On Tue, May 31, 2011 at 12:36 PM, Aristid Breitkreuz wrote:

2011/5/31 Michael Snoyman

...

On Tue, May 31, 2011 at 11:17 AM, Aristid Breitkreuz wrote:

...

Except currently only SimpleDB actually works. Is your XML cursor stuff ready soon? Because that's one of the reasons why I paused development. AWS involves a lot of XML and I felt that my current  combination of xml + xml-monad (two packages) is not optimal or even just future-proof.

The code is up on Hackage, and ready to be tested. As it stands now, I'm not actually using it for my internal code at all, so I haven't stress-tested it much yet.

The commit at least says the code is incomplete, and it is also lacking the accessor functions that would make it a superior alternative to the combination of xml + xml-monad. No pressure though: I don't have an acute need for AWS myself, right now (I started the aws package because I thought it was cool, not because I currently need it). Also, my offer to help with Cursor still stands - I'd just like to know which functions I should implement.

Aristid

The offer is appreciated. The bare minimum set of accessors I was planning on implementing were the following XPath axes: parent, ancestor, preceding-sibling, following-sibling, preceding, following, child, descendant. So based on this, preceding, following and descendant are missing, and "child" was misnamed as "children." Then we likely want to include some filtering functions, such as checking names, node type, and so on. But for that, real life usage will be more indicative than anything I say here. I think that the missing axes will require a bit of experience with difference lists to get right. I might have a chance today to take a crack at it, otherwise I guess we'll just see who gets to it first. Do you have any other ideas on what should be implemented? Michael