Wed 19 November 2014

I've learned so many fun little lessons over the past week, its time for another post!

What is a URI? Is it the same as a URL?

URI stands for uniform resource identifier. A very common form of a URI is the URL, which stands for uniform resource locator. The URL is just the address of a website - ie. the thing you type into your browser when you want to visit a webpage. Apparently, sometimes, the terms URI and URL are used interchangeably. But, if you want to be precise, know that a URL is a type of URI. Time to dig a little deaper into URL's...

What meant by the 'path' and the 'query' of a URL?

The wikipedia page, here, does a good job at explaining the URL. I'll summarize what I learned from it here. The URL consists of several different parts. I'll introduce these parts and discuss them with respect to this example: http://www.cineplex.com/Search?Query=Interstellar

• The very first part of the URL is the protocol. It specifies which application protocol is being used to obtain the resource. In our example the application protocol is 'http'.
• The domain name can be an IP address or a registered and easier to remember name associated to an IP address. In our example the domain name is 'cineplex.com'. If you are creating a server on your own computer, your domain name might be 'localhost'.
• After the domain name a port number can optionally be specified. By default we are directed to port 8000, and so it is almost always unnecessary to manually specify a port.
• A single domain name or IP address can host a multitude of files. The particular file/resource that a URL is after is specified by the URL path. It specifies the address of the particular file we are requesting and is analogous to the way in which we specify the path to a particular file within our own computer. In our example the URL path is '/Search'.
• The query string contains data (usually inputed by the user) which will be passed to some sort of software running on the server. The example I've shown is the result of typing "Interstellar" into the search bar on the cineplex homepage. "Interstellar" was passed to the servers computer and used in some program to figure out the output. Then the results were passed back to me in the webpage with URL "http://www.cineplex.com/Search?Query=Interstellar". The query string part of the URL is '?Query=Interstellar'. As far as I know, the query string always starts with a question mark.

urllib.uncode('string') is great! Here's why:

As I may have previously mentioned, I have decided to write my own web framework. Not because I think I can do a better job than the existing frameworks, but because I want to learn more about what is happening behind the scenes with web development. An issue I encountered as I developped my framework involved html forms and processing user input. If a user inputed any special characters, like '!', '\~', '?', etc., my browser would encode these characters and send my server an http request with these encodings. For example, '!' was encoded as '%21'. Why is the browser doing this? Although its tempting to think that this character encoding is just a pain in the butt, its actually pretty useful! Indeed, some characters in the URL have a special meaning. The question mark, for example, is always used to signify the begining of a query. So whenever the USER inputs characters which have (or may have) a special meaning, the browser will encode these characters to avoid confusion. The places where these characters are used by the browser to indicate a special meaning, they are NOT encoded. So, to make this clear, when a user inputs "?" into a form, it will be encoded. But if a query string is present, the "?" at the begining of the query will not be encoded. And this makes my life alot easier when I have to parse through the raw http requests. I can guarantee that whenever I see a "?" it is signifying the begining of a query string, and NOT some user input. Great! But now, I DO have to worry about decoding the encoded characters so that they can correctly be passed as a variable to an html page. How do I decode a special character from its encoded represenation? Using the python's urllib module!! Check it out:

1
2
3
4
5
6
7
8
9

>>> urllib.quote('!!!')
'%21%21%21'
>>> encoded_input = urllib.quote('&!?')
>>> print encoded_input
%26%21%3F
>>> decoded_input = urllib.unquote(
encoded_input)
>>> print decoded_input
&!?

What is a template engine?

A template engine provides a way for users to put variables and possibly even logic into their html templates. Templating engines consist of a templating language and a templating compiler. The templating language specifies syntax that will be understood to mean specific things in your html template. So, for example, you could say that two curly braces will enclose any variables and that a curly brace and a percent sign will enclose any python logic. Then, you're compiler will parse this html template and then will see which parts need to perform logic or input variables, and will ultimately translate the template into an html page void of any logic or variables, that can be displayed by your browser. Often, template engines are part of what makes up a web framework.

Generators are a kind of python object!

You know how you can write lists like:

1
2

>>> [x for x in 'abcdefg']
['a','b', 'c', 'd', 'e', 'f', 'g']

This way of specifying a list is called a list comprehension. I just found out yesterday that if we write the same thing but instead use the non-square embraces, then we create something called a generator. Check it out:

1
2

>>> (x for x in 'abcdefg')
<generator object <genexpr> at 0x10b323910>

Woah! So whats a generator? I honestly don't know. Yet. All I know is that its a different kind of python object, one which I had previously no idea about. By typing in `dir((x for x in 'abcdef')) we can see the collection of methods that generators have, and it is, obviously, a different collection than the methods associated with lists. I hope to come back to generators in the future to explain why they are used and what they are good for!

Hexadecimal Numbers in Python

Have you ever seen numbers starting with 0x in python, and not really known what they were for? I have seen these around, and always just thought they were some sort of complicated encoding. In reality, numbers begining with 0x are just hexidecimal numbers - ie. numbers with base 16 instead of 10. (If you aren't familiar with hexidecimal numbers, no worries. You can read up on it here.) So we can write:

1
2

>>> 0x11
17

since 11 is the base 16 representation of the decimal number 17. Mystery solved.

Category: Blog Tagged: Web