todo

CHM decompiler is a program that converts the internal files of CHM files back into the HHP, HHC, an

Options:
	timestamps in integer/date format
	print out the window properties that are marked invalid in comments

Features:
	recreate hhc/k without the #TOCIDX/BTree files - maybe impossible
		#TOCIDX
			children are after their parents
			can get which except the last are books, which are pages
				last can proably be guessed using #TOPICS file - no url
			some intelligent routines decide which things are parents of which nodes
			or do a dumb dump (for now)
		BTree
			can just dump out all the #TOPICS entries that are from the hhk
			not accurate in terms of parent/child etc or keywords, but good enough
				still need to warn the user though
	samples - need proper case insensitive recursive dir search
		glob( "samples", GLOB_NOSORT|GLOB_MARK, NULL )?
	[TEXT POPUPS] - "
	html file processing:
		alinks - need a html editing parser & a general ww-links-btree interface
		split - ", maybe impossible
	detect klinks added to the index by comparing the sitemap and recreated hhks
	guess names of context ids based on title etc (ala helpdeco)
	different backends: filesystem (unix, win32, mac?), chmlib, ms-its
		file search
		stat
		open, close, read, seek, tell...
		no_hard_links()
		path manipulation