Skip to content

Latest commit

 

History

History
209 lines (154 loc) · 10.9 KB

api.md

File metadata and controls

209 lines (154 loc) · 10.9 KB

Validator API

The Validator offers an API which allows you to integrate the Validator in your own applications.

Dependency Management

Currently, we do not deploy to Maven Central or similar. Hence, you need to build and optionally deploy the Validator artifacts to your own shared (or local) repository (see for example Maven Documentation).

Maven

Then you can declare the dependency as follows:

<dependency>
   <groupId>de.kosit</groupId>
   <artifactId>validationtool</artifactId>
   <version>${validator.version}</version>
</dependency>

Gradle

dependencies {
    compile group: 'de.kosit', name: 'validationtool', version: '1.1.0'
}

Usage

Prerequisite for use is a valid scenario definition and the a folder with all necessary artifacts for validation (repository) either on the filesystem or on the classpath.

The following example demonstrates loading scenario.xml and whole configuration from classpath and validating one XML document:

package org.kosit.validator.example;

import java.net.URISyntaxException;
import java.net.URL;
import java.nio.file.Path;
import java.nio.file.Paths;

import de.kosit.validationtool.api.Check;
import de.kosit.validationtool.api.Configuration;
import de.kosit.validationtool.api.Input;
import de.kosit.validationtool.api.InputFactory;
import de.kosit.validationtool.api.Result;
import de.kosit.validationtool.impl.DefaultCheck;
import org.w3c.dom.Document;

public class StandardExample {

    public void run(Path testDocument) throws URISyntaxException {
        // Load scenarios.xml from classpath
        URL scenarios = this.getClass().getClassLoader().getResource("scenarios.xml");
        // Load the rest of the specific Validator configuration from classpath
        Configuration config = Configuration.load(scenarios.toURI()).build();
        // Use the default validation procedure
        Check validator = new DefaultCheck(config);
        // Validate a single document
        Input document = InputFactory.read(testDocument);
        // Get Result including information about the whole validation
        Result report = validator.checkInput(document);
        System.out.println("Is processing succesful=" + report.isProcessingSuccessful());
        // Get report document if processing was successful
        Document result = null;
        if (report.isProcessingSuccessful()) {
            result = report.getReportDocument();
        }
        // continue processing results...
    }

    public static void main(String[] args) throws Exception {
        // Path of document for validation
        Path testDoc = Paths.get(args[0]);
        StandardExample example = new StandardExample();
        // run example validation
        example.run(testDoc);

    }
}

The Result interface has convenience methods to retrieve details about XSD validation errors and Schematron messages and other processing results. See Result.java for details.

Initializing all XML artifacts and XSLT-executables is expensive. The Check instance is threadsafe and keeps all artifacts. Therefore, we recommend the re-use of a Check instance.

Beside the validator's configuration the only input are instances of Input which can be created by various methods of the InputFactory. The InputFactory calculates a hash sum for each Input which is also written to the Report. SHA-256 from the JDK is the default algorithm. It can be changed using other read-methods of InputFactory.

The main interface Check.java allows using a batch interface (processing list of Inputs). However, there is no parallel processing implemented at the moment.

Accept Recommendation and Accept Match

A tri-state object AcceptRecommendation can be retrieved from the Result using getAcceptRecommendation().

The three defined states are:

  1. ACCEPTABLE i.e. the recommendation is to accept input based on the evaluation of the overall validation.
  2. REJECT i.e. the recommendation is to reject input based on the evaluation of the overall validation.
  3. UNDEFINED i.e. the evaluation of the overall validation could not be computed (overall processing is incomplete)

The accept recommendation is based on either:

  1. Schema and Schematron validation results
  2. or on acceptMatch configuration of the scenario (see below)

Accept match in scenario configuration

For your own configuration you can add an acceptMatch element in each scenario. It can contain an XPATH expression over your own defined Report to compute a boolean value. An XPATH expression evaluating to true will lead to an ACCEPTABLE and otherwise to a REJECT recommendation.

This allows to have control over what validation result is to be considered acceptable for your own application context. E.g. you can overrule Schematron validation errors with acceptMatch configuration and consider certain errors as acceptable. Nevertheless you can not overrule schema errors with accept match.

Building scenario configurations with the Builder API

Instead of pre-configured scenario files it is possible to create a validator configuration using a builder API. A valid configuration consists of the following:

  • at least one valid scenario configuration, this includes
    • a valid match configuration to identify/activate this scenario
    • a valid XML schema configuration
    • a valid report transformation configuration
    • valid schematron validation configurations (optional)
    • a valid accept match configuration to compute acceptance information (optional)
  • a valid fallback scenario configuration

A simple configuration looks like this:

import static de.kosit.validationtool.config.ConfigurationBuilder.*;
import de.kosit.validationtool.api.Configuration;
import java.net.URI;
import java.nio.file.Path;

public class MyValidator {

 public static void main(String[] args) {
    Configuration config = Configuration.create().name("myconfiguration")
                          .with(scenario("firstScenario")
                                          .match("//myNode")
                                          .validate(schema("Sample Schema").schemaLocation(URI.create("simple.xsd")))
                                          .validate(schematron("my rules").source("myRules.xsl"))
                                          .with(report("my report").source("report.xsl")))
                          .with(fallback().name("default-report").source("fallback.xsl"))
                          .useRepository(Paths.get("/opt/myrepository"))
                          .build();
    Check validator = new DefaultCheck(config);
    // .. run your checks
 }
}

The build API provides various methods to configure your scenarios and the validation process.

It is also possible to provide runtime artifacts like XsltExecutable, XPathExecutable or Schema to configure the validator. This gives you complete control over loading these artifacts.


Note: Creating these objects requires usage of the same instance of the saxon Processor as used during validation later. Therefore, you need to supply a custom ResolvingConfigurationStrategy or use the internal one to create these objects. See below.


Configure XML Security and Resolving

When using XML related technologies you are supposed to handle certain security issues properly. The KoSIT validator pursues a rather strict strategy. The default configuration:

  • disables DTD validation completely
  • allows loading/resolving only from a configured local content repository (a specific folder)
  • tries to prevent known XML security issues (see OWASP XML_Security_Cheat_Sheet.html)

However, you can configure certain aspects related to resolving and security yourself. The validator uses a single interface for accessing or creating the necessary XML API objects like SchemaFactory, Validator,URIResolver or Processor: ResolvingConfigurationStrategy.java

There are 3 implementations available out of the box:

  1. StrictRelativeResolvingStrategy.java which is the default, prevents known XML attacks and only allows loading from a specific local repository location
  2. StrictLocalResolvingStrategy.java which opens the first strategy to load resources from local locations
  3. RemoteResolvingStrategy.java which further opens the second to load resources also from remote locations via http and https

You can configure usage of one of these implementations using the ResolvingMode via

Conifuguration config = Configuration.load(URI.create("myscenarios.xml"))
                            .resolvingMode(ResolvingMode.STRICT_LOCAL)
                            .build();

If you decide to implement your own strategy, you can configure this via:

Conifuguration config = Configuration.load(URI.create("myscenarios.xml"))
                            .resolvingStrategy(new MyCustomResolvingConfigurationStrategy())
                            .build();

⚠️ Attention: If you decide to implement a custom strategy you need to handle XML security risks on your own. Please make sure, that you prevent XXE and other kind of attacks. Consider using BaseResolvingStrategy.java and the protected methods within to disable certain features.