I am back to writing blog after a gap of nearly two months. Past few weeks had been very busy working on new features of ColdFusion and creating a mobile app for CFSummit2013 . I along with my colleagues Asha and Rakshith had been working on this app for nearly a month, though not necessarily full time. There were two main reasons for building the app 1. to use mobile features of next ColdFusion releases to provide feedback and catch bugs 2. to create a useful app for the attendees of CFSummit.
I don’t know if attendees would find the app useful (I certainly hope so), but we were able to provide a lot of valuable feedback on ColdFusion mobile features and catch many bugs. The application took more time than it should have, because the mobile features of ColdFusion were still under developement. We would hit a bug and would have to wait (if it was a blocking issue) till it got fixed. In many case we did try to work around issues. But the whole process took a lot more time.
But this post is not about mobile features of ColdFusion and how they were used in the CFSummit app. I am going to talk about my experience in creating HTML5 UI for this app, the issues we faced and lessons learnt. Some of the UI issues are still unresolved.
Eclipse 4.3 (Kepler) was released in June this year. There have been some changes to plugin dependencies in E4 RCP applications from version 4.2 to 4.3. I had created sample projects for my book using Eclipse 4.2. If you import those projects in Kepler, then applications would not work.
I have recreated sample projects for Kepler and have submitted them to PACKT. Hopefully they should be available for download soon.
However if you follow through code examples in the book by creating your own E4 RCP projects in Keper, then things should work fine. Eclipse RCP APIs haven’t changed, as far as code samples in the book are concerned. However some of the links mentioned in the book are different for Kepler. They are –
Application project wizard for E4 has changed a bit in Kepler. Now you have an option to create sample content (menus and views) in the wizard. Earlier you did not have this option and the wizard used to created sample content. This option is turned off by default. If you do not select this, then no menu options would be created. In which case skeleton sample application would look different from what is explained in the book.
If you select the above option, then along with menus and toolbar button, a view is also created.
To try out examples in the book, I would recommend that you select the option to create sample content and then delete the ‘Sample Part’ from Application.e4xmi (it is in Application->Windows->Trimmed Window->Controls->Perspective Stack->Perspective->Controls->PartSashContainer->Part Stack). You can also delete Java source file for this part (SamplePart.java in codesnippetapp.parts package). This will make the skeleton project same as explained in the book.
I was discussing with a couple of colleagues of mine yesterday about a HTML5 mobile application that we are developing, and one of the requirements was to pre-package database with the application. A few months back I had created an application that did just that. I thought the solution might be of interest to some of the readers of this blog.
Mobile (or for that matter non-mobile) browsers can create databases for your HTML5 applications. HTML5 provides APIs to create and access the database. But how do you pre-package the database?
The solution is not really to pre-package the database, but package data that you want a database to initialize with. I don’t think you can really package a SQLite database and ask browser to use it for your HTML5 application. You need to create database and tables when the application is run the first time and then load data packaged with the application. Let’s say we want to create a table called ‘person’ and want to populate this table when the application is run the first time. Continue reading “Pre-packaging database with HTML5 Mobile Application”
I use JavaCC for generating parsers in Java. And use JJTree to create AST after parsing. JJTree creates nodes of the AST and you can configure JavaCC options to capture tokens in the node – i.e. if you want each node to contain start and end tokens. The default code generated by JavaCC creates Token class with offsets that are relative to the starting offset of the line. It has fields like beginLine, beginColumn, endLine and endColumn. Here the line numbers are absolute line numbers (starting from 1) and column fields contain offsets (again starting from 1) within the corresponding lines.
However many times you want to capture absolute offsets of tokens in the input stream, and not just relative offset in the line. I wish there was a JavaCC option to enable this. But it is not too complex if you want to do it yourself.
