I have been the maintainer of django-compressor for a couple of years now. According to PyPI it was downloaded 2306 times in the last day and 46,641 times in the last month. It’s in the top 10 of apps on Django Packages. It’s well tested, has an okay documentation and is released under a very permissive license. So far 91 contributors worked on it. It has around 4000 lines of code, an average number of code comments and a good (albeit not stellar) test coverage. There wasn’t a new release for almost a year and I’ve been too busy with other things to work on both release preparation as well as bug fixes. So there are tons of improvements that could be made. For example: Improve the integration with staticfiles Add better documentation for remote storages such as Amazon S3 or Rackspace Cloud Files Improve the beginner documentation Simplify configuration Improve the integration with state of the art tools like Grunt Finish the port to Python 3 Add asynchronous compressing based on Celery or RQ Improve the integration with Jinja2 (e.g. add offline compression support, port Peter Bengtsson’s django-jingo-offline-compressor There are many other ideas that I’m sure everyone has, but ...

Grunt is a powerful task runner with an amazing assortment of plugins. It's not limited to the frontend, but there are many frontend-oriented plugins that you can take advantage of to combine and minify your static media, compile sass and less files, watch for changes during development and reload your browser automatically, and much more. In the last several years, the amount of tooling around frontend development has expanded dramatically. Frameworks, libraries, preprocessors and postprocessors, transpilers, template languages, module systems, and more! Wiring everything together has become a significant challenge, and a variety of build tools have emerged to help ease this burden. Grunt is the current leader because of its fantastic plugin community, and it contains a wide array of plugins that can be very valuable to a Django developer. Today I'm going to talk about an easy way to integrate Grunt with Django's runserver, and highlight a few plugins to handle common frontend tasks that Django developers often deal with. Installing Grunt Grunt uses Node.js, so you'll need to have that installed and configured on your system. This process will vary depending on your platform, but once it's done you'll need to install Grunt. From the documentation: $ ...

As part of my endeavors as an Open Source developer and particular as a Python developer I spent most of my time writing documentation in a formatting language called reStructuredText. It’s a neat language originally developed for docutils to be used in Python “docstrings”. Later when Python’s own documentation was in a need to be refactored a project called Sphinx (not to be confused with the search server with the same name) was developed with many added features on top of docutils. One of which was the ability to easily extend Sphinx based documentations with extensions, e.g. such as new roles and directives. While working on my blog software I realized that while I wanted it to be database driven (to be able to render things in more dynamic ways than say when I’d only render some static HTML files) but still didn’t want to miss the flexibility of reStructuredText. I also wanted the blog software to be able to handle non-trivial content, such as images, videos, Github gists etc, to further improve my writing. But how should I combine those orthogonal topics into one system? Hardcoding HTML into the blog entries wasn’t an option neither was leaving out non-text ...

We are almost a month into the new year, time to look forward.But first a look backwards. The year of 2013 was a year of big development projects and lots of quiet in the main repository in between. Two major updates were released during the year.The first update, the "many sessions per player" update, originated in a feature request that I thought would be easy to implement but which led to far-ranging changes (and honestly, improvements) to how Players and Sessions interconnect. It was a lot more more work than I anticipated.The second update was about moving Evennia's web server from the Portal level into the Server-level. The actual moving of the server was actually considerably easier than I thought it would be. But it turned out that a truckload of other things came along with it. Not only did the cache system have to change in order to accommodate the new webs erver, I had to also finalize the Out-of-band structure, since this made use of the cache system. And while I were at it, other fixes were done and ... the update grew and grew. When it finally merged late last year it closed plenty of issues, but it ...

While Django has support for loading fixtures in its unit testing tools, I have found that maintaining fixtures over time is nothing less than a giant pain in the butt. As your model definitions change over time, remembering to re-run manage.py dumpdata myapp >fixtures.json any time your fixtures are affected is something I guarantee you will forget to do. Not to mention the fact that your database may or may not contain records suitable for testing at any given time. Then all of your unit tests have blown up, it takes you a little bit of time to figure out why, and before you know it you are cursing and out of your flow. For a time I tried using Alex Gaynor's django-fixture-generator but that only solved the problem of creating the records you want to test, not the problem of fixtures sitting on disk that are out of sync with your model schema. After being introduced to Django class-based views' mixin-heavy style, I cooked up a technique for unit test TestCase classes to create the records they need at runtime, thus bypassing the need for fixtures. In an apps tests/__init__.py I create a PopulateDbMixin: class PopulateDbMixin(object): def populate_db(self): """Assume ...

