Release 1.0.10

• Added Content Type parameter (the three important indexing parameters are now Indexing Type, Parsing Type, and Content Type)

Release 1.0.8

• Improvements in Deep Parsing
• supporting CSV in the IndexerTool
• supporting Ngram statistics
• better handling text with lots of whitespaces
• beginning to support Sentiments related to Entities

Release 1.0.7

Release 1.0.7:

• Added MetaFields field to the output of Indexer, where clients can pass metadata to the indexer, and this metadata will be passed through unchanged into the output and stored in WebResource
• In Document, changed hash algorithm to run a hash over collectionItemId, if the text field is empty (useful when a document is empty, but has a unique ID)

Release 1.0.6

Major update: Started exposing Signal shortkeys in output

Dictionaries: Corrected typo in name of Substitution dictionary
LangUtils: Addressed an edge case when text contains too many "neutral" characters
FloatVector, StringVector: Added a data element for Signal shortkeys, added handling to emotion arithmetic
ContentIndexSummary: added a deep "copy" operation
HashUtils: Added hash function to calculate Signal shortkey

Release 1.0.3

Added custom idiom dictionary support
Fixed synchronization issue in EnglishIndexerPool

Release 1.0.1

Release 1.0.1:

• Bug fix: Added text validation before calling the Base Forms dictionary to prevent extra-long execution
• Feature: Introduced parameters and the ability to return All sentiment chunks in a text, or Top sentiments (as previously). Set contentIndex.indexingType to IndexingConsts.IndexingType.FULLINDEX or IndexingConsts.IndexingType.TOPSENTIMENTS
• Added two new fields in ContentIndex - ParentUrl and ParentPubTime - to support threaded conversations.
• Feature: Added scripts/runIndexer.sh script to enable automated testing

Initial release

Initial Sirocco release