The Multi Schema Validation toolkit is a Java based toolkit consisting of 8 different submodules.
The core module is the Multi-Schema XML Validator (MSV) for the validation of XML documents against several kinds of XML schemata
The core supports RELAX NG, RELAX Namespace, RELAX Core, TREX, XML DTDs, and a subset of XML Schema Part 1.
Most outstanding is the design of MSV core using the Abstract grammar model (AGM). This is a schema-independent grammar model. All supported schemata are parsed into this internal representation. This model, coupled with the grammar reader, may be useful for other applications. For instance, two use cases are the generation of source code using schema2template or the ODF Validator.
The builds of all MSV sub-projects were tested successfully using JDK 8, JDK 11 and JDK 17 on Windows and Linux.
The MSV toolkit consists of a eight sub-projects, the main projects in bold. Each sub-projects has its own directory, its own build script, etc.
| sub-project | description & dev guide reference |
|---|---|
| xsdlib | XML Schema Datatype (XSD) Library An implementation of W3C XML Schema Part 2 (see xsdlib JavaDoc). |
| msv core | Multi-Schema XML Core Validator A schema model and validator implementation (see MSV Core JavaDoc).</br>Dependent on XSDLib and testharness. |
| generator | XML Instance Generator A tool that produces valid XML documents by reading a schema. Dependent on MSV. |
| schmit | MSV SCHema In Transformation XSLT add-on (Schmit) XSLT Extension For Schema Annotation. |
| relames | Multi-Schema XML Validator Schematron add-on An experimental implementation of RELAX NG + Schematron validation. Dependent on MSV. |
| rngconverter | RELAX NG Converter reads a schema and produces an equivalent RELAX NG schema. Dependent on MSV. |
| tahiti | Data-binding implementation |
| trexconverter | TREX Converter Reads a schema and produces an equivalent TREX pattern. |
NOTE: Not all previous forks and releases embrace all the projects below, only the latest msv does.
If you want to contribute code than
mvn test.MSV is build with Maven. To build the entire project, use Maven 3.</br> Maven is downloadable for free at Apache.. To build via command line at project root level use: ‘mvn install’ This builds the release packages for all sub-projects in a proper order. When preparing release packages, it is a good idea to use this target so that dependencies are processed correctly. (But you should run a project-local “release” first to make sure that there is no error in the repository.)
When you are working on a sub-project, you can build in the sub-project directory via ‘mvn install’ saving some time by building this alone.
Most of the sub-projects have a similar directory structure aligned to the standard directory layout of the Maven build system.
| directory | description |
|---|---|
| src/main | keeps source files. Files in this directory will be included in the release package. |
| src/test | keeps test files. Files in this directory will NOT be included in the release package. JUnit is used throughout the project. |
| target | keeps the compiled .class files. both “test” and “src” are compiled into this directory. |
| directory | description |
|---|---|
| docs | Project documentation. Files in this directory are shown at https://xmlark.github.io/msv/. |
The sources of the deliverables of the sub-projects in bold (in the subproject table above) have a BSD license. but their tests and all other sources have missing license headers. Sometimes Apache 1.1 license header do exist. The original MSV code repository from Sun/Oracle is no longer accessible. The Glassfish team as new owner is not responding https://javaee.github.io/other-migrated-projects.html but a fork exists from the former Code Owner Kohsuke Kawaguchi (KK) at Oracle.