Rule 8.8.1
Summary
This test consists in checking whether each change of language is valid and relevant
Business description
Criterion
Test
Description
Pour chaque page web, le code de langue de chaque changement de langue vérifie-t-il ces conditions ?
- Le code de langue est valide.
- Le code de langue est pertinent.
Level
AA
Technical description
Scope
Page
Decision level
Decidable
Algorithm
Selection
Set1
All the tags different from <html> that have a lang or a xml:lang attribute
Process
Test1
We extract the language code from the lang or the xml:lang attributes from elements of Set1. If these attributes
are both set and different, we keep the xml:lang attribute if the doctype of the page is found within the
“XhtmlDoctypeDeclarations” whitelist, and we keep the lang attribute instead.
For each extracted lang attribute, we check its validity regarding the following regexp : \\w{2,3}(\\-\\w{2,})?$
For each invalid element, raise a messageA.
MessageA : Malformed Language Declaration
- code : MalformedLanguageDeclaration
- status: Failed
- parameter : extracted lang
- present in source : yes
Test2
We extract the language code from the lang or the “xml:lang” attributes from elements of Set1. If these attributes
are both set and different, we keep the xml:lang attribute if the doctype of the page is found within the
“XhtmlDoctypeDeclarations” whitelist, and we keep the lang attribute instead.
For each extracted lang attribute, we check its validity regarding the “ValidLanguageCode” whitelist
- Used nomenclature : “ValidLanguageCode”, “XhtmlDoctypeDeclarations”
- Reference : All the valid language codes recommanded in http://www.loc.gov/standards/iso639-2/php/code_list.php and all the xhtml doctypes declarations recommanded in http://www.w3.org/QA/2002/04/valid-dtd-list.html
For each invalid element, raise a messageA.
MessageB : Wrong Language Declaration
- code : WrongLanguageDeclaration
- status: Failed
- parameter : extracted lang
- present in source : yes
Test3
We extract the language code from the lang or the xml:lang attributes from elements of Set1. If these attributes
are both set and different, we keep the xml:lang attribute if the doctype of the page is found within the
“XhtmlDoctypeDeclarations” whitelist, and we keep the lang attribute instead.
For each extracted text, we detect the language and compare with the declaration of the lang attribute.
If the detected language and the declared language are different, if the size of the extracted text is superior to 20 words, raise a messageA. Raise a MessageB instead.
If the detected language and the declared language are identical, but the size of the extracted text is inferior to 20 words, raise a messageC.
MessageA : Irrelevant Language Declaration
- code : IrrelevantLanguageDeclaration
- status: Failed
- parameter : extracted lang, declared lang, extracted text, snippet
- present in source : yes
MessageB : Suspected Irrelevant Language Declaration
- code : SuspectedIrrelevantLanguageDeclaration
- status: Pre-Qualified
- parameter : extracted lang, declared lang, extracted text, snippet
- present in source : yes
MessageC : Suspected relevant Language Declaration
- code : SuspectedRelevantLanguageDeclaration
- status: Pre-Qualified
- parameter : extracted lang, declared lang, extracted text, snippet
- present in source : yes
Analysis
Not Applicable
The page has no tag different from <html> with the lang or xml:lang attribute (Set1 is empty)
Passed
All the lang declarations are valid and identical to the detected lang and all the tested text size are superior to 20 words
Failed
At least one lang declaration is invalid or badly-formed (Test1 OR Test2 return false for at least one element of Set1)
or
At least one lang declaration is different from the detected lang, with a text size superior to 20 words
Pre-Qualified
In all other cases