[contents]
Copyright © 2009 W3C ® (MIT, ERCIM, Keio), All Rights Reserved. W3C liability, trademark, document use rules apply.
This document is an introductory guide to the Evaluation and Report Language (EARL) 1.0 and is intended to accompany the normative document Evaluation and Report Language 1.0 Schema [EARL-Schema]. The Evaluation and Report Language is a framework for expressing test results. Although the term test can be taken in its most widely accepted definition, EARL is primarily intended for reporting and exchanging results of tests of Web applications and resources. EARL is a vendor-neutral and platform-independent format.
EARL is expressed in the form of an RDF vocabulary. The Resource Description Framework (RDF) is a language for representing semantically information about resources in the World Wide Web. However, EARL is not conceptually restricted to these resources and may be applied in other contexts outside the Web.
This section describes the status of this document at the time of its publication. Other documents may supersede this document. A list of current W3C publications and the latest revision of this technical report can be found in the W3C technical reports index at http://www.w3.org/TR/.
Please send comments about this document to the mailing list of the ERT WG. The archives for this list are publicly available.
This is a W3C Working Draft of the Evaluation and Report Language (EARL) 1.0 Guide. This document will be published and maintained as a W3C Recommendation after review and refinement. Publication as a Working Draft does not imply endorsement by the W3C Membership. This is a draft document and may be updated, replaced or obsoleted by other documents at any time. It is inappropriate to cite this document as other than work in progress.
This document was produced under the 5 February 2004 W3C Patent Policy. The Working Group maintains a public list of patent disclosures relevant to this document; that page also includes instructions for disclosing a patent. An individual who has actual knowledge of a patent which the individual believes contains Essential Claim(s) with respect to this specification should disclose the information in accordance with section 6 of the W3C Patent Policy.
This document has been produced as part of the W3C Web Accessibility Initiative (WAI). The goals of the Evaluation and Repair Tools Working Group (ERT WG) are discussed in the Working Group charter. The ERT WG is part of the WAI Technical Activity.
This document is an introductory guide to the Evaluation and Report Language (EARL) 1.0 and is intended to accompany the normative document Evaluation and Report Language 1.0 Schema [EARL-Schema] and its associated vocabularies: HTTP Vocabulary in RDF [HTTP-RDF], Representing Content in RDF [Content-RDF] and Pointer Methods in RDF [Pointers-RDF]. The objectives of this document are:
The primary audience of this document are quality assurance and testing tool developers such as accessibility checkers, markup validators, etc. Additionally, we expect that EARL can support accessibility and usability advocates, metadata experts and Semantic Web practitioners among others. We do not assume any previous knowledge of EARL, but it is not the target of this document to introduce the reader to the intricacies of RDF and therefore, the following background knowledge is required:
Although the concepts of the Semantic Web are simple, their abstraction with RDF may bring difficulties to beginners. It is recommended to read carefully the aforementioned references and other tutorials found on the Web. It must be also borne in mind that RDF is primarily targeted to be machine processable and therefore, some of its expressions are not very intuitive for developers used to work with XML only.
This document is organized as follows. Section 2 introduces the basic concepts of EARL, its target audiences and some use cases. Section 3 presents a step-by-step introduction to write an EARL report. Section 4 presents some advanced uses of the vocabulary and how to extend it. Finally, Section 5 presents some conclusions and possible future work.
The Evaluation and Report Language (EARL) is a framework targeted to express and compare test results. EARL builds on top of the Resource Description Framework [RDF], which is the basis for the Semantic Web. Like any RDF vocabulary, EARL is a collection of statements about resources, each with a subject, a predicate (or a verb) and an object. These statements can be serialized in many ways (e.g., RDF/XML or Notation 3, also known as N3). A typical EARL report could contain the following statements (simplifying the notation and not including namespaces):
<#someone> <#checks> <#resource> . <#resource> <#fails> <#test> .
