Passing multiple parameters from AppleScript to Node.js Server

Most previous solutions to moving data from AppleScript to a server on the same machine involve MAMP/PHP and passing the parameters via a GET or POST to the URL. It works, but it is slow and awkward. I have found another solution. The concept is to use Applescript to put all of the parameters into a single JSON string, and put the string on the system clipboard. Then open the URL. If your server is set up with node.js with express.js framework, you can retrieve the JSON data, and pass all the parameters to the jade node template engine. With very little code, you can use the data for display, encryption, messaging, or storage in SQL. Tricks to help make this happen: Enclose the JSON elements (both the keys and the values) in tildes instead of quotes before passing to clipboard. (Also clean out all the illegal characters like newlines. I like having them at the other end, so I substitute triple-dashes for every newline and change it back on the other end.) In node.js, FIRST, replace the tildes with double quotes, and change the newline placeholders to \n. THEN use json = JSON.parse(jsonstr); to turn it from a JSON string to a JSON object. THEN use res.render(‘Jadefilename’, json); in the routes/index.js file. Inside the jade template, the jsons value can be called simply with #{key} To pass multiple jsons at once, use res.render('Jadefilename', {item1:json1, item2:json2}; etc, and inside jade the values correspond to #{item1.key} An important afterthought: these passed json objects received by jade are available in the regular jade template but INACCESSIBLE inside javascript functions nested in jade. The correction to this problem is add the following one line to the top of your nested javascript function to make the json object into a local variable: var local_json =!{JSON.stringify(imported_json)}; Now you can call the values from local_json, instead of imported_json, (item1 in the example above) inside the javascript that is nested inside the jade template. While on the subject of node.js, here is a refresher on how to get started with express. first time: install node from nodejs.org then “sudo npm install” the following: -g express generator -g nodemon mysql crypto-js copy-paste returning - check install, versions: node -v npm -v express -V (yes that one is capitalized) nodemon -v And here is what you really need to know: create an express project: express projectname-version start application with nodemon: cd projectname-version nodemon bin/www start application with debug: DEBUG=my-application ./bin/www IDE: Sublime Text 2, Terminal, Safari, Applescript Editor, Remote Desktop Connection

Setting up Java Eclipse IDE on Windows 7 and JRE

After installing and uninstalling both Eclipse and JRE (Java runtime environment) several times on a Windows 7 64-bit machine, I finally have a sequence I can recommend. In my case, I had to have my testing environment in eclipse be compatible with the runtime environment on a Windows Server 2008 R2 without access to the JRE on that machine. This meant it had to be JRE 1.6 and 32 bit. (Even though the default installation on the Windows 7 machine was JRE 1.7 and 64 bit). 1. Uninstall all the implementations of java from the control panel uninstall utility. 2. You can't uninstall the Eclipse IDE, you can just delete it. 3. Install the JRE (1.6.0.6) and verify it is working by going to the program file (x86) folder and/or the windows\syswow64 folder (these are where 32-bit programs reside on 64-bit machines) and from the command line java -version. You can also verify it's installed by going into control panel ->java->java->Application runtime settings->view. 4. Install the IDE but be sure to download the 32-bit version. I downloaded it to a mac, put it in dropbox, let it sync for about 20 minutes on the target machine, and then copied the whole folder (after renaming it eclipse_32_bit) from dropbox to desktop on Windows 7. You can still have eclipse_64_bit on the same machine, in another folder. 5. Launch the IDE, select a default workspace folder, open the workbench, and immediately go to Window->Preferences->Java->Installed JREs->Execution Evironments. 6. Click on JavaSE-1.6 and here is where the magic is. In the compatible JREs pane the freshly installed jre1.6.0.06 is there, and not only that (it is there for other execution environments) it says "perfect match". 7. Click the compatible jre1.6.0.06 checkbox. This should give you the correct JRE for testing your application within Eclipse.

Eclipse ADT (Build v22.2.1 for Mac OS) Pointers

When you create a new app from scratch (file->new->Android Application Project) and select the defaults, it automatically opens the java file (from the src folder) for editing, and the main xml file (from the res->layout folder) for editing. The xml file has two panes, text and graphical layout. Start with the graphical layout to drag and drop objects like fields and buttons onto your UI. Then edit the xml text and the java file to handle events with the objects. The YouTube tutorials on this, called "Android: project 1(v1)", up to (v6), by "Lecture Snippets" are very efficient and provide simple java code for this. The YouTube tutorial "Android Basic SQLite Database" by "profgustin" gives some basic SQLite code, which I have enhanced below to interact with a field from Lecture Snippets. The main thing not to forget is: although SQLite is "built-in" to the Android Developer Tools SDK, the new app cannot use the classes until they are imported in the java code AND attached to the build path. I found that typing SQLiteDatabase in the java code, and either hovering over it or double clicking on it brought up a dialog box that allowed me to both make the import statement in the code, and add the android.jar file as a Source Attachment in the Java Build Path (this is otherwise found by right-clicking the project file in the navigator pane-> Properties->Java Build Class->Android. I use the Cursor class to manipulate SQL data, so I also had to import android.database.Cursor; as well. I think import android.database.*; may also work. package com.example.foo2; import android.os.Bundle; import android.app.Activity; import android.database.Cursor; import android.database.sqlite.SQLiteDatabase; import android.view.*; import android.widget.*; public class MainActivity extends Activity { Button btn1; EditText etxt1; @Override protected void onCreate(Bundle savedInstanceState) { super.onCreate(savedInstanceState); setContentView(R.layout.activity_main); SQLiteDatabase db = openOrCreateDatabase("MyDB", MODE_PRIVATE, null); db.execSQL("CREATE TABLE IF NOT EXISTS MyTable (field1 VARCHAR, field2 VARCHAR, field3 VARCHAR);"); db.execSQL("INSERT INTO MyTable VALUES('red','green','blue');"); Cursor c = db.rawQuery("SELECT field1 FROM MyTable", null); c.moveToFirst(); final String MyString = c.getString(c.getColumnIndex("field1")); db.close(); btn1=(Button)findViewById(R.id.button1); etxt1=(EditText)findViewById(R.id.editText1); btn1.setOnClickListener(new View.OnClickListener() { @Override public void onClick(View v) { etxt1.setText(MyString); } }); }

