[
https://issues.apache.org/jira/browse/OODT-371?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=13190527#comment-13190527
]
Luca Cinquini commented on OODT-371:
------------------------------------
Hi Chris,
poor choice of words on my part... indeed, it is not a bug, because it serves a
very valuable purpose. Fact is though that the use of ',' causes metadata
fields that contain a comma to be treated as multiple values, when instead they
should be kept intact. That's why the opendapps module, which often deals with
metadata fields including a comma, overrides the default delimiter of PathUtils
to be a '&', which is a far less frequently used character in metadata fields.
Other characters would work too, but it has to be something that is not
mistaken for a regular expression special character, otherwise the split won't
work.
Would you agree, or have I got something totally wrong ?
thanks, Luca
> Improve the richness and consistency of metadata extracted from the THREDDS
> catalogs
> ------------------------------------------------------------------------------------
>
> Key: OODT-371
> URL: https://issues.apache.org/jira/browse/OODT-371
> Project: OODT
> Issue Type: Improvement
> Components: opendapps
> Reporter: Luca Cinquini
> Assignee: Chris A. Mattmann
> Fix For: 0.4
>
> Attachments: opendapps-asf-20120121.patch
>
>
> The main purpose of this patch is to improve the richness and consistency of
> metadata extracted from the THREDDS catalogs and OpenDAP streams, and to
> check that the required information is indeed present in the resulting OODT
> metadata profiles.
> Details on all classes affected follow:
> Profiler
> - invokes a profile-checking utility and prints out a summary of the most
> important metadata fields for quick review by the publisher
> DasMetadataExtractor
> - extracts variable names, long names and CF standard names from the opendap
> DAS stream
> ThreddsMetadataExtractor
> - stores additional metadata, such as the hostname
> - parses all types of <documentation> tags, including xlinks, and uses the
> "type" attribute to create different metadata elements
> - adds additional geospatial and temporal coverage elements
> - stores multiple THREDDS access URLs as OODT <resLocation> attributes: the
> OpenDAP URL, the THREDDS catalog URL, and the TDS HTML landing page. All
> <resLocation>s are ecoded as tuple, to store the multiple fields
> - does NOT parse tHE variable information in the THREDDS catalog, since this
> metadata is more reliably and consistently extracted from the opendap stream
> ProfileChecker
> - new utility class that checks an OODT Profile versus a list of
> required/optional elements.
> ProfileUtils
> - fixes bug that caused any metadata value containing a ',' to be split
> across multiple XML elements
> - checks for the string "null" before adding a value to the metadata
> - allow for multiple values of the same profile element to be provided in the
> configuration file, and includes them in the resulting OODT profile
> opendap.config.xml
> - updated example of configuration with new metadata fields
> General changes
> - changed level of log output in several classes so that relevant information
> stands out more
--
This message is automatically generated by JIRA.
If you think it was sent incorrectly, please contact your JIRA administrators:
https://issues.apache.org/jira/secure/ContactAdministrators!default.jspa
For more information on JIRA, see: http://www.atlassian.com/software/jira
