Web browser source code request..

[vanjast]

Did a search but seem to be going around in circles..

Can anyone point me towards Web browsing source code (assembly) and the background knowledge requirements to make a simple browser.
I'd imagine it'll be something like a hmtl code (and/or other) parser of sorts.

Thanks
Van

[dedndave]

i think it was Edgar that had some code for that
let me see if i can find it....

[dedndave]

ok - it was Rob - but i can't find the thread
here is the attachment
it looks like he adapted code by NaN - perhaps that's the user name we need to search

[dedndave]

here are a few related threads....

http://www.masm32.com/board/index.php?topic=7247.0
http://www.masm32.com/board/index.php?topic=10904.0
http://www.masm32.com/board/index.php?topic=15793.0

[vanjast]

Why... thank you kind sir... :U

[vanjast]

I didn't explain myself properly - Here's what I'd like to do.

When a webpage downloads, I'd like to scan the html (or whatever it is) file that downloads, looking for things like adverts, javascipt..etc
The idea is to create a dynamic user web filter. Linked to this will be a TCP/FTP/UDP/etc monitor that will trap any asp type link attempts.
AND..added to this will be a 'Junk Feeder' that will allow the user to swop a transmitted packet with 'junk'.

Essentially a privacy protector package. Maybe this has been done already... but I'll like to try it.. for phun!!
:bg

[zemtex]

ad muncher is such a program that filters ads and it was written in asm. I know the guy who maintains the ad filter database, I talk to him on a regular basis.

[vanjast]

That could be interesting...  :bg
It must be via scanning the html/asp downloaded (generally I'd block any asp type downloads), but it'll be interesting to seewhat flows up and down the net.

I'm busy playing around at the moment clicking on my usual bookmarks (KMeleon) and 99% of the webpages link to other monitoring links.
All going through port 80 - There's no other way.. and one cannot block Port-80
Besides that the 'monitors' have a sh-i,tload of IP addy's to link up on, and to manually filter through a router is... manual labour.
and my router runs on Javascript.... Why..Why do these monkeys do this ??? :tdown

I must make my own dsl router  :green2

[Gunner]

Write a proxy and filter everything that goes through it.

[vanjast]

It's a thought..
My main aim was to provide a simple 'secure' browser under the user preference control by 'hammering the page' as it downloads
A Proxy might be bypassable if any software has access at the Kernel level
:8)

[vanjast]

An idea which I think will be much better...
If I place a 'net-buffer' (so to speak) between my PC and router and connect the NetBuffer to the PC via USB, I'll have 'untamperable' (code wise -that is) filter.
In simple terms there will be no way that webmasters/programmers can circumvent the filter without access to the firmware... Also this little 'goodie' can be used for 'secure' comms.

I'll get down to making one of these right away...  :bg
:8)

[jj2007]

You could use URLDownloadToFile to get the file without risk. Then filter out everything that is "not essential" (by the way, MasmBasic is very, very good at that, with e.g. the Recall, Instr and Replace$ functions).
The problem is that nowadays an awful lot of essential things are being generated on the fly with JavaScript, using the DOM. What you get with filtering is mot likely almost unusable...

[dedndave]

yes - you aren't just talking about reading HTML files
there are so many other file types that one can lose count - lol
as Jochen mentioned, java is a big one - there are also php, css, asp/aspx, and so on
not to mention swf and all the other types of active content

when you start stripping all this stuff away, you can turn a robust webpage into a rather dull page of plain text
in many cases, the result is unusable - facebook and yahoo are good examples

[vanjast]

You could use URLDownloadToFile to get the file without risk. Then filter out everything that is "not essential" (by the way, MasmBasic is very, very good at that, with e.g. the Recall, Instr and Replace$ functions).
The problem is that nowadays an awful lot of essential things are being generated on the fly with JavaScript, using the DOM. What you get with filtering is mot likely almost unusable...

That looks usefull