From: Thorsten Ottosen (thorsten.ottosen_at_[hidden])
Date: 2006-04-19 16:44:04
Thanks for your review so far.
Matt Doyle wrote:
>>In particular, consider if you can answer the following questions:
>>- was the library suitable for your daily xml-tasks? If not, what was
>>- was the library's performance good enough? If not, can you suggest
> Performance is a concern given the recursive nature of the tree. I want write a test case parsing Usenet's group list. By it's nature it demands storage in a hierarchical tree where every node is both a leaf and a branch (like ptree :) , it's a big list, and insertion / retrieval is mostly random. Hopefully I'll have time to do this before the review is over.
>>- was the library's design flexible enough? If not, how would you
>>suggest it should be redesigned to broaden its scope of use?
> The number one place I see this library being specialized, both by the users and the maintainer, is in the parsers. Case in point, my above test requires me to write a different parser. I think it would be useful in the long run to have a basic_parser<...> that could be inherited from as an interface to the user who needs to write their own parser. All the preprocessor stuff, or other implementation details, can be hidden there so that the user just inherits and parses away. There may be other places where customization hooks make sense... Any thoughts?
Let me just say (as a review maanger) that such a test would be
immensely useful (an an obvious candidate for inclusion with the library
if it is accepted).
I know it is an easy task, though.
Boost list run by bdawes at acm.org, gregod at cs.rpi.edu, cpdaniel at pacbell.net, john at johnmaddock.co.uk