[contents]

Abstract

This document provides guidance for developers on implementing Evaluation and Report Language (EARL) 1.0 in software tools and process. EARL is a vocabulary, the terms of which are defined across a set of specifications and technical notes, that is used to describe test results. The primary motivation for developing this vocabulary is to facilitate the exchange of test results between web accessibility evaluation tools in a vendor-neutral and platform-independent format. It also provides reusable terms for generic quality assurance and validation purposes.

While this document provides developer guidance for using and implmenting EARL, Evaluation and Report Language (EARL) 1.0 Schema defines the core terms of the vocabulary, and other specifications provide additional terms for representing HTTP exchanges between clients and servers, HTTP Vocabulary in RDF 1.0, for representing web content itself, Representing Content in RDF 1.0, or for specifying particular locations within or sections of content, Pointer methods in RDF 1.0. An Evaluation and Report Language (EARL) Overview is also available.

1. Introduction

This document is a guide to the Evaluation and Report Language (EARL) 1.0 for developers of software tools and proccesses. It provides an introduction to EARL and its uses, defines conformance requirements for tools supporting EARL, and describes approaches for serializing EARL data in different formats.

EARL is a vocabulary, the terms of which are defined across a set of specifications and technical notes, that is used to describe test results in a machine-readable format. This set of specifications includes:

• Evaluation and Report Language (EARL) 1.0 Schema
• HTTP Vocabulary in RDF 1.0
• Representing Content in RDF 1.0
• Pointer Methods in RDF 1.0

An Evaluation and Report Language (EARL) Overview is also available.

1.1. Audience of this Document

The assumed audience of this document is developers of software tools and processes, who want to implement EARL. This includes developers of quality assurance tools, in particular developers of web accessibility evaluation tools, web authoring tools, and web quality assurance tools.

This document assumes that the reader is familiar with the Resource Description Framework [RDF] and can read its serializations. Readers who wish to learn more about RDF should read a general introduction or the RDF Primer [RDF-PRIMER]. This document is also written with consideration for developers who are more accustomed to XML than RDF, but reader is cautioned about notable differences between the syntax-based nature of XML and the semantic-based nature of RDF.

1.2. Document conventions

Keywords

The keywords must, required, recommended, should, may, and optional in this document are used in accordance with RFC 2119 [RFC 2119].

Namespaces

The RDF representation of the vocabulary defined by this document uses the namespace https://www.w3.org/ns/earl#. The prefix earl is used throughout this document to denote this namespace. Other prefixes used throughout this document include:

• cnt - Representing Content in RDF namespace https://www.w3.org/2011/content# (defined by [Content])
• dc - Dublin Core (DC) namespace https://purl.org/dc/terms/ (defined by [DC])
• doap - Description of a Project (DOAP) namespace https://usefulinc.com/ns/doap# (defined by [DOAP])
• foaf - Friend of a Friend (FOAF) namespace https://xmlns.com/foaf/0.1/# (defined by [FOAF])
• ht - HTTP Vocabulary in RDF namespace https://www.w3.org/2011/http# (defined by [HTTP])
• ptr - Pointer Methods in RDF namespace https://www.w3.org/2009/pointers# (defined by [Pointers])
• rdf - RDF namespace https://www.w3.org/1999/02/22-rdf-syntax-ns# (defined by [RDF])
• rdfs - RDF Schema namespace https://www.w3.org/2000/01/rdf-schema# (defined by [RDFS])
• xsd - XMLS namespace https://www.w3.org/2001/XMLSchema# (defined by [XMLS])

2. Overview

Figure 1. Stages in the testing processes.

The Evaluation and Report Language (EARL) is a vocabulary to describe test results in a machine-readable format. It supports the test reporting stage in the testing process, as described by different standards on testing, such as in [IEEE 829]. Some of the typical stages in the testing process include:

1. Test Plan, which prescribes the scope, approach, resources, and schedule of the testing activities.
2. Test Specification, which describes the test cases and the test procedures for carrying out individual tests.
3. Test Execution, which covers the actual execution of the tests according to the test plan and specification.
4. Test Reporting, which deals not only with the creation of test results (i.e. reports), but may include post-processing (e.g. filtering, aggregation, summarization, etc.).

EARL vocabulary can be used to describe resources to be tested as defined by the test plan, test cases and criteria as defined by the test specification, and outcomes of the test execution stages. But more importantly, EARL provides a format to uniformly record these and other elements in semantically rich testing reports. EARL does not provide vocabulary specifically for describing test cases and the logic of their execution, this would be part of a test case description language that could complement EARL.

2.1 Structure of EARL

[Editor note: currently this section overlaps with section "2. Classes" of EARL 1.0 Schema; this will be revised at a later stage.]

The terms of EARL are defined using the Resource Description Framework [RDF], which is technology to express semantic data in a machine-readable format. Like any RDF vocabulary, EARL is a collection of statements about resources, each with a subject, a predicate (or verb), and an object. RDF statements describe resources and relationships, such as in the following example:

<#Assertion> <#assertedBy> <#Assertor> .
<#TestResult> <#outcome> <#OutcomeValue> .

EARL provides a standardized vocabulary to describe specific resources and relationships that are relevant to test reporting. For instance, it provides vocabulary to describe Assertor, TestResult, and OutcomeValue in the example above.

The core construct of EARL is an Assertion, which describes the context and outcome of an individual test execution. It contains the following information:

Assertor

This can include information about who or what ran the test. For example human evaluators, automated accessibility checkers, or combinations of these.

Test Subject

This can include web content (such as web pages, videos, applets, etc.), software (such as authoring tools, user agents, etc.), or other things being tested.

Test Criterion

What are we evaluating the test subject against? This could be a specification, a set of guidelines, a test from a test suite, or some other testable statement.

Test Result

What was the outcome of the test? A test result could also include contextual information such as error messages or relevant locations within the test subject.

Examples of Assertions

2.2. Uses of EARL

As a standardized machine-readable format, EARL facilitates processing of test results, such as those generated by automated or semi-automated web accessibility evaluation tools. Web authoring tools and quality assurance software can aggregate such test results to support website developers in developing high quality web content. EARL has been specifically designed to support a broad variety of uses cases, including the following:

Combining results from software tools

Quality assurance testing, such as web accessibility evaluation, is often carried out by combinations of software tools and human evaluators. For instance, different evaluators may be testing different parts of a website, and single a evaluator may be using one or more software tools for testing or recording test results. Some tests might be fully automatable, and may be executed without any human intervention. Partial reports from different software tools can be combined, by using EARL as the standardized format for expressing test results.

Exchanging data between software tools

A standardized format for expressing test results also allows software tools to exchange data. In particular, testing tools such as checkers and validators can be more easily integrated into authoring tools, such as content management systems. Testing tools can also be integrated into quality assurance tools that help process and analyze the test results, or that provide customized reports of the test results for different audiences.

Querying and analyzing test reports

EARL provides fine-grained data about the context and outcome of test results, including information about the tested resources and the testing modalities, to allow many different types of queries and analysis. For instance, queries can be used to generate customized reports for managers who want a higher-level view, project managers who want information specific to the resources they are managing, and developers who want detailed bug reports about errors they need to fix. The nature of RDF also allows semantic inference and other approaches for advanced data mining.

Benchmarking software testing tools

EARL can also be used to compare the results provided by different testing tools, such as web accessibility evaluation tools. In particular, it can be used to compare the results provided from executing test suites that have normalized outcome, and so benchmark deviations such as false positives and false negative generated by different testing tools.

Evaluating dynamic and multilingual websites

EARL includes vocabulary to describe comprehensively web resources, including any parts of the entire HTTP exchange between a client and a server. This is particularly useful to record HTTP headers relevant for language and content negotiation, as well as the actual content received from the server and that has been tested. Moreover, user interaction with a website can be recorded, to help describe the particular context of the test execution.

Annotating web resources with metadata

Test results can also be used to describe the availability or lack of particular features of the resources tested. As RDF data, EARL test results are particularly useful as metadata for describing features of web content in a machine-readable format. For instance, EARL reports could be associated with web resources using RDFa [provide reference], and can be processed by RDF-aware browsers and search-engines to serve particular user preferences.

2.3. Limitations of EARL

EARL describes the context and outcomes of test executions. It does not describe test cases and the logic of their execution, or specifically support stages of the testing process other than test reporting. However, EARL provides a semantically rich vocublary the terms of which can be used in other stages too.

EARL does not specifically address application security and privacy. EARL can contain data with different levels of sensistivity and it is important to consider potential security and privacy issues when using EARL. For instance, test results expressed in EARL could contain sensitive information such as the internal directory structure of a web server, username and password information, parts of restricted web pages, or testing modalities. Security and privacy considerations for these situations need to be made at the application level. For example, certain parts of the data may be restricted to appropriate user permissions, encrypted, or obfuscated.

3. Core Vocabulary

EARL provides a core set of terms to establish a common vocabulary for describing and exchanging test results. These are primarily terms to describe the assertor, test subject, test criterion, and test result as discussed in section 2.1. Structure of EARL.

