smallcat1.gif (979 bytes) You are here: Cheshire II Home > Documentation

Cheshire II System

This page is intended as a directory of information about Cheshire II set-up and operation. Information on configuration files has its own page. The following list briefly describes the commands that are created in the bin directory when the system is built and their general purpose. There are links to more detailed information on the various commands. There are also pages available on the Cheshire Explain/Metadata services and on Datastore and virtual databases, and bitmapped indexes.

In addition the CHANGES file often includes detailed descriptions of new feature in the order they were added to the system.

Information on setting up the sytem and use of these commands in building databases and IR services is discussed in the documents contributed by Rob Sanderson of Liverpool University (now at Los Alamos).

batch_load: a utility to load unloaded batch indexes due to indexing failure.
buildassoc: a utility to build an associator file from an SGML file.
cheshire2: Main client programs for both batch(script) and interactive search operations. This page includes all of the internal scriptable TCL commands for search, including information search ranking options
cheshire_load: A utility to build a Datastore file from an SGML/XML file.
countdb: a utility to count the number of items in an index and produce a frequency count.
delete_recs: a utility to (logically) delete a record (or records) and all of its index entries.
db_archive db_checkpoint db_deadlock db_dump db_load db_recover db_stat: utility programs associated with the DBOpen system used in Cheshire II indexes. See the manual pages in the doc directory.
dtd_parser: utility program to test DTDs for parsibility -- outputs a description of the DTD contents.
dumpcomp: utility program to print information about components extracted from records in a database
dumpdb: utility program to print the contents of an index file.
dumpds: Utility program to print information about the contents of a DataStore file.
dumppost: utility program to print the contents and postings of an index file.
dumprecs: utility program to print the contents of an SGML data file or a single record from the file. NOTE: for printing DataStore files use read_datastore instead of this program.
getnumrecs: utility program to report the highest record id number in a data file.
highpost: utility program to print all entries in an index with more than some specified postings.
in_test: Test version of the indexing program (with voluminous output, useful for tracking indexing data problems.)
index_cheshire: The main index creation program. It is suggested that the batch (-b) flag be used for best performance. (NOTE: use of the batch flag requires sufficient work space on the disks where the index will be located to hold the index contents TWICE -- but indexing is MUCH faster than not using the flag)
index_clusters: If cluster files are generated during the index_cheshire run, this program is used to finish generation of the cluster files and indexes.
index_vectors: If vector index files are specified in the config file, they are generated during the index_cheshire run, then this program is used subsequently to finish generation of the vector indexes.
jserver: The Non-Z39.50 server. The server is configured by a combination of a "server.init" and the database configuration files for each database being served (see configfiles.html). This version of the server was created to interact with Java client interfaces
marc2sgml: Conversion utility to converting MARC records to SGML (using the Berkeley DTD) -- currently USMARC12.DTD. Various versions of the USMARC DTD are available in the docs directory.
parser: Utility program to test parsing of SGML/XML documents -- outputs a description of the document contents.
read_datastore: A utility to read the contents of a Datastore file.
sgml2marc: Conversion utility to convert SGML conforming to the Berkeley USMARC DTD to MARC records.
BuildContDB: Conversion utility to convert FILECONT files to a BerkeleyDB database.
staffcheshire: An X windows version of the client with combined elements of the cheshire client and server, as well as other utility code. This program is intended to be used as a scriptable program for managing the system. (see scripts in the config-gui directory as examples)
test_config: A utility program to test the validity and parsibility of database configuration files.
testsrch2: Simple line-oriented command driven interface to the search engine. useful for testing, and can also be used in non-client scripting languages (such as perl) as a way to access Cheshire databases. Note: the program is called "tstsearch" in the Windows version.
webcheshire: Combined elements of the cheshire client and server, used as a scriptable CGI driver (see scripts directory for samples)
zserver: The Z39.50 server. The server is configured by a combination of a "server.init" and the database configuration files for each database being served (see configfiles.html)
ztcl: A version of the client software without X Window support. Can be used for utilities or as a line-by-line interface (for those that know a bit about Tcl/Tk).

Typing the name alone of any of the utility programs will show the usage and required arguments for the command.

This page has been translated into Spanish language by Maria Ramos from Webhostinghub.com/support/edu.

Ray R. Larson
Associate Professor
School of Information Management and Systems
University of California, Berkeley
Berkeley, California 94720-4600