Tag: Tutorials

PHPainfree Test Drive

June 7, 2010 » Consume, Geek

What is PHPainfree

PHPainfree is a relatively young PHP framework written by Eric Ryan Harrison.

The README for the project is in first person, so I’ll let PHPainfree explain what it is:

I am an ultra-lightweight PHP framework. I am inspired by the MVC concept, but I’m too artsy to let myself be defined by labels like that. I basically do what I want.


To try the framework out I cloned the git repository. This should be roughly equivalent to version 0.6.3, in case you want to follow along at home.

My setup is a bit unique, as a trial usage in a sub-directory of my local machine, but you should be able to adjust your install to suit.

It essentially boiled down to these steps:

  1. Clone Repository
  2. Edit includes/PainfreeConfig.php
  3. Symlink includes/ and templates/ into htdocs/
  4. Tweak RewriteBase in htdocs/.htaccess

Shell Transcript

[email protected]:/var/www/localhost/PHPainfree$ ls
CHANGELOG.md  htdocs  includes  LICENSE  README.md  templates
[email protected]:/var/www/localhost/PHPainfree$ cd includes/
[email protected]:/var/www/localhost/PHPainfree/includes$ cp PainfreeConfig-GENERIC.php PainfreeConfig.php
[email protected]:/var/www/localhost/PHPainfree/includes$ vim PainfreeConfig.php
[email protected]:/var/www/localhost/PHPainfree/includes$ cd ..
[email protected]:/var/www/localhost/PHPainfree$ cd htdocs/
[email protected]:/var/www/localhost/PHPainfree/htdocs$ ln -s ../includes/ .
[email protected]:/var/www/localhost/PHPainfree/htdocs$ ln -s ../templates/ .
[email protected]:/var/www/localhost/PHPainfree/htdocs$ ls -a
.  ..  css  .htaccess  images  includes  index.php  js  templates
[email protected]:/var/www/localhost/PHPainfree/htdocs$ vim .htaccess

Listing: htdocs/.htaccess

RewriteEngine On
RewriteBase /PHPainfree/htdocs/

RewriteRule ^js/(.+)$ js/$1 [L]
RewriteRule ^css/(.+)$ css/$1 [L]
RewriteRule ^images/(.+)$ images/$1 [L]

RewriteCond %{SCRIPT_FILENAME} !-f
RewriteCond %{SCRIPT_FILENAME} !-d
RewriteRule ^(.+)$ index.php?route=$1&%{QUERY_STIRNG} [L]

After that is all done, it should happily serve up it’s welcome page.

It Works!
It Works!

Framework Basics

When PHPainfree claim’s to be ultra-lightweight, they mean it. Many of the bits and pieces you would expect on a framework just don’t exist. Many.

But more on that later. For now, let’s take apart the default files and build something out of them. What we’ll attempt to assemble is that paragon of beginner programs, the todo list.

Looking at the provided example files it really seems to me that this is a very view driven framework. The “logic” part runs first, but really just sets up things for the “view” part. Model and controller seem smashed together into the “logic” files, but this is just my interpretation of the design.

This is how the provided example files flow:

Logic Setup => Template(s) Run => Calls Logic Methods

Getting Started


According to includes/PainfreeConfig.php the BaseView “is the name of your base template inside of the templates folder. This base view generally provides the overall framework of output for your application”.

To feel out how the framework handles I created a very small stub BaseView in templates/layout.tpl

Listing: templates/layout.tpl

    <?php echo $Controller->title(); ?>

title(); ?>


According to includes/PainfreeConfig.php the ApplicationController is “…the primary controller for your application”.

I created a new controller for my test, taken almost entirely from the provided includes/Generic.php.

Listing: includes/ToDoList.php


The last step for this first exploratory version is to change some variables in our configuration file.

Listing: includes/PainfreeConfig.php

$PainfreeConfig = array(

// ApplicationController is the primary controller for your
// application. Generic.php is provided, but doesn't do anything
// except look for the view in the templates/ folder
'ApplicationController' => 'ToDoList.php',

// BaseView is the name of your base template inside of the templates
// folder. This base view generally provides the overall framework
// of output for your application
'BaseView' => 'layout.tpl',

Once that is done, it should now be serving my new files.


