不同浏览器对键盘事件的不同认知

原文地址：点击打开链接
Javascript Madness Intro

JavaScript Madness: Keyboard Events

Jan Wolter

1. Introduction

This document summarizes the results of some browser tests done while attempting to implement key stroke handling code in JavaScript. It documents inconsistancies in the way different browsers implement keyboard events.

The tests were originally done with the intention of learning just enough to write the code I needed to write. Coverage has expanded considerably since then, but the results here still are not comprehensive or authoritative and do not cover all aspects of keyboard event handling.

This data is based on tests of many, many browsers over many, many years, but is not comprehensive. I update it periodically as new browsers cross my desktop. The browser versions most recently tested are:

 WindowsMacintoshLinuxInternet Explorer9.0.8112.164215.2-Firefox5.0
(Gecko 5.0)5.0.1
(Gecko 5.0.1)4.0
(Gecko 2.0)Safari4.0.4
(WebKit 531.21.10)5.0.2
(WebKit 533.18.1)-Chrome3.0.195.33
(WebKit 532.0)-4.0.249.43 Beta
(WebKit 532.5)Opera10.539.1010.10Konqueror--4.3.1

The script used to collect the test results reported here is available at http://unixpapa.com/js/testkey.html. I mostly report only what I can test myself, so this report is necessarily incomplete:

• It primarily focuses on standard US keyboards. There are a huge range of other keyboard layouts in use in the world, which include not only different characters, but standard characters in different places. So, for example, many UK keyboards have a 3 £ key and a # ~ key, neither of which exists on US keyboards. I don't know what keycodes keys like these send.

• It does not cover the behavior of keypad keys on the Macintosh, because none of my Macs have keypads.

This document will usually refer to "Gecko" instead of "Firefox" and to "WebKit" instead of "Safari" or "Chrome". That's because browser behavior usually depends on the rendering engine, and different browsers that use the same rendering engine work the same. See the Layout Engine page for more information, including mappings of layout engine versions to browser versions.

Previous versions of this document included coverage of the iCab 3 browser, but iCab has switched to using the WebKit rendering engine, and so presumably behaves exactly like Safari. Since it is unlikely that many web developers will want to go out of their way to support iCab 3, that material has been removed from this document and archived in aseparate report on iCab 3.

2. Event Triggering

In all recent browsers, pressing a key triggers a series of Javascript events which can be captured and handled. These events, however, were not defined by any standard until DOM3 which few browsers have yet implemented.

There is strong agreement across all browsers about which events should be sent and what order they should be sent in when a key is pressed:

BrowserEvents sent when normal key is typedWebKit ≥ 525keydown
keypress
textInput
 
keyupAll Otherskeydown
keypress
 
keyup

Windows versions of Opera have a bit of buggy behavior: when you type the +, -, *, or / keys on the keypad, then two keypress events are triggered instead of one. This has been observed on Opera 11 and Opera 8.5. I don't know how long this bug has been around.

The keydown event occurs when the key is pressed, followed immediately by the keypress event, and possibly the textInput event. Then the keyup event is generated when the key is released.

To understand the difference between keydown and keypress, it is useful to distinguish between "characters" and "keys". A "key" is a physical button on the computer's keyboard. A "character" is a symbol typed by pressing a button. On a US keyboard, hitting the 4 key while holding down the Shift key typically produces a "dollar sign" character. This is not necessarily the case on every keyboard in the world. In theory, the keydown and keyup events represent keys being pressed or released, while thekeypress event represents a character being typed. In practice, this is not always the way it is implemented.

textInput is a new event defined by the the DOM3 standard. So far, only the WebKit browsers supports it. textInput is a replacement and generalization of keypress (which is deprecated in DOM3). It is supposed to fire whenever text is input, whether by keyboard or not (it could be spoken text or cut/pasted text).

There are some oddities to textInput to be aware of. When setting up event handlers for it, you must use addEventListener() or attachEvent() to set up the event handler. Older style methods won't work. Also you must capitalize it correctly when you pass the event name into the set up function. It must be "textInput" not "textinput".

