Sometimes, the stemming filters available in Solr do more than you would like them to do. For example, they can stem brand names or the second name of a person. Sometimes, you would like to protect some of the words that have a special meaning in your system or you know that some words would cause trouble to a stemmer or stemmers. This recipe will show you how to do it.
Before we continue, please remember that the method described in this recipe may not work with custom stemmers that are not provided with Solr.
In order to have the defined words protected we need a list of them. Let's say that we don't want the words cats
and dogs
to be stemmed.
To achieve that, we should start by writing the words we want to be protected from stemming into a file. Let's create the file called
dontstem.txt
with the following contents:cats dogs
Now let's put the created file in the same directory as the
schema.xml
file (usually it's theconf
directory...