Adding Captcha To Django's Built-in Comments

I recently had a good friend of mine ask for some help in adding a captcha field to his comments form. I gave him some pointers, but before he could put them into action he had to leave for a Thanksgiving roadtrip home. I didn't give much mind to the idea of putting captchas on my own site since it's not all that popular amongst spammers yet. When I woke up this morning, however, I found myself with a few spare minutes to see if my pointers were correct.

Some of the ideas I shared with my friend turned out to not work very well. As I tinkered about trying to get things to work on my own site, I think I came up with a relatively efficient way of doing things.

Installing The Captcha

The captcha field I use is quite simple and effective. I originally got it from http://django.agami.at/media/captcha/, but the project seems to be unmaintained now. Along the road to Django 1.0, some changes were made to the way form fields work, and there is a minor change required in the base code for this captcha field if you want it to work. Alternatively, you can use a copy of the field that I'm currently using.

All you need to do is extract the captcha directory somewhere on your PYTHONPATH. The author recommends putting it in django.contrib, but I usually just place it straight on the PYTHONPATH so all I need to do is from captcha import CaptchaField instead of from django.contrib.captcha import CaptchaField. Minor details...

Adding The Captcha

The first thing you'll want to do after installing the captcha field is add the field itself to your comments form. Instead of subclassing the built-in django.contrib.comments.forms.CommentForm form, I simply decorated the constructor of the form as such:

1
2
3
4
5
6
7
8
9
from django.contrib.comments.forms import CommentForm
from captcha import CaptchaField

def add_captcha(func):
    def wrapped(self, *args, **kwargs):
        func(self, *args, **kwargs)
        self.fields['security_code'] = CaptchaField()
    return wrapped
CommentForm.__init__ = add_captcha(CommentForm.__init__)

This adds a field called security_code to the CommentForm, and it works the same way as if you had done something like this:

1
2
3
4
5
6
7
from django import forms
from captcha import CaptchaField

class MyCommentForm(forms.Form):
    name = forms.CharField()
    ...
    security_code = CaptchaField()

You can put the decorating snippet from above anywhere you'd like so long as the module you put it in is loaded at some point in your project. I usually put this sort of magic in my main urls.py file so it's harder to forget about when I debug things.

Fixing the Form

The first problem with this little trick seems to be that the CaptchaField is rendered as unsafe HTML in the default form.html template in the built-in comments application. That just means that, instead of seeing the captcha, you will see the HTML necessary to render the CaptchaField directly on the page, like this:

<input type="hidden" name="security_code" value="captcha.caZ1SqQ" />
<img src="/static/captchas/caZ1SqQ/0656f09d3974850397dd4c4974f23a35.gif"
 alt="" /><br /><input type="text" name="security_code"
 id="id_security_code" />

To fix that, you can apply the safe filter to the field and make the template look something like this:

{% load comments i18n %}
<form action="{% comment_form_target %}" method="post">
<table>
{% for field in form %}
    {% if field.is_hidden %}
    {{ field }}
    {% else %}
    <tr{% ifequal field.name "honeypot" %} style="display:none;"{% endifequal %}>
        <th>{{ field.label_tag }}:</th>
        <td {% if field.errors %} class="error"{% endif %}>
            {{ field|safe }} &nbsp;
            {% if field.errors %}{{ field.errors|join:"; " }}{% endif %}
        </td>
    </tr>
    {% endif %}
{% endfor %}
</table>
<p class="submit">
    <input type="submit" name="post" class="submit-post"
     value="{% trans "Post" %}" />
    <input type="submit" name="preview" class="submit-preview"
     value="{% trans "Preview" %}" />

</form>

Notice the {{ field|safe }} in there. Also note that I prefer the table layout for the comment form over the default mode. If you change your form template as I have done, you should put the updated copy in your own project's template directory. It belongs in templates/comments/form.html, assuming that your templates directory is called templates. You'll probably also want to check out the preview.html template for the django.contrib.comments application. I changed mine to look like this:

{% extends "comments/base.html" %}
{% load i18n %}

{% block title %}{% trans "Preview your comment" %}{% endblock %}

{% block content %}
    {% load comments %}
    {% if form.errors %}
    <h1>{% blocktrans count form.errors|length as counter %}Please
     correct the error below{% plural %}Please correct the errors below
     {% endblocktrans %}</h1>
    {% else %}
    <h1>{% trans "Preview your comment" %}</h1>
        <blockquote>{{ comment|linebreaks }}</blockquote>

        {% trans "and" %} <input type="submit" name="submit"
         class="submit-post" value="{% trans "Post your comment" %}"
         id="submit" /> {% trans "or make changes" %}:

    {% endif %}

    {% include 'comments/form.html' %}
{% endblock %}

See how I just use the include tag to pull in the comments/form.html template I mentioned above? Saves a lot of typing and potential for problems... If you update the preview.html template, you should save your copy in templates/comments/preview.html, assuming your templates directory is called templates.

