[tei-council] xml:lang="eng"

Thu Dec 8 08:45:49 EST 2011

I've fixed a bunch of language codes. I have two remaining questions:

1. transpose.xml has this example:

<exemplum xml:lang="unk"><egXML 
xmlns="http://www.tei-c.org/ns/Examples">   <transpose>
       <ptr target="#ib02"/>
       <ptr target="#ib01"/>
    </transpose>

There is a language code "unk", but it is for the language Enawené-Nawé, 
which I have never heard of. I suspect what was intended in this case 
was "und", for "undetermined". Shall I go ahead and change that, or does 
someone remember having a particular fondness for an example in 
Enawené-Nawé?

2. The reason the xml:lang="eng" problems didn't show up through 
conventional XQuery- and XPath-based diagnosis is that those examples in 
ST-Infrastructure.xml are done as CDATA, like this:

  <egXML xml:lang="und" xmlns="http://www.tei-c.org/ns/Examples">
<![CDATA[<TEI>
<teiHeader  xml:lang="eng">
   <!-- ... -->
</teiHeader>
<text  xml:lang="fra">
<body>
<div>
  <!-- chapter one is in French -->
</div>
<div xml:lang="deu">
  <!-- chapter two is in German -->
</div>
<div>
  <!-- chapter three is French -->
</div>
<!-- ... -->
</body>
</text></TEI>]]></egXML>

I don't know why they're done this way, rather than putting the elements 
in the Examples namespace as normal. Does anyone know?

Cheers,
Martin

On 11-12-08 05:33 AM, Lou Burnard wrote:
> Ah, we should probably update the reference in CH to point to this in
> that case.
>
> On 08/12/11 12:24, Gabriel Bodard wrote:
>> By the way, I recently discovered that there is now a single IANA
>> registry of language codes so we no longer need to worry about looking
>> in the 2-letter list, and then if we don't find what we're looking for
>> move on to the 3-letter list. Instead, all codes, both 2- and
>> (occasional) 3-letter (and 4-letter script codes) are listed on a single
>> page at<http://www.iana.org/assignments/language-subtag-registry>, so
>> there is no longer any danger of accidentally using a 3-digit code that
>> has been deprecated in favour of its 2-digit replacement.
>>
>> (For example, "Ancient Greek (to 1453)" is now unambiguously listed as
>> "grc", so should not be confused with "el".)
>>
>> I've found this very helpful while trying to decide how to tag Nabatean
>> script in papyri...
>>
>> G
>>
>> On 2011-12-08 12:00, Lou Burnard wrote:
>>> The 3 letter codes presumably date from an older gentler time, when men
>>> had long hair and rejoiced in diversity...
>>>
>>>
>>>
>>> On 08/12/11 08:23, James Cummings wrote:
>>>>
>>>> I certainly didn't leave three letter ones there on purpose! Fix them I say.
>>>> JamesC
>>>>
>>>> --
>>>> James Cummings, InfoDev, OUCS, University of Oxford (via phone)
>>>>
>>>> Laurent Romary<laurent.romary at inria.fr>     wrote:
>>>>
>>>>
>>>> So yes, you're right there. These are things that should be changed.
>>>> Cheers,
>>>> Laurent
>>>>
>>>> Le 8 déc. 2011 à 04:15, Martin Holmes a écrit :
>>>>
>>>>> And I see<text xml:lang="fra">     too...
>>>>>
>>>>> On 11-12-07 07:14 PM, Martin Holmes wrote:
>>>>>> Hi all,
>>>>>>
>>>>>> I thought we'd come to the conclusion that two-letter lang codes should
>>>>>> be used where they exist, and three-letter ones where they don't, but
>>>>>> ST-infrastructure.xml has lots of examples of
>>>>>>
>>>>>> xml:lang="eng"
>>>>>>
>>>>>> Before I change this, I thought I should check whether I'm wrong about
>>>>>> our conclusion, especially since James edited this file the other day,
>>>>>> and might have left "eng" alone on purpose.
>>>>>>
>>>>>> Cheers,
>>>>>> Martin
>>>>> --
>>>>> tei-council mailing list
>>>>> tei-council at lists.village.Virginia.EDU
>>>>> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
>>>>>
>>>>> PLEASE NOTE: postings to this list are publicly archived
>>>>
>>>> Laurent Romary
>>>> INRIA&     HUB-IDSL
>>>> laurent.romary at inria.fr
>>>>
>>>>
>>>>
>>>> --
>>>> tei-council mailing list
>>>> tei-council at lists.village.Virginia.EDU
>>>> http://lists.village.Virginia.EDU/mailman/listinfo/tei-council
>>>>
>>>> PLEASE NOTE: postings to this list are publicly archived
>>>
>>
>