Build indexes

Syntax

Build indexes {file-name}

Description

This command rebuilds all the indexes for the specified file which have been dropped with the Drop indexes command. Drop indexes deletes all the indexes for the specified file apart from the sequence number index. Build indexes checks that all the indexes defined in the file class actually exist in the data file and builds those which are not there. This command does not build any indexes which already exist even if they are in a damaged state.

If the specified file name does not include a data file name as part of the notation, the default data file for that file is assumed. If the file is closed or memory-only, the command does not execute and returns flag false.

If you are not running in single user mode, this command automatically tests that only one user is using the data file (the command fails with the flag false if this is not true), and further users are prevented from logging onto the data until the command completes.

If a working message with a count is open while the command is executing, the count will be incremented at regular intervals. The command may take a long time to execute, and it is not possible to cancel execution even if a working message with cancel box is open.

The flag is set if at least one index is successfully rebuilt. Note that the command is not reversible.

Example

Do not flush data 
Drop indexes {fCustomers}
Repeat 
  Working message {Building indexes...}
  Build indexes {fCustomers}
Until flag true