database toolbox: SQL Query on datatype "image" is limited to 8000 1Byte values

In matlab Version 7.0 (2016b) it was possible to make a query on Image datatype in SQL for more then 8000 1Byte values. Now in matlab Version 7.1 (2017a) the queryresult is limited to the first 8000 values. In ver7.0 the result was int8 in ver7.1 the result is uint8

For example there is a database called "LocalDB" with a table called "dbo.Test" and a field called "imagedatafield" with datatype image and one column with 11000 Byte values

code with matlab ver 7.0

setdbprefs('DataReturnFormat', 'structure');

conn = database('LocalDB','','');

SqlStr = ['SELECT imagedatafield FROM dbo.Test];

result = fetch(conn,SqlStr);

result.imagedatafield

ans =

java.lang.Object[][]:

[11000×1 int8]

code with matlab ver 7.1 (the same code on the same database as with ver 7.0)

setdbprefs('DataReturnFormat', 'structure'); conn = database('LocalDB','',''); SqlStr = ['SELECT imagedatafield FROM dbo.Test]; result = fetch(conn,SqlStr); result.imagedatafield

ans =

cell

[8000×1 uint8]

Is there a way to get all data of a datatype "Image" and not just the first 8000 ??

Answer 1

[Anoop Somashekar]

From 2017a onwards, JDBC-ODBC bridge interface was migrated to Native-ODBC interface. Native-ODBC interface is not equipped to deal with long data-types such as BLOB, TEXT, NVARCHAR(MAX), VARCHAR(MAX), IMAGE etc.There are certain technical limitations to support these data types in Native-ODBC interface therefore it was decided to limit the number of bytes per column to 8000 bytes for long types (including image types). This is a known limitation in Native-ODBC interface.

[Georgio Mazarin]

Anoop Somashekar, okay Native-ODBC interface can't. Do you now which ODBC is able to? Any ideas how do we load large data amounts from blob cols? p.s. I'm use varbinary(max) instead of image, effect is the same.

[Georgio Mazarin]

Solved, JDBC can.

[Joe Kleespies]

Georgio, can you describe how you were able to solve this problem with JDBC? Using the JDBC driver still limits the columns to 8000 bytes for me.