Release with Spark 3.5.0 / Scala 2.12

Changelog:

• Versions bumps to support Spark 3.5.X and Scala 2.12

Release v0.3.10

Changelog:

• Update sbt to 1.4.7

• Create codacy-analysis.yml (#302)

• Create codacy-analysis.yml

• [sbt] patch updates (#307)

Closes #306 , #304 , #300 , #301 , #300 , #291 , #275

• Update scalatest to 3.2.8 (#317)

• Update scalactic to 3.2.8 (#316)

• Update sbt-release to 1.0.15 (#308)

• Update sbt-pgp to 2.1.2 (#309)

• Update scalactic to 3.2.9 (#325)

• Update scalatest to 3.2.9 (#326)

• Update spark-core, spark-mllib, spark-sql to 2.4.8 (#327)

• Update algebird-core to 0.13.8 (#330)

• Update jts-core to 1.18.1 (#312)

• Update sbt-scoverage to 1.8.2 (#331)

• Update lucene to 8.8.2 (#311)

• [hotfix] sbt migration (#332)

• Update spatial4j to 0.8 (#313)

• Update sbt to 1.5.3 (#333)

Release v0.3.8

See the project milestone 0.3.8.

Release v0.3.7

Breaking change:

LuceneRDDResponse is now extending RDD[Row] both for searching and linkage. LuceneRDDResponse offers now a toDF() for DataFrame convertion.

Changelog:

• LuceneRDD with DataFrame constructor handles ArrayTypes, i.e., multivalued fields
• LuceneRDDResponse extends RDD[Row] (breaking change)
• [sbt] update to spark 2.4.2
• Allow different Lucene Analyzers per field (#164)
• Update to Lucene 8 (#161)
• Allow the user to set a custom analyzer by specifying the class name.

Release v0.3.6

Breaking changes:

• Method LuceneRDD.blockEntityLinkage now takes as argument a linker of type Row => org.apache.lucene.search.Query
• Method LuceneRDD.blockDedup now takes as argument a linker of type Row => org.apache.lucene.search.Query

Changelog:

• Lucene version 7.7.1
• Fix issue (#150) on blockEntityLinkage with same name on columns.

Release 0.3.5

Changelog:

• Persist RDD to disk to reduce memory footprint of LuceneRDD, see here
• Allow users to select which fields to not analyze, i.e., by configuration file or any field ending with _notanalyzed
• Upgrade to Lucene version 7.6.0

Release 0.3.4

Changelog:

• Update to spark 2.4.0
• Remove unused parameter, thanks to @tuleism
• Allow user not to analyze fields by configuration and by naming convention

Release 0.3.1

Changelog:

• LuceneRDD and ShapeLuceneRDD linkage supports linkage using RDD.cartesian
• ShapeLuceneRDD complete kryo class registration
• Update to Lucene 7.1.0
• Minor fixes in Javadoc

v0.2.7

Changelog:

• Add .toDF() method in LuceneRDDResponse
• Update Lucene to v5.5.4
• Fix issue that numeric fields where also stored as string fields in SparkDoc (62be817)
• Use english analyzers as default analyzers.
• Add field(fieldName: String) method in SparkDoc

v0.2.6

Changelog:

• Provide more configuration options for StringFields
• Add Ngram custom analyzer
• Add termvector() method in LuceneRDD
• Add indexStats() method in LuceneRDD
• Lazily evaluate fields() method in LuceneRDD
• Add TermDocMatrix class
• Minor version updates