The Dgidx program indexes the tagged Endeca records that were prepared by Forge, and creates the proprietary indices for the Endeca MDEX Engine.

The usage of Dgidx is as follows:

dgidx [-qv] [--flags] <data export file> <output db_prefix>

where <db_prefix> specifies the path to the directory, and the prefix used for the files in your Guided Search application.

Dgidx supports the following flags:

Flag

Description

-q

Quiet mode.

-v

Verbose mode.

--compoundDimSearch

Enable compound dimension search for the application. Use of this option increases indexing time. However, if this option is not enabled at index time, compound dimension results (multiple-dimension-value results) are not returned by the MDEX Engine.

--cov

Compute and report coverage statistics for dimensions and properties.

--diacritic-folding

Ignore character accents when indexing text. For details about how characters with diacritical marks are mapped to their ASCII equivalents, see the MDEX Engine Developer's Guide.

--help

Print the help message and exit.

--lang <lang-id>-u-<collation>

Indexes documents as being in the language specified by <lang-id>, and the flag also specifies an optional collation order in the <collation> portion of the argument. If unspecified, the default for <lang-id> is en (US English). For details about using international languages, see the MDEX Engine Developer's Guide.

--nostrictattrs

Disable strict attribute checking. Allows records to retain property values for properties with no property (or PROP_REF element) defined in the XML configuration file, and in the Properties view of Developer Studio.

--numbins <num>

Limit the number of records that Dgidx reads.

--out <stdout/stderr file>

Specify file path to which stdout/stderr should be remapped (the default is to use default stdout/stderr for the process).

--sort <spec>

Specify a default sort specification for the data set. The format of <spec> is (including the quotation marks):

"key|dir"

where key is the name of a property or dimension on which to sort and dir is either asc for ascending or desc for descending (if not specified, the order will be ascending).

key can also be a geocode property, as in this example:

"Location(43,73)|desc"

You can specify multiple sort keys in the format:

"key_1[|dir_1]||key_2[|dir_2]||...||key_n[|dir_n]"

If you specify multiple sort keys, the records are sorted by the first sort key, with ties being resolved by the second sort key, whose ties are resolved by the third sort key, and so on.

Note that if you are using the Oracle Endeca Application Controller (EAC) to control your environment, you must omit the quotation marks from the --sort flag. Instead, use the following syntax:

--sort key_1|dir_1||key_2|dir_2||...||key_n|dir_n

--spellmode <mode>

Specify the spelling correction mode for the application. Supported modes are:

--spellnum

In spelling modes that enable the espell module, include non-word terms (numbers, symbols, and so on) in the espell dictionary. By default, such terms are not included.

--stemming-updates <file>

Specify an optional XML file of stemming updates to apply to a default stemming dictionary. See the MDEX Engine Developer's Guide for XML examples and file name requirements.

--threads <num>

Specify the number of sorting threads to use for the multi-threaded portion of the indexing process. The default is 1. If this flag is not specified, or if 1 is specified for it, Dgidx uses one sorting thread. If the specified value is greater than 1, Dgidx uses the specified number of threads to sort data.

Note that Dgidx runs in multithreaded mode by default. In addition to the number of sorting threads that you can control with the --threads flag, Dgidx may use additional maintenance threads that run in the background by default, and are not used for sorting data.

To improve indexing performance, Oracle recommends increasing the number of sorting threads. In deployments where a dedicated server is used for indexing the Endeca application, allocate as many threads as your server allows to the Dgidx sorting process.

For best performance, the number of sorting threads specified should correlate with the number of cores on the server. Since sorting is only part of the indexing process, using N sorting threads does not speed up Dgidx by N times.

--version

Print version information and exit.


Copyright © Legal Notices