2.1. Events Triggered by Special Keys

In addition to all the normal keys used to input ASCII characters, keyboards typically have many special purpose keys that do other things. These do not necessarily generate the same events as normal keys, and they show less consistency across browsers.

"Modifier keys" are one class of special keys. They include keys like Shift, Control and Alt, that don't send characters, but modify the characters sent by other keys. For nearly all modern browsers, both keydown and keyup events are triggered by modifier keys, but keypress events are not. This is consistant with their being "key" events not "character" events.

However, Konqueror and some older browser versions do have different behaviors:

BrowserEvents sent when modifier keys are typedGecko ≥ 1.7 
Internet Explorer
WebKit ≥ 525
Opera ≥ 10.10keydown
 
keyupOpera ≤ 9.50
Konquerorkeydown
keypress
 
keyupWebKit < 525
Gecko 1.6no events sent

Note that textInput is never fired for modifier keys, since they do not result in text entry.

Most browsers treat the Caps Lock key the same as any other modifier key, sending keydown when it is depressed and keyup when it is released, but there are exceptions. Gecko browsers generate a keypress event for Caps Lock even though they don't for other modifier keys. Macintosh versions of Safari 3 get really clever: each time you strike and release the Caps Lock key, only one event is triggered, and it is keydown if you turning on caps-lock mode and keyup if you are turning it off. Safari does not do this withNum Lock.

There are many other special keys on a typical keyboard that do not normally send characters. These include the four arrow keys, navigation keys like Home and Page Up, special function keys like Insert and Delete, and the function keys F1 through F12. Internet Explorer and WebKit 525 seem to classify all of these with the modifier keys, since they generate no text, so in those browsers there is no keypress event for them, only keyup and keydown. Many other browsers, like Gecko, do generate keypress events for these keys, however.

Old versions of WebKit had a bug that caused two identical keyup events to be triggered when arrow keys and other special keys were released. I know this existed in WebKit 312 and I know it was fixed in WebKit 525, but I don't know when it was fixed.

Standard Windows keyboards typically have two Start keys and a Menu key, while Apple keyboards have two Apple keys. I'm not going to attempt to describe the behavior of those keys in detail here. They are very inconsistent across browsers, don't exist on all keyboards, and they frequently have default actions that cannot be disabled. As such, Javascript programmers would be well advised to stay away from them.

If NumLock is off, and you hit keypad number key while holding Shift down, then Windows systems trigger some extra events. Windows browsers pretend that the Shift key was released before the key was typed, and then pressed again after it was released, and they trigger keyup, keydown and (in some browsers) keypress events to indicate this. Linux systems don't do this. I don't know if Macintoshes do.

2.2. Events Triggered on Auto-Repeat

If a key is held down long enough it typically auto-repeats, and some additional events will be triggered on each autorepeat. On Macintosh and Linux systems, modifier keys usually don't auto-repeat, but on Windows systems they do (which seems weird to me). In most browsers, an autorepeat is sensibly treated as a character event, but not a key event, so it triggers a keypress but not a keydown or keyup. But, of course, there is some variation:

BrowserEvents triggered on each autoreapeatnormal keysspecial keysInternet Explorer (Windows)keydown
keypresskeydownWebKit ≥ 525keydown
keypress
textInputkeydownGecko (Windows)keydown
keypressGecko (Some Linuxs) 
Gecko (Macintosh) 
WebKit < 525
Konqueror 
Opera keypress onlyGecko (Oher Linuxs)keyup 
keydown 
keypressInternet Explorer (Macintosh)no events triggered

Gecko's behavior seems to be different on different versions of Linux. On some versions of Linux, mostly newer versions, it generates extra events, in a manner only previously seen on iCab 3. I don't know exactly what makes the difference.

2.3. Suppressing Default Event Handling

If you are installing your own handlers for key events, then sometimes you won't want the browser default action to occur (such as having the character appear in a text entry area). To prevent this, you typically have the event handler return false, and maybe call event.preventDefault() and event.stopPropagation() if they are defined. But on which event handler must you suppress defaults? This, of course, varies from browser to browser.

