Differents in use case and implementation

text

"text" is the default context type for fulltext search. It is based on a prefixtree. The index stores all available information including the words position in the indexed content. Therefore it supports phrase search.

textSimple

"textSimple" is a special context type for fulltext search. It is based on the same prefixtree as the default "text" context type, but just stores which words occure in which documents, not their exact position. Therefore it is not possible to use phrase search on these contexts.

Benchmarks

The benchmarks show that "textSimple" uses less memory and performs equal or bedder on the main operations update and delete.

x-axis shows number of documents in index

Small values in index

Indexing a dictionary: Content of each document is just a single word from the dictionary

Memory

Insert

Deletion of 10 documents

Big values in index

Each document contains of a 100 words from the dictionary.

Memory

Insert

Deletion of 10 documents