> what is the typical usage pattern of that list? For me, it's almost always streaming through it, building up some other data structure as I run along.
> if (memory) space is very tight, then the result of the parse > could just be a list of file offsets, to be used later, > when elements are accessed. Yes, this makes more sense from a random access point of view - or perhaps if you will only examine a few elements, but you don't know which right away. > http://hackage.haskell.org/package/xml-conduit-1.2.0.2/docs/Text-XML-Stream-Parse.html Hm, thanks, maybe I should take a look? -k -- If I haven't seen further, it is by standing in the footprints of giants
