This document is a preliminary draft of a specification for the Selection API and selection related functionality. It replaces a couple of old sections of the HTML specification, the selection part of the old DOM Range specification.

This document defines APIs for selection, which allows users and authors to select a portion of a document or specify a point of interest for copy, paste, and other editing operations.

Any feedback or discussion of this specification should be sent to the public-webapps mailing list with "[selection-api]" as the Subject header prefix (archive).

This specification is based on the HTML Editing API specification edited by Aryeh Gregor.

Background

IE9 and Firefox 6.0a2 allow arbitrary ranges in the selection, which follows what this spec originally said. However, this leads to unpleasant corner cases that authors, implementers, and spec writers all have to deal with, and they don't make any real sense. Chrome 14 dev and Opera 11.11 aggressively normalize selections, like not letting them lie inside empty elements and things like that, but this is also viewed as a bad idea, because it takes flexibility away from authors.

So I changed the spec to a made-up compromise that allows some simplification but doesn't constrain authors much. See discussion. Basically it would throw exceptions in some places to try to stop the selection from containing a range that had a boundary point other than an Element or Text node, or a boundary point that didn't descend from a Document.

But this meant getRangeAt() had to start returning a copy, not a reference. Also, it would be prone to things failing weirdly in corner cases. Perhaps most significantly, all sorts of problems might arise when DOM mutations transpire, like if a boundary point's node is removed from its parent and the mutation rules would place the new boundary point inside a non-Text/Element node. And finally, the previously-specified behavior had the advantage of matching two major implementations, while the new behavior matched no one. So I changed it back.

See bug 15470. IE9, Firefox 12.0a1, Chrome 17 dev, and Opera Next 12.00 alpha all make the range initially null.

This specification defines conformance criteria that apply to a single product: the user agent that implements the interfaces that it contains.

Implementations that use ECMAScript to implement the APIs defined in this specification must implement them in a manner consistent with the ECMAScript Bindings defined in the Web IDL specification [[!WEBIDL-1]], as this specification uses that specification and terminology.

Definition

Every document ([[!DOM4]]) with a browsing context ([[!HTML5]]) has a unique selection associated with it.

This is a requirement of the HTML spec. IE9 and Opera Next 12.00 alpha seem to follow it, while Firefox 12.0a1 and Chrome 17 dev seem not to. See Mozilla bug, WebKit bug.

This one selection must be shared by all the content of the document (though not by nested documents), including any editing hosts in the document. Editing hosts (defined in [[!HTML5]]) that are not inside a document cannot have a selection

Each selection can be associated with a single range (defined in [[!DOM4]]). When there is no range associated with the selection, the selection is empty. The selection must be initially empty.

A document's selection is a singleton object associated with that document, so it gets replaced with a new object when Document.open() is called. See bug 15470. IE9 and Opera Next 12.00 alpha allow the user to reset the range to null after the fact by clicking somewhere; Firefox 12.0a1 and Chrome 17 dev do not. We follow Gecko/WebKit, because it lessens the chance of getRangeAt(0) throwing.

Once a selection is associated with a given range, it must continue to be associated with that same range until this specification requires otherwise.

For instance, if the DOM changes in a way that changes the range's boundary points, or a script modifies the boundary points of the range, the same range object must continue to be associated with the selection. However, if the user changes the selection or a script calls addRange(), the selection must be associated with a new range object, as required elsewhere in this specification.

If the selection's range is not null and is collapsed ([[!DOM4]]), then the caret position must be at that range's boundary point. When the selection is not empty, this specification does not define the caret position; user agents should follow platform conventions in deciding whether the caret is at the start of the selection, the end of the selection, or somewhere else.

This short-changes Mac users. See bug 13909.

Each selection has a direction, forwards, backwards, or directionless. If the user creates a selection by indicating first one boundary point of the range and then the other (such as by clicking on one point and dragging to another), and the first indicated boundary point is after ([[!DOM4]]) the second, then the corresponding selection must initially be backwards. If the first indicated boundary point is before ([[!DOM4]]) the second, then the corresponding selection must initially be forwards. Otherwise, it must be directionless.

Each selections also have an anchor and a focus. If the selection's range is null, its anchor and focus are both null. If the selection's range is not null and its direction is forwards, its anchor is the range's start, and its focus is the end. Otherwise, its focus is the start and its anchor is the end.

