headers
Matej Cepl | 2 Jul 2007 13:39
Picon
Favicon

Re: Haggadah in XML?

On 2007-06-27, 09:04 GMT, James Cummings wrote:
> What elements do you feel you need?  Perhaps we could make some 
> suggestions.

OK, I tried to do what I could, but I was not able to go all the 
way to the validating document, and some things look awkward to 
me.

I have created this TEI customization 
http://www.ceplovi.cz/matej/tmp/tei_drama.xml which generated 
this RNC schema http://www.ceplovi.cz/matej/tmp/TEI.rnc I begun 
to work on my document 
http://www.ceplovi.cz/matej/tmp/PCF_haggadah.xml and it is almost 
good, but I cannot validate it:

[matej <at> hubmaier haggadah]$ xmllint --relaxng TEI.rng --noout 
PCF_haggadah.xml PCF_haggadah.xml:147: element sp: Relax-NG 
validity error : Did not expect element speaker there
PCF_haggadah.xml:147: element sp: Relax-NG validity error 
: Expecting element l, got speaker
PCF_haggadah.xml:149: element p: Relax-NG validity error 
: Element p has extra content: text
PCF_haggadah.xml:149: element p: Relax-NG validity error : Did 
not expect element p there
PCF_haggadah.xml:147: element sp: Relax-NG validity error 
: Element sp failed to validate content
PCF_haggadah.xml:164: element speaker: Relax-NG validity error 
: Did not expect element speaker there
PCF_haggadah.xml:164: element speaker: Relax-NG validity error 
: Expecting element l, got speaker
(Continue reading)

Permalink | Reply | 5 comments |
headers
Sebastian Rahtz | 2 Jul 2007 18:17
Picon
Picon
Favicon

Re: Haggadah in XML?

Matej Cepl wrote:
>
> [matej <at> hubmaier haggadah]$ xmllint --relaxng TEI.rng --noout 
> PCF_haggadah.xml PCF_haggadah.xml:147: element sp: Relax-NG 
> validity error : Did not expect element speaker there
>   
beware of using xmllint as your RELAX NG validator. Its messages are
far from easy to understand. I recommend using jing or rnv or the oXygen
editor.

your file has two problems:

 a) if you use <speaker> inside <sp>, then the text  must be inside eg 
<p>. You can't
mixed text content and element content of <sp>.

 b) you misspelt cRef as cref

when I correct those, its validates fine. your use of cRef seems
to solve the Biblical problem, thats just what it is for.

--

-- 
Sebastian Rahtz      

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
Permalink | Reply | 4 comments |
headers
Matej Cepl | 2 Jul 2007 22:11
Picon
Favicon

Re: Haggadah in XML?

On 2007-07-02, 16:17 GMT, Sebastian Rahtz wrote:
> when I correct those, its validates fine. your use of cRef 
> seems to solve the Biblical problem, thats just what it is for.

And some less complicated method of biblical quotation than 
<cit><quote></quote><ref></ref></cit>?

Matěj
Permalink | Reply | 1 comment |
headers
Sebastian Rahtz | 2 Jul 2007 22:50
Picon
Picon
Favicon

Re: Haggadah in XML?

Matej Cepl wrote:
> On 2007-07-02, 16:17 GMT, Sebastian Rahtz wrote:
>   
>> when I correct those, its validates fine. your use of cRef 
>> seems to solve the Biblical problem, thats just what it is for.
>>     
>
> And some less complicated method of biblical quotation than 
> <cit><quote></quote><ref></ref></cit>?
>   
I'm leaving that to wiser heads..... :-}

--

-- 
Sebastian Rahtz      

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
Permalink | Reply | 0 comments |
headers
Matej Cepl | 3 Jul 2007 10:32
Picon
Favicon

Re: Haggadah in XML?

On Mon, 02 Jul 2007 17:17:35 +0100, Sebastian Rahtz scripst:
> when I correct those, its validates fine. your use of cRef seems to
> solve the Biblical problem, thats just what it is for.

And surely it does validate -- thanks a lot both of you!

Matěj
Permalink | Reply | 0 comments |
headers
Matej Cepl | 3 Jul 2007 13:23
Picon
Favicon

Re: Haggadah in XML?

On Mon, 02 Jul 2007 17:17:35 +0100, Sebastian Rahtz scripst:
> beware of using xmllint as your RELAX NG validator. Its messages are far
> from easy to understand. I recommend using jing or rnv or the oXygen
> editor.

OK, trying jing (and yes, I will use it because it is able to use compact 
syntax directly without conversion, which is very cool), but I don't see 
that much difference in the readibility of the outputs:

[matej <at> hubmaier haggadah]$ jing -t -c TEI.rnc PCF_haggadah.xml 
/home/matej/archiv/2007/clanky/haggadah/PCF_haggadah.xml:166:50: error: 
text not allowed here
/home/matej/archiv/2007/clanky/haggadah/PCF_haggadah.xml:167:59: error: 
text not allowed here
/home/matej/archiv/2007/clanky/haggadah/PCF_haggadah.xml:168:58: error: 
text not allowed here
/home/matej/archiv/2007/clanky/haggadah/PCF_haggadah.xml:169:43: error: 
text not allowed here
/home/matej/archiv/2007/clanky/haggadah/PCF_haggadah.xml:170:13: error: 
unfinished element
Elapsed time 545+287=832 milliseconds
[matej <at> hubmaier haggadah]$ trang TEI.rnc TEI.rng
[matej <at> hubmaier haggadah]$ xmllint --noout --relaxng TEI.rng 
PCF_haggadah.xml 
PCF_haggadah.xml:164: element sp: Relax-NG validity error : Did not 
expect element speaker there
PCF_haggadah.xml:164: element sp: Relax-NG validity error : Expecting 
element l, got speaker
PCF_haggadah.xml:164: element sp: Relax-NG validity error : Expecting an 
element got text
(Continue reading)

Permalink | Reply | 0 comments |
headers
Martin Holmes | 3 Jul 2007 20:49
Picon
Picon
Favicon

The affiliation element

Hi folks,

I'm trying to find a way to use the <affiliation> element in documents 
based on a schema generated today from Roma. The schema includes most of 
the namesdates module (including <affiliation>).

It seems that <affiliation> cannot be a child of any of these elements:

<person>
<persName>
<name>
<author>

Am I right about that? If so, is it intentional? It seems to be an odd 
limitation; I can include <email> and <address> inside <name>, but not 
<affiliation>. Surely they're the same type of thing?

Cheers,
Martin
--

-- 
Martin Holmes
University of Victoria Humanities Computing and Media Centre
(mholmes <at> uvic.ca)
Half-Baked Software, Inc.
(mholmes <at> halfbakedsoftware.com)
martin <at> mholmes.com
Permalink | Reply | 18 comments |
headers
Sebastian Rahtz | 3 Jul 2007 22:56
Picon
Picon
Favicon

Re: The affiliation element

Martin Holmes wrote:
> I'm trying to find a way to use the <affiliation> element in documents 
> based on a schema generated today from Roma. The schema includes most 
> of the namesdates module (including <affiliation>).
>
> It seems that <affiliation> cannot be a child of any of these elements:
>
> <person>
> <persName>
> <name>
> <author>
>
> Am I right about that? 
no; something has gone wrong there. <affiliation> is definitely allowed 
inside <person>. Its a sibling
of <persName> and <name>.

--

-- 
Sebastian Rahtz      

Information Manager, Oxford University Computing Services
13 Banbury Road, Oxford OX2 6NN. Phone +44 1865 283431
Permalink | Reply | 17 comments |
headers
Martin Holmes | 3 Jul 2007 23:32
Picon
Picon
Favicon

Re: The affiliation element

Sebastian Rahtz wrote:
> Martin Holmes wrote:
>> I'm trying to find a way to use the <affiliation> element in documents 
>> based on a schema generated today from Roma. The schema includes most 
>> of the namesdates module (including <affiliation>).
>>
>> It seems that <affiliation> cannot be a child of any of these elements:
>>
>> <person>
>> <persName>
>> <name>
>> <author>
>>
>> Am I right about that? 
> no; something has gone wrong there. <affiliation> is definitely allowed 
> inside <person>. Its a sibling
> of <persName> and <name>.

I don't actually seem to be able to use <person> anywhere useful (such 
as inside <author>). Ideally, what I'd like to do is this (inside a 
biblStruct):

<author>
   <name>
    <forename>Joe</forename>
    <surname>Bloggs</surname>
   </name>
   <affiliation>University of Life</affiliation>
</author>
(Continue reading)

Permalink | Reply | 16 comments |
headers
Sebastian Rahtz | 4 Jul 2007 00:00
Picon
Picon
Favicon

Re: The affiliation element

Martin Holmes wrote:
>
> I don't actually seem to be able to use <person> anywhere useful (such 
> as inside <author>).
<person> is used inside <listPerson>, primarily; that's a structured record
which is a parallel a <biblStruct>. It was not really envisaged as child 
of <author>
> Ideally, what I'd like to do is this (inside a biblStruct):
>
> <author>
>   <name>
>    <forename>Joe</forename>
>    <surname>Bloggs</surname>
>   </name>
>   <affiliation>University of Life</affiliation>
> </author>
>
> If <affiliation> is a sibling of <persName> and <name>, then that 
> should work, right? 
its a sibling inside one content model (model.personPart), but not 
universally.

I don't say your proposal above is wrong, but it's definitely not
how <person> was designed last year.

whether <affiliation> should be a member of other classes as well
is also an interesting notion.

To be truthful, time is running
out for changes in this area. To meet the deadline for the release
(Continue reading)

Permalink | Reply | 14 comments |

Gmane