Last year I blogged about creating a recordable HTML5 Canvas. I explained how to record strokes/drawings created on a HTML5 Canvas and play them back. There are a couple of comments on that post asking me to explain how to save recordings and load them back. Serialization and deserialization of recordings was on my to-do list for a long time and finally this week I got around to implement it. There are different ways to serialize and deserialize recordings, and I have implemented a simple method – using JSON. The serialized data is a bit verbose because of descriptive variable names I have used, but you can change that easily.
RecordableDrawing (in this script file) function remains unchanged. I have added a new file drawingSerializer.js . Two important functions in this file are serializeDrawing and deserializeDrawing.
serializeDrawing takes a RecordableDrawing object as argument and returns JSON string containing array of recordings.
deserializeDrawing takes a String (serialized data) as argument and returns array of Recording objects.
I am currently building a parser using JavaCC. I have used JavaCC in the past, but whenever I use it after a long gap, I have to relearn a few things about it – particularly handling warnings. So I thought this time I would blog about ways to handle some of the frequent warnings that I have seen.
If you are unfamiliar with JavaCC, then it is a parser generator. You create grammer using EBNF (Extended Backus-Naur Form) and feed it to JavaCC. JavaCC then creates Java classes for the parser. I do not want to make this post into JavaCC tutorial. There are some very good tutorials available at JavaCC Documentation page and FAQ. I especially find Lookahead MiniTutorial and Token Manager MiniTutorial very useful. If you use Eclipse IDE, then you would find JavaCC plugin for Eclipse useful – it provides wizard to create JavaCC or JJTree (JJTree creates AST, Abstract Syntax Tree, after parsing the input) files, provides code colorization, outline, code hyper link, syntax checking and compilation. You can also set JavaCC debug options easily using this plugin.
I will use following tokens that are generated by default if you use the wizard provided by JavaCC Eclipse plugin to create a JavaCC grammer file. I have created a .jjt file for examples in this blog.
I am happy to announce that my book titled Instant Eclipse 4 RCP Development How-to has been published by PACKT Publishing. It is a mini-book that describes specific tasks and solutions to build RCP applications using Eclipse 4.
Eclipse 4 has introduced new frameworks, APIs and tools to develop Rich Client Platform (RCP) Applications. In this book I have taken a sample application and explained how to implement it from start to finish using Eclipse 4 SDK. The book is divided into a number of focused tasks. Each task builds the sample application incrementally .
Here is the list of tasks covered in the book –
Setting up a development environment
Creating a skeleton E4 application
Adding menu and toolbar items
Injecting your own objects using DI
Creating a pop-up menu
Creating custom events and handlers
Adding a keyboard shortcut
Creating custom objects using DI
Creating views dynamically
Styling an application using CSS
Customizing and exporting the application
I was contacted to write this book towards the end of last year. Initially it got delayed a bit because I was busy with other things. It was finally published yesterday.
I have always configured and used Android SDK from Eclipse. In fact, ADT (Android Development Kit) bundle now comes with Eclipse IDE pre-configured for Android development. This makes using the SDK very easy. Options like configuring SDK manager and virtual device manager are available right from Eclipse menu.
However, I had a requirement where I wanted to run only Android emulator and not IDE. The Android SDK site has all the information to do this, but not in one easy to find place. So I thought I will write about it, for my own reference.
Last year I had blogged about animating sprite using Kinetic JS.
Code in that post was part of a simple game I had created. So the code specific for animating sprite was explained in snippets. A reader of that post contacted me with a request to provide a complete example. So I created a small demo of sprite animation only. If you are interested, you can download it from here. This demo animates images in the sprite sheet at a fixed location, it does not move the image. So I thought it would be interesting to add motion to sprite images when they are animated.
First, take a look at the demo. Click on the Walk button to make the person walk. You can stop any time. If the person hits right side wall, then she falls, which is also simulated using sprite animation. I know the ‘walk’ does not look natural, but creating graphics is not my string point and this is the best I could manage.
Parsing with Rhino is quite simple and well documented.