As EARL is an RDF vocabulary it can reuse existing RDF terms from other vocabularies. It reuses several terms from different vocabularies that are considered to be well-established and widely supported by the community. In particular, it makes use of these vocabularies:

• Dublin Core (DC) - Typically used to described electronic and physical resources (see [DC]). It provides many flexible terms that are used in EARL as resource descriptors and to describe some relationships between resources.
• Friend of a Friend (FOAF) - Typically used to describe people, groups, and their relationships, but also agents, documents, and other entities (see [FOAF]). EARL reuses some of the terms to primairly describe assertors.
• Description of a Project (DOAP) - Typically used to describe software development projects and their releases (see [DOAP]). EARL reuses some of the terms to describe software instances as assertors or test subjects.

EARL terms can also be reused in other RDF vocabularies. More importantly, the terms in EARL can be refined to support application-specific aspects by using the reification advantages that RDF provides. For instance, an organization may have quality management tools with detailed description of the people and roles involved in the testing process, such as tester, developer, quality assurance manager, and so on. For EARL tools it is sufficient to "know" that, for example, a tester is a type of assertor for it to function in that environment.

More discussion on the reuse of RDF vocabularies and refinements of EARL terms is provided in the description of the individual EARL classes in the following sections.

3.1. Assertion

The assertion is the heart-piece of an EARL report. An EARL report consists of at least one complete assertion but usually many more, each describing an individual test result. An assertion is merely the particular combination of an assertor, test subject, test criterion, test result, and optionally test mode instances.

Note: EARL reports coming from a single source may often have one assertor instance, such as a web accessibility evaluation tool that is generating the results, that is referenced from each assertion. Sometimes the test subject, such as a web page, may also not change frequently within a report from a single source.

Reference

Properties

• Domain of:
  • earl:assertedBy
  • earl:subject
  • earl:test
  • earl:result
  • earl:mode
• Range of: none

Examples

[Editor note: this section will be updated to provide examples that complement the ones in the Schema document.]

<earl:Assertion rdf:about="#assertion">
  <earl:assertedBy rdf:resource="#assertor"/>
  <earl:subject rdf:resource="https://www.example.org/"/>
  <earl:test rdf:resource="https://www.w3.org/TR/WCAG20-TECHS/H36"/>
  <earl:result rdf:resource="#result"/>
</earl:Assertion>

3.2. Assertor Class

Assertor - an entity such as a person, a software tool, an organization, or any other grouping that carries out a test collectively. From Assertor Class

The assertor is the entity claiming a result. In many cases this could be a software tool such as a web accessibility evaluation tool, but it can also be a person such as an accessibility tester or an entire organization such as a consultancy. EARL also provides support to describe different combinations of people and software tools working together. For instance, one can describe a group of testers that collectively claim a result or an individual tester using a set of software tools.

EARL does not prescribe the level of detail and the extent to which an assertor is described. However, it is generally good practice to keep detailed recordings of who or what made specific claims at the raw data level and filter that as needed during the aggregation of the results. For instance, a consultancy may want to keep detailed internal records of the individual test results (for instance for potential conflict resolutions), but not disclose this externally to their clients.

Therefore, it is recommended to use one of the following types, or refinement of these types, when describing an assertor:

earl:Software

Software - the assertor is a piece of software.

foaf:Agent

Agent - the assertor is an agent, as defined by [FOAF].

foaf:Person

Person - the assertor is a person, as defined by [FOAF].

foaf:Organization

Organization - the assertor is an organization, as defined by [FOAF].

foaf:Group

Group - the assertor is a group of agents, as defined by [FOAF].

Examples

<earl:Assertor rdf:about="joe">
  <dc:title>Joe Doe</dc:title>
</earl:Assertor>

<earl:Software rdf:about="https://www.example.org/tools/#cooltool">
  <doap:name xml:lang="en">Cool Tool</doap:name>
  <doap:description xml:lang="en">My favorite tool!</doap:description>
  <doap:created>2011-04-27</doap:created>
  <doap:homepage rdf:resource="https://example.org/tools/cool/"/>
  <doap:release>
    <doap:revision>1.0.3</doap:revision>
  </doap:release>
</earl:Software>

<foaf:Group rdf:about="#assertor">
  <dct:title xml:lang="en">Bob using Cool Tool</dct:title>
  <dct:description xml:lang="en">Bob doing semi-automated testing</dct:description>
  <earl:mainAssertor rdf:resource="https://www.example.org/people/#bob"/>
  <foaf:member rdf:resource="https://www.example.org/tool/#cooltool"/>
</foaf:Group>

2.3. TestSubject Class

Test Subject - the class of things that have been tested against some test criterion.

Related Classes

Rather than specifying only an earl:TestSubject type, it is recommended that one of the following types be employed in addition:

earl:Software

Software - the test subject is a piece of software being tested.

cnt:Content

Content - the test subject is a representation of the content as defined by [Content].

http:Response

HTTP Response - the test subject is the response from an HTTP server as defined by [HTTP].

foaf:Document

Document - the test subject is a document, such as electronic file, as defined by [FOAF].

[Editor's note 1: ERT WG is considering to remove foaf:Document unless compelling use-cases can be presented; feedback on this consideration is welcome.]

Related Properties

• Domain of: none
• Range of:
  • earl:subject

It is recommended to provide additional information about the Test Subject by using the following properties from external vocabularies:

dct:title

Human readable title for the subject.

dct:description

Human readable descriptions of the subject.

dct:date

Date on which the subject was created or identified.

dct:hasPart

Reference to another subject that are part of this subject.

dct:isPartOf

Reference to another subject of which this subject is a part of.

Examples

<earl:TestSubject rdf:about="https://www.example.org/">
  <dct:title xml:lang="en">example.org Web site</dct:title> 
  <dct:description xml:lang="en">Each page on the example.org Web site</dct:description> 
  <dct:hasPart rdf:resource="https://www.example.org/style.css"/> 
  <dct:hasPart rdf:resource="https://www.example.org/page1.html"/> 
  <dct:hasPart rdf:resource="https://www.example.org/page2.html"/> 
  <dct:hasPart rdf:resource="https://www.example.org/image1.png"/> 
  <dct:hasPart rdf:resource="https://www.example.org/image2.png"/> 
</earl:TestSubject>

2.4. TestCriterion Class

Test Criterion - a testable statement, usually one that can be passed or failed. It is a super class for all types of tests including things such as validation requirements, code test cases, checkpoints from guidelines such as Web Content Accessibility Guidelines [WCAG], or others.

Related Classes

Rather than specifying only an earl:TestCriterion type, it is recommended that one of the following types be employed in addition:

earl:TestRequirement

Test Requirement - a higher-level requirement that is tested by executing one or more sub-tests. For example WCAG 2.0 Success Criterion 1.1.1, which is evaluated using several Techniques for Success Criterion 1.1.1 and combining the results.

earl:TestCase

Test Case - an atomic test, usually one that is a partial test for a requirement. For example, Technique H36: Using alt attributes on images used as submit buttons provides a partial test for WCAG 2.0 Success Criterion 1.1.1.

Related Properties

• Domain of: none
• Range of:
  • earl:test

It is recommended to provide additional information about the Test Subject by using the following properties from external vocabularies:

dct:title

Human readable title for the test criterion.

dct:description

Human readable description of the test criterion.

dct:hasPart

Relationship to other test criteria that are part of this criterion.

dct:isPartOf

Relationship to other test criteria of which this criterion is a part of.

Examples

<earl:TestCase rdf:about="https://www.w3.org/TR/WCAG20-TECHS/H36">
  <dct:title xml:lang="en">H36</dct:title>
  <dct:description xml:lang="en">Technique H36 - Using alt attributes 
    on images used as submit buttons </dct:description>
  <dct:isPartOf rdf:resource="https://www.w3.org/TR/WCAG20-TECHS/"/>
  <dct:hasPart rdf:resource="https://www.w3.org/TR/WCAG20-TECHS/H36#H36-tests"/>
</earl:TestCase>

2.5. TestResult Class

Test Result - the actual result of performing the test. It includes both machine-readable values as well as human-readable description of the results (typically error messages).

Related Properties

• Domain of:
  • earl:info
  • earl:outcome
  • earl:pointer
• Range of:
  • earl:result

It is recommended to provide additional information about the Test Result by using the following properties from external vocabularies:

dct:title

Human readable title for the result.

dct:description

Human readable description of the result.

dct:date

Date on which the result was obtained (typically when the subject was tested).

Examples

<earl:TestResult rdf:about="#result">
  <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed"/>
  <dct:title xml:lang="en">Invalid Markup (code #353)</dct:title>
  <dct:description rdf:parseType="Literal" xml:lang="en">
    <div xmlns="https://www.w3.org/1999/xhtml">
      <p>The <code>table</code> element is not allowed to appear
        inside a <code>p</code> element</p>
    </div>
  </dct:description>
  <earl:pointer rdf:resource="#pointer"/>
  <earl:info rdf:parseType="Literal" xml:lang="en">
    <div xmlns="https://www.w3.org/1999/xhtml">
      <p>It seems the <code>p</code> element has not been closed</p>
    </div>
  </earl:info>
</earl:TestResult>

2.6. TestMode Class

Test Mode - describes how a test was carried out. It reflects the information provided by the Assertor and is used to simplify some commonly used queries.

