[LibX] URL-encoding & ISBN shorting

Godmar Back godmar at gmail.com
Thu Aug 4 06:45:48 PDT 2011


On Tue, Aug 2, 2011 at 8:50 AM, Alexander Wagner <a.wagner at fz-juelich.de> wrote:
> On 02.08.2011 14:29, Godmar Back wrote:
>
> Hi!
>
>>> Is there a way to prevent URL encoding while using bookmarklet? I'm
>>> using nbn-resolving.de to resolve both URN and DOI identifiers, however
>>> it doesn't like it if the / in a DOI gets encoded.
>>
>> Can you give an example where you believe LibX incorrectly encodes?
>
> I added a catalogue like:
>
> Name: DOI/URN Resolver
> URL Template: http://nbn-resolving.de/%nbn
> Search option: URN/DOI
>
> The latter defines the %nbn variable. All other fields are empty.
>

LibX assumes that the search terms go into the query string part of a
URL, not in the path part.

But when %2f is used in the path part of a URL rather than /, it
represents a different URL.

If we want to allow the use of % fields in REST-URLs such as this one,
we need to allow insertion without encoding. We could add a special
syntax for that, such as %%nbn.

 - Godmar


More information about the Libx mailing list