Re: Context URL Storage

In article <33D76EF6.31FA_at_mindspring.com>, dstraub_at_mindspring.com writes:
|> I am trying to use Context to index the HTML files on a web server. If
|> anyone has done this, I was wondering:
|>

snip

|> 2. Is there a copy of the text in context and a pointer to the web? If
|> not, how does Context remotely index the web pages?
|>

If you want to index web pages, there are several ways to do it in ConText:

1. you can get all the physical files onto your disk, then use the external datastore. In this scenario, all the text is on your disk as files.
2. you can put all the page text into the database, then use the direct datastore. In this scenario, all the text is in your db.
3. you can populate a table with URLs, then use the URL datastore (v2.0+) In this scenario, only the URLs are in your database, and ConText will index it using its built-in web client to get the actual text.

In all cases, you will need a 3PT spider to crawl the 'net, as ConText does not (yet) have web crawling abilities.

-- 
----------------------------------------------------------------------------
Garrett Kaminaga                                 Oracle ConText Server Group
----------------------------------------------------------------------------