OAINSERTINDEX
oainsertindex, when used with a native routine to add rows to a table,
updates the OMNIDEX indexes to include any key values corresponding to
the row identified by rowid. Incorporating this routine into your native
update applications lets you automatically maintain OMNIDEX indexes whenever
a row is inserted.
Syntax
oainsertindex (instance, options, status, table,
columns, buffer, rowid)
instance -- Identifies a
unique connection to an OMNIDEX environment as established
by oaconnect. Instance is a 32-bit signed integer passed by
value as it was returned by oaconnect.
options -- Is a character
string, passed by reference, that indicates the action or
actions for oainsertindex to take. Terminate the option list
with a semicolon or a null character.
status -- Indicates the success or failure
of oainsertindex. A zero status.error means a successful call to oainsertindex.
The status structure is passed by reference and contains fourteen 32-bit
signed integers, followed by a 36-character buffer.
table -- Is a character value
passed by reference, not longer than 33 bytes including a
semicolon or null terminator. Table contains the name of the
table to which you are adding the row.
columns -- Is a character
array passed by reference, not longer than 4096 bytes, that
specifies those columns from table that will contain the data
passed in buffer. Columns must contain column names separated
by commas, in the order that they are defined for the table
in the OMNIDEX environment catalog, or an asterisk ( * ) to
represent all column names.
buffer -- Is an array passed by reference
that contains the data that was added to the columns referenced in columns.
This is the data that will be indexed. Buffer must equal the combined
length of the columns referenced in columns. The data in buffer must be
in native format. If columns references two columns: one 30-byte character
column and one four-byte binary integer column, then buffer must reference
a buffer that contains 30 bytes of character data followed by four bytes
of binary integer data.
Binary data may require conversion when rows are transferred
from the server to a client. For more information about byte-ordering,
see “Byte ordering and byte boundaries”.
rowid -- The native identifying
value for the row that was added. The size and type of this
parameter depends on the database management system to which
the row was added. See your database management system’s
documentation for the correct type and size.
Options
CHAR=n -- Converts character
values to correct binary format before inserting the index
entry, where n is the byte length of the input number. The
default is 32 bytes. The alphanumeric representation of the
values is left justified and space-filled to length n.
Example
Top
|
