Uploading Attachments to SharePoint Lists Using SPServices

Easy!For years people have been asking me how they could upload files using SPServices. I’ve dodged the questions every time because I simply didn’t know how.

On a current client project,  I’m building a slick Single Page Application (SPA) to manage tasks in SharePoint 2010. It’s basically a veneer over the clunky out-of-the-box task list experience. Rather than hopping from page to page, the team using the application can accomplish everything they need to do on a single page.

Every project has specific needs, and for this one I’m using KnockoutJS. But the code snippets I give below are generic enough that you should be able to use them in any context with a little thinking.

It’s been going well, but I had been putting off implementing adding attachments because…well, as I said, I didn’t know how.

One of the benefits of the shift from all server side development to a much more significant focus on client side techniques is that some of the bright minds in the SharePoint development community have turned their eyes toward solving these things, too. Usually it’s on SharePoint 2013 or SharePoint Online at Office 365. However, at least now when I search for “SharePoint JavaScript xxx”, there are likely to be some great hits I can learn from.

In this case, there were two excellent posts, one from James Glading and one from Scot Hillier. (See the Resources section below for links.)

The first step is to enable HTML5 capabilities in IE10. This requires two small changes to the master page. This seems simple, but it can have effects that you don’t expect. In other words, don’t consider this just a “no brainer”. You need to plan for the change across your Site Collection and any impacts it may have.

The first change is to switch from using “XHTML 1.0” to “HTML” as the DOCTYPE. This is what “turns on” HTML5.

<!DOCTYPE html>

Then, in this project we set the content meta tag to IE=10 because we want to aim for IE10.

<meta http-equiv="X-UA-Compatible" content="IE=10"/>

If your browser base is different, you can, of course, set this differently. In this project, we will require all users of the application to be running IE10 or greater. Firefox and Chrome have supported the bits of HTML5 we need for so long, that there’s little concern about people who choose to use those browsers.

Once we have those two small changes in the master page (we are using a copy of v4.master with no other customizations at the moment), we “have HTML5”. It’s most obvious because some of the branding flourishes I’ve put in like rounded corners show up in IE10 now, rather than just square boxes.

Once we have HTML5 enabled, we can start to use the File API, a.k.a. the FileReader. This is such a simple little thing, that it’s hard to believe it gives us the capability it does.

<input type="file" id="attachment-file-name"/>

That’s it. That simple HTML gives us a file picker on the page. Depending on your browser, it will look something like this:

File Picker

When you make a file selection with this very familiar widget, you can query the element using jQuery.

var file = $("#attachment-file-name").files[0];

Note that we’re after a single file, so we grab the first object in the file list array. We get an object that looks something like this screen shot from Firebug:

File object from the File Picker

Once we have the file object, we can look at what the Lists SOAP Web Service needs to get in order to upload the file. Again, it’s pretty simple. Here is the list of inputs and output from the MSDN documentation page for the Lists.AddAttachment Method.

Parameters

listName
A string that contains either the title or the GUID for the list.
listItemID
A string that contains the ID of the item to which attachments are added. This value does not correspond to the index of the item within the collection of list items.
fileName
A string that contains the name of the file to add as an attachment.
attachment
A byte array that contains the file to attach by using base-64 encoding.

Return Value

A string that contains the URL for the attachment, which can subsequently be used to reference the attachment.

The first three inputs are straightforward. We pass in the name of the SharePoint list, the ID of the list item, and the name of the file we want to attach. It’s that last input parameter called “attachment” that’s a bit tricky.

When we upload a file via an HTTP POST operation – which is what all of the SOAP Web Services use – we have to pass text. But the files we want to upload are as likely as not to be binary files. That’s where the Base64 format comes in. Believe me, you don’t need to know exactly what the format actually is, but you should probably understand that it enables us to send binary files as text bytes. Those bytes are then decoded on the server end so that the file can be stored in all of its original, pristine glory.