Selection interface

Selection interface provides a way to interact with the selection associated with each document.

        interface Selection {
          readonly attribute Node? anchorNode;
          readonly attribute unsigned long anchorOffset;
          readonly attribute Node? focusNode;
          readonly attribute unsigned long focusOffset;
          readonly attribute boolean isCollapsed;
          readonly attribute unsigned long rangeCount;
          readonly attribute DOMString type;
          Range getRangeAt(unsigned long index);
          void addRange(Range range);
          void removeRange(Range range);
          void removeAllRanges();
          void empty();
          void collapse(Node? node, optional unsigned long offset = 0);
          void setPosition(Node? node, optional unsigned long offset = 0);
          void collapseToStart();
          void collapseToEnd();
          void extend(Node node, optional unsigned long offset = 0);
          void setBaseAndExtent(Node anchorNode, unsigned long anchorOffset, Node focusNode, unsigned long focusOffset);
          void selectAllChildren(Node node);
          [CEReactions]
          void deleteFromDocument();
          boolean containsNode(Node node, optional boolean allowPartialContainment = false);
          stringifier DOMString();
        };
      
anchorNode

The attribute must return the anchor node ([[!DOM4]]) of the context object ([[!DOM4]]), or null if the anchor is null.

anchorOffset

The attribute must return the anchor offset ([[!DOM4]]) of the context object, or 0 if the anchor is null.

focusNode

The attribute must return the focus node of the context object, or null if the anchor is null.

focusOffset

The attribute must return the focus offset of the context object, or 0 if the focus is null.

isCollapsed

The attribute must return true if and only if the anchor and focus are the same (including if both are null). Otherwise it must return false.

rangeCount

The attribute must return 0 if the context object is empty, and must return 1 otherwise.

type

The attribute must return "None" if the context object is empty, "Caret" if the context object's range is collapsed, and "Range" otherwise.

getRangeAt

The method must throw an IndexSizeError ([[!DOM4]]) exception if index is not 0, or if the context object is empty. Otherwise, it must return a reference to (not a copy of) the context object's range.

Thus subsequent calls of this method returns the same range object if nothing has removed the context object's range in the meantime. In particular, getSelection().getRangeAt(0) === getSelection().getRangeAt(0) evaluates to true if the selection is not empty.

IE9 and Firefox 4.0 return the same object every time, as the spec says. Chrome 12 dev and Opera 11.10 return a different object every time.

addRange

The method must follow these steps:

  1. If the root ([[!DOM4]]) of the range's boundary points are not the document associated with context object, abort these steps.
  2. If rangeCount is not 0, abort these steps.
  3. Set the context object's range to range by a strong reference (not by making a copy).

Since range is added by reference, subsequent calls to getRangeAt(0) returns the same object, and any changes that a script makes to range after it is added must be reflected in the selection, until something else removes or replaces the context object's range. In particular, the selection will contain b as opposed to a after running the following code: var r = document.createRange(); r.selectNode(a); getSelection().addRange(r); r.selectNode(b);

At Step 2, Chrome 58 and Edge 25 do nothing. Firefox 51 gives you a multi-range selection. At least they keep the exisiting range.

At Step 3, Chrome 58 and Firefox 51 store a reference, as described here. Edge 25 stores a copy. Firefox 51 changes its selection if the range is modified.

removeRange

The method must make the context object empty by disassociating its range if the context object's range is range. Otherwise, it must throw a NotFoundError.

removeAllRanges

The method must make the context object empty by disassociating its range if the context object has an associated range.

empty

The method must be an alias, and behave identically, to removeAllRanges().

collapse

The method must follow these steps:

  1. If node is null, this method must behave identically as removeAllRanges() and abort these steps.
  2. The method must throw an IndexSizeError exception if offset is longer than node's length ([[!DOM4]]) and abort these steps.
  3. If node's root is not the document associated with the context object, abort these steps.
  4. Otherwise, let newRange be a new range.
  5. Set ([[!DOM4]]) the start and the end of newRange to (node, offset).
  6. Set the context object's range to newRange.
setPosition

The method must be an alias, and behave identically, to collapse().

collapseToStart

The method must throw InvalidStateError ([[!DOM4]]) exception if the context object is empty. Otherwise, it must create a new range, set both its start and end to the start of the context object's range, and then set the context object's range to the newly-created range.