Testing It Out

At this point, you should be able to try out your newly installed captcha-fied comments. If it doesn't work, please comment on this article and perhaps we can figure out the problem!

Project Release: django-watermark 0.1.0-pre1

I've found myself in many situations where I have several images that need a watermark applied to them. Applying a watermark to more than 5 images at a time is annoying at best, even if you do have macros to do the mundane stuff for you. I don't like to do things that annoy me.

Recently, I found myself yet again confronted with a "please watermark these images" sort of situation, and I decided to take a more logical approach to the problem: program it. The result of my efforts is called django-watermark and is available on Google Code and PyPI. To be perfectly honest, I actually ganked most of the useful code for this application from a generous soul on the Internet.

This seems to be working fine for me. I think the positioning for watermarks might need more work, but there are several options to hold people over for the time being. Please let me know what your thoughts are.

ReStructured Text

I've been working on updating all of my articles to use ReStructured text formatting these past couple of days. Basically ever since I started using Django on my personal sites, I used Textile formatting to make my articles look pretty, and it treated me well. However, I found myself with a few extra minutes to do some research this weekend, and the topic I happened to research was the difference between ReStructured text, Markdown, and Textile.

After my brief research into these utilities, I decided it would be worth my while to convert all of my articles and other Textile-formatted content to use ReStructured text instead. Most of this is due to personal preferences for certain conventions--very little technical reasoning came into play when I made the decision to switch (so don't ask me why I abandoned one for the other if you expect a technical pro/con list). Suffice it to say that I consider the ReST syntax to be superior to that of Textile.

In the conversion process, I realized that I wrote a whole lot more in my articles than I thought I did! It took much longer than expected, but the work is done now, and all of the articles should be updated. Most of you probably won't be able to tell much of a difference one way or another, but if you do find a problem with the formatting of any of my articles, please notify me!

Project Release: django-pendulum 0.1

For those of us who like to know how much time we spend on various projects, I've released a little project to help you out. Pendulum is a small Django application that I've been using to keep track of the time I work on my various personal projects. It's very easy to use, and it's been working great for me for several months.

This weekend I took the time to make several improvements that I deemed necessary before I could make the project publicly accessible. While I've done some testing on my own, I'm fairly certain that I haven't tried all combinations of situations, so there might be some problems that you find as you use it.

To learn more, head on over to http://code.google.com/p/django-pendulum/. There you can learn how to download and install it and what you need to do to configure it.

Enjoy!

Another Notch for Free Software

I tend to use Amarok, one of KDE's most popular audio players, to manage my music. Amarok can usually handle synchronizing music on an iPod just fine, but it turns out that it doesn't play very well with my 3-gen iPod nano. For one reason or another, any time I tried to copy music from my linux box using Amarok, the whole iPod became useless. The device recognized that space was being used, but it wasn't recognizing any music or movies or anything. Because of this, I have been using my wife's macbook for the past while to synchronize my iPod. It's been working fine, but it sure is inconvenient to interrupt her when she needs to be doing homework or something and I want to pop into iTunes for a few minutes.

Today I was listening to my iPod and doing some homework between classes when all of the sudden my music stopped playing. I thought the playlist might have ended, so I went to start the music up again only to find out that the little guy had locked up (it does this from time to time). I couldn't remember the reboot sequence off the top of my head, so I Googled it and stumbled across a short and sweet article with the goods.

This article also referenced the use of a program called Floola in place of iTunes. I decided to investigate it briefly because my homework was boring me. Come to find out that Floola is free, works on Windows, Linux, and MacOS, and it actually does support my 3-gen nano! It also has some cool features like being able to download music videos straight from YouTube and converting them to work on the iPod. I'm really enjoying the program! One less reason to keep Windows around!