Here’s the code I ended up with, skinnied down as much as possible to make it a little clearer to follow. I draw heavily from Scot and James’ posts for this, but nuanced for SPServices. I’ve also stripped out all of the error handling, etc.

  • First, the getFileBuffer function reads the contents of the file into the FileReader buffer. readAsArrayBuffer is an asynchronous method, so we use a jQuery Deferred (promise) to inform the calling function that the processing is done.
  • The contents of the buffer are then converted from an ArrayBuffer – which is what the FileReader gives us – into a Base64EncodedByteArray. This works by passing the buffer through a Uint8Array along the way.
  • Finally, we use the toBase64String method to convert the SP.Base64EncodedByteArray to a Base64String.

Yeah, that’s a lot to swallow, but again, you don’t really need to understand how it works. You just need to know that it does work.

Finally, we call Lists.AddAttachment using SPServices to do the actual upload.

/* From Scot Hillier's post:
http://www.shillier.com/archive/2013/03/26/uploading-files-in-sharepoint-2013-using-csom-and-rest.aspx */
var getFileBuffer = function(file) {

  var deferred = $.Deferred();
  var reader = new FileReader();

  reader.onload = function(e) {
    deferred.resolve(e.target.result);
  }

  reader.onerror = function(e) {
    deferred.reject(e.target.error);
  }

  reader.readAsArrayBuffer(file);

  return deferred.promise();
};

getFileBuffer(file).then(function(buffer) {
  var bytes = new Uint8Array(buffer);
  var content = new SP.Base64EncodedByteArray(); //base64 encoding
  for (var b = 0; b < bytes.length; b++) {
    content.append(bytes[b]);
  }

  $().SPServices({
    operation: "AddAttachment",
    listName: "Tasks",
    listItemID: taskID,
    fileName: file.name,
    attachment: content.toBase64String()
  });

});

Very cool! And as I tweeted yesterday, far easier than I ever would have expected. Yes, it took me a good chunk of a day to figure out, but it definitely works, and pretty fast, too.

If you use this example as a base, you could fairly easily build out some other file uploading functions. Combined with the other attachment-oriented methods in the Lists Web Services, you can also build the other bits of the attachment experience:

  • GetAttachmentCollection – Returns a list of the lit item’s attachments, providing the full path to each that you can use in a list of links.
  • DeleteAttachment – Once you’ve uploaded an attachment and realized it was the wrong one, this method will allow you to delete it.

Moral of the story: Fear not what you do not know. Figure it out.

Resources

What’s the Story for HTML5 with SharePoint 2010? by Joe Li

Uploading Files Using the REST API and Client Side Techniques by James Glading

Uploading Files in SharePoint 2013 using CSOM and REST by Scot Hillier

Lists.AddAttachment Method

This article was also published on IT Unity on Jun 02, 2014. Visit the post there to read additional comments.

Update 2014-05-28 11:55 GMT-5

Hugh Wood (@HughAJWood) took one look at my code above and gave me some optimizations. Hugh could optimize just about anything; he is *very* good at it. I *think* I can even see the difference with larger files.

getFileBuffer(file).then(function(buffer) {
  var binary = "";
  var bytes = new Uint8Array(buffer);
  var i = bytes.byteLength;
  while (i--) {
    binary = String.fromCharCode(bytes[i]) + binary;
  }
  $().SPServices({
    operation: "AddAttachment",
    listName: "Tasks",
    listItemID: taskID,
    fileName: file.name,
    attachment: btoa(binary)
  });
});

Similar Posts

42 Comments

  1. Can someone please post more details about how you implemented this with readAsDataURL? I am having issues similar to what others here described with files above 15MB failing to load, so I am trying to switch to readAsDataURL as several people have suggested but I can’t get it working. Thanks.

  2. Alright even more weird. I got it to work using readAsDataURL with a relatively small file (~7MB zip file) but if I try it with a larger ~43MB zip file the browser (chrome) freezes for a bit then unlocks the page but the developer tools completely lock up and if you click in it at all it closes and you have to reopen it. Finally with the larger file it doesn’t get attached to the item and I never see any error anywhere as to what happened. Any ideas, I am at a complete loss here?

