Sufmeta

From SRB

Contents

Sufmeta

NAME

Sufmeta - modifies attribute-value-unit triplets of user-defined metadata information about SRB data objects.

SYNOPSIS

Sufmeta AttName AttVal [ AttUnit ] datasetName

Sufmeta -c AttName AttVal [ AttUnit ] collectionName

Sufmeta -u AttName AttVal [ AttUnit ] userName@domainName

Sufmeta -r AttName AttVal [ AttUnit ] resourceName

Sufmeta [-d] -f metadataFileName separator datasetName

Sufmeta -c -f metadataFileName separator collectionName

Sufmeta -r -f metadataFileName separator resourceName

Sufmeta -u -f metadataFileName separator

Sufmeta -D -f metadataFileName separator

Sufmeta [-c|-u|-r] -Q [-z zone ] AttName Op AttVal [ AttName Op AttVal ] ...

Sufmeta datasetName|collectionName

Sufmeta -u userName@domainName

Sufmeta -r resourceName

Sufmeta -R -d AttName AttValue datasetName

Sufmeta -R -c AttName AttValue collectionName

Sufmeta -R -r AttName AttValue resourceName

Sufmeta -R -u AttName AttValue userName@domainName

Sufmeta -e [ inSRBMetadataFileName ] inSRBExtractorStyleSheet datasetName

Sufmeta -C 1 sourceCollectionName targetDataName

Sufmeta -C 2 sourceCollectionName targetCollectionName

Sufmeta -C 3 sourceDataName targetDataName

Sufmeta -C 4 sourceDataName targetCollectionName

DESCRIPTION

Sufmeta provides the facility for inserting, deleting, updating and accessing attribute-value type user-defined meta-data on SRB objects dataName, collectionName, resourceName or userName@domainName.

The first four synopses ingest AttName AttValue [AttUnit] triplets into a single dataObject, collection user or resource.

The next four synopses is used to bulk ingest metadata from metadataFileName for a single dataObject, collection, resource or user. The separator is a single character used to separate the triples in a line in metadataFileName.

The ninth synopsis is used to bulk ingest triplets into multiple dataObjects.

The tenth synopsis allows querying of the four types of srbObjects separately using upto five conditions of the form AttName Op AttVal. AttVal can be a number or a single-quoted character strings with possible wild-card characters, * and ?. The conditions supported are =, >, <, >=, <=, <>, like, not like, in, not in, between and not between.

The next three synopses show metadata triplets for a given dataName, collectionName, resourceName or userName@domainName.

The next four synopses (-R option) used to delete AttName AttValue pair metadata from dataName, collectionName, resourceName or userName@domainName.

The next synopsis (-e option) allows for (remote) extracting of metadata using a T-language template stored in the SRB by extracting the metadata either from the dataset itself or from another metadata file which is also stored in the SRB.

The next four synopses (-C option) are used to copy metadata from one SRB object/collection to another SRB object/collection.

OPTIONS

-h
display command options
-d
operate on metadata for data objects
-c
operate on metadata for collections
-u
operate on metadata for users
-r
operate on metadata for resources
-Q
defines a query operation
-f
defines bulk load option with metadata stored in a file.
-e
defines bulk load option with metadata extracted from a srbObject or extracted from the targetObject itself.
-z
zone option applies to the second synopsis and is used to perform the Sls operation on a particular zone. If ALL is used for zone then every zone known to the current zone will be queried. zone can have wild cards.

Examples

The first four synopses [-dcur] are used to ingest metadata for srbData, srbCollections srbUsers and srbResources respectively.

The next four synopses [-dcur -f] are used to ingest bulk metadata for srbData, srbCollections,srbUsers and srbResources respectively.

The next synopsis [-D -f]are used to ingest bulk metadata for srbData for many datasets.

The next four synopses [-dcur -Q] are used to query metadata for srbData and srbCollections respectively.

The last three synopses are used to get all metadata values for srbData, srbCollections, srbUsers and srbResources. Use '*' and '?' for wildcards; assumes an and between conditions

Sample Usage: Sufmeta alpha 200 foo.dat

Sufmeta -c alpha 200 myColl

Sufmeta -c beta bar myColl

Sufmeta -c beta 200 myColl

Sufmeta -Q alpha = 200

Sufmeta -Q -c alpha = 200 beta = bar

Sufmeta -Q -c alpha = 200 beta like '*a*'

Sufmeta -Q -c alpha '>' 100 beta like '*a*'

Sufmeta foo.dat

Sufmeta -f mdFile '|' ticket123.dat

Sufmeta -c -f mdFile '|' ticket123.dat

where mdFile is of the form: 
  ticketnum|es2345  
  car name|ford escort  
  driver name|john q. public  
  numplate|123455  
  speed|89|kmph|overspeed  
  fine|55|dollars

Sufmeta -D -f mdFileMult '|'

where mdFileMultis of the form:  
  SETMINMETADATANUM=GIVENPERDATA  
  /home/srb/rajatest|Sput.c|11|ticketnum|es2345  
  /home/srb/rajatest|Sput.c|12|car name|ford escort
  /home/srb/rajatest|Sput.c|13|driver name|john q. public  
  /home/srb/rajatest|Srm.c|11|numplate|123455  
  /home/srb/rajatest|Srm.c|12|speed|89|kmph|overspeed  

The value for SETMINMETADATANUM can be GIVENPERDATA in which case, the value for the metadatanum field is to be given in the third value in each line.

The value for SETMINMETADATANUM can be GETFROMMCAT in which case the value is generated internally by the SRB system. In this case the metadatanum field in the metadatfile should not be included. The metadatanum field is needed only when using GIVENPERDATA option for the SETMINMETADATANUM parameter.

The value for SETMINMETADATANUM can be a number (say 0), then the value is used for the first triplet for each given data object and incremented for subsequent metadata triplets for that object. This is useful when you know that the metadata is being freshly loaded and you can start at 0 or some constant number. This will speedup the load compared to the GETFROMMCAT option.

SEE ALSO

Smeta