For collapseToStart/End, IE9 mutates the existing range, while Firefox 9.0a2 and Chrome 15 dev replace it with a new one. The spec follows the majority and replaces it with a new one, leaving the old Range object unchanged.

collapseToEnd

The method must throw InvalidStateError exception if the context object is empty. Otherwise, it must create a new range, set both its start and end to the end of the context object's range, and then set the context object's range to the newly-created range.

extend

The method must follow these steps:

  1. If node's root is not the document associated with the context object, abort these steps.
  2. If the context object is empty, throw an InvalidStateError exception and abort these steps.
  3. Let oldAnchor and oldFocus be the context object's anchor and focus, and let newFocus be the boundary point (node, offset).
  4. Let newRange be a new range.
  5. If node's root is not the same as the context object's range's root, set newRange's start and end to newFocus.
  6. Otherwise, if oldAnchor is before or equal to newFocus, set newRange's start to oldAnchor, then set its end to newFocus.
  7. Otherwise, set newRange's start to newFocus, then set its end to oldAnchor.
  8. Set the context object's range to newRange.
  9. If newFocus is before oldAnchor, set the context object's direction to backwards. Otherwise, set it to forwards.

Reverse-engineered circa January 2011. IE doesn't support it, so I'm relying on Firefox (implemented extend() sometime before 2000) and WebKit (implemented extend() in 2007). I'm mostly ignoring Opera, because gsnedders tells me its implementation isn't compatible. Firefox 12.0a1 seems to mutate the existing range. IE9 doesn't support extend(), and it's impossible to tell whether Chrome 17 dev or Opera Next 12.00 alpha mutate or replace, because getRangeAt() returns a copy anyway. Nevertheless, I go against Gecko here, to be consistent with collapse().

setBaseAndExtent

The method must follow these steps:

  1. If anchorOffset is longer than anchorNode's length or if focusOffset is longer than focusNode's length, throw an IndexSizeError exception and abort these steps.
  2. If the roots of anchorNode or focusNode are not the document associated with context object, abort these steps.
  3. Let anchor be the boundary point (anchorNode, anchorOffset) and let focus be the boundary point (focusNode, focusOffset).
  4. Let newRange be a new range.
  5. If anchor is before focus, set the newRange's start to anchor and its end to focus. Otherwise, set them to focus and anchor respectively.
  6. Set the context object's range to newRange.
  7. If focus is before anchor, set context object's direction to backwards. Otherwise, set it to forwards
selectAllChildren

The method must follow these steps:

  1. If node's root is not the document associated with the context object, abort these steps.
  2. Let newRange be a new range and nodeLength be the length of node.
  3. Set newRange's start to (node, 0).
  4. Set newRange's end to (node, nodeLength).
  5. Set the context object's range to newRange.
  6. Set the context object's direction to forwards.

Based mostly on Firefox 9.0a2. It has a bug that I didn't reproduce, namely that if you pass a Document as the argument, the end offset becomes 1 instead of the number of children it has. It also throws a RangeException instead of DOMException, because its implementation predated their merging.

IE9 behaves similarly but with glitches. It throws "Unspecified error." if the node is detached or display:none, and apparently in some random other cases too. It throws "Invalid argument." for detached comments (only!). Finally, if you pass it a comment, it seems to select the whole comment, unlike with text nodes.

Chrome 16 dev behaves as you'd expect given its Selection implementation. It refuses to select anything that's not visible, so it's almost always wrong. Opera 11.50 just does nothing in all my tests, as usual.

The new range replaces any existing one, doesn't mutate it. This matches IE9 and Firefox 12.0a1. (Chrome 17 dev and Opera Next 12.00 alpha can't be tested, because getRangeAt() returns a copy anyway.)

deleteFromDocument

The method must invoke deleteContents() ([[!DOM4]]) on the context object's range if the context object is not empty. Otherwise the method must do nothing.

This is the one method that actually mutates the range instead of replacing it. This matches IE9 and Firefox 12.0a1. (Chrome 17 dev and Opera Next 12.00 alpha can't be tested, because getRangeAt() returns a copy anyway.)

containsNode

The method must return false if the context object is empty or if node's root is not the document associated with the context object.

