[Yazlist] Unicode Normalization for data in pazpar2

Adam Dickmeiss adam at indexdata.dk
Fri Nov 5 12:31:12 CET 2010


On 2010-10-29 12:15, Porst, Sven wrote:
> Hi everyody,
>
> pazpar2 offers the convenient icu_chain settings to manipulate the data
> that's used for margekeys. In particular I can add an 'NFC' statement to
> the chain used for mergekeys to ensure items are merged correctly even
> when servers send records for identical titles in differing Unicode
> Normalization Forms.
>
> I couldn't discover a similar setting to process the complete records
> retrieved from the servers. As systems occasionally have problems
> displaying NFD strings correctly, I'd consider it an advantage to have
> normalized NFC strings everywhere. It seems that pazpar2 comes with all
> the technology built in to achieve that, but I either can't find the
> relevant setting or pazpar cannot use ICU in the way I need it.
>
For complete "original" records PP2 do not apply ICU transforms. Only 
character set conversion may be performed --- as offered by the 
ZOOM_record_get function.

It's a nice idea. Could be achieved within ZOOM as well.. By extending 
ZOOM_record_get perhaps.

/ Adam


> Any clues?
>
>          Sven
>




More information about the Yazlist mailing list