Building The Application

That's great and all, and we learned about the processing pipeline and stuff, but really, we didn't do anything.

So let's get down to it. I like a nice MVC pattern, with convention over configuration, so here is how I'm laying out my application. Note that you do not have to do it this way, PHPainfree is written to encourage you to do it, well, just about any way you want.

+-- includes
|   +-- Autoload
|   |   `-- BaseController.php
|   +-- controllers
|   |   +-- list.php
|   |   `-- todo.php
|   +-- main.php
|   `-- PainfreeConfig.php
`-- templates
    +-- layout.tpl
    +-- list
    |   `-- index.tpl
    `-- todo
        `-- index.tpl

Routing Requests

To make my structure work, I had to create my own routing system. I couldn't find anything built into PHPainfree that would do this for me, but that's okay because it's pretty simple. I set my ApplicationController option to "main.php" and placed this in there.

Listing: includes/main.php

route );

  if( 1 <= count( $path ) )
    $controller = preg_replace( '/[^a-z0-9_-]/', '', strtolower( $path[0] ) );

  if( 2 <= count( $path ) )
    $method = preg_replace( '/[^a-z0-9_-]/', '', strtolower( $path[1] ) );

  if( 3 <= count( $path ) )
    $arguments = array_slice( $path, 2 );

  $Controller = new Base_Controller(); // Will be replaced by a real controller
  require_once( 'controllers/' . $controller . '.php' );


Note that on line 18 I instantiate a class called BaseController. This is a stub class that I created for all of my controllers to inherit from, that way I have a consistent interface to call in my templates.

My BaseController.php file will be placed into includes/Autoload to take advantage of the loading feature of PHPainfree. Any file placed into the includes/Autoload folder will be automatically included at runtime, just after the configuration file is loaded and just before the logic file is ran. This is useful for loading libraries, or to do some request pre-processing.

Listing: includes/Autoload/BaseController.php

      // Make the controller available to the view
      $Controller =& $this;
        $PainfreeConfig['TemplateFolder'] . '/' .
        // My controller classes all end in "_controller", so I cut that off here
        strtolower( substr( get_called_class(), 0, -11 ) ) .
        '/' . $method . '.tpl'


The Database

Up to this point I haven't touched a database, which PHPainfree has some support for. Real quick I'll set up a MySQL database for our ToDo application.

MySQL Transcript

mysql> create database todolist;
Query OK, 1 row affected (0.02 sec)

mysql> grant all on todolist.* to [email protected] identified by 'password';
Query OK, 0 rows affected (0.05 sec)

mysql> flush privileges;
Query OK, 0 rows affected (0.03 sec)

mysql> use todolist;
Database changed
Query OK, 0 rows affected (0.01 sec)

mysql> CREATE TABLE todos ( id INT(6) UNSIGNED AUTO_INCREMENT, list_id INT(6) UNSIGNED, title VARCHAR(255), created DATETIME, completed DATETIME, PRIMARY KEY(id) );
Query OK, 0 rows affected (0.01 sec)

mysql> INSERT INTO lists ( name ) VALUES ( 'Yard Work' );
Query OK, 1 row affected (0.00 sec)

mysql> INSERT INTO todos ( list_id, title, created ) VALUES ( 1, 'Mow Grass', NOW() ), ( 1, 'Weed Garden', NOW() );
Query OK, 2 rows affected (0.00 sec)
Records: 2  Duplicates: 0  Warnings: 0


Configuring The Database

Configuring the database connection in PHPainfree is relatively straightforward. Just open up includes/PainfreeConfig.php and find the Database key. This is an array of MySQL connections, which cascade if they fail.

For instance, if you have a development environment and a production environment, you could place your dev configuration after the production configuration.

In development, the production connection would fail and then load the development configuration. Nothing to change, no environment variables to set, it just works.

Listing: includes/PainfreeConfig.php

  'Database' => array(
    'Primary' => array(
      'type'   => 'mysql',
      'host'   => 'localhost',
      'user'   => 'painfree',
      'pass'   => 'password',
      'schema' => 'todolist',
      'port'   => 3306

Using The Database

Using the database is easy too. The $Painfree global variable has a member called db which provides access to our configured database. But what is $Painfree->db? Well, a little bit of digging into the PHPainfree core and we find out it is just a normal MySQLi link object. Nothing fancy, no database abstractions.