Related Instances

Where applicable it is recommended to use one of the following instances of earl:TestMode, to categorize the mode in which the test was carried out:

earl:automatic

Automatic - where the test was carried out automatically by the software tool and without any human intervention.

earl:manual

Manual - where the test was carried out by human evaluators. This includes the case where the evaluators are aided by instructions or guidance provided by software tools, but where the evaluators carried out the actual test procedure.

earl:semiAuto

Semi-Automatic - where the test was partially carried out by software tools, but where human input or judgment was still required to decide or help decide the outcome of the test.

earl:undisclosed

Undisclosed - where the exact testing process is undisclosed.

earl:unknownMode

Unknown - where the testing process is unknown or undetermined.

Related Properties

• Domain of: none
• Range of:
  • earl:mode

It is recommended to provide additional information about the Test Mode by using the following properties from external vocabularies:

dct:title

Human readable title for the test mode.

dct:description

Human readable description of the test mode.

Examples

<earl:Assertion rdf:about="#assertion">
  <earl:mode rdf:resource="https://www.w3.org/ns/earl#semiAuto"/> 
</earl:Assertion>

2.7. OutcomeValue Class

Outcome Value - a value or expression that describes a resulting condition from carrying out the test.

Related Instances

Where applicable it is recommended to use one of the following instances of earl:OutcomeValue, to categorize the outcome of carrying out the test:

earl:passed

Passed - the subject passed the test.

earl:failed

Failed - the subject failed the test.

earl:cantTell

Cannot tell - it is unclear if the subject passed or failed the test.

earl:inapplicable

Inapplicable - the test is not applicable to the subject.

earl:untested

Untested - the test has not been carried out.

Related Classes

In cases where it is necessary to create further instances of earl:OutcomeValue, it is recommended that one of the following types be employed in addition:

earl:Pass

Pass - the class of outcomes to denote passing a test. Subclasses may include ordinal, nominal, or continuous values or expressions.

earl:Fail

Fail - the class of outcomes to denote failing a test. Subclasses may include ordinal, nominal, or continuous values or expressions.

earl:CannotTell

Undetermined - the class of outcomes to denote an undetermined outcome. Usually this happens when an automated test requires human judgement to make a definite decision.

earl:NotApplicable

Not applicable - the class of outcomes to denote the test is not applicable. This could be due to a mismatch between the test and the subject or for any other reason.

earl:NotTested

Not tested - the class of outcomes to denote the test has not been carried out. This is useful for reporting as well as for other uses of progress monitoring.

Related Properties

• Domain of: none
• Range of:
  • earl:outcome

It is recommended to provide additional information about the Outcome Value by using the following properties from external vocabularies:

dct:title

Human readable title for the outcome value.

dct:description

Human readable description of the outcome value.

Examples

<earl:TestResult rdf:about="#result">
  <earl:outcome rdf:resource="https://www.w3.org/ns/earl#passed"/>
</earl:TestResult>

<rdf:Description rdf:about="https://example.org/my/warning#warning">
  <rdf:type rdf:resource="https://www.w3.org/ns/earl#Pass"/>
  <dc:title xml:lang="en">Warning</dc:title>
  <dc:description xml:lang="en">the subject passed the test but there are warnings</dc:description>
</rdf:Description>
<earl:TestResult rdf:about="#result">
  <earl:outcome rdf:resource="https://example.org/my/terms#warning"/>
</earl:TestResult>

2.8. Software Class

[Editor's note 2: ERT WG is looking for feedback on the use of DOAP Project to describe Software; feedback on this issue is welcome.]

A Software is any piece of software such as an authoring tool, browser, or evaluation tool. It can be used to describe an Assertor, such as a validation or other quality assurance tool, and it can be used to describe a Test Subject (for example to test compliance of an authoring tool to Authoring Tool Accessibility Guidelines [ATAG] or of a browser to User Agent Accessibility Guidelines [UAAG]).

Note: earl:Software is a sublass of doap:Project to denote the narrower meaning of executable "Software", that is an outcome of a "Project".

Related Properties

It is recommended to provide information about the Software by using the following properties from external vocabularies:

doap:name

Human readable name of the software.

doap:description

Human readable description of the software.

doap:homepage

Homepage for the software.

doap:created

Date when the software was created or released, in YYYY-MM-DD form. e.g. 2004-04-05.

doap:release

Version information about the software release.

Examples

<earl:Software rdf:about="#cooltool">
  <doap:name xml:lang="en">Cool Tool</doap:name>
  <doap:description xml:lang="en">My favorite tool!</doap:description>
  <doap:created>2011-04-27</doap:created>
  <doap:homepage rdf:resource="https://example.org/tools/cool/"/>
  <doap:release>
    <doap:revision>1.0.3</doap:revision>
  </doap:release>
</earl:Software>

3. Using the Evaluation and Report Language (EARL)

[Editor note: this entire section will be revised and refined in later iterations.

EARL is not a standalone vocabulary and builds on top of many existing vocabularies that cover some of its needs for metadata definition. This approach avoids the re-creation of applications already established and tested like the Dublin Core elements. The referenced specifications are:

• Dublin Core Metadata Initiative (DCMI) Terms. Dublin Core is a metadata standard for describing digital resources, often expressed in XML. The aforementioned document is an up-to-date specification of all metadata terms maintained by the Dublin Core Metadata Initiative. Included are the fifteen terms of the Dublin Core Metadata Element Set, which have also been published as IETF RFC 5013 [RFC5013], ANSI/NISO Standard Z39.85-2007 [NISOZ3985] and ISO Standard 15836 [ISO15836]. RDF Schema versions of the DCMI term declarations are available at [DCMISCHEMAS].
• Friend of a Friend (FOAF) project. The FOAF project is about creating a Web of machine-readable resources describing people, the links between them, and the things they create and do [FOAF].
• Representing Content in RDF [Content-RDF]. This is an RDF vocabulary to semantically represent any type of content, either on the Web or in any storage media.
• HTTP vocabulary in RDF [HTTP-RDF]. This is an RDF vocabulary used to represent HTTP requests and responses. It is useful to identify online resources accessed via HTTP(S), which cannot be uniquely resolved via a URI [URI]. Typical examples are Web servers accessed via content negotiation, Web applications using the POST method, etc.
• Pointer Methods in RDF [Pointers-RDF]. This is an RDF vocabulary to enable pointing in an accurate way to certain parts within a document, particularly HTML and XML documents.

These vocabularies are referenced via namespaces in the corresponding RDF serialization. The list of the normative namespaces can be found in the EARL 1.0 Schema. RDF can be serialized in many equivalent ways, but its XML presentation [RDF/XML] is the preferred method and will be used throughout this document.

3.1 Basic report components

In the following sections, we will construct an EARL report with several examples of each component of the report. The root element of any EARL report is an RDF node, in which we declare the namespaces used to define additional classes and/or properties.

  <rdf:RDF
         xmlns:earl="https://www.w3.org/ns/earl#"
         xmlns:rdf="https://www.w3.org/1999/02/22-rdf-syntax-ns#"
         xmlns:rdfs="https://www.w3.org/2000/01/rdf-schema#">
    <!-- ... -->
</rdf:RDF>

Next, let us assume that we want to express the results of an XHTML validation in a given document with the W3C HTML Validator. The tested document can be found in the fictitious URL https://example.org/resource/index.html and has the following HTML code:

  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html lang="en" xmlns="https://www.w3.org/1999/xhtml" xml:lang="en">
  <head>
  <title>Example of project pages</title>
  </head>
  <body>
  <h1>Project description</h1>
  <h2>My project name</h2>
    <!-- ... -->
  </body>
</html>

This document has three errors that will constitute the basis of our EARL report:

1. Error: Line 14, column 7: document type does not allow element "li" here; missing one of "ul", "ol" start-tag.
2. Error: Line 15, column 6: end tag for "li" omitted, but OMITTAG NO was specified.
3. Error: Line 16, column 9: there is no attribute "alt".

The first step is to define who performed the test, either a human being or a software tool. This is noted in the EARL framework as an Assertor. Let us consider different use cases. First, let us assume that only the W3C HTML Validator performed the test. This could be expressed as an Assertor:

    <earl:Assertor rdf:about="https://validator.w3.org/about.html#">
    <dct:title xml:lang="en">W3C HTML Validator</dct:title>
    <dct:description xml:lang="en">
    W3C Markup Validation Service, a free service that checks Web documents in formats like HTML and XHTML for conformance to W3C Recommendations and other standards.
    </dct:description>
  </earl:Assertor>
  

Notice that the Assertor class provides a mechanism by which to specify more information by leveraging standard Dublin Core properties like dct:title and dct:description. This is not the only possible serialization of this report. An alternative, expressed in N3, could be:

  @prefix earl:    <https://www.w3.org/ns/earl#> .
@prefix dct:     <https://purl.org/dc/terms/> .
<https://validator.w3.org/about.html#>
      a       earl:Assertor ;
      dct:description """W3C Markup Validation Service, a free service that checks Web documents in formats like HTML and XHTML for conformance to W3C Recommendations and other standards."""@en ;
      dct:title "W3C HTML Validator"@en .
      