  3. Thanks for sharing this code – it helped a lot!
    In my case i got an error 500 – internal server error. The list could not be found. Writing a new listitem works – adding an attachment gives the error.
    Then i added webURL: “/sitepath” and now it works fine!

  4. Me again…
    Any idea what to do to avoid version number impact by uploading/deleting multiple files?
    When i upload 4 files to a listitem, the versionnumber jumps 4 counts up. 1 would be enough…

  5. How would you modify this procedure to update a custom column in the document library?

    I have a nintex workflow that will remove any file that does not have a custom column “Document Type” set. So adding the document through an interface like this is what I’m going for, but the workflow would remove it. Unfortunately, we had to put this in place because people were not putting the right kind of documents in the right place or not tagging them and we had to figure out what the documents were after the fact. If I could set this field at the time the document was uploaded then that would avoid this issue. I will also look to see if I can do this from within the workflow. Possibly rename the loaded file to a type that is selected on the page or something like that. Any thoughts?

    Great Work!

      1. i use that for insert

        function base64ArrayBuffer(arrayBuffer) {
        	var base64    = '',
        		encodings = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/',
        		bytes         = new Uint8Array(arrayBuffer),
        		byteLength    = bytes.byteLength,
        		byteRemainder = byteLength % 3,
        		mainLength    = byteLength - byteRemainder,
        		a, b, c, d,
        		chunk;
        	
        	// Main loop deals with bytes in chunks of 3
        	for (var i = 0; i &lt; mainLength; i = i + 3) {
        		// Combine the three bytes into a single integer
        		chunk = (bytes[i] &lt;&lt; 16) | (bytes[i + 1] &lt;&gt; 18 // 16515072 = (2^6 - 1) &lt;&gt; 12 // 258048   = (2^6 - 1) &lt;&gt;  6 // 4032     = (2^6 - 1) &lt;&gt; 2 // 252 = (2^6 - 1) &lt;&lt; 2		
        		// Set the 4 least significant bits to zero
        		b = (chunk &amp; 3)   &lt;&lt; 4 // 3   = 2^2 - 1		
        		base64 += encodings[a] + encodings[b] + &#039;==&#039;
        		} else if (byteRemainder == 2) {
        		chunk = (bytes[mainLength] &lt;&gt; 10 // 64512 = (2^6 - 1) &lt;&gt;  4 // 1008  = (2^6 - 1) &lt;&lt; 4		
        		// Set the 2 least significant bits to zero
        		c = (chunk &amp; 15)    &lt;&lt;  2 // 15    = 2^4 - 1		
        		base64 += encodings[a] + encodings[b] + encodings[c] + &#039;=&#039;
        	}		
        	return base64
        }
        
        	
        function handleUploadFile(e) {
        	e.stopPropagation();
        	e.preventDefault();
        	var files, i, f;
        	files = e.target.files
        	for (i = 0, f = files[i]; i != files.length; ++i) {
        		var reader, name, extension;
        		reader = new FileReader();
        		reader.filename = f.name;
        		extension = f.name.split(&#039;.&#039;).pop();
        		reader.onload = function (e) {
        			var data = this.result;
        			var File = base64ArrayBuffer(data);
        			$().SPServices({
        			               operation: &quot;AddAttachment&quot;,
        			               listName: &#039;List&#039;,
        			               listItemID: &#039;ID&#039;,
        			               fileName: &#039;FileName&#039;,
        			               attachment: File,
        			               completefunc: function(xData, Status) {
        			                            $(xData.responseXML).SPFilterNode(&quot;z:row&quot;).attr(&quot;ows_ID&quot;);
        			               }
        			 });
        		};
        		reader.readAsArrayBuffer(f);
        	}
        }
        
    1. @Jorge:

      That’s an ocean I haven’t crossed. Depending on the other parts of your environment, that could get really messy. If you’re on SharePoint 2013+ or Office 365, you have better methods at your disposal.

      M.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.