Sunday, May 15, 2005

scrape

my day trader friend has cool analysis tools, but wants a bigger hammer.
he wants 6 months of daily's (open, hi, low, close, volume) for all stocks on the nasdaq so he can filter ALL stocks for current trends and then only have to study the top 10 doing what he wants in a given day instead of relying on someone else giving him 'ideas'.
this seems easy enough, especially given http://finance.yahoo.com/.

who'd of thought google suggest is actually built on a tool that can be used as a screen scraper?! after some reading about ajax i discovered the Microsoft.XMLHTTP tool i was using to do the scraping actually met with some very good reviews (recent magazine i read) as MS started the foundation for making the ajax stuff possible.

so, when your own IIS server at home fails (don't ask. i've never seen one not work except this one), hope on over to http://www.pablosoftwaresolutions.com/ and get baby web server 2.6. this baby does asp out of the box! awesome. give me an ado connection string for an access database and some regular expressions for some fancy parcing (thanks dan for the heads up) and now we're getting somewhere!

everyone likes to see cool colored code, so in the absence of patience to install a real developement environment at home, i opted for notepad ++.

about 70 lines of code and i get CLEAN histories of stocks i can cram into a database. i love regular expressions. i don't understand them, but that sure is not going to keep me from using them!

i thought this was going to take me longer to 'prove' (get a working beta with all major functions done) - but it only took about 3 hours. maybe i'm actually becoming more tallented?

if anyone wants the code, the server, the notepad++ all rolled up into one nice zip file for playing with on their own, let me know.

2 comments:

k2h said...

not to be picky, but it seems your 'code' has screwed up the sidebar in IE. nice =)

forkev said...

that is some powerful code!
i'll edit the post.