An Assertor is a generic type. EARL allows the use of certain FOAF classes like Agent, Organisation, or Person to provide more semantic information on the type of assertor. Additionally, EARL defines the Software class to declare tool assertors. Thus, our W3C Validator could be described more adequately in the following way:

      <earl:Software rdf:about="https://validator.w3.org/about.html#">
    <dct:title xml:lang="en">W3C HTML Validator</dct:title>
    <dct:hasVersion>0.7.1</dct:hasVersion>
    <dct:description xml:lang="en">
    W3C Markup Validation Service, a free service that checks web documents in formats like HTML and XHTML for conformance to W3C Recommendations and other standards.
    </dct:description>
  </earl:Software>
  

Notice the aditional property, dct:hasVersion, indicating the version of the software. Let us consider now the case where the assertor is a person. This can be expressed as in the following example:

  <foaf:Person rdf:ID="john">
    <foaf:mbox rdf:resource="mailto:john@example.org"/>
    <foaf:name>John Doe</foaf:name>
  </foaf:Person>
  

We could combine assertors as well. The typical example could be an expert evaluator and a software tool, which perform the analysis. This set of assertors can be expressed under the umbrella of a foaf:Group. We should define who is the main assertor within a foaf:Group through the mainAssertor property (notice in the example how the person is defined as a blank node):

  <foaf:Group rdf:ID="assertor01">
    <dct:title>John Doe and the W3C HTML Validator</dct:title>
    <earl:mainAssertor rdf:resource="https://validator.w3.org/about.html#"/>
    <foaf:member>
      <foaf:Person>
        <foaf:mbox rdf:resource="mailto:john@example.org"/>
        <foaf:name>John Doe</foaf:name>
      </foaf:Person>
    </foaf:member>
  </foaf:Group>
  

The second step is to define what was analyzed, the tested resource. For that, EARL defines the TestSubject class. This class is a generic wrapper for things to be tested like Web resources (cnt:Content) or software (earl:Software). In this case, the Example 3.2 could be represented as:

  <rdf:Description rdf:about="https://example.org/resource/index.html">
    <dct:title xml:lang="en">Project Description</dct:title>
    <dct:date rdf:datatype="https://www.w3.org/2001/XMLSchema#date">2006-02-14</dct:date>
    <rdf:type rdf:resource="https://www.w3.org/ns/earl#TestSubject"/>
  </rdf:Description>
  

Using the Representing Content in RDF vocabulary (via the cnt:ContentAsText class), we could insert the content of the test XHTML file into the report:

  <cnt:ContentAsText rdf:about="https://example.org/resource/index.html">
    <dct:title xml:lang="en">Project Description</dct:title>
    <dct:date rdf:datatype="https://www.w3.org/2001/XMLSchema#date">2006-02-14</dct:date>
    <cnt:characterEncoding>UTF-8</cnt:characterEncoding>
    <cnt:chars>&lt;?xml version="1.0" encoding="UTF-8"?&gt;
&lt;!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"&gt;
&lt;html lang="en" xmlns="https://www.w3.org/1999/xhtml" xml:lang="en"&gt;
  &lt;head&gt;
  &lt;title&gt;Example of project pages&lt;/title&gt;
  &lt;/head&gt;
  &lt;body&gt;
  &lt;h1&gt;Project description&lt;/h1&gt;
  &lt;h2&gt;My project name&lt;/h2&gt;
  &lt;p&gt;The strategic goal of this project is to make you understand EARL.&lt;/p&gt;
  &lt;ul&gt;
    &lt;li&gt;Here comes objective 1.
    &lt;li&gt;Here comes objective 2.&lt;/li&gt;
  &lt;/ul&gt;
  &lt;p alt="what?"&gt;And goodbye ...&lt;/p&gt;
  &lt;/body&gt;
&lt;/html&gt;
</cnt:chars>
  </cnt:ContentAsText>
  

The third step is to define the criterion used for testing the resource. EARL defines test criteria under the umbrella of the TestCriterion class. This class has two subclasses, TestRequirement and TestCase, depending on whether the criterion is a high level requirement, composed of many tests, or an atomic test case. In our example, we are testing validity against the XHTML 1.0 Strict specification, which could be expressed in the following way via the TestRequirement class:

  <earl:TestRequirement rdf:about="https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
    <dct:title xml:lang="en">XHTML 1.0 Strict Document Type Definition</dct:title>
    <dct:description xml:lang="en">DTD for XHTML 1.0 Strict.</dct:description>
  </earl:TestRequirement>
  

The fourth step is to specify the results of the test. There were three errors discovered by the W3C Validator that need to be presented as TestResults. In this case, we present only the errors, but it is also possible to present positive results. In the example below, we present the message errors as text messages within XHTML snippets. We will see later how to improve the machine-readability of such results.

  <earl:TestResult rdf:ID="error1">
    <dct:description rdf:parseType="Literal">
      <div xml:lang="en" xmlns="https://www.w3.org/1999/xhtml">
        <p>Error - Line 14 column 7: document type does not allow element
           <code>li</code>here; missing one of
           <code>ul</code>, <code>ol</code> start-tag.</p>
      </div>
    </dct:description>
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
  </earl:TestResult>
  <earl:TestResult rdf:ID="error2">
    <dct:description rdf:parseType="Literal">
      <div xml:lang="en" xmlns="https://www.w3.org/1999/xhtml">
        <p>Error - Line 15 column 6: end tag for
           <code>li</code> omitted, but OMITTAG NO was specified.</p>
      </div>
    </dct:description>
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
  </earl:TestResult>
  <earl:TestResult rdf:ID="error3">
    <dct:description rdf:parseType="Literal">
      <div xml:lang="en" xmlns="https://www.w3.org/1999/xhtml">
        <p>Error - Line 16 column 9: there is no attribute
           <code>alt</code>.</p>
      </div>
    </dct:description>
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
  </earl:TestResult>
  

3.2 Putting the pieces together

The final step is to merge together the created components. The EARL statements for this purpose are called Assertions, and have four key properties: earl:assertedBy, earl:subject, earl:test and earl:result. Each of them serves to point to the corresponding assertors, test subjects, test requirements, and results, respectively. From our previous examples, we could build our first complete report with our three assertions:

  <earl:Assertion rdf:ID="ass1">
        <earl:result rdf:resource="#error1" />
        <earl:test rdf:resource="https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" />
        <earl:subject rdf:resource="https://example.org/resource/index.html" />
        <earl:assertedBy rdf:resource="#assertor01" />
    </earl:Assertion>
    <earl:Assertion rdf:ID="ass2">
        <earl:result rdf:resource="#error2" />
        <earl:test rdf:resource="https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" />
        <earl:subject rdf:resource="https://example.org/resource/index.html" />
        <earl:assertedBy rdf:resource="#assertor01" />
    </earl:Assertion>
    <earl:Assertion rdf:ID="ass3">
        <earl:result rdf:resource="#error3" />
        <earl:test rdf:resource="https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" />
        <earl:subject rdf:resource="https://example.org/resource/index.html" />
        <earl:assertedBy rdf:resource="#assertor01" />
    </earl:Assertion>
    

3.3 An accessibility example

Our next example presents the results of an accessibility test in a given Web resource. Let us consider a simple XHTML page, which presents the image of a cat:

    <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html lang="en" xmlns="https://www.w3.org/1999/xhtml" xml:lang="en">
  <head>
  <title>A cat's photography</title>
  </head>
  <body>
  <h1>A cat's photography</h1>
  <p>Image of a cat who likes
    <acronym title="carview.php?tsp=Evaluation and Report Language">EARL</acronym>, although it
    seems quite tired.
    <img src="../images/cat.jpg" alt="Image of a white cat with black spots."/>
  </p>
  </body>
</html>

We have in this case a software tool called "Cool Tool" that performs a test against the Common Failure F65 from the (X)HTML techniques for WCAG 2.0 [WCAG20]. This technique proofs the existence of the alt attribute for given (X)HTML elements like img. The software can be represented as:

<earl:Software rdf:about="https://example.org/cooltool/">
    <dct:title xml:lang="en">Cool Tool accessibility checker</dct:title>
    <dct:hasVersion>1.0.c</dct:hasVersion>
    <dct:description xml:lang="en">A reliable compliance checker for Web Accessibility</dct:description>
  </earl:Software>
  

The test requirement can be represented as:

