v2.6.2

What's Changed

• #537 Found unit test with empty name by @miroslavpojer in #538
• #516 Add support for packed unsigned by @yruslan in #546
• #543 Improve performance of basic ASCII processing. by @yruslan in #548
• #545 Add support for 'string' debug columns for ASCII file format by @yruslan in #549
• #542 Implement filler naming strategies and previous fields based strategy. by @yruslan in #554
• #553 Fix a regression bug introduced in #543 concerning variable OCCURS. by @yruslan in #555
• #544 Added an extension option 'data_paths' that replaces 'paths' for multi path load. by @yruslan in #557
• #556 Fix the option 'file_end_offset' truncates records of each parti… by @yruslan in #558
• Update Scala 2.13 to 2.13.10, Spark 3.2.3 and 3.3.1 for CI and publish builds by @yruslan in #551
• Update Scalatest to 3.2.14. by @yruslan in #552

Full Changelog: v2.6.1...v2.6.2

v2.6.1

What's Changed

• #531 Add support for CP1047 by @yruslan in #535
• #532 - Add code coverage support - Jacoco by @miroslavpojer in #533
• #529 Fix tests failing on EOL issues on Windows. by @yruslan in #530
• Update sbt version used to build the project. by @yruslan in #536

Full Changelog: v2.6.0...v2.6.1

v2.6.0

What's Changed

• Add support for Scala 2.13 by @yruslan in #515
• #192 Decouple AST transformation logic by @yruslan in #519
• Feature/517 string length metadata by @yruslan in #520
• #521 Fixed index generation for variable block (VB) files. by @yruslan in #523

Full Changelog: v2.5.1...v2.6.0

Minor bugfix release

What's Changed

• Fix FILLERs being dropped for a FILLER with OCCURS GROUP. by @yruslan in #511

Full Changelog: v2.5.0...v2.5.1

Feature Release

What's Changed

• #491 Overpunching Strictness made configurable, strict by default by @yruslan in #493
• #423 Add byte length record when generating record ids. by @yruslan in #494
• #396 Add support for copybooks without a root record by @yruslan in #496
• Feature/improve null detection on by @yruslan in #503
• Feature/502 fix signed decimal scaling down pic by @yruslan in #504

Full Changelog: v2.4.10...v2.5.0

Minor bugfix release

• #481 ASCII control characters are now ignored instead of being replaced with spaces.
  A new string trimming policy (keep_all) allows keeping all control characters in strings (including 0x00).
• #484 Fix parsing of ASCII files so that only full records are parsed. The old behavior
  can be restored with .option("allow_partial_records", "true").

Minor bugfix release

• #474 Fix numeric decoder of unsigned DISPLAY format. The decoder made more strict and does not allow sign
  overpunching for unsigned numbers.
• #477 Fixed NotSerializableException when using non-default logger implementations
  (Thanks @joaquin021).

Minor feature release

• Improved schema flattening method SparkUtils.flattenSchema() for dataframes that have arrays. Array size metadata is used to determine maximum array elements, making it much faster for dataframes produced from mainframe files.
• #324 Allow removing of FILLERs from AST when parsing using 'parseSimple()'. The signature of the method has
  changed. The boolean arguments now reflect more clearly what they do.
• #466 Added maxElements and minElements to Spark schema metadata for
  array fields created from fields with OCCURS. This allows knowing the maximum number of elements in arrays when flattening the schema.

Minor bugfix release

• #459 Fixed signed overpunch for ASCII files.

Minor bugfix release

• #451 Fixed COMP-9 (Cobrix extension for little-endian binary fields).