envjs/env-js

A pure-JavaScript browser environment.

env-js's Issues

Timers not cleared when new page is loaded

When a different page is loaded, the timers defined in the old page are still active.

For now I have added the following code to Envjs.exchangeHTMLDocument in core.js:
if (doc === window.document) {
Envjs.timers.length = 0;
}

But that may be too crude a method when there are multiple windows involved.

Use xbe4x library for e4x support on non-rhino platforms?

Project code here:

http://code.google.com/p/xbe4x/

Select first option by default

When selectedIndex is requested for the first time, I think it is best to select the first item in the dropdown if nothing was selected and an option is available.

I did not verify that this is according to specs, but at least FF, IE, Safari and Chrome will never return null in case an option is available.

One way to fix this can be found in orslumen@18d271d

But it is probably more correct to select the first option once the HTML is parsed and no options have been marked as selected yet.

src/platform/ruby-v8/xhr.js is corrupt, committed as binary...

I have no idea how this happened but it obviously breaks the rubyracer platform. See this mailing list thread for more info: https://github.com/inbox/4182005

window.toString overrides global toString

I'm currently in the process of porting env.js to Nashorn. While trying to run jQuery I discovered that the toString method in window.js (https://github.com/envjs/env-js/blob/master/envjs/window.js#L749) actually causes all objects to return '[Window]' when their toString method is called.

Chrome:

>toString()
"[object Window]"  
>toString.call("test")  
"[object String]"
>var a = {}
undefined
>a.toString()
"[object Object]"

Nashorn + env:

console.log(toString());
>>[Window]
console.log(toString.call("test"));
>>[Window]
var a = {};
console.log(a.toString());
>>[Window]

jQuery is unable to make ajax requests because the toString method is used when registering the transports (https://github.com/jquery/jquery/blob/master/src/ajax.js#L61). Since the toString method is broken, jQuery's "typeof" method will default to calling the transport functions objects which causes the "isFunction" check to fail, preventing any transports from being registered.

HTMLAnchorTag doesn't support a.hash

I get the feeling this project is long since abandoned, but I'll leave this here for what it's worth. It was a great project, BTW!

Env.js 1.2 doesn't support a.hash. That is, if you create an anchor tag, with an href, and then try to look it up via a.hash, it returns undefined. This code fixes that (albeit crudely):

HTMLElement.prototype.setAttribute = function(name, value) {
    //console.log("CSS set attribute: " + name + ", " + value);
    origSetAttribute.apply(this, arguments);
    if (name === "style") {
        updateCss2Props(this, value);
    }
    // Add support for 'HTMLAnchorTag.hash'
    if (this.tagName === "A" && name === "href" && value.indexOf( "#" ) !== -1 ) {
        this.hash = value.substring( value.indexOf( "#" ));
    }
}

jQuery 1.5.1

Some fixes were needed to get jQuery 1.5.1 working with EnvJs in Ruby Racer:

i) The jQuery.clone method uses an expando attribute that is removed in jQeury.cloneFixAttributes. But that attribute may not be present on the current node. If doc.implementation.errorChecking === false it should fail silently, see NamedNodeMap#removeNamedItem(NS).

ii) jQuery directly retrieves and sets the value attribute on the nodes, but (at least in Ruby Racer) the NamedNodeMap does not accept a call to value=. To fix this, explicit getter and setters are added, see NamedNodeMap#get/set value

iii) jQeury.cloneFixAttributes makes use of Element.clearAttributes and Element.mergeAttributes. Both methods were not defined yet.

The code changes I used to get jQuery 1.5.1 working can be found in orslumen@c3e702c

1.3 pre03 not working in Eclipse/JSDT due to path issue.

I’m trying JavaScript Development Tool and it comes with Rhino as debugging tool. It is very nice.

I was able to make envjs-1.2.13 working. Now, switching to the 1.3 pre03. Hitting issue on path setting..
Under envjs/rhino.js, at the end it calls require to load in additional modules.
require('envjs/platform/rhino');
require('envjs/window');
But inside require, it relies on the “user.dir” to search for file and it fail for my case since, user.dir will return the directory where Eclipse is started from.
var currentPath = java.lang.System.getProperty('user.dir');

Should it goes with the way to have an envjsConfig object with baseUrl property set prior to load Envjs?
Then, internally, use the object to load file accordingly.

