v1.0.0-beta

1.0.0 (2019-07-23)

Breaking Changes

• #1073 equalTo constraints no longer produce null (d05eef3)
• #1133 Re-enable schema validation by default (302e0a8)

Features

• #896 Basic upfront tree pruning (6b24852)
• #1120 Add frequency of appearance option when using setFromFile (dcfffe9)
• #701: Automatic versioning and release (8a371df)
• build: Migrate from Maven to Gradle (6a43c5e)

Bug Fixes

• #1076:
  • Prevent invalid formatting with formattedAs (8de51d1)
• #1000
  • Integers outputted to JSON as numbers not strings (d9d6f30)
• #701:
  • Fix circleci typo (c429600)
  • Update commit message checking (134a3e7)

v0.1.3 - Beta 3

Key features in this release

• Several new data types have been introduced, as follows:

  • ISIN
  • SEDOL
  • CUSIP
  • RIC
  • FirstName
  • LastName
  • FullName
    These types are all defined as data types in the profile document, which is a change to the previous partial support for ISINs.

• In cases where data must (or must not) be a member of a set, large sets can now be loaded from external files instead of being included in the profile.

• The generator can now stream to standard output in generation mode, rather than writing its output to a file.

• The codebase has been split into separate smaller packages. The build process generates individual .jar files for each package, but also still generates a complete generator.jar file which contains all dependencies of the program.

List of significant tickets

• #440 - Add ability to generate to stdout in generation mode.
• #487 - Combine financial codes with regular expression constraints.
• #909 - Generator modularisation.
• #933 - Exception thrown in ISIN generation in combination with not equal to.
• #938 - Generate SEDOL values.
• #939 - Generate CUSIP values.
• #940 - Generate RIC values.
• #943 - Generate ISIN values.
• #948 - Generate plausible real names.
• #981 - Issue with anyOf in violation mode.
• #991 - Program startup issue caused by jar manifest.
• #994 - Documentation updates.
• #1004 - Complex data types should be specified as types in the profile.
• #1006 - issue with allOf in some situations
• #1041 - Load sets from external files.
• #1042 - Financial code and real name data types.

v0.1.2 - Beta 3

• #918 Remove Routed Walker
• #912 DataBagSource is unnecessary and confusing
• #141 Allow specification of granularity for datetime fields
• #880 Update schema validator to reject invalid string length constraints
• #837 As a user I expect to get fewer unique strings as I apply more constraints
• #849 Generated strings should be shorter than 1000 characters
• #900 Update/Rework Getting Started Guide

v0.1.1 - Beta 2

• #890 Check for contradictions across fields
• #902 Update GitHub templates
• #897 Update README to use correct/updated syntax
• #889 use minimal combination strategy in random mode
• #886 fix visualise reductions
• #893 Hide exception output during cucumber runs
• #892 strip trailing zeros when checking scale
• #891 create getBigDecimal for random
• #875 check granularity against numeric
• #879 Ensure string operators cannot express strings longer than 1000 chars
• #877 Rename AutomationUtils -> AutomatonUtils
• #874 refactor numeric restrictions
• #870 recheck granular to bug
• #871 Recheck temporal bug
• #859 string length tests
• #864 Granular to tests
• #858 string docs update
• #854 json schema ux improvement
• #832 cucumber errors require message assertion
• #834 handle set restrictions with empty whitelist but can be null
• #815 profile violations tests
• #851 granularTo tests
• #840 Edit json schema
• #842 update docs
• #847 Remove codecov reports
• #838 Cucumber style standardisation
• #836 Update Readme

v0.1 - Beta 1

Initial beta release

• #606 DateTimes are always treated as UTC
• #767,#768 Fields are checked to see if they are typed, can be disabled via --allow-untyped-fields
• #35 JSON schema validation has been disabled by default, can be enabled via --enable-schema-validation
• #795 Numeric values in a set are restricted by any present granularTo constraint
• #818 Bug-fix for violating a profile with multiple rules
• #761 temporal data type has been renamed to datetime
• #667 datetime values are emitted to millisecond granularity

v0.0.6 - Alpha 3

#698- Validate the profile JSON against the JSON schema
#704- Specify numeric values via integer or decimal
#756,#710- Fixes for production of values
#707- Numeric values restricted to 20.d.p. (granularity) and between -(10²⁰)...10²⁰ (inclusive)
#771- Fixes for the visualise command
#763- Improvements to reporting of generator velocity
#672,#611- Fixes for when temporal constraints contradict

v0.0.5 - Alpha 2

Breaking change, profile schemaVersion must be 0.1 now, see #693

• #613 & #635 - Fix for contradicting temporal constraints
• #341 - Introduced ability to output data in JSON format
• #649 - Introduced (draft) JSON schema for the profile
• #714 - Updated generator to emit RANDOM data by default, limited to 1000 rows by default
• #636 - Temporal values restricted to between 0001-01-01T00:00:00.000 and 9999-12-31T23:59:59.999 (inclusive)
• #527 - Fixes for constraints embedded in an anyOf/if
• #695 - Fixes to ensure RANDOM emits temporal values from correct range
• #693 - Updated schemaVersion of profile to 0.1
• #577 & #680 - Infrastructure improvements

v0.0.4- Alpha 1

Changes from previous release:

• #499 - Release now contains dependencies
• #402 - Selective violation of operators
• #446 & #510 & #526 - String generation fixes with unicode characters
• #464 - Progress is now reported by the tool
• #431 - Constraints must be contained within a rule
• #543 & #560 - Fix to reductive iteration visualiser
• #517 - number of rows is mandatory when random mode is enabled
• #569 - fix for trace constraints
• #521 - fix for using not(allOf())
• #554 - fix for data generation with contradictory profiles
• #548 & #576 & #580 & #567 - usability improvements
• #545 - random value generation fixes
• #627 & #566 & #612 & #637- file output checks
• #579 - fixes for violation of larger profiles
• #595 - validity checks in the profile

v0.0.3- Fixes to RANDOM data generation & Package dependencies

Changes from previous release:

• Generator package now contains all dependencies required to run the generator
• Generator can now generate volumes of RANDOM
• Generator command line arguments have changed, no longer use generateTestCases <profile> ... now use generate --violate <profile> ...

Bug fixes:

• #499 - Generator JAR file now contains dependencies
• #479 - JSON and CSV outputs are now consistent (and fixed) when writing dates (default format can be overriden with formattedAs)
• #469 - Generation of violating data is now a command line switch (--violate) (see updated documentation)
• #466 & #68 - Generator can now produce rows of random data, remember to specify the -n <rows> switch otherwise the generator will run indefinitely

v0.0.2- Performance improvements to initial release

Changes from initial release:

• Significant performance improvements when working with string values

Bug fixes:

• #449 - improved memory management when generating data, increasing overall performance
• #460 - extension is not added twice to filenames
• #461 - relative path names are now supported
• #463 - null pointer errors with some profiles
• #468 - resolved some issues with sample profiles