6.1.1. Constructor Parameters

The Blob() constructor can be invoked with the parameters below:

A  blobParts  sequence

which takes any number of the following types of elements, and in any order:

• ArrayBuffer [TypedArrays] elements.

• ArrayBufferView [TypedArrays] elements.

• Blob elements.

• DOMString [WebIDL] elements.

An  optional  BlobPropertyBag

which takes one member:

• type, the ASCII-encoded string in lower case representing the media type of the Blob. Normative conditions for this member are provided in the Blob constructor steps.

6.4.1. The slice method

The slice method returns a new Blob object with bytes ranging from the optional start parameter upto but not including the optional end parameter, and with a typeattribute that is the value of the optional contentType parameter. It must act as follows :

1. Let O be the Blob object on which the slice method is being called.

2. The optional start parameter is a value for the start point of a slice call, and must be treated as a byte-order position, with the zeroth position representing the first byte. User agents must process slice with start normalized according to the following:

   1. If the optional start parameter is not used as a parameter when making this call, let relativeStart be 0.

   2. If start is negative, let relativeStart be max((size + start), 0)).

   3. Else, let relativeStart be min(start, size).

3. The optional end parameter is a value for the end point of a slice call. User agents must process slice with end normalized according to the following:

   1. If the optional end parameter is not used as a parameter when making this call, let relativeEnd be size.

   2. If end is negative, let relativeEnd be max((size + end), 0)

   3. Else, let relativeEnd be min(end, size)

4. The optional contentType parameter is used to set the ASCII-encoded string in lower case representing the media type of the Blob. User agents must process theslice with contentType normalized according to the following:

   1. If the contentType parameter is not provided, let relativeContentType be set to the empty string .

   2. Else let relativeContentType be set to contentType and run the substeps below:

      1. If relativeContentType contains any characters outside the range of U+0020 to U+007E, then set relativeContentType to the empty string and return from these substeps.
      2. Convert every character in relativeContentType to lower case using the "Converting a string to ASCII lowercase" algorithm.

5. Let span be max((relativeEnd - relativeStart), 0).

6. Return a new Blob object S with the following characteristics:

   1. S consists of span consecutive bytes from O, beginning with the byte at byte-order position relativeStart.

   2. S.size = span.

   3. S.type = relativeContentType.

      Note

      The type t of a Blob is considered a parsable MIME type if the ASCII-encoded string representing the Blob object's type, when converted to a byte sequence, does not return undefined for the parse MIME type algorithm [MIMESNIFF].

   4. Let the snapshot state of S be set to the snapshot state of O.

    // obtain input element through DOM
    
    var file = document.getElementById('file').files[0];
    if(file)
    {
      // create an identical copy of file
      // the two calls below are equivalent
      
      var fileClone = file.slice(); 
      var fileClone2 = file.slice(0, file.size);
      
      // slice file into 1/2 chunk starting at middle of file
      // Note the use of negative number
      
      var fileChunkFromEnd = file.slice(-(Math.round(file.size/2)));
      
      // slice file into 1/2 chunk starting at beginning of file
      
      var fileChunkFromStart = file.slice(0, Math.round(file.size/2));
      
      // slice file from beginning till 150 bytes before end
      
      var fileNoMetadata = file.slice(0, -150, "application/experimental");      
    }
    

6.4.2. The close method

Calling the close method must permanently neuter the original Blob object. This is an irreversible and non-idempotent operation; once a Blob has been neutered, it cannot be used again; dereferencing a Blob URL bound to a Blob object on which close has been called results in a network error. A neutered Blob must have a size of 0.

8.5.1. Multiple Reads

The FileReader interface makes available three asynchronous read methods - readAsArrayBuffer, readAsText, and readAsDataURL, which read files into memory. If multiple concurrent read methods are called on the same FileReader object, user agents must throw an InvalidStateError [DOM4] on any of the read methods that occur whenreadyState = LOADING.

