Index files

You might notice that all but one of the files observe the following pattern :

SOME_UUID . SOME_EXTENSION

Your index is in fact divided into smaller independent indexes called segments. The reason for this division is explained in Incremental Indexing. The UUID part stands as the segment name, while the extension express which data-structure is stored in the file.

• .info contains some meta information about the segment
• .term contains the term dictionary
• .idx contains the inverted lists
• .fieldnorm contains the field norms
• .pos contains the positions information
• .store contains the stored documents
• .fast contains the fast terms

The last file, meta.json, contains in a JSON format :

• The schema of your index
• The name of the segments that were committed and are available for search.