Develop Reference

Firefox add-on get the DOM window which made an HTTP request

I'm capturing the HTTP requests in a Firefox Add-on SDK extension. I need to get the DOM window associated with the request. However, I'm getting an NS_NOINTERFACE error.
Here is my code:
var httpRequestObserver = {
observe: function (subject, topic, data) {
var httpRequest = subject.QueryInterface(Ci.nsIHttpChannel);
var requestUrl = subject.URI.host;
var domWin;
var assWindow;
console.log('URL: ', requestUrl);
try {
domWin = httpRequest.notificationCallbacks.getInterface(Ci.nsIDOMWindow);
assWindow = httpChannel.notificationCallbacks.getInterface(Ci.nsILoadContext)
.associatedWindow;
console.log(domWin);
} catch (e) {
console.log(e);
}
// console.log('TAB: ', tabsLib.getTabForWindow(domWin.top));
var hostName = wn.domWindow.getBrowser().selectedBrowser.contentWindow.location.host;
console.log('HOST: ', hostName);
},
get observerService() {
return Cc['#mozilla.org/observer-service;1'].getService(Ci.nsIObserverService);
},
register: function () {
this.observerService.addObserver(this, 'http-on-modify-request', false);
},
unregister: function () {
this.observerService.removeObserver(this, 'http-on-modify-request');
}
};
httpRequestObserver.register();
I've tried both nsIDOMWindow and nsILoadContext, but NS_NOINTERFACE error always appears on an attempt to get the window object.

I have finally managed to get the data I need via
httpRequest.notificationCallbacks.getInterface(Ci.nsILoadContext).topFrameElement
For example, to get url of the document which started the request, I used
httpRequest.notificationCallbacks.getInterface(Ci.nsILoadContext).topFrameElement._documentURI.href

Since you already found how to get the <browser> from the request you can do the following to get back to SDK APIs:
let browser = ....topFrameElement
let xulTab = browser.ownerDocument.defaultView.gBrowser.getTabForBrowser(browser)
let sdkTab = modelFor(xulTab)
modelFor() is documented in the tabs module.