8.5.2. The result attribute

On getting, the result attribute returns a Blob's data as a DOMString, or as an ArrayBuffer [TypedArrays], or null, depending on the read method that has been called on theFileReader, and any errors that may have occurred. The list below is normative for the result attribute and is the conformance criteria for this attribute:

• On getting, if the readyState is EMPTY (no read method has been called) then the result attribute must return null.

• On getting, if an error in reading the File or Blob has occurred (using any read method), then the result attribute must return null.

• On getting, if the readAsDataURL read method is used, the result attribute must return a DOMString that is a Data URL [DataURL] encoding of the File or Blob's data.

• On getting, if the readAsText read method is called and no error in reading the File or Blob has occurred, then the result attribute must return a string representing the File or Blob's data as a text string, and should decode the string into memory in the format specified by the encoding determination as a DOMString.

• On getting, if the readAsArrayBuffer read method is called and no error in reading the File or Blob has occurred, then the result attribute must return an ArrayBuffer[TypedArrays] object.

If a read is successful, the result attribute must return a non-null value only after a progress event (see also [ProgressEvents]) has fired, since all the read methodsaccess Blob data asynchronously. Tasks are queued to update the result attribute as Blob data is made available.

8.5.3. The readAsDataURL(blob) method

When the readAsDataURL(blob) method is called, the user agent must run the steps below (unless otherwise indicated).

1. If readyState = LOADING throw an InvalidStateError exception [DOM4] and terminate this algorithm.

   Note: The readAsDataURL() method returns due to the algorithm being terminated.

2. If the blob has been neutered through the close method, throw an InvalidStateError exception [DOM4] and terminate this algorithm.

   Note: The readAsDataURL() method returns due to the algorithm being terminated.

3. If an error occurs during reading of the blob parameter set readyState to DONE and set result to null. Proceed to the error steps.

4. If no error has occurred, set readyState to LOADING

5. Fire a progress event called loadstart.

6. Return the readAsDataURL() method, but continue to process the steps in this algorithm.

7. Make progress notifications.

8. When the blob has been fully read into memory queue a task to do the following:

   1. Set readyState to DONE.

   2. Set the result attribute with the blob as a DataURL [DataURL] after it has been fully read into memory; on getting, the result attribute returns the (complete) data of blob as a Data URL [DataURL].

      • Use the blob's type attribute as part of the Data URL if it is available in keeping with the Data URL specification [DataURL].

      • If the type attribute is not available on the blob return a Data URL without a media-type. [DataURL].

        Data URLs that do not have media-types [RFC2046] must be treated as plain text by conforming user agents. [DataURL].

   3. Fire a progress event called load.
   4. Unless readyState is LOADING fire a progress event called loadend. If readyState is LOADING do NOT fire loadend.

9. Terminate this algorithm.

8.5.4. The readAsText(blob, label) method

The readAsText() method can be called with an optional parameter, label, which is a DOMString argument that represents the label of an encoding [Encoding Specification]; if provided, it must be used as part of the encoding determination used when processing this method call.

When the readAsText(blob, label) method is called (the label argument is optional), the user agent must run the steps below (unless otherwise indicated).

1. If readyState = LOADING throw an InvalidStateError [DOM4] and terminate these steps.

   Note: The readAsText() method returns due to the algorithm being terminated.

2. If the blob has been neutered through the close method, throw an InvalidStateError exception [DOM4] and terminate this algorithm.

   Note: The readAsText() method returns due to the algorithm being terminated.

3. If an error occurs during reading the blob parameter, set readyState to DONE and set result to null. Proceed to the error steps.

4. If no error has occurred, set readyState to LOADING

5. Fire a progress event called loadstart.

6. Return the readAsText() method, but continue to process the steps in this algorithm

7. Make progress notifications.

8. When the blob has been read into memory fully, queue a task to do the following:

   1. Set readyState to DONE

   2. Set the result attribute to the fully read blob, represented as a string in a format determined by the encoding determination.

   3. Fire a progress event called load.
   4. Unless readyState is LOADING fire a progress event called loadend. If readyState is LOADING do NOT fire loadend.

