limit crawler to defined depth

there is a need to limit crawler to some defined depth, and importance of this option is to avoid crawling of infinite loops, with dynamic generated urls, that occur in some sites, and to optimize crawler to select important urls.
an option is define a iteration limit on generate,fetch,parse,updatedb cycle, but it works only if in each cycle, all of unfetched urls become fetched, (without recrawling them and with some other considerations)
we can define a new parameter in CrawlDatum, named depth, and like score-opic algorithm, compute depth of a link after parse, and in generate, only select urls with valid depth.