BrowserWhich event handlers need to suppress defaults to prevent key from appearing in text boxInternet Explorer
Gecko
Konqueror 4.3either keydown or keypressWebKiteither keydown, keypress or textInputOpera
Konqueror 3.5keypressKonqueror 3.2keydown

Suppressing defaults on the keydown event has some odd side effects on some browsers, in that it may prevent some other events from firing. Apparantly, triggering further events is taken to be part of the default action of the keydown event in these browsers.

BrowserSide effect suppressing defaults on keydownGecko
WebKit < 525
OperaNo changeInternet Explorerkeypress event never occurs.
keyup event works normally.WebKit ≥ 525keypress and textInput events never occur.
keyup event works normally.Konquerorkeypress event only occurs on auto repeats.
keyup event works normally.

Note that WebKit also prevents textInput from firing if keypress suppresses defaults. This makes sense, since suppressing the default action on either keydown or keypressprevents text entry, so since there is no text input, there should be no textInput event. The DOM3 standards say that keyup should still occur if the default action onkeydown is suppressed, but textInput should not.

In Konqueror 4.3.1, I noticed a brand new weirdness. If you don't suppress the default action on keyup then you get two keyup events. I also seemed to sometimes get duplicatekeydown and keypress events if defaults weren't suppressed on either keydown or keypress.

Most applications will either use only keypress or use only keyup/keydown, so this all works out pretty well in most browsers. If you are handling keypress and want to suppress default handling of the key, return false from that handler. If you are handling keydown/keyup and want to suppress defaults, install a keypress handler that does nothing except return false.

2.4. Event Triggering Summary

To give a clearer side by side comparison, suppose we press the Shift key, then press the A key, holding it down long enough to auto-repeat just once, then release A, and the release Shift. The events we see are shown below for various browsers. Events marked in red do not occur if there is a keydown handler that returns false.

 Internet Explorer
(Windows)Gecko
(Windows)Gecko
(Linux/Mac)
Opera≥10.10WebKit≥525WebKit<525Opera≤9.50KonquerorInternet Explorer
(Mac)Shift pressedkeydown
 keydown
 keydown
 keydown
  keydown
keypresskeydown
keypresskeydown
 A pressedkeydown
keypresskeydown
keypresskeydown
keypresskeydown
keypress
textInputkeydown
keypresskeydown
keypresskeydown
keypresskeydown
keypressA autorepeats 
keydown
keypress 
keydown
keypress 
 
keypress 
keydown
keypress
textInput 
 
keypress 
 
keypress 
 
keypress A releasedkeyupkeyupkeyupkeyupkeyupkeyupkeyupkeyupShift releasedkeyupkeyupkeyupkeyup keyupkeyupkeyup

I used to exclaim here about no two browsers being alike here, but progress is being made. The newer versions of WebKit are extremely close to IE, differing only in being the first to support the new textInput event, and Opera now behaves almost identically with Linux/Mac versions of Gecko.

3. Identifying Keys

When you catch a keyboard event, you may wish to know which key was pressed. If so, you may be asking too much. This is a very big mess of browser incompatibilities and bugs.

3.1. Classic Values Returned on Key Events

The keydown and keyup events should return a code identifying a key, not a code identifying a character. It is not obvious how to do this. ASCII codes don't really suffice, since the same key can generate different characters (if combined with shift or control), and the same character can be generated by different keys (such as the numbers on the keyboard and the numbers on the keypad). Different browsers use different ways of assigning numeric codes to the different keys. We will call these "Mozilla keycodes", "IE keycodes", "Opera keycodes" and "psuedo-ASCII codes" and we'll explain them in more detail below.

Not only do the browsers differ in what values they return, they differ in where they return them. Three different properties of the event object may be used to return them. They are event.keyCode, event.which and event.charCode.

