[Bibjson-dev] omnidator

Fri Jun 17 06:56:51 UTC 2011

William, thanks for putting us in touch. Jim, pleased to meet you.

I know, use and love bibutils - very happy to collaborate, here!

> I don't see where you get N^2-N (maybe unicode problems, that's how I
> read N?-N, apologies if I've misunderstood), omnidator from what I can
> tell uses RDF as an internal representation like the bibutils use
> MODS.  So actually the way it works, I think, more or less, there are
> N serialisers and N deserialisers, so 2N just like bibutils - but it
> isn't limited to bibliographic data.

As far as I understood, Jim simply stated the general fact that if you  
want to translate from N languages to N languages you would, in the  
naive case, need N^2-N converters, whereas applying the Lingua Franca  
pattern [1] one can cut it down to 2N. This is true for bibutils and  
as well for omnidator. Nothing new here and no criticism implied ;)

Jim said:
> ] We need a strategy/architecture for managing conversions to/from  
> BibJSON, presumably a webservice like omnidator, and some community  
> development to support this over time.
> ] Thoughts about how to achieve this?

... and William rightly pointed out:
> To the extent that bibjson can be aligned with json-ld using a
> suitable vocabulary a lot of this can come along for free. To
> the extent that we need or want bibtex/mods/marc21/etc i/o it is
> then just as simple as writing the necessary parsers/serialisers,
> and we could just use omnidator or a only slightly modified
> variant.

Indeed. JSON-LD [2] alignment would be desirable if this doesn't break  
your downstream compatibility. Now, omnidator is nothing more than a  
GAE deployment (with a bit of an UI on top) of the Schema.org gateway  
[3] which happens to leverage the terms defined by Schema.org. Given  
that this is not a show-stopper I'd suggest you clone this repo and  
develop a fork, adapting it to your needs.

If I can be of help (re supported input/output formats) please let me  
know and I'll prioritise them - simply raising an issue on the repo  
[3] would help to keep track of it (pls. do mention Schema.org gateway  
in the issue, as we have quite some sub-projects in it ;)

Cheers,
	Michael

[1] http://c2.com/cgi/wiki?LinguaFrancaPattern
[2] http://json-ld.org/
[3] https://github.com/mhausenblas/schema-org-rdf/tree/master/tools/schema-gateway

--
Dr. Michael Hausenblas, Research Fellow
LiDRC - Linked Data Research Centre
DERI - Digital Enterprise Research Institute
NUIG - National University of Ireland, Galway
Ireland, Europe
Tel. +353 91 495730
http://linkeddata.deri.ie/
http://sw-app.org/about.html

On 16 Jun 2011, at 21:30, William Waites wrote:

> Hi Jim,
>
> The author of omnidator is a friend and collaborator, Michael
> Hausenblas, copied here.
>
> I don't see where you get N^2-N (maybe unicode problems, that's how I
> read N?-N, apologies if I've misunderstood), omnidator from what I can
> tell uses RDF as an internal representation like the bibutils use
> MODS.  So actually the way it works, I think, more or less, there are
> N serialisers and N deserialisers, so 2N just like bibutils - but it
> isn't limited to bibliographic data.
>
> I share your preference for simple unix-style pipes with simple
> programs over web interfaces in general, and you can get that just by
> installing the different necessary I/O plugins and then using rdfpipe.
>
> An interesting extension to this for bibliographic data (as well as
> other specific domains) would be to use FuXi (which also can behave as
> a simple command line filter) that could take resources expressed
> using one vocabulary and, given some domain-specific rules, translate
> them to some other vocabulary. This way you could go between
> book-as-expressed-in-schema.org to book-as-expressed-in-bibo or
> book-as-expressed-in-bibtex-rdf.
>
> To the extent that bibjson can be aligned with json-ld using a
> suitable vocabulary a lot of this can come along for free. To
> the extent that we need or want bibtex/mods/marc21/etc i/o it is
> then just as simple as writing the necessary parsers/serialisers,
> and we could just use omnidator or a only slightly modified
> variant.
>
> Cheers,
> -w
>
> * [2011-06-16 09:17:30 -0700] Jim Pitman <pitman at stat.Berkeley.EDU>  
> écrit:
>
> ] http://omnidator.appspot.com/
> ]
> ] omnidator takes whatever kind of data format that uses Schema.org  
> terms and turns it into any other kind of data format.
> ]
> ] Compare with http://www.scripps.edu/~cdputnam/software/bibutils/
> ]
> ] The bibutils program set interconverts between various  
> bibliography formats using a common MODS-format XML intermediate.  
> For example, one can convert RIS-format files to Bibtex by doing two  
> transformations: RIS->MODS->Bibtex. By using a common intermediate  
> for N formats, only 2N programs are required and not N²-N. These  
> programs operate on the command line and are styled after standard  
> UNIX-like filters.
> ]
> ] We need a strategy/architecture for managing conversions to/from  
> BibJSON, presumably a webservice like omnidator, and some community  
> development to support this over time.
> ] Thoughts about how to achieve this?
> ]
> ] --Jim
> ]
> ] _______________________________________________
> ] Bibjson-dev mailing list
> ] Bibjson-dev at lists.okfn.org
> ] http://lists.okfn.org/mailman/listinfo/bibjson-dev
>
> -- 
> William Waites                <mailto:ww at styx.org>
> http://river.styx.org/ww/        <sip:ww at styx.org>
> F4B3 39BF E775 CF42 0BAB  3DF0 BE40 A6DF B06F FD45