Ethan Quach wrote:
>
>
> Jack Schwartz wrote:
>> Hi Ethan.
>>
>>
>>>>
>>>>
>>>> General:
>>>>
>>>> In general, the parser places the requirement on the derived
>>>> profiles project that the format of the tags be in an easily
>>>> searchable format, for example that they contain a common prefix
>>>> which can be searched for.
>>>
>>> Its not clear to me that we're going to be using the
>>> parser to do syntactical validation of the base manifest
>>> on the server. I will note this possibility though.
>> OK, but there are other reasons too. In addition to use for
>> validation, this will allow you to easily search for the items you
>> want to process. Since they can be anywhere in the tree, the ability
>> to use some kind of wildcard to find all items will be useful.
>
> That's a good point; tough, doesn't the document need
> to qualify to meet the given schema before the parser
> can even read and serve the document to searches for
> a consumer?
Not if we don't structure the parser that way. The parser is planned to
load the data into memory and make it available before it is validated
(by the schema or otherwise). That way, the derived manifest data can
be plugged in and other setup done to the data tree before it is validated.
On a related note: it is up to the client to insure that it consumes
data which has been validated (e.g. that it calls the validator before
using the data, or waits for some other process to validate the data first).
Thanks,
Jack
>
>
> thanks,
> -ethan
>
