[thredds] regExp limitations in datasetScan and scan

Hi all, 

Quite new to TDS catalogs and ncML.
I am running TDS 4.3.18 and Tomcat 7.0.42
My question is about the regExp attribute used in both datasetScan and 
aggregation/scan elements.
I have not been able to get TDS to match a regExp which includes sub 
directories.
  
Is this a known limitation?
  
A very simplified example:
  
This works:
  <aggregation dimName="time" type="joinExisting">
      <scan location="/Users/eric/NERACOOS/data/buoy/netcdf/B0123/realtime" 
regExp=".*met\.realtime\.nc" subdirs="true" />

But this did not:
   <aggregation dimName="time" type="joinExisting">
      <scan location="/Users/eric/NERACOOS/data/buoy/netcdf/B0123" 
regExp="realtime/.*met\.realtime\.nc" subdirs="true" />
         
threddsServlet.log
  2013-10-14T14:50:13.287 -0400 [     11715][       6] ERROR - 
thredds.server.opendap.OpendapServlet - path= /UMO/B01/UnionScan/Agg.html
  java.lang.IllegalStateException: There are no datasets in the aggregation 
DatasetCollectionManager{ 
collectionName='/Users/eric/NERACOOS/data/buoy/netcdf/B0123**/realtime/.*met\.realtime\.nc'
 recheck=null
   dir=/Users/eric/NERACOOS/data/buoy/netcdf/B0123 
filter=realtime/.*met\.realtime\.nc

I have been able to aggregate these on the time dimension across multiple 
deployments by using symbolic links from a single directory but was hoping to 
avoid that.

I've had a similar lack of results using datasetScan regExp's. Have tried 
numerous permutations of the regExp, e.g. ^…$, various escapes, etc.

Unfortunately my TDS is not yet up on our public server only on my local 
machine.


Eric Bridger
------------------------------------------------------
Gulf of Maine Research Institute -- Ocean Data Products
350 Commercial St.
Portland, ME 04101
207 228-1662  ebridger@xxxxxxxx
http://www.gmri.org/      http://www.neracoos.org/
 







  • 2013 messages navigation, sorted by:
    1. Thread
    2. Subject
    3. Author
    4. Date
    5. ↑ Table Of Contents
  • Search the thredds archives: