|
PureBytes Links
Trading Reference Links
|
On Fri, 2004-05-14 at 21:53, Sven Napolean Montessori wrote:
> Because HTTP is a stateless protocol, you can massively parallize the
> fetches and gain an order of magnitude in speed.
True enough.
However if Yahoo data serves anyones needs, be aware that you can
download ~ 200 symbols at a time (can't remember exactly), so 7000
queries not needed.
I'm pulling down ~ 8000 symbols for TSX, NYSE, Nasdaq and a hundred or
so ETFs and HOLDRs etc on AMEX - I think it takes about 60 seconds in
total.
I have some python code that does just that, each day at 16:30ET as a
cron job on a unix machine, but could be run on any platform. The data
gets shoved into a SQL database from which I run reports. I'm not using
the data for anything else at the moment - I currently use eSignal to
drive TS2ki and also use TS8.
If interested, "google python yahoo quotes" and you'll find one or two
other folks who have done this and put code up on the web. I can't
remember if anyone is generating a Globalserver compatible import file
(I am not) but it would be fairly trivial to do that.
Seems like you already have a perl solution for same though!
--
Michael Watkins
TrendVue.com
mw@xxxxxxxxxxxx
|