  <earl:TestCase rdf:about="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65">
    <dct:title xml:lang="en">Failure of Success Criterion 1.1.1 from WCAG 2.0</dct:title>
    <dct:description xml:lang="en">Failure due to omitting the alt attribute on img elements, area elements, and input elements of type image.</dct:description>
  </earl:TestCase>
  

We can make the test result more amenable to machine processing by making use of the Pointers [Pointers-RDF] vocabulary. In this case, we identify the line number where the test was compliant:

  <ptr:LineCharPointer rdf:ID="pointer">
    <ptr:lineNumber>15</ptr:lineNumber>
    <ptr:reference rdf:resource="https://example.org/resource/index.html" />
  </ptr:LineCharPointer>
  <earl:TestResult rdf:ID="result">
    <earl:pointer rdf:resource="#pointer" />
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#passed" />
  </earl:TestResult>
  

Which leads to the following assertion:

  <earl:Assertion rdf:ID="assert">
    <earl:result rdf:resource="result" />
    <earl:test rdf:resource="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" />
    <earl:subject rdf:resource="https://example.org/resource/index.html" />
    <earl:assertedBy rdf:resource="https://example.org/cooltool/" />
  </earl:Assertion>
  

3.4 Identifying unambiguously the resource

There are cases where the identification of a resource on the Web requires more than a URL. This occurs typically when the user agent and the server exchange HTTP messages via Content Negotiation to deliver the best possible alternative to the client. A common scenario appears when the user expresses a preference for given languages with a ranking via the Accept-Language header. Under those circumstances, it is necessary to use the HTTP vocabulary in RDF [HTTP-RDF] to identify correctly the TestSubject.

Let us assume that our exemplary Web server can deliver under the URL https://example.org/resource/index.html two versions (English and Spanish) of a given XHTML page. The English version can be seen in Example 3.13. The Spanish version can be seen in the listing below:

  <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
  "https://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
<html lang="es" xmlns="https://www.w3.org/1999/xhtml" xml:lang="es">
    <head>
        <title>Fotografia de un gato</title>
    </head>
    <body>
        <h1>Fotografia de un gato</h1>
        <p>Imagen de un gato al que le gusta
        <acronym title="carview.php?tsp=Evaluation and Report Language" xml:lang="en" lang="en">EARL</acronym>, aunque aparenta estar muy cansado.
        <img src="../images/cat.jpg" />
        </p>
    </body>
</html>

The English resource can be represented as:

<rdf:RDF xmlns:rdf="https://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:earl="https://www.w3.org/ns/earl#"
    xmlns:dct="carview.php?tsp=https://purl.org/dc/terms/"
    xmlns:cnt="https://www.w3.org/2008/content#"
    xmlns:http="https://www.w3.org/2006/http#"
    xml:base="https://www.example.org/resource/content_001#">
    <cnt:ContentAsBase64 rdf:ID="content1">
        <cnt:bytes rdf:datatype="https://www.w3.org/2001/XMLSchema#base64Binary">PD94bWwgdmVyc2lv...</cnt:bytes>
    </cnt:ContentAsBase64>
    <http:Response rdf:ID="response1">
        <http:httpVersion>1.1</http:httpVersion>
        <http:statusCodeNumber>200</http:statusCodeNumber>
        <http:sc rdf:resource="https://www.w3.org/2008/http-statusCodes#200" />
        <http:reasonPhrase>OK</http:reasonPhrase>
        <http:headers rdf:parseType="Collection">
            <http:MessageHeader>
                <http:fieldName>Vary</http:fieldName>
                <http:hdrName rdf:resource="https://www.w3.org/2008/http-headers#vary" />
                <http:fieldValue>Accept-Language</http:fieldValue>
            </http:MessageHeader>
            <!-- ... -->
        </http:headers>
        <http:body rdf:resource="#content1" />
    </http:Response>
    <http:Connection rdf:ID="connection1">
        <http:connectionAuthority>www.example.org:80
        </http:connectionAuthority>
        <http:requests rdf:parseType="Collection">
            <http:Request rdf:resource="#request1" />
        </http:requests>
    </http:Connection>
    <http:Request rdf:ID="request1">
        <http:httpVersion>1.1</http:httpVersion>
        <http:methodName>GET</http:methodName>
        <http:mthd rdf:resource="https://www.w3.org/2008/http-methods#GET" />
        <http:abs_path>/resource/index.html</http:abs_path>
        <http:headers rdf:parseType="Collection">
            <http:MessageHeader>
                <http:fieldName>Accept-Language</http:fieldName>
                <http:hdrName rdf:resource="https://www.w3.org/2008/http-headers#accept-language" />
                <http:fieldValue>en</http:fieldValue>
            </http:MessageHeader>
            <!-- ... -->
        </http:headers>
        <http:resp rdf:resource="#response1" />
    </http:Request>
</rdf:RDF>

The Spanish one could be represented as:

<rdf:RDF xmlns:rdf="https://www.w3.org/1999/02/22-rdf-syntax-ns#"
    xmlns:earl="https://www.w3.org/ns/earl#"
    xmlns:dct="carview.php?tsp=https://purl.org/dc/terms/"
    xmlns:cnt="https://www.w3.org/2008/content#"
    xmlns:http="https://www.w3.org/2006/http#"
    xml:base="https://www.example.org/resource/content_002#">
    <cnt:ContentAsBase64 rdf:ID="content2">
        <cnt:bytes rdf:datatype="https://www.w3.org/2001/XMLSchema#base64Binary"
            >PD94bWwgdmVyc2lvbj...</cnt:bytes>
    </cnt:ContentAsBase64>
    <http:Response rdf:ID="response2">
        <http:httpVersion>1.1</http:httpVersion>
        <http:statusCodeNumber>200</http:statusCodeNumber>
        <http:sc rdf:resource="https://www.w3.org/2008/http-statusCodes#200" />
        <http:reasonPhrase>OK</http:reasonPhrase>
        <http:headers rdf:parseType="Collection">
            <http:MessageHeader>
                <http:fieldName>Vary</http:fieldName>
                <http:hdrName rdf:resource="https://www.w3.org/2008/http-headers#vary" />
                <http:fieldValue>Accept-Language</http:fieldValue>
            </http:MessageHeader>
            <!-- ... -->
        </http:headers>
        <http:body rdf:resource="#content2" />
    </http:Response>
    <http:Connection rdf:ID="connection2">
        <http:connectionAuthority>www.example.org:80</http:connectionAuthority>
        <http:requests rdf:parseType="Collection">
            <http:Request rdf:resource="#request2" />
        </http:requests>
    </http:Connection>
    <http:Request rdf:ID="request2">
        <http:httpVersion>1.1</http:httpVersion>
        <http:methodName>GET</http:methodName>
        <http:mthd rdf:resource="https://www.w3.org/2008/http-methods#GET" />
        <http:abs_path>/resource/index.html</http:abs_path>
        <http:headers rdf:parseType="Collection">
            <http:MessageHeader>
                <http:fieldName>Accept-Language</http:fieldName>
                <http:hdrName rdf:resource="https://www.w3.org/2008/http-headers#accept-language" />
                <http:fieldValue>es</http:fieldValue>
            </http:MessageHeader>
            <!-- ... -->
        </http:headers>
        <http:resp rdf:resource="#response2" />
    </http:Request>
</rdf:RDF>

Strictly speaking, for the representation of the TestSubject, only the http:Response object is needed. However, it is recommended to use the http:Request and http:Connection objects to facilitate the replicability of the results. The replicability of the results is also time-dependent as the resources may change over time. Therefore, timestamps or modification dates in the reports are also recommended.

We are now in the situation to allow our Cool Tool accessibility checker (see Example 3.14) to produce accurate reports on both versions of the page. The evaluation report for the English resource (assuming the same test requirement of Example 3.15) looks like the following snippet:

<rdf:RDF xmlns:rdf="https://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:earl="https://www.w3.org/ns/earl#"
  xmlns:dct="carview.php?tsp=https://purl.org/dc/terms/"
  xmlns:cnt="https://www.w3.org/2008/content#"
  xmlns:ptr="https://www.w3.org/2009/pointers#"
  xml:base="https://www.example.org/earl/report1#">
  <earl:Assertion rdf:ID="assert">
    <earl:result rdf:resource="result" />
    <earl:test rdf:resource="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" />
    <earl:subject rdf:resource="https://www.example.org/resource/content_001#response1" />
    <earl:assertedBy rdf:resource="https://example.org/cooltool/" />
  </earl:Assertion>
  <earl:Software rdf:about="https://example.org/cooltool/">
    <dct:title xml:lang="en">Cool Tool accessibility checker</dct:title>
    <dct:hasVersion>1.0.c</dct:hasVersion>
    <dct:description xml:lang="en">A reliable compliance checker for Web Accessibility</dct:description>
  </earl:Software>
  <earl:TestCase rdf:about="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65">
    <dct:title xml:lang="en">Failure of Success Criterion 1.1.1 from WCAG 2.0</dct:title>
    <dct:description xml:lang="en">Failure due to omitting the alt attribute on img elements, area elements, and input elements of type image.</dct:description>
  </earl:TestCase>
  <ptr:LineCharPointer rdf:ID="pointer">
    <ptr:lineNumber>15</ptr:lineNumber>
    <ptr:charNumber>5</ptr:charNumber>
    <ptr:reference rdf:resource="https://www.example.org/resource/content_001#content1a" />
  </ptr:LineCharPointer>
  <earl:TestResult rdf:ID="result">
    <earl:pointer rdf:resource="#pointer" />
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#passed" />
  </earl:TestResult>
</rdf:RDF>