9. Terminate this algorithm.

8.5.5. The readAsArrayBuffer(blob) method

When the readAsArrayBuffer(blob) method is called, the user agent must run the steps below (unless otherwise indicated).

1. If readyState = LOADING throw an InvalidStateError exception [DOM4] and terminate these steps.

   Note: The readAsArrayBuffer() method returns due to the algorithm being terminated.

2. If the blob has been neutered through the close method, throw an InvalidStateError exception [DOM4] and terminate this algorithm.

   Note: The readAsArrayBuffer() method returns due to the algorithm being terminated.

3. If an error occurs during reading the blob parameter, set readyState to DONE and set result to null. Proceed to the error steps.

4. If no error has occurred, set readyState to LOADING

5. Fire a progress event called loadstart.

6. Return the readAsArrayBuffer() method, but continue to process the steps in this algorithm.

7. Make progress notifications.

8. When the blob has been read into memory fully, queue a task to do the following:

   1. Set readyState to DONE

   2. Set the result attribute to the value of the fully read blob as an ArrayBuffer [TypedArrays] object.

   3. Fire a progress event called load.

   4. Unless readyState is LOADING fire a progress event called loadend. If readyState is LOADING do NOT fire loadend.

9. Terminate this algorithm.

8.5.6. Error Steps

1. Fire a progress event called error. Set the error attribute; on getting, the error attribute must be a a DOMError object that indicates the kind of file error that has occurred.

2. Unless readyState is LOADING, fire a progress event called loadend. If readyState is LOADING do NOT fire loadend.

3. Terminate the algorithm for any read method.

   Note: The read method returns due to the algorithm being terminated.

8.5.7. The abort() method

When the abort() method is called, the user agent must run the steps below:

1. If readyState = EMPTY or if readyState = DONE set result to null and terminate this overall set of steps without doing anything else.

2. If readyState = LOADING set readyState to DONE and result to null.

3. If there are any tasks from the object's FileReader task source in one of the task queues, then remove those tasks.

4. Terminate the algorithm for the read method being processed.

5. Fire a progress event called abort

6. Unless readyState is LOADING fire a progress event called loadend. If readyState is LOADING do NOT fire loadend.

8.5.8. Blob Parameters

The three asynchronous read methods, the three synchronous read methods, and createObjectURL take a mandatory Blob parameter. This section defines this parameter.

blob

This is a Blob argument and must be a reference to a single File in a FileList or a Blob object not obtained from the file system that is in scope of the global object from which the method call was made.

8.5.9. Determining Encoding

When reading blob objects using the readAsText() read method, the following encoding determination steps must be followed:

1. Let encoding be null

2. If the label argument is present, set encoding to the result of "getting an encoding" using the Encoding Specification [Encoding Specification] for label.

3. If the "getting an encoding" steps above return failure, then set encoding to null.

4. If encoding is null, and the blob argument's type attribute is present, and it uses a Charset Parameter [RFC2046], set encoding to the result of "getting an encoding" using the Encoding Specification [Encoding Specification] for the portion of the Charset Parameter that is a label of an encoding.

   Example

   If blob has a type attribute of text/plain;charset=UTF-8 then "getting an encoding" is run using UTF-8 as the label. Note that user agents must parse and extract the portion of the Charset Parameter that constitutes a label of an encoding.

5. If the "getting an encoding" steps above return failure, then set encoding to null.

6. If encoding is null, then set encoding to utf-8.

7. "Decode" [Encoding Specification] this blob using fallback encoding encoding, and return the result. On getting, the result attribute of the FileReader object returns a string in encoding format. The synchronous readAsText method of the FileReaderSync object returns a string in encoding format.

8.5.10. Events

When this specification says to make progress notifications for a read method while the read method is processing, queue a task to fire a progress event called progressat the FileReader object about every 50ms or for every byte read into memory, whichever is least frequent. At least one event called progress must fire before load is fired, and at 100% completion of the read operation; if 100% of blob can be read into memory in less than 50ms, user agents must fire a progress event called progress at completion.

