1.25

• Add support for vector indexes (hnsw, ivf and vec_bf), see
• Add support for KNN search - a filtering operation of float vector fields (Query.whereKnn())
• Add support for new response value rank - a metric by which the results of KNN and full-text queries are sorted.

1.24

• Support new builtin connect signature (reindexer v5 compatibility)

1.23

• Add supporting enum type for index using String or Integer (ordinal) mapping (gh-107).

1.22

• Add lombok dependency to avoid writing boilerplate code
• Remove the limit of 63 indexes per namespace (actual limit is 255)
• Fix the index appendable property
• Implement addIndex, updateIndex, dropIndex operations (gh-105, for testing purposes only!)

1.21

• Fixed isArray flag in nested scalar field indexes for composite field (gh-101).

Attention!
For correct work you have to either update or delete old array-indexes, created over nested scalar fields before connector's updating. Alternatively, you may change types for such fields in you object model from 'scalar' to 'collection'. Otherwise you will get conflict on openNamespace() call.

1.19

• Added use of subqueries as arguments in the query.where() method (gh-99)

1.18

• CTAG format changed, backwards compatible with the old one
• Added indexing of UUID for UUID and String fields, type of index is HASH only
• Fix query.where() for null value
• Tune CPROTO connection

1.17

• Add new params parsing for sharding support
• Remove item IDs request (those IDs were not used by connector anyway)

1.16

• Fixed logging of Query.whereComposite() (gh-91)
• Add calculation of the total count of matching documents in Query (gh-93):
  • Interface CloseableIterator renamed to ResultIterator,
  • add Query.reqTotal(),
  • add ResultIterator.getTotalCount()

1.15

• Add support of RX cluster for running of cproto tests (gh-89)