From these two simple statements, it can be inferred already the main components of an EARL Report (wrapped up in an assertion):
This structure shows the universal applicability of EARL and its ability to refer to any type of test: bug reports, software unit tests, test suite evaluations, conformance claims or even tests outside the world of software and the World Wide Web (although in such cases, there might be open issues for its full applicability). It must be stressed again the semantic nature of EARL: its purpose is to facilitate the extraction and comparison of test results by humans and especially by tools (the Semantic Web paradigm); it is not an application optimized for information storage, for which some other XML applications might be more suitable.
Initially, EARL was conceived as a way to create, merge and compare Web accessibility reports from different sources (tools, experts, etc.). However, this original aim has been expanded to cover wider testing scenarios. Summarizing, EARL enables the:
The extensibility of RDF allows tool vendors or developers the addition of new functionalities to the vocabulary, without losing any of the aforementioned characteristics. Other testers might ignore the extensions that they do not understand when processing third-party results.
It is also 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. The scope of this document is limited to the use of the EARL vocabulary: security and privacy considerations 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.
The keywords must, required, recommended, should, may, and optional in this document are used in accordance with RFC 2119 [RFC2119].
The applicability of EARL to different scenarios can be seen in use cases such as the following:
The list of use cases is not limited to the Web. EARL could also be applied to generic Software Quality Assurance processes, where it will be required to map real objects, actors and processes to URIs.
EARL is as well flexible enough to respond to the needs of a variety of audiences involved in testing or quality assurance processes. Typical profiles include:
In software testing and quality assurance environments there are several typical steps that are followed. These are:
Figure 1 displays graphically the aforementioned elements:
Figure 1. Steps in software testing processes.
The previous steps can be matched to existing standards like IEEE 829 [IEEE-829], which defines a set of basic software tests documents.
EARL is not an 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:
RDF can be serialized in different ways, but its XML representation [RDF/XML] is the preferred method and will be used throughout this document. However, even when selecting this approach, there are many equivalent ways to express an RDF model. 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.
In the following sections, we will make an step-by-step introduction to EARL with several examples. The root element of any EARL report is an RDF node. There we declare the namespaces used to define additional classes and/or properties.
Example 3.1. The root element of an EARL report [download file].
<rdf:RDF xmlns:earl="http://www.w3.org/ns/earl#" xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:rdfs="http://www.w3.org/2000/01/rdf-schema#"> <!-- ... --> </rd: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 in
EARL. The tested document can be found in the fictitious URL
http://example.org/resource/index.html
, and has the following HTML
code:
Example 3.2. An XHTML document to be validated [download file].
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html lang="en" xmlns="http://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:
li
"
here; missing one of "ul
", "ol
" start-tag.li
" omitted, but OMITTAG NO
was specified.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
:
Example 3.3. A generic tool as an
Assertor
[download
file].
<earl:Assertor rdf:about="http://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 how in the Assertor
class, EARL
offers the possibility to specify more information by using 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:
Example
3.4. An Assertor
expressed in N3 notation [download file].
@prefix earl: <http://www.w3.org/ns/earl#> . @prefix dct: <http://purl.org/dc/terms/> . <http://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:
Example 3.5. A
Software
assertor [download
file].
<earl:Software rdf:about="http://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 how we inserted a new property, 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:
Example 3.6. A
Person
as an EARL assertor [download file].
<foaf:Person rdf:ID="john"> <foaf:mbox rdf:resource="mailto:john@example.org"/> <foaf:name>John Doe</foaf:name> </foaf:Person>
EARL offers the possibility of defining combination
of assertors. The typical example is 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):
Example 3.7. A
foaf:Group
(software tool and person) as an assertor [download file].
<foaf:Group rdf:ID="assertor01"> <dct:title>John Doe and the W3C HTML Validator</dct:title> <earl:mainAssertor rdf:resource="http://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:
Example
3.8. A TestSubject
with some Dublin Core properties
(non-abbreviated RDF/XML serialization) [download file].
