[
Lists Home |
Date Index |
Thread Index
]
Bob Foster wrote:
>> Not in my experience. Having access to a syntax tree can be quite
>> useful.
>
>
> Yes, it can. But the most common use of syntax trees is as a parse
> result. A corollary in this domain would be if one could read a stored
> (or user-entered) query and obtain a syntax tree from it. But I didn't
> get that from the sketched proposal, which seemed to me to be
> output-only. Granted, given that there is no actual proposal, one is
> free to attach all sorts of wonderful things one would like to be in
> it, but it would be good to know what use cases you or others are
> thinking an OOPified API would address.
>
Here is the beginning sketchy proposal
=== from http://lists.xml.org/archives/xml-dev/200405/msg00346.html
>With algebraic types, it would be possible to build up an abstract
>syntax tree, which would be passed to the driver before shipping to the
>database.
>
=== end
The algebraic types bit does not matter, it really boils down to having
an ast or a similar representation.
All this started to avoid having only strings.
I don't think asking for use cases makes much sense: you can include
o all use cases of the string based API (as all of this is possible,
either with the objects, or using some wrappers that parse strings), plus
o having a high-level representation of queries (introspection,
reflection, (schema?)validation)
It is not obvious to do any of the latter use cases with strings (unless
you parse them), and good design does not a priori make it more
difficult to get work done.
cheers,
Burak
|