I would like to do some operation on my filesystem through a Thunderbird plugin.
For example create a folder at a specific location and a text file,
containing some data from thunderbird, in this folder.
As you know, Mozilla Extensions consist of javascript code.
So I looked for this, found some code about ActiveXObject, which is not working for Thunderbird.
Any ideas what should I do about it?
First, read up on extensions in general (Firefox docs apply to Thunderbird as well, except for the Add-on SDK, which does not really work with Thunderbird; go the XUL overlay route).
Then there are multiple ways to perform File I/O, in particular XPCOM stuff and OS.File:
https://developer.mozilla.org/en-US/Add-ons/Code_snippets/File_I_O
https://developer.mozilla.org/en-US/docs/JavaScript_OS.File/OS.File_for_the_main_thread
Here is a code snippet from my extension. I create text file in Profile directory and then I add some text into this file.
var path = Components.classes["#mozilla.org/file/directory_service;1"].getService( Components.interfaces.nsIProperties).get("ProfD", Components.interfaces.nsIFile).path + "\\";
var file = Components.classes["#mozilla.org/file/local;1"].createInstance(Components.interfaces.nsILocalFile);
file.initWithPath(path);
file.append("settings.txt")
file.create(Components.interfaces.nsIFile.NORMAL_FILE_TYPE, 0664)
var outputStream = Components.classes["#mozilla.org/network/file-output-stream;1"].createInstance( Components.interfaces.nsIFileOutputStream );
outputStream.init( file, 0x04 | 0x10, 0664, 0 );
var output = "some text here"
var result = outputStream.write( output, output.length );
outputStream.close();
Related
So, after having not programmed for 23 years, I decided to start learning Javascript.
I am trying to write a program to read through my music files and create a HTML page based on the files found in a specific directory.
It goes well until I hit filenames containing diacritics in it (like é, ü, ø etc).
For Example: André Hazes turns into : André Hazes
For Example: Andrea Bocelli & Sarah Brightman - Time to Say Goodbye [Con Te Partirò] (single) turn into Andrea Bocelli & Sarah Brightman - Time to Say Goodbye [Con Te PartiroÌ€] (single)
The link I have created doesn't work anymore
The command I use to create the HTML statement is:
<td>${item.vFilename}</td>
This is the code I use to read the files from the filesystem. I work on a Mac, OS Catalina, so basically an Unix variant.
// List all files in a directory in Node.js recursively in a synchronous fashion
var ReadDirFiles = function(pdir, pfilelist) {
files = vFileSystem.readdirSync(pdir,"utf-8");
filelist = pfilelist;
files.forEach(function(file) {
if (vFileSystem.statSync(pdir + '/' + file).isDirectory()) {
filelist = ReadDirFiles(pdir + '/' + file, filelist);
}
else {
vstats = vFileSystem.statSync(pdir + '/' + file);
// debug info
// console.log(vstats);
filelist.push({vFilename: file, vDir: pdir, vBirthtime: formatDate(vstats.birthtime), vSize: vstats.size});
}
});
return filelist;
};
This is the statement I use to write the output to disk and it turns out the problem is in the write statement:
fs.writeFileSync(buildPathHtml.buildPathHtml(), html);
When the output is written back to disk, the conversion of the diacritics happens.
Anyone knows the trick how to work diacritics?
Try using encoding and decoding functions in your script. There are many functions doing this in Javascript, just use what you need. For a complete encoding/decoding you can use (or only copy & paste) into your script the code suggested in https://www.strictly-software.com/htmlencode/. There is a little encoding Javascript library doing the job (encoder.js).
Maybe your file system isn't utf8 based. I read somewhere it's Western for Mac.
I want to identify few properties during my run and form a json object which I would like to write to a ".json"file and save it on the disk.
var target = UIATarget.localTarget();
var properties = new Object();
var jsonObjectToRecord = {"properties":properties}
jsonObjectToRecord.properties.name = "My App"
UIALogger.logMessage("Pretty Print TEST Log"+jsonObjectToRecord.properties.name);
var str = JSON.stringify(jsonObjectToRecord)
UIALogger.logMessage(str);
// -- CODE TO WRITE THIS JSON TO A FILE AND SAVE ON THE DISK --
I tried :
// Sample code to see if it is possible to write data
// onto some file from my automation script
function WriteToFile()
{
set fso = CreateObject("Scripting.FileSystemObject");
set s = fso.CreateTextFile("/Volumes/DEV/test.txt", True);
s.writeline("HI");
s.writeline("Bye");
s.writeline("-----------------------------");
s.Close();
}
AND
function WriteFile()
{
// Create an instance of StreamWriter to write text to a file.
sw = new StreamWriter("TestFile.txt");
// Add some text to the file.
sw.Write("This is the ");
sw.WriteLine("header for the file.");
sw.WriteLine("-------------------");
// Arbitrary objects can also be written to the file.
sw.Write("The date is: ");
sw.WriteLine(DateTime.Now);
sw.Close();
}
But still unable to read and write data to file from ui automation instruments
Possible Workaround ??