keydown and keyup events event.keyCodeevent.whichevent.charCodeIE <9.0 (Windows)IE keycodeundefinedundefinedInternet Explorer (Mac)IE keycodeundefinedextended ASCII codeIE ≥ 9.0
WebKit ≥ 525IE keycodeIE keycodezeroWebKit < 525IE keycodeIE keycodeASCII code if ASCII character,
zero otherwiseGeckoMozilla keycodeMozilla keycodezeroOpera ≥ 9.50 (all platforms)
Opera 7 (Windows)Mozilla keycode except keypad and branded keys give Opera keycodesMozilla keycode except keypad and branded keys give Opera keycodesundefinedOpera 8.0 to 9.27 (Windows)Opera keycodeOpera keycodeundefinedOpera < 9.50 (Linux & Macintosh)Pseudo-ASCII codePseudo-ASCII codeundefinedKonqueror 4.3Pseudo-ASCII codePseudo-ASCII codezeroKonqueror 3.5Pseudo-ASCII codePseudo-ASCII code if key has an ASCII code,
zero otherwisezeroKonqueror 3.2Pseudo-ASCII codePseudo-ASCII codeundefined

In version 9.50, Opera abandoned Opera keycodes and Pseudo-ASCII keycodes in favor of Mozilla keycodes for most keys (thus reverting to the behavior of Windows Opera 7). WebKit has modified their Konqueror-derived code to use IE keycodes, and I expect Konqueror will follow. Thus there seems to be a convergences on the IE and Mozilla keycodes, which are pretty similar. This is kind of encouraging.

On keydown and keyup, the event objects also have flags that indicate which modifier keys were being pressed when the key was typed. These are:

   event.shiftKey   event.ctrlKey   event.altKey   event.metaKey

These all have true or false values. According to the DOM 3 standard, on the Macintosh, the Option key should activate event.altKey and the Command key should activateevent.metaKey. These attributes seem to work correctly on all modern browsers tested, except event.metaKey is undefined in all versions IE. There is some freakishness in obsolete browsers that can probably be ignored these days. In Macintosh versions of IE, the Command key sets event.ctrlKey and the Control key does nothing. In Netscape 4, none of these attributes existed and the event.modifiers attribute needed to be used instead.

One would think that if a key is typed when Caps Lock is on, then event.shiftKey would be true, but this is not the case in any browser tested. There is also a lot of inconsistency in the values these flags take on the keydown and keyup events actually associated with pressing and releasing the modifier keys, but I can't imagine anyone would care enough to justify documenting the details.

3.2. Classic Values Returned on Character Events

For keypress events, it is pretty clear that the ASCII code of the typed character should be returned, and pretty much all browsers do that.

But what if there is no ASCII code associated with the key? Arrow keys and keys like Page Down and F1 don't have ASCII codes. We call these "special" keys in contrast to the "normal" keys that have ASCII codes. Note that Esc, Backspace, Enter, and Tab are "normal" because they have ASCII codes.

When keypress events are generated for special keys, the browser needs to return some non-ASCII value to indicate which key ways pressed. We'll see that various different browsers do this in different ways.

Some browsers avoid this problem by not generating keypress events for special keys. A good case can be made that this is the right thing to do, since these keystrokes are arguably not character events. But such arguments are weakened by the arbitrariness of the division between normal and special keys. Why should the keyboard Backspace key have a keypress event, but not the keypad Delete key? Is Tab really fundamentally different than right arrow?

keypress events event.keyCodeevent.whichevent.charCodeIE < 9.0 (Windows)normal:ASCII codeundefinedundefinedspecial:no keypress events for special keysIE (Mac)normal:ASCII codeundefinedASCII codespecial:no keypress events for special keysGeckonormal:zeroASCII codeASCII codespecial:Mozilla keycodezerozeroIE ≥ 9.0
WebKit ≥ 525normal:ASCII codeASCII codeASCII codespecial:no keypress events for special keysWebKit < 525normal:ASCII codeASCII codeASCII codespecial:extended ASCII codeextended ASCII codeextended ASCII codeOpera ≥ 10.50 (all platforms)normal:ASCII codeASCII codeundefinedspecial:Mozilla keycode, except keypad and branded keys give Opera keycodeszeroundefinedOpera ≥ 9.50 (all platforms)
Opera 7 (Windows)normal:ASCII codeASCII codeundefinedspecial:Mozilla keycode, except keypad and branded keys give Opera keycodeszero for arrows, function keys, PageUp, PageDown
same as event.keyCode otherwiseundefinedOpera 8.0 to 9.27 (Windows)normal:ASCII codeASCII codeundefinedspecial:Opera keycodezero for arrows, function keys, PageUp and PageDown,
same as event.keyCode otherwiseundefinedOpera < 9.50 (Linux & Macintosh)normal:ASCII codeASCII codeundefinedspecial:Opera keycodezero for arrows, function keys, PageUp and PageDown,
same as event.keyCode otherwiseundefinedKonqueror 4.3normal:ASCII codeASCII codeASCII codespecial:Pseudo-ASCII codePseudo-ASCII codezeroKonqueror 3.5normal:ASCII codeASCII codeASCII codespecial:Pseudo-ASCII codezerozeroKonqueror 3.2normal:ASCII codeASCII codeundefinedspecial:no keypress events for special keys