Listing: includes/core/DBD/mysql.php

Bringing It Together

Applying all this knowledge and configuration, let's start our first controller, the List_Controller. This first version will simply fetch all the active lists from the database and get them ready for the template.

Listing: includes/controllers/list.php

db->prepare( "SELECT id, name FROM lists" );
      $stmt->bind_result( $id, $name );
      while( $stmt->fetch() )
        $lists[] = array( 'id' => $id, 'name' => $name );

      return $lists;


  $Controller = new List_Controller();

Now we need to make our template to use this controller. Again, very basic.

Listing: templates/list/index.tpl

ToDo Lists

    lists() as $list ): ?>

At this point, we should be able to render this view:

A List of Lists
A List of Lists

Finishing Up

From here it is a short work to finish the application with another controller and a few more actions. Rather than post a bunch of reptitive code snippets I'll provide my completed source, here.

A List of ToDo's
A List of ToDo's


So, there is my first PHPainfree application. As with any new tool, my usage is probably flawed until I learn more about it. So take this review with a grain of salt.


PHPainfree is a young framework, and it's a thin one. Coming from a heavier framework background, it feels too thin to me. I missed having an ORM, and built in helpers (think Form, Link, Validation). Also, there is no real exception stack that I could find, just $Painfree->debug() for you to use.

MySQL is the only option right now, though it is easily extended. For example, I wrote this in just a few seconds to add SQLite3 support.

Listing: includes/core/DBD/sqlite.php

However, having multiple drivers is shallow when there is no abstraction element. Since it uses native driver objects, I can't just switch from MySQL to SQLite3, because I would then have to switch all of my method calls. Using PDO would be a good option for PHPainfree, IMHO.

My other qualm is the rendering stream. I'm used to the standard MVC pattern, where the controller fetches data with models and publishes it via views. There may be a way to work like that in PHPainfree, but it's not readily apparent.


It's light, at the cost of including minimal features. And it's fairly easy to understand. According to SLOCCount there are only 101 lines of source (after removing the config file and default controller). You can read the whole framework in a few minutes.

Really, I think this is a framework to build your own framework. The core idea of PHPainfree is to stay out of your way. If I intended to use PHPainfree on a regular basis, I would set it up the way I like it, dumping libraries into includes/Autoload and then keep a tracking version in git with all my addons.


I think that where you can draw the most value from this framework is building something you love on top of this common core code. So give it a try at http://github.com/februaryfalling/PHPainfree

Looking up words in a Dictionary using Python

March 1, 2010 » Consume, Geek

First off, I do not mean dictionary in the Python sense of the word. I mean dictionary in the glossary sense, like Merriam-Webster. This collision of terminology makes Googling for this functionality particularly difficult and frustrating.

I came across three useful Python solutions, and I’m going to detail usage of two of them in this post.

Option 1: NLTK + Wordnet

First up is accessing Wordnet.

“Wordnet is a large lexical database of English…”

The only Python way of accessing this (that I came across) is NLTK, a set of

“Open source Python modules, linguistic data and documentation for research and development in natural language processing…”

Getting NLTK Installed

For various reasons, NLTK is not packaged by Debian, so I had to install it by hand. Even if your distro does package NLTK, you might want to read this bit anyway. Installing was a cinch with easy_install nltk. However, this does not install the corpus (where wordnet is stored). As shown below:

>>> from nltk.corpus import wordnet
>>> wordnet.synsets( 'cake' )
Traceback (most recent call last):
  File "<stdin>", line 1, in <module>
  File "/usr/lib/python2.5/site-packages/nltk-2.0b8-py2.5.egg/nltk/corpus/util.py", line 68, in __getattr__
  File "/usr/lib/python2.5/site-packages/nltk-2.0b8-py2.5.egg/nltk/corpus/util.py", line 56, in __load
    except LookupError: raise e
  Resource 'corpora/wordnet' not found.  Please use the NLTK
  Downloader to obtain the resource: >>> nltk.download().
  Searched in:
    - '/home/jmhobbs/nltk_data'
    - '/usr/share/nltk_data'
    - '/usr/local/share/nltk_data'
    - '/usr/lib/nltk_data'
    - '/usr/local/lib/nltk_data'

So what we need to do is run the NLTK installer, as shown here:

>>> import nltk
>>> nltk.download()
NLTK Downloader
    d) Download      l) List      c) Config      h) Help      q) Quit
Downloader> d

Download which package (l=list; x=cancel)?
  Identifier> wordnet
    Downloading package 'wordnet' to /home/jmhobbs/nltk_data...
      Unzipping corpora/wordnet.zip.

    d) Download      l) List      c) Config      h) Help      q) Quit
Downloader> q

Using NLTK + Wordnet

Now that we have everything installed, using wordnet from Python is straight forward.

