Okay, this is for [EMAIL PROTECTED] -- you guys were my primary audience, but a didn't realize there was a cross-posting rule, but it makes sense in retrospect. Anyway: Lately I've been putting the finishing touches on a set of modules that retrieve historical stock quotes from various locations on the web. I have a number of questions and concerns. It seems obvious that this is desirable functionality, but there are some issues that are not completely within my control. All of the freely available historical quote data I've seen so far is only available in HTML. My modules allow for the specification of several characteristics of each source in order to effectively and reliably gather the requested data. (note that the modules do not REQUIRE html sources, but so far that's all they source). These modules necessarily rely on specific URLs and page layouts. I have done several things to minimize dependencies on page layout, so for the most part when web sites alter their pages the quotes will still be fetched (briefly, a couple of the tricks are context independent table extraction and daisy chain failover from site to site). But it is always possible that something might change with the sources that will break the information flow. Is this sort of reliability level acceptable for CPAN? I will of course update the site-specific modules to the best of my ability, plus they are always user-tunable. And they will of course have a big fat DISCLAIMER that indicates you'd better know some perl if you want to use them for any mission critical activities. The next issue is the names. The ones I picked are big and fat, but quite descriptive. I'd like to know what you think (so far I've only got two specific site instances): Finance::HistoricalQuotes Finance::HistoricalQuotes::MotleyFool Finance::HistoricalQuotes::FinancialWeb The first is of course the base class, which is an LWP::UserAgent derived creature. The others are specific instances which tailor the behavior based on the quirks of that data source. In typical usage, you would use the top level module, and if desired specify a failover order for the subclasses. (Failover can occur for a variety of reasons -- the site could be down, or you could be requesting a symbol that no longer exists. I've found that the Motley Fool is very reliable, but does not provide quotes for defunct ticker symbols. FinancialWeb is slower, and a bit less reliable, but does have defunct tickers...I am, of course, wide open to suggestions on other reliable sources) Any thoughts? Thanks, Matt Sisk [EMAIL PROTECTED]
