Creating a Pears Database Description Configuration File Based on an Existing Newton .dsc File

Contents

Introduction
General Guidelines
[Bartlett] Section
[LockServer] Section
[DB] Section
[Handleinput_record_type] Section
[index_definition] Sections
   Examples from the DC Database Framework
   Examples from the MARCCat Database Framework

Introduction

You must have a Pears database description configuration file to convert a Newton database to a Pears database. This document provides some guidelines for creating a Pears database description configuration file with the same characteristics as an existing Newton database description (.dsc) file.

You may also find these guidelines useful when you are creating a database description file for a new Pears database and you want to base the database on an existing Newton database.

Since this document supplements the Pears Database Description Configuration File document, it organizes these guidelines according to the structure of the database description configuration file.

General Guidelines

• The sample Pears databases available on the SiteSearch FTP site include database description configuration files for each database. You may find these useful as examples.
• As with WebZ configuration files:
  • use the # character to comment out all or part of a line
  • use #include to incorporate the contents from another configuration file into a file

Return to Contents

[Bartlett] Section

• See the [Bartlett] section in Pears Database Description Configuration File.

Return to Contents

[LockServer] Section

• See the [LockServer] section in Pears Database Description Configuration File.

Return to Contents

[DB] Section

• The input record type for converting a Newton database to a Pears database is DB. For a new Pears database or for subsequently adding new records to a converted Pears database, use the record handler appropriate to the type of input records you wish to add to the database.
• The Name variable is equivalent to the database identification in the .dsc file.
• The blocksize and filename variables take the place of the database file definitions in the .dsc file.
• The RecordIDIndex variable denotes the index for the field that contains the unique identifier in each record. This is equivalent to the Newton internalaccnindx value.

Return to Contents

[Handleinput_record_type] Section

• See the [Handleinput_record_type] section in Pears Database Description Configuration File and Pears Record Handlers.

Return to Contents

[index_definition] Sections

• See Pears-Newton Indexing Routine Comparison to create Pears index definitions that duplicate Newton indexing routines.
• You do not need to define sparse indexes in Pears. Pears creates the equivalent of a sparse index automatically without any action on your part.
• You can define term adjacency for any keyword index by adding this variable to its [index_definition] section:

  OccurrenceRoutine = ORG.oclc.pears.Bartlett.wordfield

• You can define plural indexes for keyword indexes by using ORG.oclc.pears.IndexRoutines.PluralWords indexing routine.
• You can define a restrictor index by using ORG.oclc.pears.Bartlett.termrest as the indexing routine, specifying the index ID for the index for which you want to create a restrictor index, and providing a list of the restrictor values. You do not need to specify the size of the bitmaps that hold all restrictor values (Newton restrictsize value). See Examples from the MARCCat Database Framework.
• If you make a Newton database available to patrons through the WebZ interface, keep the same index numbers in its Pears database description configuration file as those in the existing .dsc file. Then you won't have to change the alternateID variables for the database's index definitions in its WebZ database configuration file.
• Place each tagpath reference (which denotes a field to be indexed) on a separate line. You can use either a numeral in each tagpath variable (tagpath1, tagpath2, and so on), or an asterisk (tagpath*), but you cannot mix numerals and asterisks in the same index definition. If you use numerals, each numeral must be in sequential order and you cannot skip any numbers.
• You define stopwords in a special index definition, usually named [stopwords]. See the example below and the [stopwords] section of Pears Database Description Configuration File.

The remainder of this section provides side-by-side examples of the Pears and Newton index definitions from the Record Builder DC(2) and MARCCat database frameworks provided with the Record Builder application. Even if you don't use Record Builder, these examples compare index definitions in Pears and Newton. Look at the examples from both database frameworks; they illustrate principles that apply to many types of databases.

Examples from the DC Database Framework

• Overall keyword index, which includes all fields, with term adjacency
• Keyword index for the DC:Type field with term adjacency
• Phrase index for the DC:Description field with handling for plural endings
• Keyword index for the DC:Identifier field with term adjacency
• Phrase index for the DC:Title field
• Stopwords definition for all indexes

Examples from the MARCCat (MARC Catalog) Database Framework

• Keyword index for the ID number in subfield 1
• Phrase index for titles combining subfields a and b from the 245 field into a single term
• Keyword index for titles with term adjacency and handling for plural endings
• LC call number index
• Dewey decimal number index
• Publication date index
• Restrictor for the publication date index
• Language index
• Restrictor for the language index
• Format restrictor
  

Return to Contents

See Also

Converting a Newton Database to a Pears Database
Creating a New Pears Database
Pears Database Description Configuration File
Pears-Newton Indexing Routine Comparison
Pears Record Handlers
Pears Indexing Routines