I did a simple hack & get all the files loaded but I’m still seeing weird things (newbie to java & javascript)
Somewhere the file path are messed up...

org.mozilla.javascript.WrappedException: Wrapped java.net.URISyntaxException: Illegal character in path at index 10: file://C/:\Eclipse-indigo/C:\workspace\client\scripts\example\simple.js (file:/C:/workspace/client/scripts/envjs/platform/rhino.js#251)
at org.mozilla.javascript.Context.throwAsScriptRuntimeEx(Context.java:1773)
at org.mozilla.javascript.MemberBox.newInstance(MemberBox.java:202)
at org.mozilla.javascript.NativeJavaClass.constructSpecific(NativeJavaClass.java:281)
at org.mozilla.javascript.NativeJavaClass.construct(NativeJavaClass.java:200)
at org.mozilla.javascript.Interpreter.interpretLoop(Interpreter.java:3377)
at script(file:/C:/workspace/client/scripts/envjs/platform/rhino.js:251)
at script(file:/C:/workspace/client/scripts/envjs/platform/core.js:1021)
at script(file:/C:/workspace/client/scripts/envjs/platform/core.js:2403)
at script(file:/C:/workspace/client/scripts/envjs/platform/core.js:1007)
at script(file:/C:/workspace/client/scripts/envjs/platform/core.js:943)
at script(file:/C:/workspace/client/scripts/envjs/window.js:855)
at script(file:/C:/workspace/client/scripts/envjs/window.js:34)
at script.require(/C:/workspace/client/scripts/envjs/rhino.js:85)
at script(/C:/workspace/client/scripts/envjs/rhino.js:108)
at script(/C:/workspace/client/scripts/envjs/rhino.js:105)
at org.mozilla.javascript.Interpreter.interpret(Interpreter.java:2487)
at org.mozilla.javascript.InterpretedFunction.exec(InterpretedFunction.java:178)
at org.mozilla.javascript.tools.shell.Main.evaluateScript(Main.java:564)
at org.mozilla.javascript.tools.shell.Main.processFileSecure(Main.java:486)
at org.mozilla.javascript.tools.shell.Main.processFile(Main.java:452)
at org.mozilla.javascript.tools.shell.Global.load(Global.java:244)
at sun.reflect.NativeMethodAccessorImpl.invoke0(Native Method)
at sun.reflect.NativeMethodAccessorImpl.invoke(Unknown Source)
at sun.reflect.DelegatingMethodAccessorImpl.invoke(Unknown Source)
at java.lang.reflect.Method.invoke(Unknown Source)
at org.mozilla.javascript.MemberBox.invoke(MemberBox.java:161)
at org.mozilla.javascript.FunctionObject.call(FunctionObject.java:413)
at org.mozilla.javascript.Interpreter.interpretLoop(Interpreter.java:3330)
at org.mozilla.javascript.Interpreter.interpret(Interpreter.java:2487)
at org.mozilla.javascript.InterpretedFunction.call(InterpretedFunction.java:164)
at org.mozilla.javascript.ContextFactory.doTopCall(ContextFactory.java:398)
at org.mozilla.javascript.ScriptRuntime.doTopCall(ScriptRuntime.java:3065)
at org.mozilla.javascript.InterpretedFunction.exec(InterpretedFunction.java:175)
at org.mozilla.javascript.tools.shell.Main.evaluateScript(Main.java:564)
at org.mozilla.javascript.tools.shell.Main.processFileSecure(Main.java:486)
at org.mozilla.javascript.tools.shell.Main.processFile(Main.java:452)
at org.eclipse.wst.jsdt.debug.rhino.debugger.shell.DebugMain$Evaluator.run(DebugMain.java:89)
at org.mozilla.javascript.Context.call(Context.java:515)
at org.mozilla.javascript.ContextFactory.call(ContextFactory.java:507)
at org.eclipse.wst.jsdt.debug.rhino.debugger.shell.DebugMain.main(DebugMain.java:218)
Caused by: java.net.URISyntaxException: Illegal character in path at index 10: file://C/:\Eclipse-indigo/C:\workspace\client\scripts\example\simple.js
at java.net.URI$Parser.fail(Unknown Source)
at java.net.URI$Parser.checkChars(Unknown Source)
at java.net.URI$Parser.parseHierarchical(Unknown Source)
at java.net.URI$Parser.parse(Unknown Source)
at java.net.URI.(Unknown Source)
at sun.reflect.NativeConstructorAccessorImpl.newInstance0(Native Method)
at sun.reflect.NativeConstructorAccessorImpl.newInstance(Unknown Source)
at sun.reflect.DelegatingConstructorAccessorImpl.newInstance(Unknown Source)
at java.lang.reflect.Constructor.newInstance(Unknown Source)
at org.mozilla.javascript.MemberBox.newInstance(MemberBox.java:194)
... 28 more