And the evaluation report for the Spanish resource looks like the following:

  <rdf:RDF xmlns:rdf="https://www.w3.org/1999/02/22-rdf-syntax-ns#"
  xmlns:earl="https://www.w3.org/ns/earl#"
  xmlns:dct="carview.php?tsp=https://purl.org/dc/terms/"
  xmlns:cnt="https://www.w3.org/2008/content#"
  xmlns:ptr="https://www.w3.org/2009/pointers#"
  xml:base="https://www.example.org/earl/report2#">
  <earl:Assertion rdf:ID="assert">
    <earl:result rdf:resource="result" />
    <earl:test rdf:resource="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" />
    <earl:subject rdf:resource="https://www.example.org/resource/content_002#response2" />
    <earl:assertedBy rdf:resource="https://example.org/cooltool/" />
  </earl:Assertion>
  <earl:Software rdf:about="https://example.org/cooltool/">
    <dct:title xml:lang="en">Cool Tool accessibility checker</dct:title>
    <dct:hasVersion>1.0.c</dct:hasVersion>
    <dct:description xml:lang="en">A reliable compliance checker for Web Accessibility</dct:description>
  </earl:Software>
  <earl:TestCase rdf:about="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65">
    <dct:title xml:lang="en">Failure of Success Criterion 1.1.1 from WCAG 2.0</dct:title>
    <dct:description xml:lang="en">Failure due to omitting the alt attribute on img elements, area elements, and input elements of type image.</dct:description>
  </earl:TestCase>
  <ptr:LineCharPointer rdf:ID="pointer">
    <ptr:lineNumber>16</ptr:lineNumber>
    <ptr:charNumber>9</ptr:charNumber>
    <ptr:reference rdf:resource="https://www.example.org/resource/content_002#content2a" />
  </ptr:LineCharPointer>
  <earl:TestResult rdf:ID="result">
    <earl:pointer rdf:resource="#pointer" />
    <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
  </earl:TestResult>
</rdf:RDF>

Notice how both the result and the location of the element analyzed (in this case the <img> element in the page) is different in both reports.

3.5 Advance usage

This section presents some advanced use of the vocabularies. In particular, we will show an example demonstrating the extensibility of the vocabulary (without losing its interoperability) and another example showing how to merge reports from different sources.

3.5.1 Extending the vocabularies

Let us assume a software product (Cool Validator 2.0) that validates XML documents on the Web against given DTDs or XML Schemas. According to the XML specification [XML], there are two types of errors:

Fatal errors

Errors after which the parser must not continue processing. Typically, these are well-formedness problems.

Errors

Violations of the specification. These are normally violations of the validity constraints.

The product defines an additional category, warning, which are errors reported by the underlying SAX parser. These are basically violations not included in the XML specification, and allow the product to continue its normal processing work. With these elements in mind, the following RDF Schema was developed:

    <rdfs:Class rdf:ID="FatalError">
        <rdfs:label rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> Fatal error when processing the XML file (well-formedness)</rdfs:label>
        <owl:versionInfo rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo>
        <rdfs:subClassOf rdf:resource="https://www.w3.org/ns/earl#Fail" />
    </rdfs:Class>
    <rdfs:Class rdf:ID="Error">
        <rdfs:label rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> Error when processing the XML file (validation constraint)</rdfs:label>
        <owl:versionInfo rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo>
        <rdfs:subClassOf rdf:resource="https://www.w3.org/ns/earl#Fail" />
    </rdfs:Class>
    <rdfs:Class rdf:ID="Warning">
        <rdfs:label rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> Warning when processing the XML file (parser issues)</rdfs:label>
        <owl:versionInfo rdf:datatype="https://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo>
        <rdfs:subClassOf rdf:resource="https://www.w3.org/ns/earl#Fail" />
    </rdfs:Class>

A user of the aforementioned validator defines her own XML Schema (see Example 4.2) for an e-commerce application. The schema defines some restrictions in an order element, against which running Web Services payloads must be verified. To facilitate this process and provide via the Web Service a more user-friendly error feedback to her customers, she uses this validator.

<xsd:schema xmlns:xsd = "https://www.w3.org/2001/XMLSchema" elementFormDefault = "qualified">
    <xsd:element name = "order">
        <xsd:complexType>
            <xsd:sequence>
                <xsd:element ref = "item" maxOccurs = "unbounded"/>
            </xsd:sequence>
            <xsd:attribute name = "orderid" use = "required" type = "xsd:ID"/>
            <xsd:attribute name = "customer" use = "required" type = "xsd:integer"/>
        </xsd:complexType>
    </xsd:element>
    <xsd:element name = "item">
        <xsd:complexType>
            <xsd:sequence>
                <xsd:element ref = "quantity"/>
                <xsd:element ref = "unitprice"/>
            </xsd:sequence>
            <xsd:attribute name = "itemid" type = "xsd:ID"/>
        </xsd:complexType>
    </xsd:element>
    <xsd:element name = "quantity" type = "xsd:unsignedLong"/>
    <xsd:element name = "unitprice">
        <xsd:complexType>
            <xsd:simpleContent>
                <xsd:extension base = "xsd:float">
                    <xsd:attribute name = "currency" use = "required" type = "currencyType"/>
                </xsd:extension>
            </xsd:simpleContent>
        </xsd:complexType>
    </xsd:element>
    <xsd:simpleType name = "currencyType">
        <xsd:restriction base = "xsd:string">
            <xsd:enumeration value = "euros"/>
            <xsd:enumeration value = "dollars"/>
            <xsd:enumeration value = "pounds"/>
        </xsd:restriction>
    </xsd:simpleType>
</xsd:schema>

Customer X sends as a SOAP payload the following order:

<order xmlns:xsi = "https://www.w3.org/2001/XMLSchema-instance"
    xsi:noNamespaceSchemaLocation = "order.xsd"
    orderid = "oj_384" customer = "12345">
    <item itemid = "cat_34894">
        <quantity>2</quantity>
        <unitprice currency = "dollars">40.88</unitprice>
    </item>
</order>

Which is evaluated through the Cool Validator, producing the following EARL report:

<earl:Software rdf:about="https://example.org/coolvalidator/20/">
        <dct:title xml:lang="en">Cool Validator</dct:title>
        <dct:hasVersion>2.0</dct:hasVersion>
        <dct:description xml:lang="en">The best XML validator of the world.</dct:description>
    </earl:Software>
    <earl:TestCase rdf:about="https://example.org/customers/schemas/order.xsd">
        <dct:title xml:lang="en">Ordering Web Service Schema</dct:title>
    </earl:TestCase>
    <earl:TestResult rdf:about="#result">
        <earl:info>The end-tag for element type "quantity" must end with a '&gt;' delimiter.</earl:info>
        <earl:pointer rdf:resource="#pointer" />
        <earl:outcome rdf:resource="https://example.org/ns/xmlval#FatalError" />
    </earl:TestResult>
    <ptr:LineCharPointer rdf:ID="pointer">
        <ptr:charNumber>9</ptr:charNumber>
        <ptr:lineNumber>7</ptr:lineNumber>
        <ptr:reference rdf:resource="#order" />
    </ptr:LineCharPointer>
    

This customer is aware of the EARL extensions of the Cool Validator, and can interpret the results from the perspective of the XML specification, correcting accordingly her SOAP client. Customer Y, who sent the following payload:

    <order xmlns:xsi = "https://www.w3.org/2001/XMLSchema-instance"
    xsi:noNamespaceSchemaLocation = "order.xsd"
    orderid = "oj_490" customer = "67890">
    <item itemid = "cat_30922">
        <quantity>4.0</quantity>
        <unitprice currency = "euro">783.30</unitprice>
    </item>
</order>

cannot interpret this extension of the vocabulary sent in another report. However, by supporting the EARL standard and standard subclassing mechanisms of Semantic Web vocabularies, this customer is still in the position of interpreting the outcome of the error messages and can act accordingly.