<rdf:Description rdf:about="http://example.org/resource/index.html"> <dct:title xml:lang="en">Project Description</dct:title> <dct:date rdf:datatype="http://www.w3.org/2001/XMLSchema#date">2006-02-14</dct:date> <rdf:type rdf:resource="http://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:
Example 3.9. A test subject expressed as
cnt:ContentAsText
(notice that the special XML characters have been escaped
because the document is not well-formed to be expressed as an XML Literal) [download file].
<cnt:ContentAsText rdf:about="http://example.org/resource/index.html"> <dct:title xml:lang="en">Project Description</dct:title> <dct:date rdf:datatype="http://www.w3.org/2001/XMLSchema#date" >2006-02-14</dct:date> <cnt:characterEncoding>UTF-8</cnt:characterEncoding> <cnt:chars><?xml version="1.0" encoding="UTF-8"?> <!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html lang="en" xmlns="http://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> <p>The strategic goal of this project is to make you understand EARL.</p> <ul> <li>Here comes objective 1. <li>Here comes objective 2.</li> </ul> <p alt="what?">And goodbye ...</p> </body> </html> </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:
Example
3.10. A TestRequirement
with some Dublin Core properties [download file].
<earl:TestRequirement rdf:about="http://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 TestResult
s. In this case, we
present only the errors, but within EARL 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.
Example 3.11. Results of the tests with the validator [download file].
<earl:TestResult rdf:ID="error1"> <dct:description rdf:parseType="Literal"> <div xml:lang="en" xmlns="http://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="http://www.w3.org/ns/earl#failed" /> </earl:TestResult> <earl:TestResult rdf:ID="error2"> <dct:description rdf:parseType="Literal"> <div xml:lang="en" xmlns="http://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="http://www.w3.org/ns/earl#failed" /> </earl:TestResult> <earl:TestResult rdf:ID="error3"> <dct:description rdf:parseType="Literal"> <div xml:lang="en" xmlns="http://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="http://www.w3.org/ns/earl#failed" /> </earl:TestResult>
The final step is to merge together the created
components. The EARL
statements for this purpose are called Assertion
s, 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. From our previous
examples, we could build our first complete report with our three assertions:
Example 3.12. Results of the tests with the validator [download file].
<earl:Assertion rdf:ID="ass1"> <earl:result> <earl:TestResult rdf:ID="error1" /> </earl:result> <earl:test> <earl:TestRequirement rdf:about="http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" /> </earl:test> <earl:subject> <cnt:TextContent rdf:about="http://example.org/resource/index.html" /> </earl:subject> <earl:assertedBy> <foaf:Group rdf:about="#assertor01" /> </earl:assertedBy> </earl:Assertion> <earl:Assertion rdf:ID="ass2"> <earl:result> <earl:TestResult rdf:ID="error2" /> </earl:result> <earl:test> <earl:TestRequirement rdf:about="http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" /> </earl:test> <earl:subject> <cnt:TextContent rdf:about="http://example.org/resource/index.html" /> </earl:subject> <earl:assertedBy> <foaf:Group rdf:ID="assertor01" /> </earl:assertedBy> </earl:Assertion> <earl:Assertion rdf:ID="ass3"> <earl:result> <earl:TestResult rdf:ID="error3" /> </earl:result> <earl:test> <earl:TestRequirement rdf:about="http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd" /> </earl:test> <earl:subject> <cnt:TextContent rdf:about="http://example.org/resource/index.html" /> </earl:subject> <earl:assertedBy> <foaf:Group rdf:about="#assertor01" /> </earl:assertedBy> </earl:Assertion>
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:
Example 3.13. An XHTML document to be verified [download file].
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html lang="en" xmlns="http://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="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:
Example
3.14. A Software
assertor [download file].
<earl:Software rdf:about="http://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:
Example
3.15. A TestCase
for a WCAG 2.0 technique [download file].
<earl:TestCase rdf:about="http://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 make now the test result more verbose and amenable to machine processing by making use of the EARL Pointers [Pointers-RDF] vocabulary. In this case, we identify the line number where the test was compliant:
Example 3.16. A TestResult
with a
pointer [download file].
