I would not trust a crawler because it won't pick up orphaned pages. If the site is published through a CMS there should be an administrative function that tells you how many pages or whatever are published. If the site is a static HTML site you'll have to count the published files some other way.