Proposal Draft for SoC: XPath and DOM Selectors

[Carlos Licea]

Hello everybody. 
	I've been working in my proposal for this GSoC. I'm about to post it but I 
wanted to let everyone have a look and criticize it :). Specially Maks 
Orlovich since he appears as the proposed mentor in the SoC Techbase page.
	Carlos

[Abstract]
With the crescent usage of the javascript applications in the web, the more 
recommendations you can implement the more interoperability you'll get across 
different browsers. This proposal involves the implementation of the proper 
interfaces for the XPath DOM 3, which is in essence a javascript API for r 
XPath; and the DOM Selectors recommendations.

[Content]
Name:
Carlos Licea

Email Address:
carlos.licea at kdemail.net

Freenode IRC Nick:
Carutsu

Location (City, Country and/or Time Zone):
Ciudad Juárez, Chihuahua, México. GMT-7

Proposal Name:
Implementing the XPath DOM Level 3 and DOM Selectors specifications

Motivation for Proposal / Goal:
Implement the most recent standars for my favorite web browser. As can be read 
in the summary, and in title, the goal of this project would be implementing 
the required functionality so that KHTML can properly, and as fast as 
possible, execute the most recent XPath DOM Level 3 funcions and Selectors 
that would enable KHTML to run the, likely, next generation of the Web 
applications.

Implementation Details:
I'm not properly aware of the KHTML internals and as such this is moustly a 
guess:

First the needed functions should be created inside the internals of KHTML. 
Those should behave and be named following the W3C recommendations. Proper 
unity test should be written so that it can be proben that the functions 
behave as expected during the SoC and ensure that the behavior is not modified 
in future changes. Later on, the appropiate bindigs to the JavaScript language 
should be made.

Finally, speed ups should be made, with the test making sure that no 
regresions have been done.

Tentative Timeline:
<I'm not sure how to proceed here since I lack the appropiate knowloadge of 
the KHTML's internals, and as such I'm not sure exactly which tasks and 
exactly how long they'll take, I might maka a rough, and really just a guess, 
estimate according to what I believe it could be. What do you think?>

Do you have other obligations from late May to early August (school, work, 
etc.)?:
No that I know yet. I might get a part time job, but I still don't really 
know. In any case I can easily put 28-30 hours per week in the project.

About Me (let us know who you are!):
My name is Carlos Licea, I'm 21 years old and I've been involved with Open 
Source Software since 2006. I've participed with two other Succesful GSoC 
projects, for the Marble and KOffice teams: "Implementing 2D projection for 
Marble" and "Creating a Filter for KPR to ODF files" respectively. Thanks to 
those great ptojects I have much more experience in how the development of the 
FOSS application occurs.

Finally I should state that I like Konqueror, and hence KHTML, so much that I 
wanted to get involved in it for quite some time. The final reason why I was 
drawn to this project is because it doesn't involve UI which I find either 
boring or hard to get right.