<earl:TestResult rdf:ID="result1">
        <earl:pointer rdf:resource="#pointer1" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#Fail" />
        <earl:info>The value '4.0' of element 'quantity' is not valid.</earl:info>
    </earl:TestResult>
    <earl:TestResult rdf:ID="result2">
        <earl:pointer rdf:resource="#pointer2" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#Fail" />
        <earl:info>The value 'euro' of attribute 'currency' on element 'unitprice' is not valid with respect to its type, 'currencyType' [euros, dollars, pounds].</earl:info>
    </earl:TestResult>
    <ptr:LineCharPointer rdf:ID="pointer1">
        <ptr:charNumber>33</ptr:charNumber>
        <ptr:lineNumber>6</ptr:lineNumber>
        <ptr:reference rdf:resource="#order" />
    </ptr:LineCharPointer>
    <ptr:LineCharPointer rdf:ID="pointer2">
        <ptr:charNumber>38</ptr:charNumber>
        <ptr:lineNumber>7</ptr:lineNumber>
        <ptr:reference rdf:resource="#order" />
    </ptr:LineCharPointer>
    

3.5.2 Merging reports from different sources

Using EARL, reports from different sources can be combined to obtain more information or refine existing ones. We take as starting point an XHTML file, which contains two images. One of them lacks of an alternative text attribute. In the other one, the attribute is present, but it reflects the size of the image in bytes.

    <html lang="en" xmlns="https://www.w3.org/1999/xhtml" xml:lang="en">
  <head>
  <title>My photo album</title>
  </head>
  <body>
  <h1>My photo album<</h1>
  <p>These are two nice photos I took yesterday:</p>
  <ul>
  <li>Image of a cat who likes
    <acronym title="carview.php?tsp=Evaluation and Report Language">EARL</acronym>, although it seems quite tired:
    <img src="../images/cat.jpg" />
  </li>
  <li>Image of a fir tree:
    <img src="../images/fir_tree.jpg" alt="98211 bytes" />
  </li>
  </ul>
  </body>
</html>

An accessibility evaluator who wants to verify the compliance of this page against success criteria 1.1.1 from WCAG 2.0 [WCAG20] is using for its accessibility test two tools:

• The already known Cool Tool checker (see Example 3.14) and
• The Exemplary Compliance checker (see Example 4.8 below).

    <earl:Software rdf:about="https://example.org/excompliance/">
        <dct:title xml:lang="en">Exemplary Compliance checker</dct:title>
        <dct:hasVersion>3.2</dct:hasVersion>
        <dct:description xml:lang="en">The compliance checker for Web Accessibility</dct:description>
    </earl:Software>
    

The selected tools test, among others, the following WCAG 2.0 techniques:

• Common Failure F65
• Common Failure F30

The Cool Tool checker provides the following report:

    <earl:TestResult rdf:ID="result1">
        <earl:pointer rdf:resource="#pointer1" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
    </earl:TestResult>
    <earl:TestResult rdf:ID="result2">
        <earl:pointer rdf:resource="#pointer2" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#cantTell" />
    </earl:TestResult>
    <ptr:LineCharPointer rdf:ID="pointer1">
        <ptr:lineNumber>17</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    <ptr:LineCharPointer rdf:ID="pointer2">
        <ptr:lineNumber>20</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    

In this section of the report, we can observe that the tool is able to identify correctly the error in the first image, but it is unable to discern whether the alternative attribute of the second image corresponds to its size. However, the Exemplary Compliance checker is able to download the image, check its size, and compare it to the content of the alternative attribute. This tool produces the following report:

    <earl:TestResult rdf:ID="result1">
        <earl:pointer rdf:resource="#pointer1" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
    </earl:TestResult>
    <earl:TestResult rdf:ID="result2">
        <earl:pointer rdf:resource="#pointer2" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
    </earl:TestResult>
    <ptr:LineCharPointer rdf:ID="pointer1">
        <ptr:lineNumber>17</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    <ptr:LineCharPointer rdf:ID="pointer2">
        <ptr:lineNumber>20</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    

Finally, our evaluator creates a new assertor group, which members include the evaluator and the two tools. The report that she delivers to her customer contains only the assertions that are final, substituting the undefined outcomes by those from the tool that is able to verify adequately the technique. Our evaluator can take decisions on this regard because the use of the EARL Pointers vocabulary allows her to compare exactly the location of the assertion.

    <foaf:Group rdf:ID="assertgroup">
        <dct:title>John Doe and the W3C HTML Validator</dct:title>
        <earl:mainAssertor rdf:resource="https://example.org/persons/jdoe/" />
        <foaf:member rdf:resource="https://example.org/excompliance/" />
        <foaf:member rdf:resource="https://example.org/cooltool/" />
    </foaf:Group>
    <foaf:Person rdf:about="https://example.org/persons/jdoe/">
        <foaf:mbox rdf:resource="mailto:jane@example.org" />
        <foaf:name>Jane Doe</foaf:name>
    </foaf:Person>
    <earl:Software rdf:about="https://example.org/cooltool/">
        <dct:title xml:lang="en">Cool Tool accessibility checker</dct:title>
        <dct:hasVersion>1.0.c</dct:hasVersion>
        <dct:description xml:lang="en">A reliable compliance checker for Web Accessibility</dct:description>
    </earl:Software>
    <earl:Software rdf:about="https://example.org/excompliance/">
        <dct:title xml:lang="en">Exemplary Compliance checker</dct:title>
        <dct:hasVersion>3.2</dct:hasVersion>
        <dct:description xml:lang="en">The compliance checker for Web Accessibility</dct:description>
    </earl:Software>
    <earl:TestCase rdf:about="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65">
        <dct:description xml:lang="en">Failure due to omitting the alt attribute on img elements, area elements, and input elements of type image.</dct:description>
        <dct:title xml:lang="en">Failure of Success Criterion 1.1.1 from WCAG 2.0</dct:title>
    </earl:TestCase>
    <earl:TestCase rdf:about="https://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F30">
        <dct:description xml:lang="en">Failure of Success Criterion 1.1.1 and 1.2.1 due to using text alternatives that are not alternatives.</dct:description>
        <dct:title xml:lang="en">Failure of Success Criterion 1.1.1 and 1.2.1 from WCAG 2.0</dct:title>
    </earl:TestCase>
    <earl:TestResult rdf:ID="result1">
        <earl:pointer rdf:resource="#pointer1" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
    </earl:TestResult>
    <earl:TestResult rdf:ID="result2">
        <earl:pointer rdf:resource="#pointer2" />
        <earl:outcome rdf:resource="https://www.w3.org/ns/earl#failed" />
    </earl:TestResult>
    <ptr:LineCharPointer rdf:ID="pointer1">
        <ptr:lineNumber>17</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    <ptr:LineCharPointer rdf:ID="pointer2">
        <ptr:lineNumber>20</ptr:lineNumber>
        <ptr:charNumber>5</ptr:charNumber>
        <ptr:reference rdf:resource="https://example.org/resource/index.html" />
    </ptr:LineCharPointer>
    

This example demonstrates how the use of simple Semantic Web technologies enables the combination of EARL assertions to produce improved and more accurate reports.

4. Conformance for EARL 1.0 Tools and Reports

[Editor's note: ERT WG is looking for feedback on this entire section.]

This section defines conformance requirements for software tools and processes, to ensure a consistent implementation and exchange of the EARL 1.0 vocabulary. The following applies to tools conforming with EARL 1.0:

1. Conforming EARL 1.0 reports adhere to the requirements listed in 4.1 Conforming EARL 1.0 Reports
2. Software tools that produce conforming EARL 1.0 reports can provide them in valid RDF/XML notation
3. Software tools that process conforming EARL 1.0 reports can accept them in valid RDF/XML notation

4.1 Conforming EARL 1.0 Reports

Conforming EARL 1.0 reports are valid RDF graphs with:

1. At least one Assertion
2. Exactly one Assertor, referenced by earl:assertedBy, for each Assertion
   1. Exactly one identifying name (per language), referenced by dct:title , foaf:name , or doap:name for each Assertor
   2. At most one description (per language), referenced by dct:description or doap:description for each Assertor
   3. Any number of attributes, referenced by foaf:nick , foaf:mbox , or foaf:homepage for each Assertor that is also of type foaf:Agent
   4. Any number of members, referenced by foaf:member , for each Assertor that is also of type foaf:Group
   5. At most one main assertor, referenced by earl:mainAssertor, for each Assertor that is also of type foaf:Group
3. Exactly one Test Subject, referenced by earl:subject, for each Assertion
   1. Exactly one identifying title (per language), referenced by dct:title , foaf:name , or doap:name for each Test Subject
   2. At most one description (per language), referenced by dct:description or doap:description for each Test Subject
   3. At most one date (as defined by XML Datatypes), referenced by dct:date , for each Test Subject
   4. Any number of relationships, referenced by dct:hasPart or dct:isPartOf , between any instances of Test Subject
4. Exactly one Test Criterion, referenced by earl:test, for each Assertion
   1. Exactly one identifying title (per language), referenced by dct:title , for each Test Criterion
   2. At most one description (per language), referenced by dct:description or doap:description for each Test Criterion
   3. Any number of relationships, referenced by dct:hasPart or dct:isPartOf , between any instances of Test Criterion
5. Exactly one Test Result, referenced by earl:result, for each Assertion
   1. Exactly one date (as defined by XML Datatypes), referenced by dct:date , for each Test Result
   2. Exactly one Outcome Value, referenced by earl:outcome, for each Test Result
      1. Exactly one identifying title (per language), referenced by dct:title , for each Outcome Value
      2. Exactly one description (per language), referenced by dct:description , for each Outcome Value
   3. At most one identifying title (per language), referenced by dct:title , for each Test Result
   4. At most one description (per language), referenced by dct:description or doap:description for each Test Result
   5. At most one additional information (per language), referenced by earl:info for each Test Result
   6. Any number of pointer methods, referenced by earl:pointer for each Test Result
6. At most one Test Mode, referenced by earl:mode, for each Assertion
   1. Exactly one identifying title (per language), referenced by dct:title , for each Test Mode
   2. Exactly one description (per language), referenced by dct:description , for each Test Mode
7. Exactly one identifying name (per language), referenced by doap:name , for each Software
8. Conforming HTTP-in-RDF graphs, for each Test Subject that is also of type http:Response
9. Conforming Content-in-RDF graphs, for each Test Subject that is also of type cnt:Content

Note: subclasses or subproperties of terms share the same type. They are therefore considered to be equivalent entities in adhering to any of the above requirements. Also, instances in multiple languages of the same entity (such as title or description) are considered to be a single occurrence of the entity.