When this specification says to fire a progress event called e (for some ProgressEvent e at a FileReader reader), the following are normative:

• The progress event e does not bubble. e.bubbles must be false [DOM4]

• The progress event e is NOT cancelable. e.cancelable must be false [DOM4]

• The term "fire an event" is defined in DOM Core [DOM4]. Progress Events are defined in Progress Events [ProgressEvents].

9.1.1. Constructors

When the FileReaderSync() constructor is invoked, the user agent must return a new FileReaderSync object.

In environments where the global object is represented by a WorkerGlobalScope object, the FileReaderSync constructor must be available.

10.1.2. The readAsText method

When the readAsText(blob, label) method is called (the label argument is optional), the following steps must be followed:

1. If an error occurs during reading of the blob parameter, throw the appropriate exception. Terminate these overall steps.

2. If no error has occurred, read blob into memory. Return the data contents of blob using the encoding determination algorithm.

9.1.3. The readAsDataURL method

When the readAsDataURL(blob) method is called, the following steps must be followed:

1. If an error occurs during reading of the blob parameter, throw the appropriate exception. Terminate these overall steps.

2. If no error has occurred, read blob into memory. Return the data contents of blob as a Data URL [DataURL]

   • Use the blob's type attribute as part of the Data URL if it is available in keeping with the Data URL specification [DataURL] .
   • If the type attribute is not available on the blob return a Data URL without a media-type. [DataURL].

     Data URLs that do not have media-types [RFC2046] must be treated as plain text by conforming user agents. [DataURL].

9.1.4. The readAsArrayBuffer method

When the readAsArrayBuffer(blob) method is called, the following steps must be followed:

1. If an error occurs during reading the blob parameter, throw the appropriate exception. Terminate these overall steps.

2. If no error has occurred, read blob into memory. Return the data contents of blob as an ArrayBuffer [TypedArrays]

11.3.1. The Opaque String

Opaque strings must NOT include any reserved characters from [RFC3986] without percent-encoding them. Opaque strings must be globally unique. Such strings should only use characters in the ranges U+002A to U+002B, U+002D to U+002E, U+0030 to U+0039, U+0041 to U+005A, U+005E to U+007E [Unicode], and should be at least 36 characters long. UUID is one potential option available to user agents for use with Blob URLs as opaque strings, and their use is strongly encouraged. UUIDs are defined in [RFC4122]. For an ABNF of UUID, see Appendix A.

11.5.1. 200 OK

This response [RFC2616] must be used if the request has succeeded, namely the blob: URL has been requested with a GET, and satisfies the lifetime requirement. If this response code is used, the user agent must also use a Content-Type header [RFC2616] with a value equal to the Blob object's type attribute. See blob: protocol examples.

11.5.2. Network Errors

Responses that do not succeed with a 200 OK must act as if a network error has occurred [XHR]. Network errors must be used when:

• Any request method other than GET is used to dereference the URL.

• The request does not satisfy the lifetime requirement.

• The underlying resource has changed, moved, been deleted or has become invalid.

• The permissions surrounding the underlying resource do not permit access.

• Any error conditions occur with respect to accessing or reading the file.

• A security error has occurred.

11.5.3. Request and Response Headers

This section provides sample exchanges between web applications and user agents using the blob: protocol. A request can be triggered using HTML markup of the sort <img src="blob:550e8400-e29b-41d4-a716-446655440000">, after a web application calls URL.createObjectURL on a given blob, which returns blob:550e8400-e29b-41d4-a716-446655440000 to dereference that blob. These examples merely illustrate the protocol; web developers are not likely to interact with all the headers, but thegetAllResponseHeaders() method of XMLHttpRequest, if used, will show relevant response headers [XHR].

	GET 550e8400-e29b-41d4-a716-446655440000
	

  
  200 OK
  Content-Type: image/jpeg
  Content-Length: 21897
  
  ....
	