Project Release: Seasonal Stylesheets

The other day at work I was involved in a humorous discussion about having our organization's Web site change colors depending on what day of the year it was. We threw around all sorts of fun ideas, knowing full well that it would never happen for our organization. It would be too unprofessional.

That night my allergies were bothering me to the point where I couldn't sleep, so I decided to see what I could do about making our discussion more possible, at least for others who can take a little more creative license than us. By morning, I had a working application that could effectively change the colors on a Web page depending on the date.

Today I released the source code on Google Code, hoping that someone more creative than me might take it and experiment to make something both attractive and useful.

You can learn more at the project's homepage: http://code.google.com/p/django-seasonal-stylesheets/

Dexpot

I often find myself in need of more screen space when I'm using my computer. I have so many different things up on my screen at all times. Since Linux is my default operating system, I take advantage of the virtual desktop functionality that is built into most every popular window manager to make it seem like I have more screen space.

This is all fine and dandy, but it makes using Windows all the more frustrating to me. I've tried some utilities in the past that offered virtual desktop capabilities, but they always seemed weak and crashed a lot. For years I had given up on virtual desktops in Windows.

Today I decided to look into the current options for such functionality in Windows. In my searching, I came across Dexpot. I installed it, and it seems to be doing exactly what I want and more! It seems to be stable, and it only takes up 8,776K of my memory. That part could be better, but I'm not complaining :)

Anyway, I thought some of you might benefit from this utility as well.

Installing Django on Shared Hosting (Site5)

This article is a related to my previously posted article about installing Django, an advanced Web framework for perfectionists, on your own computer. Now we will learn how to install Django on a shared hosting account, using Site5 and fastcgi as an example. Depending on your host, you may or may not have to request additional privileges from the support team in order to execute some of these commands.

Note: Django requires at least Python 2.3. Newer versions of Python are preferred.

Note: This HOWTO assumes familiarity with the UNIX/Linux command line.

Note: If the wget command doesn't work for you (as in you don't have permission to run it), you might try curl [url] -O instead. That's a -O as in upper-case o.

Install Python

Site5 (and many other shared hosting providers that offer SSH access) already has Python installed, but you will want to have your own copy so you can install various tools without affecting other users. So go ahead and download virtual python:

mkdir ~/downloads
cd ~/downloads
wget http://peak.telecommunity.com/dist/virtual-python.py

Virtual Python will make a local copy of the installed Python in your home directory. Now you want to make sure you execute this next command with the newest version of Python available on your host. For example, Site5 offers both Python 2.3.4 and Python 2.4.3. We want to use Python 2.4.3. To verify the version of your Python, execute the following command:

python -V

If that displays Python 2.3.x or anything earlier, try using python2.4 -V or python2.5 -V instead. Whichever command renders the most recent version of Python is the one you should use in place of python in the next command. Since python -V currently displays Python 2.4.3 on my Site5 sandbox, I will execute the following command:

python ~/downloads/virtual-python.py

Again, this is just making a local copy of the Python installation that you used to run the virtual-python.py script. Your local installation is likely in ~/lib/python2.4/ (version could vary).

Make Your Local Python Be Default

To reduce confusion and hassle, let's give our new local installation of Python precedence over the system-wide Python. To do that, open up your ~/.bashrc and make sure it contains a line similar to this:

export PATH=$HOME/bin:$PATH

If you're unfamiliar with UNIX-based text editors such as vi, here is what you would type to use vi to make the appropriate changes:

  • vi ~/.bashrc to edit the file
  • go to the end of the file by using the down arrow key or the j key
  • hit o (the letter) to tell vi you want to start typing stuff on the next line
  • type export PATH=$HOME/bin:$PATH
  • hit the escape key
  • type :x to save the changes and quit. Don't forget the : at the beginning. Alternatively, you can type :wq, which works exactly the same as :x.

Once you've made the appropriate changes to ~/.bashrc, you need to make those changes take effect in your current SSH session:

source ~/.bashrc

Now we should verify that our changes actually took place. Type the following command:

which python

If they output of that command is not something like ~/bin/python or /home/[your username]/bin/python, something probably didn't work. If that's the case, you can try again, or simply remember to use ~/bin/python instead of python throughout the rest of this HOWTO.

Install Python's setuptools

Now we should install Python's setuptools to make our lives easier down the road.

cd ~/downloads
wget http://peak.telecommunity.com/dist/ez_setup.py
python ez_setup.py

This gives us access to a script called easy_install, which makes it easy to install many useful Python tools. We will use this a bit later.