To redirect to the stdout if we can execute a terminal command from my ui automation script. So can we execute a terminal command from the script ?
Haven't Tried :
1. Assuming we can include the library that have those methods and give it a try .
Your assumptions are good, But the XCode UI Automation script is not a full JavaScript.
I don't think you can simply program a normal browser based JavaScript in the XCode UI Automation script.
set fso = CreateObject("Scripting.FileSystemObject");
Is not a JavaScript, it is VBScript which will only work in Microsoft Platforms and testing tools like QTP.
Scripting.FileSystemObject
Is an ActiveX object which only exists in Microsoft Windows
Only few JavaScript functions like basic Math, Array,...etc..Are provided by the Apple JavaScript library, so you are limited to use only the classes provided here https://developer.apple.com/library/ios/documentation/DeveloperTools/Reference/UIAutomationRef/
If you want to do more scripting then Try Selenium IOS Driver http://ios-driver.github.io/ios-driver/
Hey so this is something that I was looking into for a project but never fully got around to implementing so this answer will be more of a guide of what to do than step by step copy and paste.
First you're going to need to create a bash script that writes to a file. This can be as simple as
!/bin/bash
echo $1 >> ${filename.json}
Then you call this from inside your Xcode Instruments UIAutomation tool with
var target = UIATarget.localTarget();
var host = target.host();
var result = host.performTaskWithPathArgumentsTimeout("your/script/path", ["Object description in JSON format"], 5);
Then after your automation ends you can load up the file path on your computer to look at the results.
EDIT: This will enable to write to a file line by line but the actual JSON formatting will be up to you. Looking at some examples I don't think it would be difficult to implement but obviously you'll need to give it some thought at first.
I'm used to selecting, opening and reading a text file (CSV, JSON) from my local directory. I now have access to our data feeds online but don't know how to do the same thing with a URL instead of a local file.
For a local file I simply use something like this:
var myFile = File.openDialog();
myFile.open();
myFile.read();
Is there a way to do this with a feed from a URL in javascript for AE?
Here is one of the feeds: feeds.nfl.com/feeds-rs/schedules.json
No need to save file, just one line do the trick =)
data = JSON.parse(system.callSystem('curl -s "https://path_to_json"'));
I think in After Effects the easiest way is to do a system call.
The Socket object from ExtendScript is complicated. Look at GetURLs.jsx by Rorohiko.
You need to allow your script to access the network. See this stackoverflow
# Mac Osx
var curlcmd = "curl feeds.nfl.com/feeds-rs/schedules.json > ~/Desktop/test/schedules.json";
var stdout = system.callSystem(curlcmd);
$.writeln(stdout);
var file = File("~/Desktop/test/schedules.json");
file.open();
var content = file.read();
file.close();
$.writeln(content);
Is there a possible way to read a local file in JavaScript.
MyFolder:
db.csv
Parse.js
Trying to fetch the contents of file db.csv in Parse.js, But in vain.
Can you share some links where I can get enough knowledge how to read a file.
Running Instruments in Xcode5, with test scripts in .js file where I have to feed in some values from a .csv file.
iOS UIAutomation, apple provides an api for running a task on the target's host.
performTaskWithPathArgumentsTimeout
Using this, we can have a bash script to printout the contents of a file that we wanted to fetch in the first case.
Bash script can be as simple as this for this requirement.
#! /bin/bash
FILE_NAME="$1"
cat $FILE_NAME
Save it as for example FileReader.sh file.
And in your automation script,
var target = UIATarget.localTarget();
var host = target.host();
var result = host.performTaskWithPathArgumentsTimeout(executablePath,[filePath,fileName], 15);
UIALogger.logDebug("exitCode: " + result.exitCode);
UIALogger.logDebug("stdout: " + result.stdout);
UIALogger.logDebug("stderr: " + result.stderr);
where in,
executablePath is where the command need to be executed.
var executablePath = "/bin/sh";
filePath is the location of the created FileReader.sh file. When executed, outputs the content to standard output (in our requirement).
[give full absolute path of the file]
fileName is the actual file to fetch contents from.
[give full absolute path of the file] In my case I had a Contents.csv file, which I had to read.
and the last parameter is the timeout in seconds.
Hope this helps others, trying to fetch contents (reading files) for performing iOS UIAutomation.
References:
https://stackoverflow.com/a/19016573/344798
https://developer.apple.com/library/iOS/documentation/UIAutomation/Reference/UIAHostClassReference/UIAHost/UIAHost.html
If the file is on the same domain as the site you're in, you'd load it with Ajax. If you're using Ajax, it's be something like
$.get('db.csv', function(csvContent){
//process here
});
Just note that the path to the csv file will be relative to the web page you're in, not the JavaScript file.
If you're not using jQuery, you'd have to manually work with an XmlHttpRequest object to do your Ajax call.
And though your question doesn't (seem to) deal with it, if the file is located on a different domain, then you'd have to use either jsonP or CORS.
And, just in case this is your goal, no, you can't, in client side JavaScript open up some sort of Stream and read in a file. That would be a monstrous security vulnerability.
