arensb@cvl.umd.edu (Andrew Arensburger) (12/02/90)
For various reasons, I need a method of constructing diff-files for tree structures, or a reference to an article describing such a scheme. The setup: some set of data is represented as an N-ary tree, indexed on some field (an ASCII string, for example). Both terminal and non-terminal nodes may contain data. Data are stored as a record, and all nodes on the same level (distance from the root) contain the same kind of data, although record types vary from level to level. There is one true and complete "reference" tree, containing all of the available data. In addition, there are several auxiliary trees which are similar to, but usually not identical to the reference tree. The auxiliary trees may be missing large branches, or the data in a node of an auxiliary tree may be different from the data in the corresponding node of the reference tree. Auxiliary trees will usually have no "extra" nodes (ones which have no counterpart in the reference tree). The problem: I need a method of storing all of these trees in files. Since the auxiliary trees are fairly similar to the reference tree, it makes sense to store a) the reference tree, and b) a series of diff files that contain the differences between the reference tree and the auxiliary tree, which make it possible to reconstruct the auxiliary tree. I am primarily interested in saving disk space, but I would like to see some "fast" solutions, or simply clever ones. Please reply via e-mail to arensb@cvl.umd.edu. Thank you for any and all help. -- -------------------------------------------------------------------\\\\^ Andrew Arensburger | K&R C! | I hate Lisp functions o\\\\\- arensb@cvl.umd.edu | ANSI no! | that start with / ...!uunet!mimsy!cvl!arensb | | "(catch (mapcon (throw" \_/