I have a very large video and I need to chunk this video to upload it to Dropbox.
I tried to use NSData, but because this file is too large, my application always crashes, so I don't know what I can do now.
For smaller videos, I used this:
NSData(contentsOfURL: self.newAsset.URL)!.subdataWithRange(NSMakeRange(0, 10000000))
and I didn't have any problem with that, but when the video is too large I have an error:
Cannot allocate memory
So, what can I do to chunk the data of large videos?
For best practice go with NSURLSession if you want to implement custom otherwise lots for third party library are there like RESTKit or AFNetworking. For NSURLSession the session NSURLSession supports three types of tasks: data tasks, download tasks, and upload tasks. All it support the background uploads/downloads as well. source(apple developer)
Data tasks send and receive data using NSData objects. Data tasks are intended for short, often interactive requests from your app to a server. Data tasks can return data to your app one piece at a time after each piece of data is received, or all at once through a completion handler.
Download tasks retrieve data in the form of a file, and support background downloads while the app is not running.
Upload tasks send data in the form of a file, and support background uploads while the app is not running.
Image Source raywenderlich.com
You should use video file url to upload large data using NSURLSession
- (NSURLSessionUploadTask *)uploadTaskWithRequest:(NSURLRequest *)request fromFile:(NSURL *)fileURL;
Related
I'm using Alamofire to download large files in my ios project.
When i cancel a download request (that is currently downloading a large file to disk) - this request produces resumeData with the data downloaded so far - but I would like a responseURL with the URL of the file with the partially downloaded data. I plan to use the responseURL serializer at the end to never load the entire data in memory - but if I want to suspend and resume downloads - then this is forcing me to load the data in memory.
there's a fileURL in the download request - but the documentation states that this is populated after the download has finished.
any pointers/suggestions would be appreciated?
As explained on GitHub, resumeData only includes the data necessary to resume a download, not the actual download data itself, so it's perfectly safe to keep it around in memory. You can parse the value to get the URL of the partially downloaded data, but it's not a formatted encoding, so it's not really appropriate for Alamofire to parse it directly.
My app downloads and display GIFs from the Internet in UIImageViews. The GIFs don't need to be cached/saved at all. However, the app takes up lots of storage on the iPhone with Documents & Data being enormous.
I'd like to be able to clear the Documents and Data folder each time the user opens the app. Is this possible to do with Swift?
The extension you use uses the following code:
Data(contentsOf: url)
https://developer.apple.com/documentation/foundation/nsdata/1413892-init
This method will cache stuff in the system cache.
If you want to keep using that extension you should modify that code to use this one instead:
init(contentsOf:options:)
https://developer.apple.com/documentation/foundation/nsdata/1407864-init
And pass "uncached" as an option to avoid the cache.
https://developer.apple.com/documentation/foundation/nsdata.readingoptions/1412417-uncached
However I think that extension is poorly written, for instance, you should only use this method to load SHORT local files, for bigger files a stream is needed, for network requests you shouldn't be using it at all.
Important
Don't use this synchronous method to request network-based
URLs. For network-based URLs, this method can block the current thread
for tens of seconds on a slow network, resulting in a poor user
experience, and in iOS, may cause your app to be terminated. Instead,
for non-file URLs, consider using the
dataTask(with:completionHandler:) method of the NSURLSession class.
See URL Session Programming Guide for details.
Data(contentsOf: url, options: .uncached)
I want to use NSURLSession to upload photos to cloud server(like OneDrive) and I hope the app can do works in background.
But NSURLSession just support "fromFile" not "fromData" in background mode, and I don't know how to get the NSURL of photos, so I write the photo's data into a file...
I think this is not a good way.
From the comments above, it seems like you'll be uploading several photos at a time and for that a "fromData" approach does not seem like a good idea.
Firstly, high resolution photos have several megabytes in size. If your loading a 5mb photo into memory, using NSData, you're probably fine. But loading 10 photos would then mean you'd need 50mb of memory to hold them all.
Secondly, uploading large chunks of data, such as photos, is also a computationally intensive task. You should most definitely not try to upload them all at the same time. This means you're going to have to build an upload queue that only starts uploading the next photo once the last one has finished. What happens then if internet connectivity is lost midway through the process? And what happens if the user closes the app all of the sudden?