Email validation has always been a bit of a nightmare. On the one hand there are all of the local domain names (sometimes even a single word can be a valid email) meanwhile there is an explosion of top level domain names confusing the entire internet for no reason I've been able to work out.Anyway my problem today is that my payment processor SagePay validates emails and will decline the transaction if the email validation fails. I am using Django's EmailField to valid the addresses but unfortunately SagePay and Django have different views on what is a valid email with SagePay being more strict.In a fit of optimism I emailed SagePay support to try to find out the algorithm, after a brief exchange of mails I got to it as "RFC532n" which completely mystified me (and Google!) However on reading their documentation I got referred to RFC3696 which made more sense although it is one of the RFC's which leaves things quite open with a lot of advice rather than firm rules. So I did a bit of testing to try to reverse engineer their algorithm and it looks like it is the Django algorithm with the addition of a ...

Today the Django project released the first alpha for Django 1.7. As such, the branch for 1.7 has been created in the git repository, and we can start mirroring it. Of course, this is still an alpha and the clone shouldn’t be used yet for prodution. The purpose is to test 1.7 early and to […]

In my previous blog post I covered using awesome-slugify to capture slugs in both ASCII and unicode. Today I'm covering the definition custom language slugify translation functions. Defining Custom Language slugify Translation Functions For those times we need ASCII representation of unicode characters, we can't always use the default unicode-to-ASCII mappings. A powerful feature of awesome-slugify is we can quickly and easily create our own translation functions. Just follow these two steps: Define a translation dictionary. Keys are the names of things you want translated, and the associated values are what the keys are translated into. Generate a translation function using slugify.main.get_slugify(). Explaining this in depth will take paragraphs of text, so I'll just demonstrate it using emoji: # -*- coding: utf-8 -*- # test_slugify_emoji.py # The way the awesome-slugify API works, you can't directly import from # the 'slugify' package. You have to import from the 'slugify.main' # module. See https://github.com/dimka665/awesome-slugify/issues/2 from slugify.main import get_slugify import pytest # Step 1: Define the translation dictionary ALT_EMOJI = { u'ʘ‿ʘ': u'smiling', u'ಠ_ಠ': u'disapproval', u'♥‿♥': u'enamored', u'♥': u'love', } # Step 2: Generate a translation function slugify_emoji = get_slugify(pretranslate=ALT_EMOJI) def test_basic_emoji(): # awesome-slugify capitalizes strings. # see https://github.com/dimka665/awesome-slugify/issues/3 assert slugify_emoji(u"ʘ‿ʘ") == u"Smiling" ...

note: The introduction mentions Django and Plone. However, this is not an article about Django or Plone. Introduction Years ago, when I was working with Plone at NASA, one thing I dreaded was when content editors would copy-and-paste from Microsoft Word into the title bar. All kinds of funny characters would appear in the title bar and URL. I would have to go into the database (ZODB) and fix things. Things didn't get better until Reed O'Brien turned on a title validator (probably in Plone.i18n). When we started using Django, one thing that made it nice was the presence of it's slugify() function and template filter. Inspired by the newspaper industry, this function it easier on both content editors and software engineers. In any case, using slugify() we completed a number of projects, with NASA Science being the only public one I worked on. As much as the slugify() function was useful, there were problems. As I discovered time and time again over the years, it didn't handle unicode. Or rather, it handled them by simply vanishing non-ASCII unicode characters. For example: >>> from django.utils.text import slugify >>> slugify(u"straße") # German for road u"strae" If you read German, you'll know ...

We are proud to release: Django-Extensions Version 1.3.3 This brings the usual tons of fixes and improvements Get it now: https://pypi.python.org/pypi/django-extensions/1.3.3 Changelog: Docs: Made it clearer that Django Extensions requires Django 1.4 or higher Translations: FR Updated Python3: Fix for shell_plus

Years ago, when I was working with Plone at NASA, one thing I dreaded was when content editors would copy-and-paste from Microsoft Word into the title bar. All kinds of funny characters would appear in the title bar and URL. I would have to go into the database (ZODB) and fix things. Things didn't get better until Reed O'Brien turned on a title validator (probably in Plone.i18n). When we started using Django, one thing that made it nice was the presence of it's slugify() function and template filter. Inspired by the newspaper industry, this function it easier on both content editors and software engineers. In any case, using slugify() we completed a number of projects, with NASA Science being the only public one I worked on. As much as the slugify() function was useful, there were problems. As I discovered time and time again over the years, it didn't handle unicode. Or rather, it handled them by simply vanishing non-ASCII unicode characters. For example: >>> from django.utils.text import slugify >>> slugify(u"straße") # German for road u"strae" If you read German, you'll know that the default Django slugify() function is converting the word 'road' to nonsense. For sites dealing with internationalization, ...

