the answer from Imran Raheem is now (as far as I can tell) obsolete due to changes in RestKit.
I am using this method for the POST
[[RKObjectManager sharedManager] postObject:nil path:#"/rest/search?ip=255.255.255.0" parameters:search success:nil failure:nil];
Here is what the RestKit objectManager says about the postObject method
/**
Creates an `RKObjectRequestOperation` with a `POST` request for the given object, and enqueues it to the manager's operation queue.
#param object The object with which to construct the object request operation. If `nil`, then the path must be provided.
#param path The path to be appended to the HTTP client's base URL and used as the request URL. If nil, the request URL will be obtained by consulting the router for a route registered for the given object's class and the `RKRequestMethodPOST` method.
#param parameters The parameters to be reverse merged with the parameterization of the given object and set as the request body.
#param success A block object to be executed when the object request operation finishes successfully. This block has no return value and takes two arguments: the created object request operation and the `RKMappingResult` object created by object mapping the response data of request.
#param failure A block object to be executed when the request operation finishes unsuccessfully, or that finishes successfully, but encountered an error while parsing the resonse data. This block has no return value and takes two arguments:, the created request operation and the `NSError` object describing the network or parsing error that occurred.
If I have the MIMEType set to JSON the Trace shows my request.body is being populated like so request.body={"Search":"Trending"}.
However if I set the MIMEType to XML the Trace shows the request.body=(null)
Here is the line I use to change the MIMEType
[RKObjectManager sharedManager].requestSerializationMIMEType = RKMIMETypeJSON;
I am pretty new to iOS and Objective-C so I may be setting up the NSDictionary that is used in the parameters of the `postObject' method wrong. Here it is just in case....
NSArray *objects =[NSArray arrayWithObjects:#"trending", nil];
NSArray *keys =[NSArray arrayWithObjects: #"Search",nil];
NSDictionary *params = [NSDictionary dictionaryWithObjects:objects forKeys:keys];
self.search=params;
Any help would be greatly appreciated! Given that I am new Snippets are especially helpful!
Oh and BTW if anyone can point me to REST method that accepts JSON input I would gladly pass it on to my server guy so I can just avoid XML all together.
As feared and suggested in other posts on this topic writing XML is not supported in version of RestKit 0.20.1. Here is a link to my conversation with Blake Watters on the subject.
https://github.com/RestKit/RestKit/issues/1430#issuecomment-19150316
I never did clear up why the request.body= (null) when the MIMEType is set to XML.
The manager of the server has agreed to set it up to receive JSON. That is how I plan to work around this.
I'm using NSURLConnection with sendAsynchronousRequest method (and handing the data in block).
Using above method, I'm downloading a video file from server and saving locally in cache folder.
This is working fine if the URL is valid. If the URL is invalid (that is the video is not available at given URL), it is still saving a file with the file name that I'm giving. NSError is nil and NSData object is not nil. How can I check whether the URL is valid or catch the error if URL is not there?
Check the NSURLResponse in the block, if it is an http request it is really an NSHTTPURLResponse. Check the statusCode to see the HTTP error. It may be a 404, 302, etc.
Also check the returned data, convert it to a string and see if there is anything interesting there. Sometimes there is a redirect in the returned html.