On 14/09/2010 12:16 AM, Justin Johansson wrote:
Is D2 suitable to implementing the W3C specifications
for XPath 2.0, and particularly the XPath & XQuery Data
Model (XDM), with both coding succinctness and runtime
time and space efficiency?
Now, by suitability, I don't mean if it can be done. Of course
it can be done in COBOL, assembler or any Turing machine
if one tries hard enough.
The challenge for D2 is to show that these W3C specs
can be implemented in D2 with ease (as say compared to
implementation in C++ or Java).
The biggest challenge that I see in any intelligent
implementation of these W3C specifications is to
produce an implementation that
(1) makes good balance of space and time complexity
from an algorithmic point of view
(2) demonstrates a well-read codification in the target
language which mirrors well, that is traceability,
with the W3C XPath/XDM specifications.
(3) implicit in (1) and (2) the target language lives
up to its expectations as advertised to be capable of
producing a concise and succinct implementation of
something as complex as XPath 2.0.
Is D2 up to a challenge a steep as this?
To make this challenge a bit more realistic in what is
humanly possible for a $64K question, let's reduce this
challenge to an XPath 2.0 implementation which ignores
XML node types as possible XDM item types.
I resubmit this challenge to address only the atomic
item types (i.e. such as xs:anyAtomicType, xs:decimal,
xs:integer, xs:string and xs:boolean) ignoring anything
else that is related to XML.
Me thinks this diluted challenge remains difficult to
implement in D (or C++ or Java for that matter).
-- Justin Johansson