Otherwise, if allowPartialContainment is false, the method must return true if and only if start of its range is before or visually equivalent to the first boundary point in the node and end of its range is after or visually equivalent to the last boundary point in the node.

If allowPartialContainment is true, the method must return true if and only if start of its range is before or visually equivalent to the first boundary point in the node or end of its range is after or visually equivalent to the last boundary point in the node.

stringifier
See W3C bug 10583.

See also nsISelection.idl from Gecko. This spec doesn't have everything from there yet, in particular selectionLanguageChange() and containsNode() are missing. They are missing because I couldn't work out how to define them in terms of Ranges.

Originally, the Selection interface was a Netscape feature. The original implementation was carried on into Gecko (Firefox), and the feature was later implemented independently by other browser engines. The Netscape implementation always allowed multiple ranges in a single selection, for instance so the user could select a column of a table However, multi-range selections proved to be an unpleasant corner case that web developers didn't know about and even Gecko developers rarely handled correctly. Other browser engines never implemented the feature, and clamped selections to a single range in various incompatible fashions.

This specification follows non-Gecko engines in restricting selections to at most one range, but the API was still originally designed for selections with arbitrary numbers of ranges. This explains oddities like the coexistence of removeRange() and removeAllRanges(), and a getRangeAt() method that takes an integer argument that must always be zero.

All of the members of the Selection interface are defined in terms of operations on the range object (if any) represented by the object. These operations can raise exceptions, as defined for the Range ([[!DOM4]]) interface; this can therefore result in the members of the Selection interface raising exceptions as well, in addition to any explicitly called out below.

Extensions to Other Interfaces

This specification extends several interfaces to provide entry points to the interfaces defined in this specification.

Extensions to Document interface

          partial interface Document {
            Selection ? getSelection();
          };
        
getSelection

The method must return the selection associated with context object if the context object has an associated browsing context, and it must return null otherwise.

If we create a Document object with no browsing context (say via document.implementation.createHTMLDocument("") and call getSelection on it), IE9 seems to return a different Selection object. Firefox 12.0a1 and Opera Next 12.00 alpha return the same object as for the current window. Chrome 17 dev returns null. See discussion. There's no meaningful selection associated with such a document, so we follow WebKit and require returning null.

Extensions to Window interface

          partial interface Window {
            Selection? getSelection();
          };
        
getSelection

The method must invoke and return the result of getSelection on the context object's document ([[!HTML5]]) property.

Extensions to GlobalEventHandlers

          partial interface GlobalEventHandlers {
            attribute EventHandler onselectstart;
            attribute EventHandler onselectionchange;
          };
        
onselectstart

The attribute must be an event handler IDL attribute for the selectstart event supported by all HTML elements, Document objects, and Window objects.

onselectionchange

The attribute must be an event handler IDL attribute for the selectionchange event supported by all HTML elements, Document objects, and Window objects.

User Interactions

The user agent should allow the user to change the selection associated with the active document (defined in [[!HTML5]]). If the user makes any modification to a selection, the user agent must create a new range with suitable start and end of the range ([[!DOM4]]) and associate the selection with this new range (not modify the existing range), and set update selection's direction to forwards if the start is before or equal to the end, backwards if if the end is before the start, or directionless if the start and the end cannot be ordered due to the platform convention.

The user agent must not make a selection empty if it was not already empty in response to any user actions (e.g. clicking on a non-editable region).

See bug 15470. IE9 and Opera Next 12.00 alpha allow the user to reset the range to null after the fact by clicking somewhere; Firefox 12.0a1 and Chrome 17 dev do not. I follow Gecko/WebKit, because it lessens the chance of getRangeAt(0) throwing.

selectstart event

When the user agent is about to associate a new range newRange to the selection in response to a user initiated action, the user agent must fire ([[!DOM4]]) an event with the name selectstart, which bubbles and is cancelable, at the node associated with the boundary point of newRange's start prior to changing the selection if the selection was previously empty or the previously associated range was collapsed.

If the event is canceled, the user agent must not change the selection.

The user agent must not fire the event when the user agent sets the selection empty.

selectionchange event

When the selection is dissociated with its range, associated with a new range or the associated range's boundary point is mutated either by the user or the content script, the user agent must queue a task to fire an event with the name selectionchange, which does not bubble and is not cancelable, at the document associated with the selection.

Acknowledgements

Many thanks to