Download Django

Let's now download the most recent development version of Django. SSH into your account and execute the following commands (all commands shall be executed on your host).

svn co http://code.djangoproject.com/svn/django/trunk ~/downloads/django-trunk

Now we should make a symlink (or shortcut) to Django and put it somewhere on the Python Path. A sure-fire place is your ~/lib/python2.4/site-packages/ directory (again, that location could vary from host to host):

ln -s ~/downloads/django-trunk/django ~/lib/python2.4/site-packages
ln -s ~/downloads/django-trunk/django/bin/django-admin.py ~/bin

Now verify that Django is installed and working by executing the following command:

python -c "import django; print django.get_version()"

That command should return something like 1.0-final-SVN-8964. If you got something like that, you're good to move onto the next section. If, however, you get something more along the lines of...

Traceback (most recent call last):
    File "<string>", line 1, in ?
ImportError: No module named django

...then your Django installation didn't work. If this is the case, make sure that you have a ~/downloads/django-trunk/django directory, and also verify that ~/lib/python2.4/site-packages actually exists.

Installing Dependencies

In order for your Django projects to become useful, we need to install some other packages: PIL (Python Imaging Library, required if you want to use Django's ImageField), MySQL-python (a MySQL database driver for Python), and flup (a utility for fastcgi-powered sites).

easy_install -f http://www.pythonware.com/products/pil/ Imaging
easy_install mysql-python
easy_install flup

Sometimes, using easy_install to install PIL doesn't go over too well because of your (lack of) permissions. To circumvent this situation, you can always download the actual PIL source code and install it manually.

cd ~/downloads
wget http://effbot.org/downloads/Imaging-1.1.6.tar.gz
tar zxf Imaging-1.1.6.tar.gz
cd Imaging-1.1.6
ln -s ~/downloads/Imaging-1.1.6/PIL ~/lib/python2.4/site-packages

And to verify, you can try this command:

python -c "import PIL"

If that doesn't return anything, you're good to go. If it says something about "ImportError: No module named PIL", it didn't work. In that case, you have to come up with some other way of installing PIL.

Setting Up A Django Project

Let's attempt to setup a sample Django project.

mkdir -p ~/projects/django
cd ~/projects/django
django-admin.py startproject mysite
cd mysite
mkdir media templates

If that works, then you should be good to do the rest of your Django development on your server. If not, make sure that ~/downloads/django-trunk/django/bin/django-admin.py exists and that it has a functioning symlink (shortcut) in ~/bin. If not, you'll have to make adjustments according to your setup. Your directory structure should look something like:

  • projects
    • django
      • mysite
        • media
        • templates
        • __init__.py
        • manage.py
        • settings.py
        • urls.py

Making A Django Project Live

Now we need to make your Django project accessible from the Web. On Site5, I generally use either a subdomain or a brand new domain when setting up a Django project. If you plan on having other projects accessible on the same hosting account, I recommend you do the same. Let's assume you setup a subdomain such as mysite.mydomain.com. On Site5, you would go to ~/public_html/mysite for the next few commands. This could differ from host to host, so I won't go into much more detail than that.

Once you're in the proper place, you need to setup a few things: two symlinks, a django.fcgi, and a custom .htaccess file. Let's begin with the symlinks.

ln -s ~/projects/django/mysite/media ~/public_html/mysite/static
ln -s ~/lib/python2.4/site-packages/django/contrib/admin/media ~/public_html/mysite/media

This just makes it so you can have your media files (CSS, images, javascripts, etc) in a different location than in your public_html.

Now for the django.fcgi. This file is what tells the webserver to execute your Django project.

#!/home/[your username]/bin/python
import sys, os

# Add a custom Python path.
sys.path.insert(0, "/home/[your username]/projects/django")

# Switch to the directory of your project. (Optional.)
os.chdir("/home/[your username]/projects/django/mysite")

# Set the DJANGO_SETTINGS_MODULE environment variable.
os.environ['DJANGO_SETTINGS_MODULE'] = "mysite.settings"

from django.core.servers.fastcgi import runfastcgi
runfastcgi(method="threaded", daemonize="false")

And finally, the .htaccess file:

1
2
3
4
5
6
RewriteEngine On
RewriteBase /
RewriteRule ^(media/.*)$ - [L]
RewriteRule ^(static/.*)$ - [L]
RewriteCond %{REQUEST_URI} !(django.fcgi)
RewriteRule ^(.*)$ django.fcgi/$1 [L]

The .htaccess file makes it so that requests to http://mysite.mydomain.com/ are properly directed to your Django project. So, now you should have a directory structure that something that looks like this:

  • public_html
    • mysite
      • media
      • static
      • .htaccess
      • django.fcgi

If that looks good, go ahead and make the django.fcgi executable and non-writable by others:

chmod 755 ~/public_html/mysite/django.fcgi

After that, head over to http://mysite.mydomain.com/ (obviously, replace the mydomain accordingly). If you see a page that says you've successfully setup your Django site, you're good to go!

