1.5.0

Fixed

• (CLI) #93 Remove usage information, when validation failed
• (CLI) #95 NPE when using empty repository definition (-r "")
• (CORE) #101 Role is null in FailedAssert

Added

• (CLI) Support for multiple configurations and multiple repositories. See cli documentation for details
• (API) Possibility to use preconfigured Saxon Processor instance for validation

Changed

• (CORE) #100 Make createReport optional

• (DAEMON) UI rewrite based on Docusaurus

• (API) ResolvingConfigurationStrategy.java#getProcessor() is removed.

• (CORE) Bump Saxon HE to 11.4

• (CORE) Bump jaxb-ri to 2.3.7

• (CORE) Various other dependency updates. See pom.xml

• (CORE) CLI parsing based on pico-cli, commons-cli is removed

v1.4.2

Fixed

• [CLI] #74 fix ansi output of the cli version
• #80 using classloader to initialize jaxb context (to support usage in OSGi environments)
• #75 Improve logging on invalid documents

v1.4.1

Fixed

• Allow more than 3 customLevel elements in scenarios (see xrechnung
  configuration issue 49)
• Remove saxon signature from java8 uber-jar (see 67)

v1.4.0

Fixed

• date conversion when using ConfigurationBuilder#date(Date)
• [CLI] #51 Suffix of report xml is missing
• #53 Fix copyright and licensing information
• #56 namespace element content needs trimming
• [DAEMON] #57 Reading large inputs correctly

Added

• read saxon XdmNode with InputFactory
• [CLI] custom output without the various log messages
• [CLI] options to set the log level (-X = full debug output, -l <level> set a specific level)
• [CLI] return code ist <> 0 on rejected results
• [CLI] read (single) test target from stdin
• [DAEMON] name inputs via request URI

Changed

• InputFactory has methods to read any java.xml.transform.Source as Input not only StreamSources
• InputFactory uses a generated UUID as name for SourceInput, if no "real" name can be derived
• saxon dependency update (minor, 9.9.1-7)
• [DAEMON] proper status codes when returning results (see daemon documentation)

v1.3.1

Fixed

• getFailedAsserts() and isSchematronValid() in DefaultResult.java do not reflect actual schematron validation result
• processing aborts on schematron execution errors (e.g. errors within schematron logic); the validator now generates a report in such cases.
• exception while resolving when using XSLT's unparsed-text() function within report generation

Added

• [CLI] add summary report

Changed

• engine info contains version number of the validator (configurations can output this in the report for maintainance puposes)
• options to customize serialized report file names (cmdline only) via --report-prefix and --report-postfix
• remove unused dependency Apache Commons HTTP

v1.3.0

Added

• Added a builder style configuration API to configure scenarios
• Added an option to configure xml security e.g. to load from http sources or not from a specific repository (so loading is configurable less restrictive, default strategy is to only load from a local repository)
• Support java.xml.transform.Source as Input

Changed

• Inputs are NOT read into memory (e.g. Byte-Array) prior processing within the validator. This reduces memory consumption.
• CheckConfiguration is deprecated now. Use Configuration.load(...) or Configuration.build(...)
• Overall processing of xml files is based on Saxon s9api. No JAXP or SAX classes are used by
  the validator (this further improves performance and memory consumption)

v1.2.1

Fixed

• Validator was creating invalid createReportInput xml in case of no scenrio match

v1.2.0

Added

• Provide access to schematron results through Result.java
  • Result#getFailedAsserts() returns a list of failed asserts found by schematron
  • Result#isSchematronValid() convinience access to evaluate whether schematron was processed without any FailedAsserts

Changed

• Result#getAcceptRecommendation() does not only work when acceptMatch is configured in the scenario
  • schema correctness is a precondition, if the checked instance is not valid, this evaluates to REJECTED
  • if acceptMatch is configured, the result is based on the boolean result of the xpath expression evaluated against the generated report
  • if no acceptMatch is configured, the result is based on evaluation of schema and schematron correctness
  • UNDEFINED is only returned, when processing is stopped somehow
• Result#isAcceptable() can now evaluate to true, when no acceptMatch is configured (see above)

v1.1.3

Fixed

• XXE vulnerability when reading xml documents with Saxon #44
  • Thanks to Landeshauptstadt München IT@M for reporting
• validator unintentionally stopped when schematron processing has errors. See #41.

v1.1.2

Fixed

• NPE in Result.getReportDocument for malformed xml input