Neoview Character Sets Administrator's Guide (R2.3)
Table 5-1 Troubleshooting Symptoms, Causes, and Recommended Corrective Actions for Users
(continued)
Recommended Corrective ActionsProbable CausesSymptomsProblem Type
Make sure you always provide prefixes
to character string literals as instructed
in “Rules for Encoding SQL Language
Elements” (page 23).
If you cannot generate prefixes with
your character string literals, contact
your HP service provider and ask them
to resolve the problem.
This error is displayed when a
user fails to explicitly specify the
correct prefix for a character string
literal (for example, _ISO88591
or _UCS2). For example, in the
ISO88591 configuration, all
newly-created columns default to
the ISO88591 column unless the
column explicitly specifies
CHARACTER SET UCS2. If it
does, the user must also specify
the UCS2 prefix (_UCS2) for every
character string literal in the SQL
statement that is associated with
a UCS2 column. If literal prefixes
are not specified for columns that
are not the default column for the
configuration, the SQL statement
will fail and return an error.
An incompatible
character sets error
(4039) is displayed at a
client workstation.
Correct
character string
literal prefixes
not provided
All SQL metadata must be stored in
ISO88591 columns, regardless of the
Neoview character set configuration.
An attempt was made to store
character data in a metadata table
column using a UCS2 column.
An error reporting an
incorrectly-prefixed
literal is received.
Using UCS2
columns to
query SQL
metadata
In the Unicode configuration, SQL
identifiers are stored in UTF8
characters, where multibyte characters
can quickly consume limited space.
Reduce the number of characters in the
SQL identifier or use 7-bit ASCII
characters instead of multibyte
characters.
an SQL identifier exceeded the
128-byte limit.
An error reporting that
an SQL identifier is too
long is received.
SQL identifier
too long
Remove the SQL string function that
results in the truncated characters or
do not use that string function with
those characters. Perform the string
manipulation from the application
side, not the SQL side.
Character strings retrieved from
the Neoview database can be
truncated or return a translation
error when:
• In the SJIS configuration,
character truncation can occur
in SQL string functions when
a SJIS character is using two
bytes and the string function
stops after the first byte.
• In the Unicode configuration,
character truncation can occur
in SQL string functions when
surrogate characters are used.
Translation errors can occur
when invalid character data is
received from an application.
• When a UTF8 character is
using more than one byte and
the string function stops before
the last byte of the character.
Symptoms can be:
• A retrieved
character string
function has
truncated characters
on byte boundaries
instead of character
boundaries, cutting
the characters in
half
• A character
translation error
caused by an
invalid or illegal
character
During character
translation, the
detection of invalid
character output is
treated as a translation
error. Invalid
characters are not
defined in the client
and server mapping
tables and so cannot be
successfully mapped.
Truncated
characters in
SQL string
functions
36 Troubleshooting Guidelines for Neoview Character Sets Users