On Dec 5, 2007, at 2:40 AM, Timo Nentwig wrote:
On Friday 30 November 2007 19:28:12 Grant Ingersoll wrote:
I guess the question becomes what is the nature of your fields? Do
you have some really large fields that you want to avoid loading b/c
they are not shown initially? That is the main use case, I guess.
I wonder why there's not LazyFieldSelector that loads all fields
lazyly in
general. Is this for some reason slow or why is this obviously
reasonable
class missing?
Probably because it wasn't obvious to me :-) That is, my use case was
almost always to load small fields right away and large fields
lazily. I haven't tested the cost of this, but you would essentially
be doing all the work and then some to load them all lazily, so I am
not sure I see the point. For example, if you have 4 fields, 3 small
and 1 large, the FieldsReader more or less does:
1. move the pointer to the start of the doc
2. Check the field selector
3. if Lazy, store a pointer, else load the field or skip
4. Skip ahead to the next field, repeat 2-4 until all fields in the
doc have been addressed
Then, when you go to access those 4 fields, which you most certainly
will at some point soon, otherwise why did you get the document to
begin with, it then has to go reset the pointer to that position and
then load the info. Chances are, b/c of the way OSes handle disk,
etc. you've already loaded all the small info (and probably in some
cases the big data too, depending on size) so why not just return it
the first time? In cases of really small fields (i.e. integers, etc.)
the cost of storing the pointer is probably really close if not the
same as storing the field, if not more b/c now you have to store the
pointer plus do a seek.
Cheers,
Grant
---------------------------------------------------------------------
To unsubscribe, e-mail: [EMAIL PROTECTED]
For additional commands, e-mail: [EMAIL PROTECTED]