Does it work with Ruby 1.9 and V8 on Linux box ?

Hello,

I try to write a little script with Ruby on a Linux Fedora 64 bits box.
I am using the gem 'therubyracer-heroku-0.8.1' for the V8 machine because the gem 'therubyracer' fails to compile on 64 bits with gcc RedHat 4.5.1.

So, in my script, i load the 'envjs/rubyracer.rb' and it success with Ruby 1.8.7 but fails with Ruby 1.9.2 for two reasons :

-1) 'configure_context' method fails for compatibility reason on 'eval'. Just add to_s method to the parameter of 'eval'.
-2) Exception raised with that stack :

👎 Uncaught TypeError: Cannot convert object to primitive value

ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:118:in block in v8': undefined methodSetHiddenValue' for #V8::C::Value:0x0000000154fa28 (NoMethodError)
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:117:in tap' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:117:inv8'
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:111:in block (2 levels) in v8' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:110:ineach'
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:110:in block in v8' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:109:intap'
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:109:in v8' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/object.rb:19:inblock (2 levels) in []='
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:75:in block in open' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/context.rb:79:inenter'
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/portal.rb:74:in open' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/object.rb:18:inblock in []='
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/object.rb:17:in tap' ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/object.rb:17:in[]='
ruby-192p136/lib/ruby/gems/1.9.1/gems/therubyracer-heroku-0.8.1.pre3/lib/v8/context.rb:58:in []=' envjs/rubyracer.rb:47:inconfigure_context'
from ./go.sh:7:in load' from ./go.sh:7:in

'

Any idea ?

Regards.

Looks like the MouseEvents are not bound

When I have a clickable div on an html page like:

Waiting to be clicked

And I send a mouse-click event to the #clickme div, nothing happens.

Example: http://jsfiddle.net/EHezK/

create setting for honoring image requests

a lot of use case require envjs at least make the request and honor cookies in response even if it doesnt load the image. See this thread,

http://stackoverflow.com/questions/5171765/google-analytics-in-java-with-rhino-and-envjs
and
http://groups.google.com/group/envjs/browse_thread/thread/25310a35347cb5a?hl=en

NodeList.toArray() returns the incorrect type

When parsing a Document Fragment, the appendChild method calls Array.prototype.push.apply(nodelist, newChild.childNodes.toArray() );

But as NodeList.toArray() returns self, you will receive Exception TypeError: Function.prototype.apply: Arguments list has wrong type.

I got around this issue by putting back the original toArray() method:
toArray: function () {
var children = [];
for ( var i=0; i < this.length; i++) {
children.push (this[i]);
}
return children;
},

See commit orslumen@bb3c8db

Do not reload page if location.href has not changed

If the document location is set to a similar URL, do not reload the page.

To fix this I added the following line of code, but I can imagine the URL comparison can be optimized (e.g. by removing blanks at the end of the url):

set href(url) {
  if (url === $url || url+'#' === $url || url === $url+'#') { return; }
  ...
}

See orslumen@a72ff2e

Call function synchronously when timeout is set to zero

Well, this is more of an enhancement request than a bug.

I am running my integration tests asynchronously. Next to that I make sure that (most) calls to setTimeout use a timeout of 0 (zero) during the integration tests, to perform well.

But timer.js by default adds all timers and intervals to the queue, regardless of the timeout value.

