| Subject: | bug |
| Date: | Mon, 23 Aug 2010 09:55:58 -0400 |
| To: | bug-DBD-DB2 [...] rt.cpan.org |
| From: | Henry Turco <henrytu [...] us.ibm.com> |
Hello Praveen,
We are in the process of migrating from DB2 v8 to DB2 v9.5 on an s390x
GNU/Linux system. I installed the latest DBD module (DBD-DB2-1.79) and we
are running perl v5.8.8.
One of the problems I ran into was with LOB data. If you select a LOB
column in an outer loop and then do another query inside the loop, the LOB
ID was getting invalidated. It worked for the first row fetched, then all
subsequent rows were messed up (i.e. the LOB data was coming back
undefined but the other non-LOB columns were coming back fine.) After much
trial and error, I decided to switch all the LOB processing to work like
you had the XML processing. So I used SQLGetData() instead of
SQLGetLength() and SQLGetSubString(). That way I eliminated using the LOB
IDs altogether. But I was wondering if that was a bad approach? It seems
to be working ok in our initial tests. But I was wondering why you didn't
do it this way from the start.
static SQLRETURN bind_lob_column_helper( imp_fbh_t *fbh, SQLINTEGER
col_num ) {
switch( fbh->dbtype ) {
case SQL_CLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_BLOB:
fbh->ftype = SQL_C_BINARY;
break;
case SQL_DBCLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_XML:
fbh->ftype = SQL_C_BINARY;
break;
}
fbh->rlen = fbh->bufferSize = fbh->dsize = 0;
return SQL_SUCCESS;
}
in static int dbd_describe()
I removed this:
else if(SQL_BLOB == fbh->dbtype ||
SQL_CLOB == fbh->dbtype ||
SQL_DBCLOB == fbh->dbtype) {
fbh->ftype = fbh->dbtype;
fbh->rlen = bufferSizeRequired = fbh->dsize =
fbh->prec;
}
else if (SQL_XML == fbh->dbtype) {
fbh->ftype = SQL_C_BINARY;
fbh->rlen = bufferSizeRequired = fbh->dsize = -1;
}
and replaced it with this:
else if(SQL_CLOB == fbh->dbtype||
SQL_DBCLOB == fbh->dbtype) {
fbh->ftype = SQL_C_CHAR;
fbh->rlen = bufferSizeRequired = fbh->dsize = -1;
}
else if (SQL_XML == fbh->dbtype ||
SQL_BLOB == fbh->dbtype) {
fbh->ftype = SQL_C_BINARY;
fbh->rlen = bufferSizeRequired = fbh->dsize = -1;
}
static SQLRETURN get_lob_length( imp_fbh_t *fbh, SQLINTEGER col_num,
SQLHANDLE hdbc ){
SQLHANDLE new_hstmt;
SQLRETURN rc;
if( fbh->loc_ind == SQL_NULL_DATA ) { /* If column value is NULL then
set rlen to -1 and return SQL_SUCCESS */
fbh->rlen = -1;
return SQL_SUCCESS;
}
switch( fbh->dbtype ) {
case SQL_CLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_BLOB:
fbh->ftype = SQL_C_BINARY;
break;
case SQL_DBCLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_XML:
fbh->ftype = SQL_C_BINARY;
}
rc = SQLGetData( fbh->imp_sth->phstmt,
col_num,
fbh->ftype,
NULL,
0,
&fbh->rlen );
return rc;
}
static SQLRETURN get_lob_data( imp_fbh_t *fbh, SQLINTEGER col_num,
SQLHANDLE hdbc ){
SQLRETURN rc;
SQLINTEGER out_length = 0;
switch( fbh->dbtype ) {
case SQL_CLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_BLOB:
fbh->ftype = SQL_C_BINARY;
break;
case SQL_DBCLOB:
fbh->ftype = SQL_C_CHAR;
break;
case SQL_XML:
fbh->ftype = SQL_C_BINARY;
}
rc = SQLGetData( fbh->imp_sth->phstmt,
col_num,
fbh->ftype,
fbh->buffer,
fbh->bufferSize+1,
&out_length );
fbh->rlen = out_length;
return rc;
}
I was wondering if anybody else ran into this problem? I know it's a bit
obscure...
-Henry
Message body is not shown because it is too large.