XPathEvaluator in Firefox addon

I am attempting to follow this article to evaluate an XPath expression. My code is copy/pasted from the article:
// Evaluate an XPath expression aExpression against a given DOM node
// or Document object (aNode), returning the results as an array
// thanks wanderingstan at morethanwarm dot mail dot com for the
// initial work.
function evaluateXPath(aNode, aExpr) {
var xpe = new XPathEvaluator();
var nsResolver = xpe.createNSResolver(aNode.ownerDocument == null ?
aNode.documentElement : aNode.ownerDocument.documentElement);
var result = xpe.evaluate(aExpr, aNode, nsResolver, 0, null);
var found = [];
var res;
while (res = result.iterateNext())
found.push(res);
return found;
}
However, I'm getting this error:
Message: ReferenceError: XPathEvaluator is not defined
Is Mozilla's article out of date, perhaps? Is there a more up-to-date article available on parsing XML in an SDK add-on?
Edit. When I tried it this way:
var {Cc, Ci} = require("chrome");
var domXPathEvaluator = Cc["#mozilla.org/dom/xpath-evaluator;1"].createInstance(Ci.nsIDOMXPathEvaluator);
I got a long error message:
- message = Component returned failure code: 0x80570019 (NS_ERROR_XPC_CANT_CREATE_WN) [nsIJSCID.createInstance]
- fileName = undefined
- lineNumber = 14
- stack = #undefined:14:undefined|#resource://helloworld-addon/index.js:14:25|run#resource://gre/modules/commonjs/sdk/addon/runner.js:145:19|startup/</<#resource://gre/modules/commonjs/sdk/addon/runner.js:86:7|Handler.prototype.process#resource://gre/modules/Promise-backend.js:920:23|this.PromiseWalker.walkerLoop#resource://gre/modules/Promise-backend.js:799:7|this.PromiseWalker.scheduleWalkerLoop/<#resource://gre/modules/Promise-backend.js:738:39|Promise*this.PromiseWalker.scheduleWalkerLoop#resource://gre/modules/Promise-backend.js:738:7|this.PromiseWalker.schedulePromise#resource://gre/modules/Promise-backend.js:762:7|this.PromiseWalker.completePromise#resource://gre/modules/Promise-backend.js:705:7|handler#resource://gre/modules/commonjs/sdk/addon/window.js:56:3|
- toString = function () /* use strict */ toString
edit 2. Here, I'll just post my whole code, because it's clear something stranger than I thought is going on. I've created a hello-world addon using the Mozilla tutorials including this one to display a popup. I've modified that further so that it will append text to a file, and modified that further to, I hope, parse and modify XML. So the resulting add-on is supposed to take text entered in the popup and append it to an XML file.
var data = require("sdk/self").data;
var text_entry = require("sdk/panel").Panel({
contentURL: data.url("text-entry.html"),
contentScriptFile: data.url("get-text.js")
});
const fooFile = "/Users/sabrina/Documents/addon/foo.xml";
var {Cc, Ci} = require("chrome");
var parser = Cc["#mozilla.org/xmlextras/domparser;1"].createInstance(Ci.nsIDOMParser);
//var domXPathEvaluator = Cc["#mozilla.org/dom/xpath-evaluator;1"].createInstance(Ci.nsIDOMXPathEvaluator);
var foo = parser.parseFromString(readTextFromFile(fooFile), "application/xml");
// Create a button
require("sdk/ui/button/action").ActionButton({
id: "show-panel",
label: "Show Panel",
icon: {
"16": "./icon-16.png",
"32": "./icon-32.png",
"64": "./icon-64.png"
},
onClick: handleClick
});
// Show the panel when the user clicks the button.
function handleClick(state) {
text_entry.show();
}
text_entry.on("show", function() {
text_entry.port.emit("show");
});
text_entry.port.on("text-entered", function (text) {
console.log(text);
// appendTextToFile(text, "/Users/sabrina/Documents/addon/output.txt");
appendFoo(text);
text_entry.hide();
});
function appendFoo(text) {
var newNode = foo.createElement("blah");
newNode.innerHTML = text;
var mainFoo = evaluateXPath(foo, '/foo')[0];
mainFoo.appendChild(newNode);
foo.save(fooFile);
}
function evaluateXPath(aNode, aExpr) {
var xpe = new XPathEvaluator();
var nsResolver = xpe.createNSResolver(aNode.ownerDocument == null ?
aNode.documentElement : aNode.ownerDocument.documentElement);
//var result = domXPathEvaluator.evaluate(aExpr, aNode, null,
// domXPathResult.ORDERED_NODE_ITERATOR_TYPE, null);
var found = [];
var res;
while (res = result.iterateNext())
found.push(res);
return found;
}
function readTextFromFile(filename) {
var fileIO = require("sdk/io/file");
var text = null;
if (fileIO.exists(filename)) {
var TextReader = fileIO.open(filename, "r");
if (!TextReader.closed) {
text = TextReader.read();
TextReader.close();
}
}
console.log(arguments.callee.name + ": have read " + text + " from " + filename);
return text;
}
function writeTextToFile(text, filename) {
var fileIO = require("sdk/io/file");
var TextWriter = fileIO.open(filename, "w");
if (!TextWriter.closed) {
TextWriter.write(text + "\n");
console.log(arguments.callee.name + ": have written " + text + " to " + filename);
TextWriter.close();
}
function appendTextToFile(text, filename) {
var textplus = readTextFromFile(filename) + text;
writeTextToFile(textplus, filename);
}
I run at the command line using jpm run which opens Firefox Developer Edition. I click the addon button, the popup comes up, I enter text, I hit return, and I see this in the console:
JPM undefined Starting jpm run on Sabrina's Helloworld Addon
Creating XPI
JPM undefined XPI created at /var/folders/gg/r_hp4hzs0gdfy70f__l18fmr0000gn/T/#helloworld-addon-0.0.1.xpi (46ms)
Created XPI at /var/folders/gg/r_hp4hzs0gdfy70f__l18fmr0000gn/T/#helloworld-addon-0.0.1.xpi
JPM undefined Creating a new profile
console.log: helloworld-addon: readTextFromFile: have read <?xml version="1.0" encoding="UTF-8"?>
<foo><blah>eek</blah><foo>
from /Users/sabrina/Documents/addon/foo.xml
console.log: helloworld-addon: ook
console.error: helloworld-addon:
JPM undefined Message: ReferenceError: XPathEvaluator is not defined
Stack:
evaluateXPath#resource://gre/modules/commonjs/toolkit/loader.js -> resource://helloworld-addon/index.js:63:9
appendFoo#resource://gre/modules/commonjs/toolkit/loader.js -> resource://helloworld-addon/index.js:57:19
#resource://gre/modules/commonjs/toolkit/loader.js -> resource://helloworld-addon/index.js:50:2
emitOnObject#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/event/core.js:112:9
emit#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/event/core.js:89:38
portEmit#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/content/sandbox.js:343:7
emitOnObject#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/event/core.js:112:9
emit#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/event/core.js:89:38
onContentEvent/<#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/content/sandbox.js:384:5
delay/<#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/lang/functional/concurrent.js:38:20
notify#resource://gre/modules/commonjs/toolkit/loader.js -> resource://gre/modules/commonjs/sdk/timers.js:40:9

