Question

I have an integer value in SOLR/Lucen.
When I search for values in it, people might add a prefix to the query:
The field is an ID. So for item with id=900, people might search “900” or “it900” or “it-900” Since this is how we represent items (categort +’-‘+id).

Right now, I strip all none numeric characters in my code. Seems to me logically there should be a simple way in the SOLR schema to make this happen.

I tried defining a query analyzer for an integer field. But, Lucen does not like analyzers for INT fields
error message:

FieldType: TrieIntField (int) does not support specifying an analyzer

The field is (The analyzer is not working, as written above)

        <fieldType name="int" class="solr.TrieIntField" precisionStep="0" omitNorms="true" positionIncrementGap="0">
            <analyzer type="query">
                <tokenizer class="solr.WhitespaceTokenizerFactory"/>
                <filter class="solr.PatternReplaceFilterFactory" pattern="(\d+)" replacement="$1" />
            </analyzer> 
        </fieldType>

Any idea how I define this?

Answer 1

Can you specify the filter on (say) a text field instead, and then copy the contents of that string field to an int field? That seems like the simplest way of handling it.