> Return to the post, I would like to know about whether the > lucene support > the substring search or not. > As you can see, one field of my document is long string > filed without any > spaces. It means the token doesn't work here. Suppose I want > to search a > string "TARCSV" in my documents. I want to return the sample > record from my > document set. I try the Wildcard search and Fuzzy search > both. But neither > seems work. I am very sure whether I do all things right in > the index and > parse stage. Do you any one has the experience in the > substring search?
Yes it is possible. Two different approaches are described in a recent thread. http://search-lucene.com/m/Wicj8UB0gl2 One of them uses both trailing and leading wildcard, e.g. q=*TARCSV* Other approach makes use of NGramFilterFactry at index time only. It seems that you will be dealing with extremely long tokens. It is a good idea to increase maxTokenLength (default value is 255) SOLR-2188 Tokens longer than this are silently ignored.