I would like setTimeout to synchronously call the given function when the timeout is set to zero, see e.g. orslumen@fe499c2

Support href '#' and 'javascript:void(0);'

When the 'href' attribute of a HTMLAnchorElement is set to '#' or 'javascript:void(0);', a click should not lead to a location change.

A workaround can be found in orslumen@80c79e3

Ruby Racer Requests with Form Data

Posting form-data using Ruby Racer did not work for me.

I got it working by passing the form-data as the data argument to the HTTPConnection.go method.

Next to that my Rails App was not receiving some HTTP_xxx request headers and the response status code was not parsed correctly.

The codefixes that I used can be found in commit
orslumen@b294e11

I am not sure though whether or not this fix would work for everyone.

timer event loop exit enhancement

http://groups.google.com/group/envjs/browse_thread/thread/4d9bd737f68cf78b

This requires allowing the Envjs.eventLoop to run in 'an indefinite loop' which means running an exact callstack n-times before providing an Envjs.unresponsive hook.

Am I on the right track here?

Thatcher

HTMLAnchorTag doesn't support URLUtils.pathname

The implementation of HTMLAnchorTag doesn't support URLUtils.pathname.

I discovered this issue while attempting to utilize sbt-jasmine-plugin to unit test an angular.js application. I have tracked down this issue with env.js as the cause of this issue I reported. I also opened this Stackoverflow question.

I intend to implement it per this documentation.

cant add require.paths because of how envjs 1.3 puts requires before user can add paths

currently the require definition is followed directly by use of require to load envjs. this prevents the user from being able to add paths to the require directive for where they want envjs to load from.

scan meta tag for character encoding

currently envjs treats all html pages as utf-8, we should be scanning for a meta tag which declares character encoding

jQuery 1.6.4: Cross-domain requests fail due to lack of CORS support

In jQuery 1.6.4, the Ajax Transport checks if the request is "crossDomain" and, if so, requires jQuery.support.cors to be true:

if ( !s.crossDomain || jQuery.support.cors ) {

CORS support checks for the existence of the 'withCredentials' property (default boolean false), which is not present in Envjs.

I added it in this commit: vanm@2f485c0

Since most requests in our test suite are classified as "crossDomain" by jQuery, this is required to get $.ajax requests working properly.

CSS Properties mixed up

In CSS2Properties of css.js , the setProperty method stores updated CSS properties in the styleIndex variable, but that object is shared between multiple (all?) nodes.

So when a CSS property is dynamically updated in one node, it also changes in all other nodes.

One way to fix is is by storing all changed properties in a separate variable, see orslumen@aa0318d

spyd not running simple script

window.location = 'http://www.zope.com';

envjs just exits...

dom elements are missing style property

please see

http://groups.google.com/group/envjs/browse_thread/thread/ebbeaec489e958c1?hl=en

xhr.js has a type error in getAllResponseHeaders

In xhr.js line 453 refers to responseHeader, that should be responseHeaders
if(this.responseHeaders.hasOwnProperty(header)){
see orslumen@44fb38d

Note that when working with jQuery 1.5.1 all XHR requests will fail on this, as jQuery uses the standard XMLHttpRequest that is defined in xhr.js

Run AJAX callback

I tried to use $.getJSON (with jQuery) to test some AJAX callbacks, but doesnt work.

getElementsByTagName out of synch with innerHTML

One of my jQuery plugins uses the $.wrap() function. After that method is called, the getElementsByTagName function returns incorrect results. It completely ignores the new element and moves the wrapped element to the end of the list.