Thus, I'd suggest you grab all the images' data and write them to a temporary folder somewhere. Then you start the upload process, which would find a file on the temporary folder, upload it, delete it once the upload finishes successfully and start over with the next file.
This way you're minimising your app's memory footprint, making it more resilient in the face of the many adversities an app may face (no internet connection, being prematurely closed by the user and so on) and also abiding by the APIs that are provided to you.
I have a problem regarding the best approach to make an App that has to download and show pdf's, It is fed by a JSON that has links to 147 pdf files, sized between 1 and 2 MB.
Questions:
What is the best approach to download all the files to an iPad?
Shall I use AFNetworking 2.0?
Is NSFileManager the way to save all the files?
Problems I may encounter:
With an asynchronous download, if lost connection or no more space on the iPad, what are the counter mesures?
Are there tutorials or examples that deal with this situation?
Sorry for all the questions but I'm new to this.
Best Regards.
What is the best approach to download all the files to an iPad?
This is really broad as #rmaddy suggested. Specific questions are more easily answered. There are lots of ways you could download a file via an HTTP request to your device each with pros/cons depending on your situation.
Shall I use AFNetworking 2.0?
Sure. You'll get no argument from me. This is a widely used and solid API to interface with HTTP-based resources.
Is NSFileManager the way to save all the files?
Yes. NSFileManager is the class you use to read/write files from/to your app's sandbox.
With an asynchronous download, if lost connection or no more space on the iPad, what are the counter measures?
I'm not 100% certain so I can't speak to exactly what happens in this case. AFNetworking may provide some help by writing to a temporary file during a download, etc....
Are there tutorials or examples that deal with this situation?
I have a sample project on Github that shows a table of files that you can download. You can watch the progress of your downloads, pause each request, resume and cancel as well. When you're done you can view the file you downloaded. It uses AFNetworking and might be useful to you:
https://github.com/chefnobody/StreamingDownloadTest
When downloading large files, the main counsel would be to avoid trying to load these into memory as you download them. Instead, make sure you stream them directly to persistent storage. In terms of handling space-specific errors, just make sure you check NSError objects that are returned to you in completion handlers or the appropriate delegate methods.
If using AFNetworking, you can specify the outputStream of the AFURLConnectionOperation to reference a NSOutputStream that you create, referencing some path in your persistent storage.
See Memory pressure issue while downloading multiple files using AFNetworking for example.
Alternatively, you can use NSURLSession (whether via AFNetworking or you do it yourself) and instantiate a NSURLSessionDownloadTask, which does the same sort of thing.
Google "NSURLSessionDownloadTask example" and you'll find tons of references. The block-based rendition of downloadTaskWithURL is incredibly simple. To do background downloads is a little more complicated and requires delegate-based implementation (see Downloading Files and Handling iOS Background Activity sections of URL Loading System Programming Guide: Using NSURLSession or watch the WWDC 2013 video, What’s New in Foundation Networking.)
Either way, you avoid some of the memory consumption challenges associated with downloading large files.
I am currently developing a application which has to be able to show offline videos which need to be downloaded first.
The problem was that these videos can be bigger that the memory that I can allocate to my application. So parts that are downloaded have to be saved immediately instead of saved in a NSData object. I'm hearing conflicting stories on whether or not RESTKit should work, and ASIHTTPRequest seems to be deprecated.
I will follow the suggestion from this thread as it seems to be the best option.
NSURLConnection download large file (>40MB)
Consider using NSURLConnection to download the video file and write the data directly to a file (NSFileHandle).
One advantage of using this method is that the NSURLConnection didReceiveData delegate method is continuously called as data is received, so you can update a progress bar.
Check out AFNetworking for network managing. I am not sure if they have video downloading, but the framework works great for images and other types of downloads that I have down before.
Without explaining all the hasle with dealing with HTTP responses by chunks and streams I would recommend using AFDownloadRequestOperation. It supports resuming downloads and has callbacks for showing download progress. I love it and use it in most of my projects.
P.S. It uses AFNetworking, which is a great framework for making all kinds of HTTP requests.