Archive for the ‘web development’ Category

Databud – Startup Weekend Kampala – April 27 to 29

I will be attending my first startup weekend in Kampala, on April 27, 2012 to April 29, 2012 and well I thought that why not share my pitch and get advice on how to refine it. No idea is great unless shared right?

In the absence of #opendata in Uganda, there is a whole lot of data locked up within individual government systems, documents, in non standard formats which needs to be unlocked, the data set free so that it can grow (Data Bud) – the data buds and grow

A picture is worth a thousand words right – below is the whole concept

Data Bud Concept

Data Bud Concept

Comments, additions, advice? Looking forward to seeing ya this weekend

The Poor Man’s Job Queue

Not all software development projects are treated the same, some have access to modern tech Virtual Private Servers (VPS), Zend Server (http://www.zend.com/products/server/), Memcached, Gearman and all the other goodies I can only dream of. You have a box with LAMP, and you cannot install anything else.

This is an example of how we got around a limitation, using available tools. Problem: I have a list of tasks to execute within my application, however I need to ensure that the tasks are executed and completed, but some are more important than others, and the execution may slow down the performance of the box we are running on. Well in this case we were loading 6 different types of XML files which were FTPed into a location on the box regularly, every 35 minutes and had to be loaded in a specific order. This was further complicated by the fact that we had to reload historical data in case of issues (1 weeks worth of uploads ~ 2100 files) without interrupting the current loading processes.

The approach used the following components:

a) Job Queue – based on the Zend Server Job Queue but simplified for our needs (see data model of tables below)

Job Queue Data Model

Job Queue Data Model

b) Queue Loader Script – loads the jobs into the job queue by scanning the location containing the files to be loaded and adds the files to the queue (since the queue is a database table, duplicates are discarded without errors) This keeps this file simple and honest

c) Job Executor Script – reads a message from a queue, reads the message body which contains the file name to be processed, could be made more complex

d) Queue Loader Cron Job – calls the Queue Loader Script to add new files to the queue

e) Job Executor Cron Job – calls the job executor script. This job has no effect if a lock file exists, and is not expired which means the script is valid and running. However if the lock file is expired, it means that the process crashed, so the lock file is deleted, a new process is started with a lock file. Basically this keeps the job executor script running indefinitely as long as there are messages to process. 

Please feel free to leave a comment on what your experiences are with similar problems. 

HTML Web Toolkits – Twitter Bootstrap

Well it seems like I have been doing a rip and replace of most of the “homegrown” components that we are using in our environments as we gear up for the new year. The latest to fall to the axe is our HTML + CSS + widget collection that has been growing over the years which was a collection of code from different places:

  1. reset.css by Tantek Celik (http://www.cssreset.com/scripts/undohtml-css-tantek-celik/) – to provide a base for initial browser compatibility
  2. Multiple column layouts from the Dynamic Drive (http://www.dynamicdrive.com/style/layouts/category/C9/
  3. JQuery and JQuery UI widgets – here and there 
  4. Input Button and Link styles inspired by Particle Tree (http://particletree.com/features/rediscovering-the-button-element/)

Well like all web development shops, you pick your poison and stick with it, but with the troubles we have been having with different IE versions, I set out to find out whether we had to go through all the pain of tweaking each project, or whether we could re-use existing solutions. Well my search led me to Twitter Bootrap (http://twitter.github.com/bootstrap/index.html) which was at version 1.4 at the time, and 2.x was in beta and HTML5 Boilerplate (http://html5boilerplate.com/) both of which seemed to be great starting points.

However Bootstrap won out, in the end, because it provided a responsive UI for different screen sizes, pre-built basic elements which we use a lot, buttons, forms, tables, dialogs, message alerts which were easily setup. Okay JQuery UI has great widgets but they are a pain to setup configure and use, the code for a dialog is ~20 lines with CSS, etc, very powerful but no everybody needs a V-8 engine.

First on the agenda was to understand the grid layouts, okay I have heard about 960 and Blueprint grids, but I never understood how they worked, but wow!!! This is amazing, all I need to do is add a few pre-defined grid classes and you can get 2 column, 3 column layouts, our home grown frameworks would never do this … I liked it, coming from an engineering background and lots of database work I think rows, columns and relationships, and the grid well, had me at Hello World 🙂 

The built with bootstrap site (http://builtwithbootstrap.com/) provides a great starting point to see what is possible and available. We decided to use a free theme for a project we are starting and its working for us. 

As we walk this journey, the next challenge is how to handle the tables vs table-less designs for forms. Please drop me a line and let me know what your take either here or on my Stack Overflow question at http://stackoverflow.com/questions/10072991/responsive-html-page-design-pc-tablet-smartphone-with-table-or-table-less-fo 

Doctrine2 Day 3 – Proxies, Associations, Relationships

Well if you are following this series, then by know you are aware that we have the validators setup, and we are almost ready to go. Well not quite so. I ran into an issue with proxies and class loaders which took a while to resolve, but what I did was:

a) Changed the Zend Framework-Doctrine2 integration to the Bisna integration (https://github.com/guilhermeblanco/ZendFramework1-Doctrine2) – note the additional configurations before

b) Learnt that DO NOT SAVE ANY MODELS IN SESSION OR CACHES due to Proxy auto loading issues

c) Develop unit tests for the model validations and saving as you go along because they will save you as you shift things around . I am currently trying to save an association to the database but due to the tests that I have running I can test out the different association mappings without issues because I track the changes using my unit tests.