Non-authoritative, speculative answer
In a different question, Wladimir Palant (author of Adblock Plus, presumably he has good knowledge of firefox) said:
Yes, a lot of global classes available in the window context aren't there in SDK modules which are sandboxes.
Source: https://stackoverflow.com/a/10522459/3512867
This could explain why XPathEvaluator is not defined in the SDK addon.
The logical conclusion would be to use Firefox's Components object to access the nsIDOMXPathEvaluator interface. Which brings up the following error:
NS_ERROR_XPC_CANT_CREATE_WN
Looking into it takes us to this, from mozillazine's forums user "lithopsian":
That means it can't create a wrapper for a non-javascript interface.
Source: http://forums.mozillazine.org/viewtopic.php?f=19&t=2854793
I am unable to judge the credibility of that statement and while the linked bug reports seem to be relevant, I can not attest they actually are:
https://bugzilla.mozilla.org/show_bug.cgi?id=994964
https://bugzilla.mozilla.org/show_bug.cgi?id=1027095
https://bugzilla.mozilla.org/show_bug.cgi?id=1029104
Unless those informations are confirmed (or dispelled) by people with a deeper knowledge of Firefox's internal workings, I can only hesitantly conclude that the nsIDOMXPathEvaluator interface can simply not work in an SDK addon.

How to handle invalid URL / IP's in websockets?

I'm using HTML / Javascript web sockets to communicate with a python server program. Now I have the option to change the server's IP via clean UI and I have a .onerror function that handles with connection errors, however this doesn't handle initial errors. What I mean by this is if I were to enter a completely invalid address, it wont even attempt to connect with it (which is fine) and spit out and error like: [Error] WebSocket network error: The operation couldn’t be completed. (kCFErrorDomainCFNetwork error 2.). How can I handle this error so I can say, popup a message for example?
Here's a brief overview of my JS script.
function updateDevice(id, ipUI){
if ("WebSocket" in window){
var ws = new WebSocket(serverIP);
// Here is where I need to handle the bad address right?
ws.onopen = function(){
ws.send(id);
};
ws.onmessage = function (evt){
var received_msg = evt.data;
};
// This function ws.onerror doesnt handle bad addresses.
ws.onerror = function(){
document.getElementById("error_msg").style.display='block';
};
}else{
alert("This site doesnt support your browser...");
};
};

You could wrap the new WebSocket in a try/catch:
try {
new WebSocket(serverIP);
} catch (e) {
if (e instanceof DOMException) {
alert('Invalid address!');
} else {
throw e;
}
}

File API writing file doesn't work Samsung smart tv SDK

I have this javascript code which makes possible writing in a file
{
var fileSystemObj = new FileSystem();
var fileObj = fileSystemObj.openCommonFile(curWidget.id +
‘/testFile.data’, ‘w’);
fileObj.writeLine(‘something to write.’);
fileSystemObj.closeCommonFile(fileObj);
}
but it doesn't work. Doesn't even display any error!

samsung developer forum (you may not see unless you sign in... )
I am quoting it.
case tvKey.KEY_RED:
alert('RED BUTTON!');
alert('CWID: '+curWidget.id);
try {
var fileSystemObj = new FileSystem();
var fileObj = fileSystemObj.openCommonFile(curWidget.id+'/testFile.data','w');
fileObj.writeLine('something to write.');
fileSystemObj.closeCommonFile(fileObj);
} catch (e) {
alert('Error: file handling: '+e);
}
break;
lead to error:
alert() : Error: file handling: TypeError: 'null' is not an object
(evaluating 'fileObj.writeLine')
Reading cause same problem.
and solution accepted in that link is:
I suppose that problem is that you have to create common dir (if does not exist ) at first :
var fileObj = fileSystemObj.openCommonFile(filePath, 'w');
if(!fileObj){
var bValid = fileSystemObj.isValidCommonPath(curWidget.id);
if (!bValid) {
fileSystemObj.createCommonDir(curWidget.id);
}
}
fileObj = fileSystemObj.openCommonFile(filePath, 'w');
fileObj.writeLine('something to write.');
fileSystemObj.closeCommonFile(fileObj);

Reading file contents on the client-side in javascript in various browsers

I'm attempting to provide a script-only solution for reading the contents of a file on a client machine through a browser.
I have a solution that works with Firefox and Internet Explorer. It's not pretty, but I'm only trying things at the moment:
function getFileContents() {
var fileForUpload = document.forms[0].fileForUpload;
var fileName = fileForUpload.value;
if (fileForUpload.files) {
var fileContents = fileForUpload.files.item(0).getAsBinary();
document.forms[0].fileContents.innerHTML = fileContents;
} else {
// try the IE method
var fileContents = ieReadFile(fileName);
document.forms[0].fileContents.innerHTML = fileContents;
}
}
function ieReadFile(filename)
{
try
{
var fso = new ActiveXObject("Scripting.FileSystemObject");
var fh = fso.OpenTextFile(filename, 1);
var contents = fh.ReadAll();
fh.Close();
return contents;
}
catch (Exception)
{
return "Cannot open file :(";
}
}
I can call getFileContents() and it will write the contents into the fileContents text area.
Is there a way to do this in other browsers?
I'm most concerned with Safari and Chrome at the moment, but I'm open to suggestions for any other browser.
Edit: In response to the question, "Why do you want to do this?":
Basically, I want to hash the file contents together with a one-time-password on the client side so I can send this information back as a verification.