ViewSets and Routers are very useful for slimming up your code and providing a lot of default functionality out of the box. They are powerful features in django-rest-framework allowing a lot of flexibility in your code while keeping things clean.Watch Now...

Modern web sites and mobile applications frequently showcase a rich set of images with various sizes and dimensions. Such services can even have their graphics, image resolutions and overall look & feel vary between different devices (desktop, mobile, etc.), and may change dramatically with constant upgrades. One of our key requirements with Cloudinary was to help make image modifications as easy, intuitive and flexible as possible. Using our dynamic URLs, you can indeed resize, crop, alter shape and apply effects using simple URL parameters. New images are transformed on-the-fly in the cloud and the resulting thumbnails are optimized and delivered via a fast CDN. As a result, viewing different derived images of a single hi-res source is a snap. Here's an example of a dynamic image manipulation URL in action: http://.../image/upload/w_100,h_80,c_fill,r_30,e_fill_light/yellow_tulip.jpg You can also use Cloudinary's dynamic URLs to fetch, manipulate and deliver Facebook & Twitter profile pictures, or any image available via a public URL. For example: http://.../demo/image/facebook/w_100,h_80,c_fill,r_30,e_fill_light/madonna.jpg Using dynamic URLs for image manipulations is both powerful and flexible. However, for some use cases, this method may be too permissive. For example, if you add a watermark to an image via a dynamic URL parameter, a viewer may tweak ...

I'm currently working on a site which uses Bootstrap 3, and makes significant use of glyphicons. As of Bootstrap 3, glyphicons are back to being web-font based. They look fine in Chrome (which is my main desktop browser), and in Safari on iOS, but I recently noticed they didn't work on my Android phone. Initially, I thought it might be an Android problem, but then I spotted that they didn't work in desktop Firefox either. Debugging in desktop Firefox is rather easier than its mobile cousin, and loading up the web console, I saw this error: Error: downloadable font: download failed (font-family: ...): bad URI or cross-site access not allowed Some Googling around got me to this Stack Overflow question, which told me I needed to add Access-Control-Allow-Origin * to the headers returned in the HTTP responses for the font files. The answer was for Apache, where I was using nginx, for which the relevant snippet is: location ~* \.(eot|ttf|woff)$ { add_header Access-Control-Allow-Origin *; } I just added that to the location block that was responsible for serving my font files, reloaded nginx, and the icons appear correctly. Here's hoping this post helps someone else figure out how to get ...

推荐一篇Django的开发及部署的文章：Django Best Practices－－A guide to d […]

In my previous blog post I covered test discovery and writing basic tests using pytest. Today I'm going to cover a few more features that I really enjoy: raises and fixtures. The Intuitively Named raises context manager When using pytest, you can assert whether or not an exception occurred via the following: # test_exceptions.py from pytest import raises def test_an_exception(): with raises(IndexError): # Indexing the 30th item in a 3 item list [5, 10, 15][30] Fixtures as Function Arguments When writing tests, it's not uncommon to need common objects used between tests. However, if you have a complicated process to generate these common objects, then you have to write tests for your tests. When using Python's venerable unittest framework, this always causes a spaghetti-code headache. However, via the virtue of simplicity, pytest helps keep our test code cleaner and more maintainable. Rather than try and explain that further, I'll write some code to get my point across: # test_fixtures.py from pytest import fixture @fixture # Registering this function as a fixture. def complex_data(): # Creating test data entirely in this function to isolate it # from the rest of this module. class DataTypes(object): string = str list = list return ...

When I first encountered Holger Krekel's pytest this summer on Jeff Knupp's blog I felt like I had been living under a rock for years. I've been using Python's unittest framework since 2006 and nose to find tests since 2008, but here was another test framework dating back to at least January 24, 2007. pytest is a very mature testing tool for testing Python. My favorite features: It can run unittest, doctest, and nose, style tests suites, making it ideal for new and legacy projects. It includes an intuitively named raises context manager for testing exceptions. You can define fixture arguments to generate baseline data. This is very, very different from Django-style fixtures. Via pytest.ini you can change the behavior of pytest. Integrates nicely with setup.py. Alright, lets dive into usage. Test Discovery The first thing that pytest provides is test discovery. Like nose, starting from the directory where it is run, it will find any Python module prefixed with test_ and will attempt to run any defined unittest or function prefixed with test_. pytest explores properly defined Python packages, searching recursively through directories that include __init.py__ modules. Since an image is probably easier to read, here's a sample directory ...

