Search

User Guide

User Guide

10.0
- 10.5 HotFix 2
- 10.5 HotFix 1
- 10.5
- 10.2 HotFix 1
- 10.2
- 10.1
- 10.0 HotFix 1

Back Next

IDX Definition

IDX Definition

This begins with the

IDX-Definition

keyword and defines a DCE Identity Index (IDX).

The IDX Definition fields are as follows:

Field	Description
NAME=	A character string that specifies the name of the IDX. This is a mandatory parameter.
COMMENT=	An optional free-form description of this IDX.
ID=	A two-letter character string used to generate the names of the actual database table that represents the IDX. Each IDX must have a unique ID. This is a mandatory parameter.
KEY-INDEX	Optional parameter to specify the name of the index file. If omitted, the value of the ID= field will be used.
KEY-INDEX-PATH=	Optional parameter to specify the location of the index file. If omitted, the index will be created in the default work directory. DCE does not support spaces in file or PATH names
IDT-NAME=	This is the name of the IDT (as defined in the File-Definition or User-Source-Table sections) that this IDX will index. This is a mandatory parameter.
AUTO-ID-FIELD=	This field is not required if loading data from a User Source Table. This is the name of a field defined in the Files Section that contains a unique record label referred to as the Record Source If no such field exists in the IDT, DCE can generate one. If DCE is being asked to generate an Id, the user can choose the name of the AUTO-ID-FIELD , however that name must be defined as a field in the Files section (if using a transform clause, this happens automatically). Refer to the AUTO-ID-NAME= section under Logical File Definition .
KEY-LOGIC=	This is a mandatory parameter describing the key-logic to be used to generate keys for the IDT. It may differ from the SEARCH-LOGIC= defined in the Search-Definition or Clustering-Definition . For more details refer to the Search Logic section.
PARTITION=(field[,length,offset[,null-field-value]]),. . .	For very large files, the key may not be selective enough (retrieves too many records). This option instructs DCE to build a concatenated key from the Key-Field (defined by KEY-LOGIC= ) and up to five fields/sub-fields taken from the record. The field , length and offset represent the field name, number of bytes to concatenate to the key and offset into the field (starting from 0) respectively. The length and offset are optional. If omitted, the entire field is used. null-field-value is used to specify the value which represents an empty field. It defaults to spaces. Records that contain a null partition value can be ignored by specifying the NO-NULL-PARTITION option (defined below).
KEY-DATA=(field[,length,offset]),. . .	Key-data is used to append redundant information from the data record to the key generated from the key-field. Specifying key-data enables the search logic routine to access and test that data, and therefore be able to quickly reject some records from the search candidate list rather than pass them to the more expensive score logic routine. It is a performance option. Key-data uses the same syntax as the PARTITION= parameter to specify up to five fields, or parts thereof, to be appended to the key. The Key-Score-Logic module can examine this data and return a result without the need to read the IDT record. If key-scoring rejects records at this early stage, the scoring overhead is reduced and the clustering’s performance is improved. See also FULL-KEY-DATA option below.
OPTIONS=	This is used to specify various options to control the keys and data stored in the IDX. ALT - Generate and store multiple keys for each record. This is the default value. When disabled only the first key returned by SSA-NAME3 is stored in the IDX. FULL-KEY-DATA - Store all IDT fields in the IDX. The data is stored in uncompressed form unless you specify the Compress-Key-Data option. This option is an alternative to specifying KEY-DATA (see above). COMPRESS-KEY-DATA(n) - Store all IDT fields in compressed form using a fixed record length of n bytes. n cannot exceed (256 - PartitionLength - KeyLength - 4). COMPRESS-KEY-DATA implies FULL-KEY-DATA . COMPRESS-METHOD(n) Methods are: Method 0 (the default) will store overflow data as multiple adjacent index records. This improves performance as it takes advantage of the locality of reference in the DBMS’ cache. Method 1 will truncate compressed Key Data if its size exceeds the limit defined by the COMPRESS-KEY-DATA option. This will cause additional I/O to access the data record, when necessary. NO-NULL-FIELD do not store records that contain a Null-Field, as defined in the Key-Logic section. NO-NULL-KEY do not store Null-Keys. A Null-Key is defined in the Key-Logic section. NO-NULL-PARTITION do not store keys that contain a Null-Partition value, as defined by the PARTITION= keyword.

Watch

Comments

0 COMMENTS

We’d like to hear from you! Log in to comment.