I am wondering why the Suggester component should return duplicate
suggestions.
Taking as an example a Suggester configured over fieldA with
the AnalyzingInfixLookupFactory, I will see in my results the entire stored
content for that field for all the documents, which means the possibility
of duplicates :
eg.
"Region":{
"co":{
"numFound":10,
"suggestions":[{
"term":"<b>Co</b>sta Blanca",
"weight":0,
"payload":""},
{
"term":"<b>Co</b>sta Blanca",
"weight":0,
"payload":""},
{
"term":"<b>Co</b>sta Smeralda",
"weight":0,
"payload":""},
{
"term":"<b>Co</b>sta Smeralda",
"weight":0,
"payload":""},
...
This sounds to me as redundant data, because when talking about
suggestions we are actually showing only the term ( so we have lost
any other attribute of the original document).
It's true we have the weight and payload that possibly could
differentiate the suggestions ( basically in the sorting or if any
payload present for term).
I am just curious what are the benefit of this duplication when we
could return a set of suggestions and avoid the client to do the extra
step of filtering duplicates ( usually you don't want to show
duplicates suggestion , that basically will produce the same search
when clicked.
Cheers
--
--------------------------
Benedetti Alessandro
Visiting card : http://about.me/alessandro_benedetti
"Tyger, tyger burning bright
In the forests of the night,
What immortal hand or eye
Could frame thy fearful symmetry?"
William Blake - Songs of Experience -1794 England
