Re: Very strange Error in Updates

Now as suggested we are working with a UNICODE database.
With lastest dev Jdbc driver , don't work.
With and older mammoth Jdbc driver they work but the original problem return.

A java string of 30 chars can't be updated over a varchar(30) field, even with a UNICODE db.

Something in the middle has a wrong count of chars even if a database encoding is Unicode. 
Server or Interface ?

Dario.

Dario V. Fassi wrote:

Worst cases:

The receiving field is a Varchar(30) .
Sample data :

CALLE=[ENFERMERA CLEMON. B- ALTO ALBR]  len=30    : Is Updated Ok.
CALLE=[ENFERMERA CLEMON. Bº ALTO ALBR]  len=30    : Is NOT Updated
CALLE=[ENFERMERA CLEMON. Bº ALTO ALB]  len=29    : Is Updated Ok.

Dario V. Fassi wrote:

I found a very strange error/behavior in a PreparedStatement  for a simple SQL Update over a VARCHAR field.
The code is like:

   Statement stmt = con.createStatement();
   PreparedStatement pstIns = con.prepareStatement("update userid.t set calle = ? ");         ResultSet rs = stmt.executeQuery( "select calle from userid.t2" );
   while ( rs.next() ) {
               pstIns.clearParameters();
               String x = rs.getString("CALLE");
               pstIns.setString(1, x );
               int nrows = pstIns.executeUpdate();
               System.out.println( "Filas afectadas "+ nrows );
    }
         
When the parameter fill the full-length of  receiving field and has any non common character, the update throw a exception like:

java.sql.SQLException: ERROR:  value too long for type character varying(30)
   at org.postgresql.core.v2.QueryExecutorImpl.receiveErrorMessage(QueryExecutorImpl.java:420)
   at org.postgresql.core.v2.QueryExecutorImpl.processResults(QueryExecutorImpl.java:345)
   at org.postgresql.core.v2.QueryExecutorImpl.execute(QueryExecutorImpl.java:251)
   at org.postgresql.core.v2.QueryExecutorImpl.execute(QueryExecutorImpl.java:159)
   at org.postgresql.jdbc1.AbstractJdbc1Statement.execute(AbstractJdbc1Statement.java:342)
   at org.postgresql.jdbc1.AbstractJdbc1Statement.executeWithFlags(AbstractJdbc1Statement.java:290)
   at org.postgresql.jdbc1.AbstractJdbc1Statement.executeUpdate(AbstractJdbc1Statement.java:245)
   at com.sistemat.app.rsmaster.reloadable.RepImportar.PrDw_Cuentas(RepImportar.java:2471)
   at com.sistemat.app.rsmaster.reloadable.RepImportar.Importar(RepImportar.java:260)
   at com.sistemat.app.rsmaster.reloadable.RepMasterImpl.Ciclo_Replicacion(RepMasterImpl.java:955)
   at com.sistemat.app.rsmaster.reloadable.RepMasterImpl.runWorker(RepMasterImpl.java:748)
   at com.sistemat.app.rsmaster.reloadable.RepMasterImpl.run(RepMasterImpl.java:427)
   at java.lang.Thread.run(Unknown Source)

This field/value produce the Exception:   CALLE=[ENFERMERA CLEMON. B§ ALTO ALBR]

If this field is changed to:                        CALLE=[ENFERMERA CLEMON. B# ALTO ALBR]
then  the update is performed without
any problem.

Although the first value can be Inserted  with a prepared statement  without problems.

A have no explanation to this case , and any help will be appreciated.

Dario Fassi.