Indexing your metadata (Iindex)
The Iindex command (case-sensitive) of the Isite package creates an efficient index of the occurence of numbers or text within delimited "fields" or file structures as well as within the full-text of the document. This is analogous to creating a database based on document (text file) contents rather than requiring that a separate searchable database be built in a package such as Oracle or Access. For metadata that are managed as "documents" this type of indexing makes sense.
Command syntax
CNIDR Iindex, Version X.XX.XXThe basic elements you need to execute this command are:
Copyright (c) 1995, 1996, 1997, 1998, 1999 MCNC/CNIDR
Iindex [-d (X)] // Use (X) as the your_index_name for database files.
[-m (X)] // Load (X) megabytes of data at a time for indexing
// (default=1, suggest multiples of 4).
[-t (X)] // Index as files of document type (X).
[-o (X)] // Document type specific option.
(X) (Y) (...) // Index files (X), (Y), etc.
- a database or index rootname name to create (e.g. fgdcmeta)
- a set of files to index (*.xml)
- a document definition file (fgdc.fields)
- a document type (fgdc)
Iindex -d /index_dir/your_index_name -t fgdc -o fieldtype=fgdc.fields *.xmlwhere
- the indexer will create a set of index files with the rootname 'your_index_name' in the "index_dir" directory
- the document type is specified as type FGDC (-t fgdc)
- certain fields will be specified for indexing as numbers rather than as words, the default (-o fieldtype=fgdc.fields)
- all targets in the current directory ending with .xml will be indexed
-o filter="\some_dir\mp"This option will save diskspace and simplify metadata maintenance.
The Iindex command will report each document it processes. Under modern architectures, several entries will be indexed per second.