The traditional method to distinguish special keys from normal keys on keypress events is to first check event.which. If it is undefined or non-zero, then the event is from a normal key, and the ASCII code for that key is in event.keyCode. If it is defined as zero, the the event is from a special key, and the keycode is in event.keyCode. This works for almost every browser, but there are two exceptions:

• The newest version of Konqueror that I have tested, version 4.3.1, returns non-zero event.which values for all special keys. The only way to distinguish an up arrow from an ampersand is to check event.charCode.
• Versions of Opera before 10.50 messes up by returning non-zero event.which values for four special keys (Insert, Delete, Home and End).

So, I guess with this new botched version of Konqueror, we have to make our tests more complex. If neither event.which nor event.charCode is defined as zero, then it is a normal key event.

The DOM 3 standard makes a half-hearted attempt to suggest standards for these "legacy" attributes. Konqueror's annoying non-zero event.which values for special keys actually kind of comply with what it suggests. Except by that standard, keypress shouldn't be firing at all for special keys.

If you are actually interested in special key events, then probably you should be hooking your code into keydown and keyup, which work more consistently across browsers. So the main practical importance of this is that keypress handlers should not treat event.keyCode as an ASCII code if either event.which or event.charCode is defined as zero.

The flags event.shiftKey, event.ctrlKey, event.altKey and event.metaKey are typically defined on keypress events, just as they are on keydown and keyup events. WebKit seems to define event.keyIdentifier on keypress as well, but I wouldn't count on future browsers doing that.

On textInput events, event.data contains the text that was input. On key inputs, this is typically a one character string. Since there are no textInput events on special keys, we don't have to worry about such cases. In WebKit, at least, none of the old traditional values mentioned above are defined on textInput events.

3.3. Key Code Values

Now for the actual values being returned for different keys. Some people refer to the Mozilla/IE keycodes as "scan codes". Scan codes are returned from the keyboard, and are converted to ASCII by the keyboard drivers. They typically vary with different kinds of keyboards. As far as I can tell, these are NOT scan codes. They vary with the browser type rather more than with keyboard type, and they don't seem to match with any keyboard scan codes that I've seen documented.

The table below lists values for keys commonly found on US keyboards. If a single value is given, then that value is sent whether the Shift key is held down or not. If two values x/y are given, the the first is sent when the key is unshifted, and the second is sent when the key is shifted. (Ideally there should be slashed values only in the ASCII column of this table, since the other codes are all used only on keyup and keydown events, which are key events not character events.)

Keys highlighted in green are consistent across all browsers tested. Keys highlighted in yellow are consistent for recent versions of IE, Gecko, WebKit and Opera. Keys highlighted in red aren't.

KeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptionsAlphabetic keys
A to Z97/65 to 122/90ASCII code of uppercase version of the letter
65 to 90Space3232323232Enter1313131313Tab99999Esc2727272727Backspace88888Modifier KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptionsShift-16161616Linux Opera < 9.0: 0Control-17171717Linux Opera < 9.0: 0
Mac Opera: 0Alt-18181818Linux Opera < 9.0: 0Caps Lock-20202020Linux Opera: 0Num Lock-144144144144Linux Opera < 9.50: 0
Win Opera < 9.00: 0Keyboard Number KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptions1 !49/3349494949/33Mac Gecko < 1.8: 49/02 @50/6450505050/64Mac Gecko < 1.9: 50/03 #51/3551515151/35Mac Gecko < 1.9: 51/04 $52/3652525252/36Mac Gecko < 1.9: 52/05 %53/3753535353/37Mac Gecko < 1.9: 53/06 ^54/9454545454/94Mac Gecko < 1.9: 54/07 &55/3855555555/38Mac Gecko < 1.9: 55/08 *56/4256565656/42Mac Gecko < 1.9: 56/09 (57/4057575757/40Mac Gecko < 1.9: 57/00 )48/4148484848/41Mac Gecko < 1.9: 48/0Symbol KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptions; :59/58591865959/58Mac Gecko: 59/0= +61/43611876161/43Mac Gecko ≥ 1.9: 61/107
Mac Gecko < 1.9: 61/0, <44/601881884444/60Mac Gecko: 188/0- _45/951091894545/95Mac Gecko ≥ 1.9: 109/0
Mac Gecko < 1.9: 0. >46/621901904646/62Mac Gecko: 190/0/ ?47/631911914747/63Mac Gecko: 191/0` ~96/1261921929696/126Mac Gecko: 192/0[ {91/1232192199191/123\ |92/1242202209292/124Mac Gecko: 220/0] }93/1252212219393/125' "39/342222223939/34Arrow KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptionsleft-arrow-37373737up-arrow-38383838right-arrow-39393939down-arrow-40404040Special KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptionsInsert-45454545Konqueror: 0
Opera < 9.0: 0Delete-46464646Konqueror: 127
Opera < 9.0: 0Home-36363636Opera < 9.0: 0End-35353535Opera < 9.0: 0Page Up-33333333Page Down-34343434Function Keys
F1 to F12-112 to 123112 to 123112 to 123112 to 123Keypad Keys
If Num Lock is on, unshifted/shifted values are returned as shown below. If Num Lock is off, Linux browsers reverse the shifted/unshifted values, while Windows browsers always return the shifted value. None of my Macintoshs have a keypad, so I don't know what they do.KeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptions. Del46/-110/46110/4678/4678/46Opera < 9.0: 78/0
Linux Opera 11.5: 190/460 Ins48/-96/4596/4548/4548/45Opera < 9.0: 48/01 End49/-97/3597/3549/3549/35Opera < 9.0: 49/02 down-arrow50/-98/4098/4050/4050/403 Pg Dn51/-99/3499/3451/3451/344 left-arrow52/-100/37100/3752/3752/37553/-101/12101/1253/1253/12Linux Opera: 53/06 right-arrow54/-102/39102/3954/3954/397 Home55/-103/36103/3655/3655/36Opera < 9.0: 55/08 up-arrow56/-104/38104/3856/3856/389 Pg Up57/-105/33105/3357/3357/33+431071074343Linux Opera 11.5: 61-451091094545Linux Opera 11.5: 109*421061064242Linux Opera 11.5: 56/471111114747Linux Opera 11.5: 191Keypad Enter1313131313Branded KeysKeyASCIIMozilla keycodesIE keycodesOpera keycodespseudo ASCII codesexceptionsLeft Apple Command-224?17?WebKit ≥ 525: 91Right Apple Command-224?17?WebKit ≥ 525: 93Left Windows Start-91912190Linux Gecko: 0Right Windows Start-92922200Linux Gecko: 0Windows Menu-939300

Note that all four encodings agree on most of the common keys, the ones highlighted in green in this table. For the letters and numbers and for spaces, tabs, enters, and arrows the codes are all the same. In fact, they are all standard ASCII values (except for the arrows).

For symbols, things are a fair mess. IE and Mozilla don't entirely agree on what the codes should be. Three keys, ; :, = + and - _, have different values in IE and Mozilla keycodes. Furthermore, there are long standing bugs in Macintosh versions of Gecko that have caused zero keyCodes to be returned for many symbols.

The Opera keycodes have been abandoned by Opera, but they had a certain simple charm. They were always the ASCII code of the character that the key sends when it is not modified by shift or control. They don't allow you to distinguish numbers typed on the keypad from numbers typed on the keyboard, and such like things, but they are, at least, fairly intuitive.