<ptr:LineCharPointer rdf:ID="pointer"> <ptr:lineNumber>15</ptr:lineNumber> <ptr:reference rdf:resource="http://example.org/resource/index.html" /> </ptr:LineCharPointer> <earl:TestResult rdf:ID="result"> <earl:pointer rdf:resource="#pointer" /> <earl:outcome rdf:resource="http://www.w3.org/ns/earl#passed" /> </earl:TestResult>
Which leads to the following assertion:
Example
3.17. Accessibility Assertion
[download file].
<earl:Assertion rdf:ID="assert"> <earl:result rdf:resource="result" /> <earl:test rdf:resource="http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" /> <earl:subject rdf:resource="http://example.org/resource/index.html" /> <earl:assertedBy rdf:resource="http://example.org/cooltool/" /> </earl:Assertion>
There are cases where the
identification of a resource on the Web requires more than a URL. This typically occurs
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 correctly identify the
TestSubject
.
Let us assume that our exemplary Web server can deliver
under the URL http://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:
Example 3.18. An XHTML file resource in Spanish [download file].
<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN" "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd"> <html lang="es" xmlns="http://www.w3.org/1999/xhtml" xml:lang="es"> <head> <title>Fotografía de un gato</title> </head> <body> <h1>Fotografía de un gato</h1> <p>Imagen de un gato al que le gusta <acronym title="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:
Example 3.19. RDF representation of Example 3.13 [download file].
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:earl="http://www.w3.org/ns/earl#" xmlns:dct="http://purl.org/dc/terms/" xmlns:cnt="http://www.w3.org/2007/content#" xmlns:http="http://www.w3.org/2006/http#" xml:base="http://www.example.org/resource/content_001#"> <cnt:ContentAsBase64 rdf:ID="content1"> <cnt:bytes rdf:datatype="http://www.w3.org/2001/XMLSchema#base64Binary" >ajrq9qguojbglj48z..........</cnt:bytes> </cnt:ContentAsBase64> <http:Response rdf:ID="response1"> <http:httpVersion>1.1</http:httpVersion> <http:statusCodeNumber>200</http:statusCodeNumber> <http:sc rdf:resource="http://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="http://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="http://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="http://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:
Example 3.20. RDF representation of Example 3.18 [download file].
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:earl="http://www.w3.org/ns/earl#" xmlns:dct="http://purl.org/dc/terms/" xmlns:cnt="http://www.w3.org/2007/content#" xmlns:http="http://www.w3.org/2006/http#" xml:base="http://www.example.org/resource/content_002#"> <cnt:ContentAsBase64 rdf:ID="content2"> <cnt:bytes rdf:datatype="http://www.w3.org/2001/XMLSchema#base64Binary" >oeuh058h3ui..........</cnt:bytes> </cnt:ContentAsBase64> <http:Response rdf:ID="response2"> <http:httpVersion>1.1</http:httpVersion> <http:statusCodeNumber>200</http:statusCodeNumber> <http:sc rdf:resource="http://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="http://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="http://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="http://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 it is
recommended to include timestamps or modification dates in the reports.
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:
Example 3.21. Evaluation report for the English XHTML resource [download file].
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:earl="http://www.w3.org/ns/earl#" xmlns:dct="http://purl.org/dc/terms/" xmlns:cnt="http://www.w3.org/2007/content#" xmlns:ptr="http://www.w3.org/WAI/ER/EARL/pointers#" xml:base="http://www.example.org/earl/report1#"> <earl:Assertion rdf:ID="assert"> <earl:result rdf:resource="result" /> <earl:test rdf:resource="http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" /> <earl:subject rdf:resource="http://www.example.org/resource/content_001#response1" /> <earl:assertedBy rdf:resource="http://example.org/cooltool/" /> </earl:Assertion> <earl:Software rdf:about="http://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="http://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="http://www.example.org/resource/content_001#content1" /> </ptr:LineCharPointer> <earl:TestResult rdf:ID="result"> <earl:pointer rdf:resource="#pointer" /> <earl:outcome rdf:resource="http://www.w3.org/ns/earl#passed" /> </earl:TestResult> </rdf:RDF>
And the evaluation report for the Spanish resource looks like the following example:
Example 3.22. Evaluation report for the Spanish XHTML resource [download file].