ADT - Android Developer Tools - with Eclipse: Starters.

Here is a refresher on how to start a new project by copying an old one. In my case, I keep my projects on a Mac in Documents->Android->Workspace folder. That is my default folder for new projects. For this example I have a same, working project that compiles with no errors and runs on a Kindle via USB cable called "Notepad". You must be sure to open the Eclipse IDE correctly. I select Eclipse from Spotlight, but I have more than one installation. It has to be the ADT version. Once Eclipse is open, in the Navigator bar on the left one has to be sure the project being copied is "open" (either double click it, or ctrl-click->open project. Can close projects the same way, for faster loading of Eclipse). 1. Highlight open project -> cmd-c -> unhighlight -> cmd-v -> select name for the project. 2. ctrl-click project -> Android Tools -> Rename Application Package -> follow instructions. 3. Navigate to res->values->strings.xml->app.name to rename the app for the device. This creates a new project in the default folder, opens it in Eclipse, allows you to independently name the folder it is in, the application package, and the app on the target device. All of 3 of these names can be the same or be named independently. This allows one to revert to a project in a skeleton form, but has all the right settings, jars, and so on to allow error-free compilation and installation to the desired target device. To run, click the run button. To stop from running, go to the app page on the target device, hold your finger on the app icon to delete the app.

SSH Server for Windows XP

For a Mac, turning on SSH server is built-in and is just one click inside System Preferences -> Sharing Pane -> Remote Login -> just check the checkbox and it even tells you the command line syntax to use to SSH in from the client. For Windows it is more work: MobaSSH Server is the software I selected. Install it last. First, go into control panel -> Windows Firewall and add exception: name it SSH and use port 22. (I also turned on remote desktop in the same pane, not sure it was necessary but I read that on a forum somewhere) Second, don't make the mistake I made. I changed my username in control panel -> Users to make it all one word with no spaces, thinking that the two word name would mess up the SSH client at the command line. WRONG! Because when you change your user name, it does not change your user file directory. (If it did it would mess up everything you have installed on the computer). Well it turns out MobaSSH uses your user name from the file pathway, instead of from the login, if they do not match. That problem had me stumped for two weeks. Make sure you have a password for your Windows login. Third, install MobaSSH. From the Client (I used Terminal (Unix) from a Mac): ssh "firstusername lastusername@192.168.1.xxx" With the quotes!!!

Creating Icon for iPhone app in Xcode 4.4

Create a png image with name below resize it to 57x57 using from command prompt (terminal): sips -z 57 57 icon.png Drag and drop to the project by clicking on the target, opening the summary panel, and dragging and dropping to the app icon section. Will test this with other image file names, sizes, and formats, but for now, this works

Click on Element within a Frame using AHK

The syntax is tricky and threw me off this project for months. I finally have cleared it up, and here it is for posterity:

The syntax for plain javascript, executable from internet explorer DebugBar v5.4.1 is:
document.parentWindow.actualframename.document.getElementByID("actualID").click()

In distinction, the syntax for AHK COM SCRIPTS is:
COM_Invoke(pwb, "document.parentWindow.actualframename.document.getElementByID[actualID].click")

NOTICE (I usually don't shout, but in this case I will make an exception): THE ARGUMENT OF getElementById IN JAVASCRIPT IS ENCLOSED IN PARENTHESES AND QUOTES. THE SAME ARGUMENT IN AHK IS ENCLOSED INSTEAD IN BRACKETS AND NO QUOTES. ALSO THE METHOD IN JAVASCRIPT HAS PARENTHESES AFTER IT click() WHEREAS IN AHK IT DOES NOT click (THE REASON FOR THESE DIFFERENCES IS THE AHK COM_Invoke FUNCTION, WHICH REQUIRES ARGUMENTS CONTAINED IN PARENTHESES AND QUOTES.)

Also it is worth saying what is the quickest way to find the actual frame name (yes it is the name, not the ID) and the actual element ID (yes ID, not the name). Answer: A magnificent AHK script called iWebBrowser2 Learner Build 2.5, by "Tank, Jethrow, and Sinkfaze" of the AHK forum. Just hover over the element until it has a red box around it, and read the answer in the dialog box.

More info can be found at these forum posts: (primer on COM functions, and link to the all-important GetWebBrowser() function the COM functions depend on)
http://www.autohotkey.com/forum/topic51020.html
http://www.autohotkey.com/forum/topic24930.html

Thank you Sean, Tank, Jethrow, and Sinkfaze, whoever you are.