# Load the wordnet corpus
from nltk.corpus import wordnet

# Get a collection of synsets (synonym sets) for a word
synsets = wordnet.synsets( 'cake' )

# Print the information
for synset in synsets:
  print "-" * 10
  print "Name:", synset.name
  print "Lexical Type:", synset.lexname
  print "Lemmas:", synset.lemma_names
  print "Definition:", synset.definition
  for example in synset.examples:
    print "Example:", example

The output of that is:

Name: cake.n.01
Lexical Type: noun.artifact
Lemmas: ['cake', 'bar']
Definition: a block of solid substance (such as soap or wax)
Example: a bar of chocolate
Name: patty.n.01
Lexical Type: noun.food
Lemmas: ['patty', 'cake']
Definition: small flat mass of chopped food
Name: cake.n.03
Lexical Type: noun.food
Lemmas: ['cake']
Definition: baked goods made from or based on a mixture of flour, sugar, eggs, and fat
Name: coat.v.03
Lexical Type: verb.contact
Lemmas: ['coat', 'cake']
Definition: form a coat over
Example: Dirt had coated her face



There are some caveats to using WordNet with NLTK. First is that the definitions aren’t always ordered in the way you would expect. For instance, look at the “cake” results above. Cake, as in the confection, is the third definition, which feels wrong. You can of course order and filter on the synset name to correct this to some degree.

Second, there is a major load time for getting WordNet ready to use. Your first call to wordnet.sysnsets will take considerably longer than the next ones. On my machine the difference was 3.5 seconds versus 0.0003 seconds.

Last, you are constrained to the English language, as analyzed by Pinceton. I’ll address this issue in the next section.

Option 2: SDict Viewer

As I said above, using WordNet is simple, but restrictive. What if I want to use a foreign language dictionary or something? WordNet is only in English. This is where the SDict format comes in. It has lots of free resource files available at http://sdict.com/en/. The best existing parser I found was SDict Viewer which is a dead project, but remarkably complete.

SDict Viewer is an application

SDict Viewer is an application, so it’s not an easy to install library. However, it is very well written and extracting what you need is simple. You can get my “library” version from http://github.com/jmhobbs/sdictviewer-lib.

Here is an example when it’s all finished:

import sys

import sdictviewer.formats.dct.sdict as sdict
import sdictviewer.dictutil

dictionary = sdict.SDictionary( 'webster_1913.dct' )

start_word = sys.argv[1]

found = False

for item in dictionary.get_word_list_iter( start_word ):
    if start_word == str( item ):
      instance, definition = item.read_articles()[0]
      print "%s: %s" % ( item, definition )
      found = True

if not found:
  print "No definition for '%s'." % start_word


Here is a sample run:

[email protected]:~$ python okay.py Cat
Cat: (n.) An animal of various species of the genera Felis and Lynx. The domestic cat is Felis domestica. The European wild cat (Felis catus) is much larger than the domestic cat. In the United States the name wild cat is commonly applied to the bay lynx (Lynx rufus) See Wild cat, and Tiger cat.
wrote /home/jmhobbs/.sdictviewer/index_cache/webster_1913.dct-1.0.index

As you can see, it gives a nice definition (thank you Webster 1913) and then it has a little junk on the end. This is the index cache, a lookup table for finding words faster. You can avoid saving it by calling dictionary.close(False) instead.

Option 3: Aard Format

In option 2 I said that SDict Viewer was a dead project, this is because the development has been moved to the Aard Dictionary project. I chose not to pursue this format, as most of the existing resources are stored in HTML formats and I needed plain text. This might be ideal for you though, as they also provide access to Wikipedia archives.

All Done

So there you have it. Two viable ways of extracting a plain text definition for a word in Python. Best of luck to you!