[openbiblio-dev] Feedback on my BibJSON record

Sylvain Hellegouarch sylvain.hellegouarch at gmail.com
Thu Feb 21 18:52:23 UTC 2013


2013/2/21 Tom Morris <tfmorris at gmail.com>

> On Thu, Feb 21, 2013 at 11:56 AM, Mark MacGillivray <mark at cottagelabs.com>wrote:
>
>> On Thu, Feb 21, 2013 at 4:15 PM, Sylvain Hellegouarch <
>> sylvain.hellegouarch at gmail.com> wrote:
>>
>>>
>>>
>>> 2013/2/21 Karen Coyle <kcoyle at kcoyle.net>
>>>
>>>> I'm not by any means a JSON expert, but doesn't JSON allow use of
>>>> utf-8? Was "Biblioth\u00e8que" used mainly because of sending this in email?
>>>>
>>>> kc
>>>
>>>
>>> Might be. I can't recall the exact serialization. I will have to look it
>>> up.
>>>
>>
>> It does. I just assumed that was what the raw data looked like.
>>
>
> JSON supports full Unicode (UTF-8 by default) in strings, but it also
> allows (but does not require) the \uNNNN notation.  You can also see
> escaped newlines (\n) in the dc:publisher field.
>


Indeed.



>
> Speaking of the publisher field, there are three things that catch my eye
> as needing improvement:
>
> 1. dc:publisher should be "Seuil" and not include publication place and
> year  "Paris\n : Seuil\n, 1986",
> 2. the author's name is not in natural order "Palom\u00e9ra, Marie-France
> de"
> 3. the title should be  "Princesse Finemouche" not "Princesse Finemouche
> / Babette Cole #59; [trad. par] Marie-France de Palom\u00e9ra",
>
> It also seems strange that the author Babette Cole is not included
> anywhere and that the translator is listed as the author.
>
>
The trouble is that my data source is not clean or use a format that
doesn't easily map to DC. Records have been created over a span of decades,
and, when my local council decided to move to a dedicated ILS, they had
tons of records to import by hand so things aren't uniform within the
dataset.

As for the name, it's a (French?) convention with libraries to specify the
lastname first. It seems by the way that national directives suggest this
format is carried on when using DC.

Finally for the title, well it's the same issue. Things were hand written
and I assume the import couldn't easily clean things up.

The difficulty is that it might not be easy to automate the cleanup. For
now, I've dumped the dataset from going through the catalog.
-- 
- Sylvain
http://twitter.com/lawouach
http://www.defuze.org
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.okfn.org/pipermail/openbiblio-dev/attachments/20130221/c609a3c8/attachment.html>


More information about the openbiblio-dev mailing list