In addition, it is strongly recommended that EARL 1.0 reports are also valid RDF graphs with:

1. Each Assertor is also one of the following types:
   • earl:Software
   • foaf:Agent
   • foaf:Person
   • foaf:Organization
   • foaf:Group
2. Each Test Subject is also one of the following types:
   • earl:Software
   • cnt:Content
   • http:Response
   • foaf:Document
3. Each Test Criterion is also one of the following types:
   • earl:TestRequirement
   • earl:TestCase
4. Each Test Mode is one of the following instances:
   • earl:automatic
   • earl:manual
   • earl:semiAuto
   • earl:undisclosed
   • earl:unknownMode
5. Each Outcome Value is one of the following instances:
   • earl:passed
   • earl:failed
   • earl:cantTell
   • earl:inapplicable
   • earl:untested
   Or is an instance of one of the following classes (or sublcasses thereof):
   • earl:Pass
   • earl:Fail
   • earl:CannotTell
   • earl:NotApplicable
   • earl:NotTested

4.2 Conforming HTTP-in-RDF Graphs

Conforming HTTP-in-RDF graphs are valid RDF graphs with:

1. Exactly one connection authority, specified by http:connectionAuthority, for each Connection
2. At most one RDF collection, referenced by http:requests, with any number of Request instances, for each Connection
3. Exactly one HTTP version, specified by http:httpVersion, for each Message
4. At most one RDF collection, referenced by http:headers, with any number of Message Header instances, for each Message
5. Exactly one message body, referenced by http:body, for each Message
6. At most one date, specified by dct:date , for each Message
7. Exactly one method name, specified by http:methodName, for each Request
8. Exactly one request URI, specified by http:requestURI, for each Request
9. At most one Method, referenced by http:mthd, for each Request
10. At most one Response, referenced by http:resp, for each Request
11. Exactly one status code value, specified by http:statusCodeValue, for each Response
12. Exactly one reason phrase, specified by http:reasonPhrase, for each Response
13. At most one Status Code, referenced by http:sc, for each Response
14. Exactly one field name, specified by http:fieldName, for each Message Header
15. Exactly one field value, specified by http:fieldValue, for each Message Header
16. At most one Header Name, referenced by http:hdrName, for each Message Header
17. At most one RDF collection, referenced by http:headerElements, with any number of Header Element instances, for each Message Header
18. Exactly one header element name, specified by http:elementName, for each Header Element
19. At most one header element value, specified by http:elementValue, for each Header Element
20. At most one RDF collection, referenced by http:params, with any number of Parameter instances, for each Header Element
21. Exactly one parameter name, specified by http:paramName, for each Parameter
22. Exactly one parameter value, specified by http:paramValue, for each Parameter

4.3 Conforming Content-in-RDF Graphs

Conforming Content-in-RDF graphs are valid RDF graphs with:

1. At most one character encoding, specified by cnt:characterEncoding, for each Content
2. Any number of relationships, referenced by dct:hasFormat or dct:isFormatOf , between any instances of Content
3. Exactly one byte sequence, specified by cnt:bytes, for each ContentAsBase64
4. Exactly one character sequence, specified by cnt:chars, for each ContentAsText
5. Exactly one XML rest, specified by cnt:rest, for each ContentAsXML
6. At most one leadingMisc, specified by cnt:leadingMisc, for each ContentAsXML
7. At most one document type delcaration, referenced by cnt:dtDecl, for each ContentAsXML
8. Exactly one XML version, specified by cnt:version, for each ContentAsXML
9. At most one XML character encoding, specified by cnt:declaredEncoding, for each ContentAsXML
10. At most one XML standalone declaration, specified by cnt:standalone, for each ContentAsXML
11. Exactly one document type name, specified by cnt:doctypeName, for each DoctypeDecl
12. At most one public identifier, specified by cnt:publicId, for each DoctypeDecl
13. At most one system identifier, specified by cnt:systemId, for each DoctypeDecl
14. At most one internal subset, specified by cnt:internalSubset, for each DoctypeDecl

5. Serializations of EARL Reports

Note: this section will be added to refer the reader to best practices and existing references in RDF/XML serializations (possibly providing a DTD or XML Schema for EARL); RDF->JSON conversion (in particular if we do end up providing an XML Schema or DTD); binary RDF (work in progress at W3C); or other formats that may be useful to tool developers.

Appendix A: References

[Content-RDF]

Representing Content in RDF. https://www.w3.org/TR/Content-in-RDF10/

[DCMISCHEMAS]

DCMI term declarations represented in RDF schema language. https://dublincore.org/schemas/rdfs/

[EARL-Schema]

Evaluation and Report Language 1.0 Schema. https://www.w3.org/TR/EARL10-Schema/

[FOAF]

FOAF Vocabulary Specification 0.91. Namespace Document 2 November 2007 - OpenID Edition. https://xmlns.com/foaf/spec/

[HTTP-RDF]

HTTP Vocabulary in RDF. https://www.w3.org/TR/HTTP-in-RDF10/

[IEEE-829]

IEEE Standard for Software Test Documentation (IEEE Std 829-1998). ISBN 0-7381-1444-8 SS94687. https://ieeexplore.ieee.org/servlet/opac?punumber=5976

[ISO15836]

Information and documentation - The Dublin Core metadata element set. ISO 15836:2009. https://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=52142

[NISOZ3985]

ANSI/NISO Z39.85 - The Dublin Core Metadata Element Set. NISO, May 2007. https://www.niso.org/standards/z39-85-2007/

[Pointers-RDF]

Pointer Methods in RDF. https://www.w3.org/TR/Pointers-in-RDF10/

[RDF]

Resource Description Framework (RDF) Model and Syntax Specification. W3C Recommendation, 22 February 1999. https://www.w3.org/TR/1999/REC-rdf-syntax-19990222/

[RDF-PRIMER]

RDF Primer. W3C Recommendation, 10 February 2004. https://www.w3.org/TR/rdf-primer/

[RDFS]

RDF Vocabulary Description Language 1.0: RDF Schema. W3C Recommendation, 10 February 2004. https://www.w3.org/TR/rdf-schema/

[RDF-XML]

RDF/XML Syntax Specification (Revised). W3C Recommendation 10 February 2004. https://www.w3.org/TR/rdf-syntax-grammar/

[RDF-XML-DIFFS]

Why RDF model is different from the XML model. Paper by Tim Berners-Lee, September 1998. https://www.w3.org/DesignIssues/RDF-XML

[RFC2119]

Key words for use in RFCs to Indicate Requirement Levels. IETF RFC, March 1997. https://www.ietf.org/rfc/rfc2119.txt

[RFC5013]

The Dublin Core Metadata Element Set. IETF RFC, August 2007. https://www.ietf.org/rfc/rfc5013.txt

[OWL]

OWL Web Ontology Language. W3C Recommendation, 10 February 2004. https://www.w3.org/TR/owl-features/

[WCAG10]

Web Content Accessibility Guidelines 1.0. W3C Recommendation, 5 May 1999. https://www.w3.org/TR/WCAG10/

[WCAG20]

Web Content Accessibility Guidelines 2.0. W3C Recommendation, 11 December 2008. https://www.w3.org/TR/WCAG20/

[XML]

Extensible Markup Language (XML) 1.0 (Fifth Edition). W3C Recommendation 26 November 2008. https://www.w3.org/TR/REC-xml/

[URI]

Uniform Resource Identifier (URI): Generic Syntax. IETF RFC, January 2005. https://www.ietf.org/rfc/rfc3986.txt

Appendix D: Contributors

Shadi Abou-Zahra, Carlos Iglesias, Michael A Squillace, Johannes Koch and Carlos A Velasco.

Appendix C: Document changes

The following is a list of changes with respect to the previous internal version:

• Adopted common definitions of EARL.
• Disambiguated the links to examples.
• Minor editorial corrections.
• Examples clarified.