<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#" xmlns:earl="http://www.w3.org/ns/earl#" xmlns:dct="http://purl.org/dc/terms/" xmlns:cnt="http://www.w3.org/2007/content#" xmlns:ptr="http://www.w3.org/WAI/ER/EARL/pointers#" xml:base="http://www.example.org/earl/report2#"> <earl:Assertion rdf:ID="assert"> <earl:result rdf:resource="result" /> <earl:test rdf:resource="http://www.w3.org/TR/2008/NOTE-WCAG20-TECHS-20081211/F65" /> <earl:subject rdf:resource="http://www.example.org/resource/content_002#response2" /> <earl:assertedBy rdf:resource="http://example.org/cooltool/" /> </earl:Assertion> <earl:Software rdf:about="http://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="http://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="http://www.example.org/resource/content_002#content2" /> </ptr:LineCharPointer> <earl:TestResult rdf:ID="result"> <earl:pointer rdf:resource="#pointer" /> <earl:outcome rdf:resource="http://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.
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.
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:
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:
Example 4.1. RDF Schema in the namespace
http://example.org/ns/xmlval#
for the error extensions of Cool Validator,
which contains new classes, extensions of earl:Fail
[download file].
<rdfs:Class rdf:ID="FatalError"> <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> Fatal error when processing the XML file (well-formedness) </rdfs:label> <owl:versionInfo rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo> <rdfs:subClassOf rdf:resource="http://www.w3.org/ns/earl#Fail" /> </rdfs:Class> <rdfs:Class rdf:ID="Error"> <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> Error when processing the XML file (validation constraint) </rdfs:label> <owl:versionInfo rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo> <rdfs:subClassOf rdf:resource="http://www.w3.org/ns/earl#Fail" /> </rdfs:Class> <rdfs:Class rdf:ID="Warning"> <rdfs:label rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> Warning when processing the XML file (parser issues)</rdfs:label> <owl:versionInfo rdf:datatype="http://www.w3.org/2001/XMLSchema#string"> 1.0</owl:versionInfo> <rdfs:subClassOf rdf:resource="http://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.
Example 4.2. XML Schema for the ordering Web Service [download file].
<xsd:schema xmlns:xsd = "http://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:
Example 4.3. SOAP payload for Customer X [download file].
<order xmlns:xsi = "http://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:
Example 4.4. First XML validation report [download file].
<earl:Software rdf:about="http://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="http://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 '>' delimiter.</earl:info> <earl:pointer rdf:resource="#pointer" /> <earl:outcome rdf:resource="http://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:
Example 4.5. SOAP payload for Customer Y [download file].
<order xmlns:xsi = "http://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 act accordingly.
Example 4.6. Second XML validation report translated to standard EARL [download file].
<earl:TestResult rdf:ID="result1"> <earl:pointer rdf:resource="#pointer1" /> <earl:outcome rdf:resource="http://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="http://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>
This example shows how via the use of EARL reports from different sources can be combined to obtain more information or refine the existing one. 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.
Example 4.7. An XHTML document to be tested [download file].
<html lang="en" xmlns="http://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="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 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:
Example
4.8. Exemplary Compliance as a Software
assertor [download file].
<earl:Software rdf:about="http://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:
The Cool Tool checker provides the following report:
Example 4.9. Extract from the Cool Tool report [download file].
