7 - should transforms for a given stylesheet be thread clean,
8 or can a stylesheet be enriched with document specific
9 informations and cleaned up later ?
10 - seems that saving back XSLT stylesheet from a compiled form might
15 -> provide functions to circulate in the import tree of stylesheets
18 -> make a separate module.
19 -> document() should not be a problem since Result Tree Fragments are
24 -> Id should be simple, key will probably requires some hash tables.
27 -> try to optimize for ID scan and tests.
30 -> add error checks on all returns
34 -> check the version stuff, design a separate module for error interfacing
35 and default handling, parsing vs. runtime, fatal / compat / warning,
36 and lack of optionnal features.
38 Support Attribute value templates:
39 -> starts to be urgent. Design it in flexible ways but try to optimize
40 to handle most of it at the stylesheet parse time ...
41 => Done for the most part need to check all attributes in XSLT constructs
42 using them and use the dedicated readin function.
45 -> add support for imbricated sorts
46 -> add lang and case-order
47 -> add foreign sorting functions (interfaces ?).
50 -> should we add validation by default ? Make this an option
51 -> redirrect validity errors
53 Contextual error reporting:
54 -> provide a couple of functions providing context analysis, not urgent
63 -> macros, config, verbosity ?
66 Support for disable-output-escaping="yes":
67 -> looks problematic, libxml has no support for anything like this,
68 and unless adding a new node type :-( or tweaking text node and
69 output routines this is gonna be messy ... must be handled at libxml
71 => Done with a trick, text node name is different, requires > 2.2.11
78 -> also put fast lookup for "text()", "comment()", "node()"