Now onto relationships, well what I found was a follows:

  1. I have been having a tough time dealing with relationships from Doctrine 1 which autoloaded relationships, however with with the Doctrine2 data mapper, you have to auto load the associations your self
  2. In Doctrine 2, you cannot define the foreign key column and relationship as this creates problems for the ORM mapper, so since I need to access the foreign key value for example personid, and the related object person, my approach has been as follows:
    • Added a person instance and relationship mapping to the person as per the Doctrine
    • Add a setter setPersonID() which loads the Person from the database using the provided ID and sets it to the person relationship provided
    • Added a getter getPersonID() which obtains the id of the person for use on the screen
  3. As recommended avoid bi-directional relationships where possible, try to keep them as uni-directional as possible.

Next I will be implementing a nested hierarchical structure using the tree nested set implementation at http://www.gediminasm.org/

Doctrine 2 – Day 2 – Model Validation using Symfony Validator Service in Zend Framework

It is just Day 2 of my experiences in the trenches, regular work, had kept me from this but I managed to get some time to keep digging. As a followup to my Doctrine 2 – Day 1 – Commentary from the Trenches. The models are up and configured, and the unit testing is setup following steps from Michelangelo van Dam’s presentation (http://www.slideshare.net/DragonBe/unit-testing-zend-framework-apps).

One of the major process that we are implementing with this migration is detailed unit testing which was tougher with the Doctrine1. With the unit test infrastructure setup, the next item on the agenda is model validation. From previous experience (before Doctrine 1) and with Doctrine1 it is critical to be able to specify validation rules using annotations without having to write PHP if statements. Being a ZF person, the first step to look was the ZF validator classes. While they seem to be well integrated with the forms, they would prove to be too verbose to use for validation of models, since I also need to be able to specify multiple validators per column, and this would not cut it for me.

Next stop was Symfony2 validator service (http://symfony.com/doc/current/book/validation.html) provides validators with annotations support. So that was the easy part, the hard part was yet to come integration. The integration followed the steps below:

a) Add Symfony Validator service to library folder, easy, just download a package from https://github.com/symfony/Validator

b) Register the Symfony Validator annotations – this is where I had problems (more later)

c) Add the Symfony Validators to the model properties

d) Add validation code which needed a validate() method in the base class from which all entities are derived, which requires @ORM\HasLifecycleCallbacks (so that the model can hook into the lifecyle call back models) @ORMPrePersist and @ORM\PreUpdate for the validate method to ensure its called before the models are saved (first time) or updated. More details on the Doctrine annotations can be found at http://docs.doctrine-project.org/projects/doctrine-orm/en/latest/reference/annotations-reference.html

So what problems did I face.

1. From the Doctrine documentation, you use the annotations directly, for example @Id, however experience has shown that you need to namespace and alias them see note from Symfony integration http://symfony.com/doc/current/book/doctrine.html . So I had to change all the Doctrine annotations to use @ORM namespace

2. The default annotation driver only supports a single namespace so you will need to update as per the pastebin below

http://pastebin.com/embed_iframe.php?i=feiKsVxg

Now I am a happy camper, got my models working using Symfony validations, we only have to write code for custom validations which happens only about 20 – 30% of the time.

As a parting shot, the Symfony team and community have done a great job for PHP, why because they provide standalone components (similar to Zend Framework), but each of their components can be used without the rest of the framework. As I was investigating the validator usage and issues, I found a thread where Fabien Potencier and team were discussing annotation support in Symfony. However they also noted that Doctrine Commons had better support, so they stopped the work on Symfony annotation support and just used the services of the Doctrine team. This is how all software development should be done, and is a torch to the rest of us. I am a convert, and happy to be a proud member of the PHP community.

Update May 8, 2012

I had promised to provide some sample snippets of what I am using for the integration with Bisna integration and Symfony validation that I ended up using so here we go https://gist.github.com/2638526 The files are as follows:

a) application.ini – there is nothing special here from Bisna. Includes the cache configuration, prod/staging/dev/test environments all of which inherit from production

b) index.php – from the public folder or htdocs – this may not be perfect but it works and am looking for ways to simplify it

c) Bootstrap – this is the file we use, highlight:

– Storing the entity manager instance in the Zend_Registry, we have a utility method which loads it from the registry and another which also provides a connection from the entity manager so its fully encapsulated

– intialization of the Zend_DB adapter, we need this since we are using the Zend_Session to save the sessions in the database

– the last config is for other resources we use. We have a dependency on the Zend_Registry class as it hides a lot of complexity

d) Document.php – a sample model class

– it extends BaseEntity which provides automatic getters and setters through the __call method, some required fields like id, datecreated, last  update date and last updated by (not all models extend this class only those which need those auditing fields)

– Why do we have getCategoryID(), setCategoryID() for the category property instead of mapping the categoryid field from the database see the next post in the series https://ssmusoke.wordpress.com/2012/03/25/doctrine2-day-3-proxies-associations-relationships/

– Unlike the Doctrine2 defaults we do not use tablename_id but rather tablenameid for the foreign keys so we have define them in each relationship.

Please let me know what I can do to make this any clearer, thanks for reading

%d bloggers like this: