{

Notes: I am assuming you already know how to set up mod_rewrite.

In your conf, do the following

# If the following is true, run the rule following it directly
RewriteCond %{HTTP_HOST} the.original.destination
RewriteRule ^(.*) your.new.destination/$1 [R=301,L]

}

The data provided is in the form of a string “option1=1 option2=2 option3=3″

The solution is quite neat and simple too…

value = “option1=1 option2=2 option3=3″

dict( map( lambda x: x.split(“=”), value.split()) ) )

Essentially this produces a list of lists which is then fed to the dict function as follows:

[['option1','1'],['option2','2'],['option3','3']]

Simple

}

So is it all lost?  I love code elegance, I find it exciting; and yes, I know that is a very geeky thing to say.  Even though Java is potentially one of the most elegant languages around nowadays, the myriad of frameworks and modular approaches available lead to levels of abstraction that definitely have a performance hit on the vm.

But not all is lost, Kirk Pepperdine, who as most Java people know, is a top Java Champion, wrote an interesting article about this on sun.com {

I invite you all to hop over and have a look.  The tips in here are not totally specific to Java, so you can learn a lot from this.

http://java.sun.com/developer/technicalArticles/Interviews/community/pepperdine_qa.html

}

More info on my friend’s blog {

http://marlongrech.wordpress.com/2008/08/11/visual-studio-2008-net-35-sp1-is-out-now/

}

I am therefore going to jot down a quick ‘teach yourself in 5 minutes’ for the basic regEx expressions.  I will not go into too much detail, since most advanced features are not used anyway on a day to day basis, yet the most simple of regEx can be really helpful in coding, even for simple search and replace operations in your IDE.  I assure you that once you get the hang of them, you will fall in love (as I have) {

The Basics

regEx is basically a pattern search, most would already know this, but I’m just stating it just in case.  As in the find menu, you would enter the word you are looking for, in actuality, that word is a pattern.

For this tutorial, I’ll use a mnemonic which I like -> ‘Californians Like Girls in Small Bikinis’ (well not only Californians maybe )

I will be using Javascript notation for the regEx here, yet you can use this in any other engine by omitting the / from the start and end of the expression.

I’m going to try not to overload my tutorial with tech explanations like lexicons and tokens, or how we could break this down using BNF, since I’m pretty sure that anyone familiar with them can figure them out for themselves.

The structure I’m using is:    /Regular Expression/   ->   Matches highlighted in my mnemonic
Followed by a brief explanation.

Simple literal regEx match

1. /al/ -> Californians Like Girls in Small Bikinis
2. /ni/ -> Californians Like Girls in Small Bikinis

In this case, this is a simple match, returning the instances of the searched token.  The expression will match the first instance, the blue instance is returned on subsequent matches.

Simple wildcard regEx match

1. /.al/ -> Californians Like Girls in Small Bikinis
2. /ni./ -> Californians Like Girls in Small Bikinis

Here we introduce the . wildcard.  This means ‘any character’.   So the first term matches ‘any character followed by a and l.

Unbound repetition of patterns

1. /.*al/ -> Californians Like Girls in Small Bikinis
2. /ni.*/ -> Californians Like Girls in Small Bikinis
3. /.*Cal/ -> Californians Like Girls in Small Bikinis

The * symbol is what we call the ‘Kleene Closure’, this means ’0 or more instances of’, so what we are searching for here is ’0 or more instances of any character followed by al’.

Note that in the first instance, the second al is matched, not the first, since the regEx acts on the whole line.  Also note the third example, since we state ‘0 or more’, ‘Cal’ is still matched.

One-or-More repetition

1. /.+al/ -> Californians Like Girls in Small Bikinis
2. /ni.+/ -> Californians Like Girls in Small Bikinis
3. /.+Cal/ -> Californians Like Girls in Small Bikinis

This is exactly like the previous example, with one difference, the + means ’1 or more’, so in the third example, nothing is matched, since there is no instance of any character followed by ‘Cal’

Character classes

1. /[Cm]al/ -> Californians Like Girls in Small Bikinis
2. /ni[ai]/ -> Californians Like Girls in Small Bikinis

The [ ] denote a character class.  This means ‘any one of these characters’.  So in the first case, we are looking for a ‘C’ or an ‘m’ followed by ‘al’, which gives us two matches.  You can see why the second one only returns one match.

Negated character classes

1. /[^Cx]al/ -> Californians Like Girls in Small Bikinis
2. /ni[^ai]/ -> Californians Like Girls in Small Bikinis

This is the negation of the previous example.  The ^ at the start of character class means ‘anything that isn’t on of these characters’.

It is important that the ^ is within the [  ]

Start of sentence

1. /^Californians/ -> Californians Like Girls in Small Bikinis
2. /^Bikinis/ -> Californians Like Girls in Small Bikinis

The ^ at the start of a regEx means ‘start of string’, so we are looking for the ‘start of a string followed by a C, and a, etc…

End of sentence

1. /Californians$/ -> Californians Like Girls in Small Bikinis
2. /Bikinis$/ -> Californians Like Girls in Small Bikinis

The $ in this case matches the ‘end of a string’

}

That will be all for this tutorial.  There will be other tutorials to follow this, which I will add to this post as links once I create them.

Any feedback would be appreciated.

Tutorials ToC:

1. The Basics (this tutorial)
2. Shortcuts
3. Search and Replace (matching groups)
4. Saving memory, and shortening your regEx
5. Advanced: Positive and Negative lookahead and lookbehind.
6. A quick’n dirty way to test regEx, as well as examples in C#, Java, VB and JavaScript

p.s. for anyone curious as to what that mnemonic is for, it is basically used to remember the 5 layers of the epidermis, starting from the lower layer. i.e. corneum, licidum, granulosum, spinosum and basale.

In most cases, this is more due to the fact that our IDE‘s tend to make life so easy for us, that we tend to lose touch with the most routine stuff, like create a WAR file manually, or creating mappings for our ORM of choice {

I shall therefore start jotting down notes on how to perform these most mundane of tasks, so as to keep a handy reference which I can come back to should I wake up one morning and not find any coffee available throughout the entire island

For starters this will be mostly focused on Java technologies, since this is my daily lingo, yet there will also be excerpts of other lingos such as C#, PHP and JavaScript… amongst others (though I seriously doubt there will be any/much VB)

}

Well, consider it done, this regex will now correctly validate days in a month as it should be (next I’ll try to cater for leap years) {

Date format is dd/mm/yy or dd/mm/yyyy (if you need US support let me know).  Also note, that this regex does not use any lookbehind, so it should also work perfectly in JavaScript (which is probably where regex is used most anyway)

^(?:(?:0?[1-9])|(?:[12]\d)|(?:3(?:(?:0)|(?:1(?![\\.-/](?:(?:0?[2469])|(?:11))))))(?!(?:[\\.-/]\d{1,2}[\\.-/])(?:(?:[02468][048]00)|(?:(?:\d{2})?(?:(?:[2468][048])|(?:[13579][26])|(?:0[48]))))))[\\.-/](?:(?:0?[1-9])|(?:1[0-2]))[\\.-/](?:(?:\d{2})|(?:1\d{3})|(?:\d1\d{2})|(?:\d{2}1\d)|(?:\d{3}1))$

}

p.s. inevitably there will be bugs… so if you find any, let me know!

That’s when cheat sheets come in handy {

My Portuguese friend João posted this very handy link on his blog, and I thought it worthy of sharing

Cheat Sheets

}

Note, this allows / \ . or – as delimiters {

^(?:(?:0?[1-9])|(?:[1-2]\d)|(?:3[0-1]))[-.\\/](?:(?:0?[1-9])|(?:1[0-2]))[-.\\/](?:(?:\d{3}[1-9])|(?:\d{2}[1-9]\d)|(?:\d[1-9]\d{2})|(?:[1-9]\d{3})|(?:\d{2}))$

}

p.s. this regex has now been superseded with an improved version here

There’s a myriad of solutions out there on the web discussing this very issue, and I never gave much thought to it until yesterday. {

I came across an excrept of code at my client which was trying to send a csv file to the client (i.e. the client pc), yet the browser (all flavours) insisted on trying to render it, rather than download it.

The solution turned out to be particularly simple; something we all learnt how to do very well when we wanted some sweets as kids… i.e. LIE (to the browser this time)

All you have to do is set the content type to application/binary and the browser will simply not bother trying to do anything with it.

Here’s the magic expression in Java, yet of course this also applies to any other lingo (and if you have a problem, ask me, tell me your language, and if I can, I’ll show you how).

response.setContentType("application/binary");
NOTE: IE works slightly different from other browsers (what's new?), in that if you cancel the download, it will close the socket throwing an exception on the server.  You of course will need to cater for this.
}

I’m working on some code right now, fixing up some system changes, and I have come across lots of crap, but this one tops it all IMHO {

400 lines of undocumented code in one method!

Go figure… and here I was thinking I was a crappy programmer when I was 20 since I used to commonly put around 50 to a 100 lines of code in my methods.

}

Oh well… just one of those nuisances you know.

So, in the hope that some will actually read this and take advantage of it, here goes {

Well, it is no rocket science really.  All you need to do is to enclose your code in a set of

put your code here

where ‘lingo’ can be any one of {

• cpp
• csharp
• css
• delphi
• html
• java
• jscript
• php
• python
• ruby
• sql
• vb
• xml

}

}

Hope that helps

<h2>Happy blogging</h2>

I found tons of posts on-line debating what was the best font, and for one, I always felt that Microsoft’s Consolas was by far the best programming font around (and technically it is almost, but not quite, entirely like the best coding font around).

To summarize what sparks the requirement of such a font, it basically boils down to {

1. Programmers like fonts that look good at SMALL point sizes.  Not because we are masochists (well some of us probably are), but because we want to see as much code as possible on one screen.  Which would also explain the phenomenon of why all programmers push their graphics cards to the highest resolutions possible (oh yes, and it also looks gorgeous on a hi-def lcd).
2. We spend hours on end reading funny words that make sense solely to us and to lexical analyzers, so we need the font to be VERY legible, and to distinguish clearly among similar characters like 0,O,o, ilL1… among others (oh yes, and the myth is true, most developers, after a full day of code, go home to see…more…code!)
3. We use words like ‘sexy’, ‘perfect’ and ‘elegant’ to describe an algorithm, so we should also be able to use such terms to describe the font we use to write such algorithms in.

}

So, to cut it short, here are the contestants {

1. 

   Proggy Tiny 10pt

2. 

   ProFont 7pt

3. 

   Envy Code R 7pt

4. 

   Droid Sans Mono 7pt

5. 

   Courier New 8pt

6. 

   Consolas 8pt

7. 

   Bit Stream Vera Sans Mono 8pt

8. 

}

And that is it really.  I have concluded that for now, after testing all these fonts, Consolas remains the one of choice for me.  ProFont was terribly inviting, yet the boldness of the font is just way too heavy in comparison to the normal formatting.

Now I’m just hoping to find a font which looks as good as Consolas does in 8pt, but at 7pt, since 8 is just a bit too big for my likes.