Afterthoughts

I've noticed that I need to "restart" my Django sites on Site5 any time I change the .py files. There are a couple methods of doing this. One includes killing off all of your python processes (killall ~/bin/python) and the other simply updates the timestamp on your django.fcgi (touch ~/public_html/mysite/django.fcgi). I find the former to be more destructive and unreliable than the latter. So, my advice is to use the touch method unless it doesn't work, in which case you can try the killall method.

Good luck!

Andy Rutledge: Don't Walk; Run

This morning I've been doing some research for work, and I stumbled upon an article that is very helpful for freelancers like me (relatively new to the freelance arena) and people looking for freelancers/design agencies.

Given the recent freelance opportunities that I have been involved with, I found this article to be particularly beneficial. There are many pieces of advice that, taken into consideration, really do distinguish the good freelancers/design agencies from the bad ones.

If the agency has any clue at all, your company will basically have to "interview for the job" in order to be accepted as a client. Agencies that know what they're doing work hard to take on only those projects and those clients that are a good fit for their skill set, even their personality. They'll want to get to know you a bit and they'll want to learn much about your project, its scope, and your specific expectations before they agree to work with you. If you find that they immediately suggest you work with them and start talking about sending you a contract, you're probably talking to the wrong agency. Hang up the phone.

I know that I am guilty of such behavior. In the past, I have quickly jumped at any opportunity to make some extra cash as a freelancer. It usually doesn't take a long time before I begin regretting my hasty acceptance of a job. This is extremely sound advice.

If anyone other than the designer who did the work presents the designs to you, something's amiss and there is likely more yet to go astray. It is an unfortunate habit among some agencies to ever get between the client and the designer. This is a terrible mistake and a clear sign that the agency, large or small, is not well run. As a result, the work produced by the agency is likely to be flawed for the voids in understanding and effectiveness that result. In short, if your project's design phase does not begin and end with direct contact between yourself and the designer, you're getting short shrift. Find another agency.

I'd like to twist this one around a little. I completely agree with what Andy has said here, but I have also been on the receiving end of this particular type of situation. On one recent project, I was receiving instructions from a person who apparently didn't have the same goals as the "real" client. The end result was a product the "real" client didn't really want. Working directly with the client/agency is critical to a successful project.

There are several other important points in Andy's article, and I wholeheartedly recommend that you read it if you find yourself in such a situation. It will only help. In conclusion, I just want to share another quote from Andy's article:

And if you do notice too many of these telltale signs of incompetence in your current agency — don't walk; run.

Using Django to Design Your Database Schema

Last night I had a buddy of mine ask me how I would approach a particular database design problem. I get similar questions quite often from my peers--suggests there is something important lacking from the database classes out there. Instead of answering him directly, I decided to come up with this tutorial for using Django to design your database schema.

For those of you new to Django, this article might seem a bit advanced. In time I will have more introductory-level Django tutorials, but I hope this one is easy enough.

Create a Django Project

The first step is to create a Django project. If you already have a project that you can play with, you can skip this step. To create a project, go to a place where you want to keep your code (like C:\projects or /home/me/projects) in a command prompt/terminal and run the following command:

django-admin.py startproject myproject

This will create a new directory in your current location called myproject (you can replace myproject with whatever you'd like so long as you're consistent). This new directory will contain a few files:

  • __init__.py
  • manage.py
  • settings.py
  • urls.py

If you get an error message when running the above command, you might not have Django installed properly. See Step-by-Step: Installing Django for details on installing Django.

Create An Application

Once you have a Django project setup, you should create a new application.

Note: If you're using Windows, you will probably need to omit the ./ on the ./manage.py commands. I will include them here for everyone else who's using Linux or a Mac.

cd myproject
./manage.py startapp specialapp

This will create a new directory in your myproject directory. This new directory will contain three files: __init__.py, models.py, and views.py. We are only concerned with the models.py file in this article.