11.6.1. Methods and Parameters

The  createObjectURL static method

Returns a unique Blob URL each time it is called on a valid blob argument, which is a non-null Blob in scope of the global object's URL property from which this static method is called. This method must act as follows:

1. If called with a Blob argument that is NOT valid, then user agents must return null.

2. If this method is called with a valid Blob argument, user agents must run the following sub-steps:

   1. Return a unique Blob URL that can be used to dereference the blob argument.
   2. Add an entry to the Blob URL Store for this Blob URL.

The  createFor static method

Returns a unique Blob URL each time it is called on a valid blob argument. Blob URLs created with this method are said to be auto-revoking since user-agents are responsible for the revocation of Blob URLs created with this method, subject to the lifetime stipulation for Blob URLs. This method must act as follows:

1. If called with a Blob argument that is NOT valid, then user agents must return null.

2. If this method is called with a valid Blob argument, user agents must run the following steps:

   1. Return a unique Blob URL that can be used to dereference the blob argument.
   2. Add an entry to the Blob URL Store for this Blob URL.
   3. Add an entry in the Revocation List for this Blob URL.

Example

In the example below, after obtaining a reference to a Blob object (in this case, a user-selected File from the underlying file system), the static methodURL.createObjectURL() is called on that Blob object.

ECMAScript

	var file = document.getElementById('file').files[0];
	if(file){
	  blobURLref = window.URL.createObjectURL(file);
	  myimg.src = blobURLref;

    ....

	}
	

The  revokeObjectURL static method

Revokes the Blob URL provided in the string url argument, and removes the corresponding entry from the Blob URL Store. This method must be processed as follows:

1. If the url refers to a Blob that is both valid and in the same origin of the global object's URL property on which this static method was called, user agents must remove the entry that for the URL on which URL.revokeObjectURL was called from the Blob URL Store; subsequent attempts to dereference this URL must return a network error response code.
2. If the url refers to a Blob that is NOT valid OR if the value provided for the url argument is not a Blob URL OR if the url argument refers to a Blob that is NOT in the same origin as the global object's URL property, this method call does nothing. User agents may display a message on the error console.

The url argument to the revokeObjectURL method is a Blob URL string.

Example

In the example below, window1 and window2 are separate, but in the same origin; window2 could be an iframe [HTML] inside window1.

ECMAScript

	myurl = window1.URL.createObjectURL(myblob);
	window2.URL.revokeObjectURL(myurl);
	

Since window1 and window2 are in the same origin, the URL.revokeObjectURL call ensures that subsequent dereferencing of myurl results in a the user agent acting as if a network error has occurred.

11.7. Lifetime of Blob URLs

Each unit of related similar origin browsing contexts [HTML] must maintain a Blob URL Store which is a list of Blob URLs that have been created using URL.createObjectURLor URL.createFor, and the blob resource that each Blob URL refers to. Additionally, each unit of related similar origin browsing contexts must maintain a Revocation List, consisting of Blob URLs that have been created with the URL.createFor method. When this specification says to add an entry to the Blob URL Store the user-agent must add the Blob URL created using URL.createObjectURL or URL.createFor, and a reference to the blob it refers to, to the list. When this specification says to add an entry to the The Revocation List, user agents must add any Blob URL created with the URL.createFor method; the Revocation List is only for auto-revoking Blob URLs. TheRevocation List and the Blob URL Store must be processed together as follows:

1. Add revoking all the URLs in the Revocation List to the global script cleanup jobs list, such that the following occurs when processing the global script cleanup jobs:

   1. For each Blob URL in the Revocation List, remove the corresponding entry from the Blob URL Store. Subsequent attempts to dereference any removed Blob URL results in a network error.

   2. When all the Blob URLs in the Revocation List have had their corresponding entries in the Blob URL Store removed, remove all the Blob URLs in the Revocation List.

2. This specification adds an additional unloading document cleanup step [HTML]: user agents must remove all Blob URLs from the Blob URL Store within that document. If these Blob URLs are dereferenced, user agents must respond with a network error.