Edited to add information about the File API
Since I originally wrote this answer, the File API has been proposed as a standard and implemented in most browsers (as of IE 10, which added support for FileReader API described here, though not yet the File API). The API is a bit more complicated than the older Mozilla API, as it is designed to support asynchronous reading of files, better support for binary files and decoding of different text encodings. There is some documentation available on the Mozilla Developer Network as well as various examples online. You would use it as follows:
var file = document.getElementById("fileForUpload").files[0];
if (file) {
var reader = new FileReader();
reader.readAsText(file, "UTF-8");
reader.onload = function (evt) {
document.getElementById("fileContents").innerHTML = evt.target.result;
}
reader.onerror = function (evt) {
document.getElementById("fileContents").innerHTML = "error reading file";
}
}
Original answer
There does not appear to be a way to do this in WebKit (thus, Safari and Chrome). The only keys that a File object has are fileName and fileSize. According to the commit message for the File and FileList support, these are inspired by Mozilla's File object, but they appear to support only a subset of the features.
If you would like to change this, you could always send a patch to the WebKit project. Another possibility would be to propose the Mozilla API for inclusion in HTML 5; the WHATWG mailing list is probably the best place to do that. If you do that, then it is much more likely that there will be a cross-browser way to do this, at least in a couple years time. Of course, submitting either a patch or a proposal for inclusion to HTML 5 does mean some work defending the idea, but the fact that Firefox already implements it gives you something to start with.

In order to read a file chosen by the user, using a file open dialog, you can use the <input type="file"> tag. You can find information on it from MSDN. When the file is chosen you can use the FileReader API to read the contents.
function onFileLoad(elementId, event) {
document.getElementById(elementId).innerText = event.target.result;
}
function onChooseFile(event, onLoadFileHandler) {
if (typeof window.FileReader !== 'function')
throw ("The file API isn't supported on this browser.");
let input = event.target;
if (!input)
throw ("The browser does not properly implement the event object");
if (!input.files)
throw ("This browser does not support the `files` property of the file input.");
if (!input.files[0])
return undefined;
let file = input.files[0];
let fr = new FileReader();
fr.onload = onLoadFileHandler;
fr.readAsText(file);
}
<input type='file' onchange='onChooseFile(event, onFileLoad.bind(this, "contents"))' />
<p id="contents"></p>

There's a modern native alternative: File implements Blob, so we can call Blob.text().
async function readText(event) {
const file = event.target.files.item(0)
const text = await file.text();
document.getElementById("output").innerText = text
}
<input type="file" onchange="readText(event)" />
<pre id="output"></pre>
Currently (September 2020) this is supported in Chrome and Firefox, for other Browser you need to load a polyfill, e.g. blob-polyfill.

Happy coding!
If you get an error on Internet Explorer, Change the security settings to allow ActiveX
var CallBackFunction = function(content) {
alert(content);
}
ReadFileAllBrowsers(document.getElementById("file_upload"), CallBackFunction);
//Tested in Mozilla Firefox browser, Chrome
function ReadFileAllBrowsers(FileElement, CallBackFunction) {
try {
var file = FileElement.files[0];
var contents_ = "";
if (file) {
var reader = new FileReader();
reader.readAsText(file, "UTF-8");
reader.onload = function(evt) {
CallBackFunction(evt.target.result);
}
reader.onerror = function(evt) {
alert("Error reading file");
}
}
} catch (Exception) {
var fall_back = ieReadFile(FileElement.value);
if (fall_back != false) {
CallBackFunction(fall_back);
}
}
}
///Reading files with Internet Explorer
function ieReadFile(filename) {
try {
var fso = new ActiveXObject("Scripting.FileSystemObject");
var fh = fso.OpenTextFile(filename, 1);
var contents = fh.ReadAll();
fh.Close();
return contents;
} catch (Exception) {
alert(Exception);
return false;
}
}

This works fine
function onClick(event) {
filecontent = "";
var myFile = event.files[0];
var reader = new FileReader();
reader.addEventListener('load', function (e) {
filecontent = e.target.result;
});
reader.readAsBinaryString(myFile);
}