Create Your Models

Models are usually a direct representation of what your database will be. Django makes creating these models extremely easy, and Python's syntax makes them quite readable. The Django framework asks for models to be defined in the models.py file that was created in the last step. Here's an example (for my buddy who prompted the creation of this article):

from django.db import models

class Component(models.Model):
    item_number = models.CharField(max_length=20)
    name = models.CharField(max_length=50)
    size = models.CharField(max_length=10)
    quantity = models.IntegerField(default=1)
    price = models.DecimalField(max_digits=8, decimal_places=2)

class Project(models.Model):
    name = models.CharField(max_length=50)
    components = models.ManyToManyField(Component)
    instructions = models.TextField()

(for more information about models, see the Django Model API Reference)

I don't know about you, but that code seems pretty straightforward to me. I'll spare you all the details about what's going on (that can be a future article).

Install Your New Application

Once you have your models setup, we need to add our specialapp to our list of INSTALLED_APPS in order for Django to register these models. To do that, open up settings.py in your myproject directory, go to the bottom of the file, until you see something like

1
2
3
4
5
6
INSTALLED_APPS = (
    'django.contrib.auth',
    'django.contrib.contenttypes',
    'django.contrib.sessions',
    'django.contrib.sites',
)

When you find that, add your specialapp to the list

1
2
3
4
5
6
7
INSTALLED_APPS = (
    'django.contrib.auth',
    'django.contrib.contenttypes',
    'django.contrib.sessions',
    'django.contrib.sites',
    'specialapp'
)

Setup Your Database

Now you need to let Django know what kind of database you're using. Django currently supports MySQL, SQLite3, PostgreSQL, and Oracle natively, but you can get third-party tools that allow you to use other database (like SQL Server).

Still in your settings.py, go to the top until you see DATABASE_ENGINE and DATABASE_NAME. Set that to whatever type of database you are using:

DATABASE_ENGINE = 'sqlite3'
DATABASE_NAME = 'myproject.db'

Save your settings.py and go back to your command prompt/terminal.

Get Django's Opinion For Your Schema

Make sure you're in your myproject directory and run the following command:

./manage.py sqlall specialapp

This command will examine the models that we created previously and will generate the appropriate SQL to create the tables for your particular database. For SQLite, we get something like this for output:

BEGIN;
CREATE TABLE "specialapp_component" (
      "id" integer NOT NULL PRIMARY KEY,
      "item_number" varchar(20) NOT NULL,
      "name" varchar(50) NOT NULL,
      "size" varchar(10) NOT NULL,
      "quantity" integer NOT NULL,
      "price" decimal NOT NULL
)
;
CREATE TABLE "specialapp_project" (
      "id" integer NOT NULL PRIMARY KEY,
      "name" varchar(50) NOT NULL,
      "instructions" text NOT NULL
)
;
CREATE TABLE "specialapp_project_components" (
      "id" integer NOT NULL PRIMARY KEY,
      "project_id" integer NOT NULL REFERENCES "specialapp_project" ("id"),
      "component_id" integer NOT NULL REFERENCES "specialapp_component" ("id"),
      UNIQUE ("project_id", "component_id")
)
;
COMMIT;

Notice how Django does all sorts of nifty things, like wrapping the table creation queries in a transaction, setting up indexes, unique keys, and defining relationships between tables. The output also offers a solution to the original problem my buddy had: an intermediate table that just keeps track of relationships between projects and components (the specialapp_project_components table).

Notice that the SQL above may not work with database servers other than SQLite.

Enhancing The Intermediate Table

After my buddy reviewed this article, he asked a very interesting and valid question: What if a project needs 3 of one component? In response, I offer the following models (this requires a modern version of Django--it doesn't work on Django 0.96.1 or earlier):

from django.db import models

class Component(models.Model):
    item_number = models.CharField(max_length=20)
    name = models.CharField(max_length=50)
    size = models.CharField(max_length=10)
    quantity = models.IntegerField(default=1)
    price = models.DecimalField(max_digits=8, decimal_places=2)

class Project(models.Model):
    name = models.CharField(max_length=50)
    components = models.ManyToManyField(Component, through='ProjectComponent')
    instructions = models.TextField()

class ProjectComponent(models.Model):
    project = models.ForeignKey(Project)
    component = models.ForeignKey(Component)
    quantity = models.PositiveIntegerField()

    class Meta:
        unique_together = ['project', 'component']

