Hi folks,
I've noticed an unexpected behavior while working with the various
built-in integer field types (int, tint, pint). It seems as the first
two ones are subject to type checking, while the latter one is not.
I'll give you an example based on the example schema that is shipped out
I suspect you missed this comment in the schema file:
***
Plain numeric field types that store and index the text
value verbatim (and hence don't support range queries, since the
lexicographic ordering isn't equal to the numeric ordering)
***
So what's happening is that the field is
Hi Erick,
I see the point. But what is pint (plong, pfloat, pdouble) actually
intended for (sorting is not possible, no type checking is performed)?
Seems to me as it is something very similar to the string type (both
store and index the value verbatim).
-Sascha
On 18.01.2011 14:38, Erick
These are legacy types that aren't, frankly, very useful in recent Solr. So
you can probably safely ignore them.
BTW, you probably want to go with Trie fields (tint, tfloat, etc) as a first
choice unless you have a definite reason not to.
Hope this helps
Erick
On Tue, Jan 18, 2011 at 10:35 AM,