<earl:TestResult rdf:ID="result1"> <earl:pointer rdf:resource="#pointer1" /> <earl:outcome rdf:resource="http://www.w3.org/ns/earl#failed" /> </earl:TestResult> <earl:TestResult rdf:ID="result2"> <earl:pointer rdf:resource="#pointer2" /> <earl:outcome rdf:resource="http://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="http://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="http://example.org/resource/index.html" /> </ptr:LineCharPointer>
In it 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:
Example 4.10. Extract from the Exemplary Compliance checker report [download file].
<earl:TestResult rdf:ID="result1"> <earl:pointer rdf:resource="#pointer1" /> <earl:outcome rdf:resource="http://www.w3.org/ns/earl#failed" /> </earl:TestResult> <earl:TestResult rdf:ID="result2"> <earl:pointer rdf:resource="#pointer2" /> <earl:outcome rdf:resource="http://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="http://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="http://example.org/resource/index.html" /> </ptr:LineCharPointer>
Finally, our evaluator creates a new assertor group, which members are herself 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.
Example 4.11. Extract from the final accessibility report [download file].
<foaf:Group rdf:ID="assertgroup"> <dct:title>John Doe and the W3C HTML Validator</dct:title> <earl:mainAssertor rdf:resource="http://example.org/persons/jdoe/" /> <foaf:member rdf:resource="http://example.org/excompliance/" /> <foaf:member rdf:resource="http://example.org/cooltool/" /> </foaf:Group> <foaf:Person rdf:about="http://example.org/persons/jdoe/"> <foaf:mbox rdf:resource="mailto:jane@example.org" /> <foaf:name>Jane Doe</foaf:name> </foaf:Person> <earl:Software rdf:about="http://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="http://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="http://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="http://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="http://www.w3.org/ns/earl#failed" /> </earl:TestResult> <earl:TestResult rdf:ID="result2"> <earl:pointer rdf:resource="#pointer2" /> <earl:outcome rdf:resource="http://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="http://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="http://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.
This guide presented a thorough overview of the Evaluation and Report Language (EARL). As mentioned in the introduction, EARL must be seen as a generic framework that can facilitate the creation and exchange of test reports. In this generality lies its strength, as it can be applied to multiple scenarios and use cases, which may even lay outside the world of software development and compliance testing.
The EARL framework allows as well merging and aggregation of results in a semantic manner, thus enabling different testing actors to share and improve results.
Of course, there could be scenarios where EARL might not be able to cope with their underlying complexity. However, its semantic nature allows its extensibility via proprietary vocabularies based upon RDF, without endangering the interoperability of the reports.
The Working Group is looking forward to receiving feedback on the current version of the schema, and expects from implementers of compliance tools issues and suggestions for improvement.
http://www.w3.org/TR/Content-in-RDF/
http://dublincore.org/schemas/rdfs/
http://www.w3.org/TR/EARL10-Schema/
http://xmlns.com/foaf/spec/
http://www.w3.org/TR/HTTP-in-RDF/
http://ieeexplore.ieee.org/servlet/opac?punumber=5976
http://www.iso.org/iso/iso_catalogue/catalogue_tc/catalogue_detail.htm?csnumber=52142
http://www.niso.org/standards/z39-85-2007/
http://www.w3.org/TR/Pointers-in-RDF/
http://www.w3.org/TR/1999/REC-rdf-syntax-19990222/
http://www.w3.org/TR/rdf-primer/
http://www.w3.org/TR/rdf-schema/
http://www.w3.org/TR/rdf-syntax-grammar/
http://www.w3.org/DesignIssues/RDF-XML
http://www.ietf.org/rfc/rfc2119.txt
http://www.ietf.org/rfc/rfc5013.txt
http://www.w3.org/TR/owl-features/
http://www.w3.org/TR/WCAG10/
http://www.w3.org/TR/WCAG20/
http://www.w3.org/TR/REC-xml/
http://www.ietf.org/rfc/rfc3986.txt
The following is a list of changes with respect to the previous internal version:
Shadi Abou-Zahra, Carlos Iglesias, Michael A Squillace, Johannes Koch and Carlos A Velasco.