Xmlfs

Un article de HurdFr_Wiki.

xmlfs was begun by Marc de Saint Sauveur in 2002, as a netfs example. It was discontinued due to strange bugs occuring in Mach and a few SIGILL, and because Marc lacked time. It has been released in 2005 after Manuel Menal spent some time on it.

Its aim is simple : xmlfs appears as a filesystem tree matching as closely as possible a XML document. It is quite simple, which makes it a good example for explaining netfs translators.

How does it work ?

Simple as it seems, there are some tricks you have to be aware of before using xmlfs. For example, remember that XML nodes do not have unique names (distinguished names). They are sorted alphabetically and we add an index to the name to make them unique. Sorting is supposedly stable, meaning that running xmlfs twice with the same XML document should give them the same index.

Current state

xmlfs is in alpha state. It only works in read-only mode an does not do anything fancy (like size computing, XSLT, or whatever). It has support for attributes (prefixed by @) and CDATA (named .textN). It hasn't been optimized so it might be a bit slow, although it shouldn't be too resource-consuming. Anything else you want to know is in the TODO list. =)

Downloads

You can download it from the repository, via CVS.