Running ./manage.py sqlall specialapp now generates the following SQL:

BEGIN;
CREATE TABLE "specialapp_component" (
    "id" integer NOT NULL PRIMARY KEY,
    "item_number" varchar(20) NOT NULL,
    "name" varchar(50) NOT NULL,
    "size" varchar(10) NOT NULL,
    "quantity" integer NOT NULL,
    "price" decimal NOT NULL
)
;
CREATE TABLE "specialapp_project" (
    "id" integer NOT NULL PRIMARY KEY,
    "name" varchar(50) NOT NULL,
    "instructions" text NOT NULL
)
;
CREATE TABLE "specialapp_projectcomponent" (
    "id" integer NOT NULL PRIMARY KEY,
    "project_id" integer NOT NULL REFERENCES "specialapp_project" ("id"),
    "component_id" integer NOT NULL REFERENCES "specialapp_component" ("id"),
    "quantity" integer unsigned NOT NULL,
    UNIQUE ("project_id", "component_id")
)
;
CREATE INDEX "specialapp_projectcomponent_project_id" ON "specialapp_projectcomponent" ("project_id");
CREATE INDEX "specialapp_projectcomponent_component_id" ON "specialapp_projectcomponent" ("component_id");
COMMIT;

As you can see, most of the SQL is the same. The main difference is that the specialapp_project_components table has become specialapp_projectcomponent and it now has a quantity column. This can be used to keep track of the quantity of each component that a project requires. You can add however many fields you want to this new intermediate table's model.

Using This SQL

There are several ways you can use the SQL generated by Django. If you want to make your life really easy, you can have Django create the tables for you directly. Assuming that you have specified all of the appropriate database information in your settings.py file, you can simply run the following command:

./manage.py syncdb

This will execute the queries generated earlier directly on your database, creating the tables (if they don't already exist). Please note that this command currently will not update your schema if the table exists but is missing a column or two. You must either do that manually or drop the table in question and then execute the syncdb command.

Another option, if you want to keep your DDL(Data Definition Language) in a separate script (maybe if you want to keep it in some sort of version control) is something like:

./manage.py sqlall specialapp > specialapp-ddl-080813.sql

This just puts the output of the sqlall command into a file called specialapp-ddl-080813.sql for later use.

Benefits of Using Django To Create Your Schema

  • Simple: I personally find the syntax of Django models to be very simple and direct. There is a comprehensive API that explains and demonstrates what Django models are capable of.
  • Fast: Being that the syntax is so simple, I find that it makes designing and defining your schema much faster than trying to do it with raw SQL or using a database administration GUI.
  • Understandable: Looking at the model code in Django is not nearly as intimidating as similar solutions in other frameworks (think about Java Persistence API models).
  • Intelligent: Using the same model code, Django can generate proper Data Definition Language SQL for several popular database servers. It handles indexes, keys, relationships, transactions, etc. and can tell the difference between server types.

Downfalls of Using Django To Create Your Schema

  • The Table Prefix: Notice how all of the tables in the SQL above were prefixed with specialapp_. That's Django's safe way of making sure models from different applications in the same Django project do not interfere with each other. However, if you don't plan on using Django for your end project, the prefix could be a major annoyance. There are a couple solutions:
    • A simple "search and replace" before executing the SQL in your database
    • Define the db_table option in your models
  • Another Technology: Django (or even Python) may or may not be in your organization's current development stack. If it's not, using the methods described in this article would just become one more thing to support.

Other Thoughts

I first thought about doing the things mentioned in this application when I was working on a personal Java application. I like to use JPA when developing database-backed applications in Java because it abstracts away a lot of the database operations. However, I don't like coming up with the model classes directly, so I usually reverse engineer them from existing database tables.

Before thinking about the things discussed in this article, I created the tables by hand, making several modifications to the schema before I was satisfied with my JPA models. This proved to be quite bothersome and time-consuming.

After using Django to develop my tables, the JPA models turned out to be a lot more reliable, and they were usually designed properly from the get-go. I haven't created tables manually ever since.

If you find yourself designing database schemas often, and you find that you have to make several changes to your tables before you/the project requirements are satisfied, you might consider using Django to do the grunt work. It's worked for me, and I'm sure it will work for you too.

Good luck!