Hi, Thanks to all for your valuable input! Metadata is such a controversial, contradictionary topic...
I think the most reusable parts of such a harvester/search engine approach are 1. the CSW harvester 2. the search indexer (like Lucene) 3. a simple metadata editor My metadata structure is a simple table with much less attributes than ISO 19115/19119/ebRIM, but with practical fields like "internal file path to own share". If anybody for example knows a software snippet/tool to access the documentation column in a GDB_Items table, pls. notify me. Yours, Stefan 2013/5/31 Christian Willmes <[email protected]>: > Am 30.05.2013 22:21, schrieb Alex Mandel: >> On 05/30/2013 01:12 PM, Stefan Keller wrote: >>> Hi all >>> >>> One of the first thing governement (should) do with it's geodata is to >>> get an overview. And one of the almost immediate proposals you get is >>> to do it ISO19115/ISO19119 and OGC's CSW 2.0. >>> >>> But like Ed Parsons says "the best policy is not to have a policy ..." >>> [1] and in fact nobody really likes to maintain metadata. And a >>> computer science pattern says "keep things close together which belong >>> together. >>> >>> So what I'm looking for is a solution quite different to most >>> geocatalog SW: The user simply get's a search page. Most of the job >>> does a crawler which collects geo-metadata from internal disk shares >>> and (geo-)databases. Of course some information - like data owner, >>> data origin, data quality - need to be delivered by the data owner >>> (e.g. in a text file with same own extension or as comments inside the >>> database property). >>> >>> Did anybody implement such a thing already somewhere? >>> >>> The only thing which comes close to this is OKFN [2]. But I'm missing >>> for example a Lucene reader for geo-metadata (ideally written in >>> Python :-) [3]). >>> >>> Yours, Stefan >>> >>> [1] https://twitter.com/GIStrobl/status/334648984415109120 >>> [2] http://lists.okfn.org/pipermail/okfn-labs/2013-April/000717.html >>> [3] http://code.google.com/p/nlupy/ >>> >> >> http://pycsw.org ? (Which is in CKAN from OKFN) Another implementation >> is geonode. >> >> Enjoy, >> Alex >> >> >> > As far as I know, pyCSW is still not fully integrated in CKAN, they are > working on it, or at least have it on their todo (for quite some time > now), but what you are looking for should be already doable with CKAN + > ckanext-spatial [1]. CSW harvesting is also implemented via extension > since some time in CKAN. What is not implemented is the CSW > Transactional stuff, input of ISO19115/19119 via CKAN (API), and not via > CSW harvesting won't be exposed via CSW endpoint from CKAN. Basic > location search (BBox) is implemented via solr in CKAN. > > Cheers, > Christian > > [1] https://github.com/okfn/ckanext-spatial > > -- > Christian Willmes > AG GIS & Fernerkundung | GIS & RS Group > Geographisches Institut | Institute of Geography > Universität zu Köln | University of Cologne > Tel.: +49 (0)221 470 6234 > Fax.: +49 (0)221 470 2280 > http://www.geographie.uni-koeln.de/willmes-christian.695.en.html > http://www.sfb806.de > http://crc806db.uni-koeln.de > > > _______________________________________________ > Geowanking mailing list > [email protected] > http://geowanking.org/mailman/listinfo/geowanking_geowanking.org _______________________________________________ Geowanking mailing list [email protected] http://geowanking.org/mailman/listinfo/geowanking_geowanking.org