The pseudo ASCII codes weren't really keycodes at all. They were just the ASCII code for the character except that for lower case letters the upper case ASCII code is sent. So those browsers really entirely abandoned the idea of keycodes, instead returning character codes slightly modified for partial IE compatibility. There is much to be said for abandoning keycodes, since the concept really gets you in trouble as you try to handle international keyboards, but something is lost when you do that. You can't, for example, tell if a number was typed on the main keyboard or the keypad. I prefer WebKit's approach, where they keep the keycodes (making them entirely compatible with IE keycodes) butalso return the character code on all key events.

Using pseudo-ASCII codes causes another problem: you can't always recognize the arrow keys on keydown and keyup events. These browsers send the same codes as IE does for arrow keys: the values 37, 38, 39, and 40. These happen to be the ASCII codes for "%", "&", "'" and "(". On U.S. keyboards all those five characters are sent by shifted keys, so you'll never see them as keycodes under any of the three keycode schemes. (Some foreign keyboards do create these characters from unshifted keys, but I don't know what keycodes are sent by those keys.) But when pseudo-ASCII keycodes are used these same values are also sent when you type those keys, so you can't tell those symbols from arrow keys. Similar problems occur with some of the other special keys like Home which sends the same values as "$".

For browsers that generate keypress events for special keys, it is also generally true that event.keyCode will have the same value for "left-arrow" and "%", however we can usually tell which it is because event.which is zero for special keys (there are problems with this in Opera and Konqueror, see above). Versions of WebKit before 525 took a different approach. They invented unique values to return instead of ASCII codes for special keys, and returned the same value in event.keyCode, event.which, and event.charCode. The table below gives the extended ASCII codes returned by old WebKit versions, and also the ones returned in event.charCode on keydown and keyup events in Macintosh versions of IE.

keyExtended ASCII codes for Special Keysup arrowdown arrowleft arrowright arrowfunction keys
F1 to F12HomeEndPage UpPage DownWebKit < 5256323263233632346323563236 to 6324763273632756327663277Macintosh IE3031282916 for all keysno events triggered

To complete the thoroughness of the mess, keycode generation in current Macintosh versions of Gecko remains buggy. For many keys, no keycodes are returned on keydown andkeyup events. Instead the keyCode value is just zero. Some of these problems were fixed in Gecko 1.9, but not all, and the keyboard plus key started returning the value that is supposed to be returned by the number pad plus key.

charactersKeycodes on Gecko keyup and keydown eventsLinux and Windows
Gecko (correct)Macintosh Gecko 1.8 and older (buggy)Macintosh Gecko 1.9 and later (buggy)! @ # $ % ^ & * ( )Same as number keys these symbols appear onzeroSame as number keys these symbols appear on-109zero109_ ~ | < > ? :Same as unshifted symbol keys these symbols appear onzerozero+61zero107Any key typed with ALT key held downSame code as without ALT keyzeroSame code as without ALT key

Macintosh Gecko does give correct charCode values on keypress events, but to a keydown or keyup handler, all the keys that return zero above are indistinguishable. This bug was reported to Mozilla (bug 44259) in June 2000, and it took eight years to get the partial fixes out. Who knows when the rest (48434) will be fixed.

3.4. New Standard Key and Character Events

The DOM3 standard abandons all hope of creating order among event.keyCode, event.which and event.charCode, and instead defines new values for keydown and keyup events. It deprecates the keypress event and replaces it with the textInput event. Unfortunately, the standard changed recently, so that a few browsers implemented an older version of the standard which is quite different from the current version.

Earlier versions of the specification defined attributes named event.keyIdentifier and event.keyLocation. The keyIdentifier was a string that in most cases looked like "U+0041" where the "0041" part is the unicode value of the character sent by the key when it is typed without modifiers, in this case the letter "A". For keys that didn't send unicode characters, or where the unicode value is not standardized, it was a string like "Enter", "Shift", "Left" or "F9". The keyLocation attribute gave values to distinguish among multiple keys that had the same identifier, like the left and right shift keys, or the keypad number keys. It was 0 for standard keys, 1 or 2 for left or right versions of a keys like Shift which appear twice on the keyboard, and 3 for keys on the numeric keypad.