Recently we wanted a way to let users create real-life events which could occur in any time zone that the user desired. By default, Django interprets any date/time that the user enters as being in the user’s time zone, but it never displays that time zone, and it converts the time zone to UTC before storing it, so there is no record of what time zone the user initially chose. This is fine for most purposes, but if you want to specifically give the user the ability to choose different time zones for different events, this won’t work. One idea I had was to create a custom model field type. It would store both the date/time (in UTC) and the preferred time zone in the database, and provide a form field and some kind of compound widget to let the user set and see the date/time with its proper time zone. We ended up with a simpler solution. It hinged on considering the time zone separately from a time. In our case, we would set a time zone for an event. Any date/time fields in that event form would then be interpreted to be in that time zone. Now, displaying ...

We are proud to release: Django-Extensions Version 1.3.0 This brings official Django 1.6 support and the usual tons of fixes and improvements Get it now: https://pypi.python.org/pypi/django-extensions/1.3.0 Changelog: Feature: SQLDiff much better notnull detection Feature: reset_db add option to explicit set the PostGreSQL owner of the newly created DB Feature: shell_plus added support for MongoEngine Feature: sync_s3 enable syncing to other cloud providers compatible with s3 Improvement: ForeignKeyAutocompleteAdmin add option to limit queryset Bugfix: graph_models fix issue with models without primary key Bugfix: graph_models fix UnicodeDecodeError using --verbose-names Bugfix: dumpscript fix problems with date/datetimes by saving them now as ISO8601 Docs: many improvements Docs: Chinese translation !!! Python3: various improvements Tests: add Django 1.6

Learn how to use basic authentication with your new API, and setup custom permissions to get fine grained precision of what can be done with data.Watch Now...

Making News Year's Resolutions is something I do every year. It's important to me. I like to set goals and see how well I do. Resolutions Write and publish fiction. This is a childhood dream that I would like to do. Write and publish as least one new technical book, preferably on Python. Release the second edition of Two Scoops of Django. Learn Haskell or some other interesting programming language. Figure out a way I can make money working in open source efforts. As a consultant, this explains why I don't do it more. Visit the three continents I've yet to see. That means South America, Africa, and Antarctica. Visit at least one new nation. It's hard to come up with preferences since the whole world is so exciting. Take road trips around the USA. See the Grand Canyon. Visit more family. Do 1000 push-ups or similar exercises in a single day. I'm at 300 now. Pull off an Aú sem Mão (no-handed cartwheel). Blog once a week. That is at least 52 blog entries! Take a fun class with Audrey. Upload all my outstanding pictures to Flickr! Learn how to surf or snowboard. See all my friends. All of ...

Inspired by Alex Clark's meme, here is my own entry into this end-of-year fun. What’s the Coolest Thing You Discovered This Year? Alex had his own focused on Python applications, frameworks, and libraries. Me, I'm going for a more general approach. This year I discovered: LaTeX That book I co-wrote with Audrey Roy? Written in LaTeX. It wasn't easy to get started, even with Audrey as my instructor, but now that I grok it, I feel extremely empowered. That's because I can do all kinds of awesome, insane hacks when Sphinx generates PDFs. I hope I get the chance to show off some of my new madness to the world in 2014. How to Tie A Bow Tie If you are going to be formal, the bow tie adds an enormous amount of class. However, clip-ons are considered tacky, especially as it's considered stylish to loosen the bow tie later in the evening to demonstrate you aren't wearing that clip. Of all the stuff I watched on the subject of bow tying, this video was the longest and the best of them all. Awesome Foods Around the World Here's the short-list: Oscypek is a Polish smoked, salted cheese that should ...

Chartkick is an application, a library for making JavaScript based charts. There is JavaScript library, Ruby wrapper and Python wrapper for Django and Flask. Let see what Chartkick can offer us in Django applications.

In this post I will explain yet another simple number theoretic function, called the Euler Totient Function. This is a simple function which is used indirectly in many problems in programming contests. The function is denoted as phi(n). phi(n) basically gives the number of integers greater than 1 and lesser than n which are co-prime to the integer n. In other words, it is the cardinality of the set S = { a: 1<a<n , gcd(a,n) = 1 }. To calculate this function, we will use a straight forward brute force approach. With a simple analysis, we can show that the running time of this algorithm is O(sqrt(n)). - Start with an estimate of the totient function as n - Iterate for i from 2 to sqrt(n) - If i | n , we know that all factors of i less than equal to n will have gcd with n as atleast i. Hence, they can never be in the set S. Hence remove them from the result. - While i|n, divide n by i. - After the iteration of i from 2 to sqrt(n), if n is a number greater than 1, remove all its multiples less than n ...