To reproduce (without jQuery):

<html lang='en-US' xml:lang='en-US' xmlns='http://www.w3.org/1999/xhtml'>
  <head/>
  <body>
    #container
      #before
      #to_be_wrapped
      #after
    <script type="text/javascript">
      //<![CDATA[
        // Print the current state to the console, this is OK
        var container = document.getElementById('container');
        console.log('\r\nBefore:\r\n'+container.innerHTML);
        var all = container.getElementsByTagName('*'); var ids = ''; for(var i=0; i < all.length; i++) { ids = ids + (String.isBlank(ids) ? '' : ', ') + all[i].getAttribute('ID'); } console.log('getElementsByTagName("*") returns%selements:%s', all.length, ids);

        // Wrap the #to_be_wrapped node in a new #wrapper node
        var wrapper = document.createElement('div');
        wrapper.id = "wrapper";
        var to_be_wrapped = document.getElementById('to_be_wrapped');
        container.insertBefore(wrapper, to_be_wrapped);
        wrapper.appendChild(to_be_wrapped);

        // Print the results of the incorrect state of getElementsByTagName
        console.log('\r\nAfter:\r\n'+container.innerHTML);
        all = container.getElementsByTagName('*'); ids = ''; for(var i=0; i < all.length; i++) { ids = ids + (String.isBlank(ids) ? '' : ', ') + all[i].getAttribute('ID'); } console.log('getElementsByTagName("*") returns%selements:%s', all.length, ids); 

        // Show that childNodes still reflects the correct state
        all = container.childNodes; ids = ''; var id_count = 0; for(var i=0; i < all.length; i++) { if (all[i].nodeType == Node.ELEMENT_NODE) { id_count++; ids = ids + (String.isBlank(ids) ? '' : ', ') + all[i].getAttribute('ID'); } } console.log('childNodes returns%selements:%s', id_count, ids); 
      //]]>
    </script>
  </body>
</html>

This produces the following results:
Before:

<div id="before"/>
<div id="to_be_wrapped"/>
<div id="after"/>
getElementsByTagName("*") returns 3 elements: before, to_be_wrapped, after 

After:

<div id="before"/>
<div id="wrapper"><div id="to_be_wrapped"/></div>
<div id="after"/>
getElementsByTagName("*") returns 3 elements: before, after, to_be_wrapped 
childNodes returns 3 elements: before, wrapper, after 

As you can see the innerHTML is correct, but the container.getElementsByTagName("*") is missing the #wrapper div and has incorrectly moved the #to_be_wrapped div to the end.

Fortunately the childNodes property does contain the correct state, so as a workaround I switched back to the old recursive implementation of getElementsByTagName using childNodes, which seems to work fine for the moment.

Log statement causing trouble

The method insertBefore in dom.js starts with the following debug statement:

log.debug('insert %s Before %s', newChild.nodeName, refChild.nodeName);

But results in a TypeError: Cannot read property 'nodeName' of null when the refChild (or newChild is undefined). This can be fixed as follows:

log.debug('insert %s Before %s', newChild && newChild.nodeName, refChild && refChild.nodeName);

See commit orslumen@c22faaf

But would it not be better (performance wise) to comment out all log statements, or to add something like if (Log.level <= debug) { log.debug(...) } to all debug statements?

window.unload event not emitted

When a different page is loaded, the window.unload event is not emitted.

For now I have added the following code at the top of Envjs.exchangeHTMLDocument in core.js:
try {
if (doc === window.document) {
event = doc.createEvent('HTMLEvents');
event.initEvent('unload', false, false);
window.dispatchEvent( event, false );
}
} catch (eee) {
log.debug('window unload event failed %s', eee);
//swallow
}

I am note sure though whether or not this code is in the right location... see orslumen@2c2027e

timer event loop exit enhancement

http://groups.google.com/group/envjs/browse_thread/thread/9a0f4fc1a1cbc501

insertCell throws "TypeError: Cannot read property 'nodeName' of null"

First of all, thank you for the great work on envjs!

While upgrading our integration tests from envjs 0.3.7 to 1.3.pre1 I noticed that when insertCell (html.js) is called with idx === numCells an error is thrown, because node is null when passed to insertBefore.

To fix this, the line this.insertBefore(cell, node); needs to be placed 2 lines higher within the else block. So the new cell is either appended at the end of the row, or inserted before an existing cell.

Kind regards,
Orslumen

PS. If you prefer a pull request, pls let me know.

requirejs error

please see

http://groups.google.com/group/requirejs/browse_thread/thread/973db6b54ab6b1e2

and

http://groups.google.com/group/envjs/browse_thread/thread/189aee681b33e19b?hl=en_US

Thatcher

improve dynamic script loading tests

one test is commented out in specs/html/spec.js but runs just fine