WebKit implemented support for keyIdentifier and got it mostly right. Older versions conformed to an older version of the standard and returned two extra zeros (eg, "U+000041") but this was corrected in version 525. Windows versions of Safari and Linux versions of Chrome return bad keyIdentifier values for all of the non-number symbol keys (WebKit Bug19906 reported in July 2008). The keyLocation attribute is always 0 or 3, so it does not distinguish between left and right modifier keys.

Konqueror returns keyIdentifier values like "Shift" and "Enter" correctly, but instead of returning the Unicode values, it returns the typed character itself, "a" or "A" instead of "U+0041". All keyLocation values are zero, except for modifiers key, which are always one, regardless of whether the left or right one was pressed.

We cannot, however expect any more browsers to implement that standard, since it has now changed. The DOM 3 standard no longer mentions event.keyIdentifier or event.keyLocation. Instead we have event.key, event.char, event.location.. So far as I know, no browser has yet implemented this new version of the DOM 3 standard.

In this standard event.char is defined only when you type a printable character, or another character with a defined code (like tab or backspace). It's basically likeevent.charCode except that it is the character, not the character code and can be any unicode character not just an ASCII code. Event.key is the same as event.char for printable keys. For other keys, even ones like tab or backspace that have character encodings, it is a string like 'Tab', 'Left' or 'F9'. These values are supposed to be the same onkeypress events as they are on keyup and keydown events, though keypress would not be fired for those cases where event.char is null.

Note that neither of these pretends to be a keycode identifying a particular physical key on the keyboard. If you press the /? key on a US keyboard while shift is off, but press the shift key before releasing the /? key, then then on keydown you'll get event.key=='/' and on keyup you'll get event.key=='?'. The only way your Javascript program will know that those two events go together is if it happens to know that those two characters are on the same key. There is an event.locale value that is supposed to give you some clue on what type of keyboard is being used, but figuring out what keys go with what on a particular keyboard is up to you.

Clearly this abandonment of the idea of keycodes is going to cause problems, but is still probably justified. In many (most?) operating systems, I don't think the browser can actually tell which key was pressed. In the browser source code I've seen, the keycodes are generated from the the character codes, not vice versa, by simply assuming that the character came from a US keyboard. So the keycode values never really worked for non-US keyboards.

So while the keycode concept was a handly one, it isn't really practically extensible in the real world. If you want a keycode in the DOM 3 universe, you'll have to go on using the legacy event.keyCode value, which, standards or no standards, isn't going away. The DOM 3 standard seems to recognize this, and reluctantly provides an appendix with some standards for event.keyCode and the like. It casts a rather weak vote for what I called "IE keycodes" above.

Conclusions

It's truely impressive what a hash has been made of a simple thing like recognizing a key pressed on the keyboard. You'd think computer technology would have advanced far enough by now to have this all worked out better.

The keypress events are generally the easiest to work with. They are likely to cause substantially fewer problems with non-US keyboard layouts and it's not too hard to identify which key was pressed. You can get the character typed by doing:

  if (event.which == null)     char= String.fromCharCode(event.keyCode);    // old IE  else if (event.which != 0 && event.charCode != 0)     char= String.fromCharCode(event.which);  // All others  else     // special key

What to do with keypress events on special keys is a problem. I recommend pretending they never happened. If you really want to process special key events, you should probably be working with keydown and keyup instead.

For keydown and keyup events, you can identify most common keys (letters, numbers, and a few others) by just looking at the event.keyCode and more or less pretending that it is an ASCII code. However, it isn't really, and the many Javascript manuals that say it can be converted to a character by doing "String.fromCharCode(event.keyCode)" are wrong. Onkeydown and keyup events, the keycodes are not character codes, and this conversion will give wild results for many keys. There is no general portable way to convert keycodes to characters. You pretty much have to sense the browser type and base the key mapping on that. I don't have information on keycodes sent by international keyboards.

Because of bugs, many keys cannot be distinguished on keydown and keyup in Macintosh Gecko.

Hope for sanity exists, with the new key event handling specifications in DOM3, but so far only WebKit implements them.

---

Last Update: Fri Aug 12 14:07:10 EDT 2011