Bad data in JDBC ResultSet

[Keith_McCully@xxxxxxxxxxxxxxx]

--




I have written a generic class that takes any iSeries Physical File in the QSYS
hierarchy and converts it to a csv file in the IFS.

I use the ResultSetMetaData to determine the data type for each column eg. CHAR,
INTEGER, DECIMAL etc. Based on this, I use the appropriate ResultSet retrieval
method eg. getBigDecimal for DECIMAL.

If the row data is good then everything runs fine.
--

 However, problems occur when
a decimal column contains non-numeric data. Given that the file could originate
from any one of many 3rd parties, I have no control over the validity of the
data.  Also it is neither possible or desirable to run a data cleansing process
up front - the conversion class must handle the bad data at column level.

I've tried :

BigDecimal  bd = rs.getBigDecimal(j) //- rs is ResultSet and j column index.

followed by catch SQLException.

Case 1. DECIMAL length 4 scale 2.
File Data = Hex D5E8D5 (YNY)
Is caught by above and handled OK by setting to zero.

Case 2. DECIMAL length 8 scale 2.
File Data = Hex 4040404040 (Blank)
No exception thrown.

I am looking for a test for case 2 that can identify blank so I can set the
result  to zero.

Any suggestions much appreciated.

Keith
--