This is a fairly simple function in Illuminator's host functions library:
function readFromFile(path) {
var result = target.host().performTaskWithPathArgumentsTimeout("/bin/cat", [path], 10);
// be verbose if something didn't go well
if (0 != result.exitCode) {
throw new Error("readFromFile failed: " + result.stderr);
}
return result.stdout;
}
If you are using Illuminator, this is host().readFromFile(path).
If you are developing an extension for one of the mozilla applications (e.g. Firefox, Thunderbird, etc.) you define a extension id in the install.rdf.
If for some reason you need to know the extension id e.g. to retrieve the extension dir in local file system (1) or if you want to send it to a webservice (useage statistic) etc. it would be nice to get it from the install.rdf in favour to have it hardcoded in your javascript code.
But how to access the extension id from within my extension?
1) example code:
var extId = "myspecialthunderbirdextid#mydomain.com";
var filename = "install.rdf";
var file = extManager.getInstallLocation(extId).getItemFile(extId, filename);
var fullPathToFile = file.path;
I'm fairly sure the 'hard-coded ID' should never change throughout the lifetime of an extension. That's the entire purpose of the ID: it's unique to that extension, permanently. Just store it as a constant and use that constant in your libraries. There's nothing wrong with that.
What IS bad practice is using the install.rdf, which exists for the sole purpose of... well, installing. Once the extension is developed, the install.rdf file's state is irrelevant and could well be inconsistent.
"An Install Manifest is the file an Add-on Manager-enabled XUL application uses to determine information about an add-on as it is being installed" [1]
To give it an analogy, it's like accessing the memory of a deleted object from an overflow. That object still exists in memory but it's not logically longer relevant and using its data is a really, really bad idea.
[1] https://developer.mozilla.org/en/install_manifests
Like lwburk, I don't think its available through Mozilla's API's, but I have an idea which works, but it seems like a complex hack. The basic steps are:
Set up a custom resource url to point to your extension's base directory
Read the file and parse it into XML
Pull the id out using XPath
Add the following line to your chrome.manifest file
resource packagename-base-dir chrome/../
Then we can grab and parse the file with the following code:
function myId(){
var req = new XMLHttpRequest();
// synchronous request
req.open('GET', "resource://packagename-base-dir/install.rdf", false);
req.send(null);
if( req.status !== 0){
throw("file not found");
}
var data = req.responseText;
// this is so that we can query xpath with namespaces
var nsResolver = function(prefix){
var ns = {
"rdf" : "http://www.w3.org/1999/02/22-rdf-syntax-ns#",
"em" : "http://www.mozilla.org/2004/em-rdf#"
};
return ns[prefix] || null;
};
var parser = CCIN("#mozilla.org/xmlextras/domparser;1", Ci.nsIDOMParser);
var doc = parser.parseFromString(data, "text/xml");
// you might have to change this xpath expression a bit to fit your setup
var myExtId = doc.evaluate("//em:targetApplication//em:id", doc, nsResolver,
Ci.nsIDOMXPathResult.FIRST_ORDERED_NODE_TYPE, null);
return myExtId.singleNodeValue.textContent;
}
I chose to use a XMLHttpRequest(as opposed to simply reading from a file) to retrieve the contents since in Firefox 4, extensions aren't necessarily unzipped. However, XMLHttpRequest will still work if the extension remains packed (haven't tested this, but have read about it).
Please note that resource URL's are shared by all installed extensions, so if packagename-base-dir isn't unique, you'll run into problems. You might be able to leverage Programmatically adding aliases to solve this problem.
This question prompted me to join StackOverflow tonight, and I'm looking forward participating more... I'll be seeing you guys around!
As Firefox now just uses Chrome's WebExtension API, you can use #serg's answer at How to get my extension's id from JavaScript?:
You can get it like this (no extra permissions required) in two
different ways:
Using runtime api: var myid = chrome.runtime.id;
Using i18n api: var myid = chrome.i18n.getMessage("##extension_id");
I can't prove a negative, but I've done some research and I don't think this is possible. Evidence:
This question, which shows that
the nsIExtensionManager interface
expects you to retrieve extension
information by ID
The full nsIExtensionManager interface
description, which shows no
method that helps
The interface does allow you to retrieve a full list of installed extensions, so it's possible to retrieve information about your extension using something other than the ID. See this code, for example:
var em = Cc['#mozilla.org/extensions/manager;1']
.getService(Ci.nsIExtensionManager);
const nsIUpdateItem = Ci.nsIUpdateItem;
var extension_type = nsIUpdateItem.TYPE_EXTENSION;
items = em.getItemList(extension_type, {});
items.forEach(function(item, index, array) {
alert(item.name + " / " + item.id + " version: " + item.version);
});
But you'd still be relying on hardcoded properties, of which the ID is the only one guaranteed to be unique.
Take a look on this add-on, maybe its author could help you, or yourself can figure out:
[Extension Manager] Extended is very
simple to use. After installing, just
open the extension manager by going to
Tools and the clicking Extensions. You
will now see next to each extension
the id of that extension.
(Not compatible yet with Firefox 4.0)
https://addons.mozilla.org/firefox/addon/2195