Skip to main content
Solved

varbinary(max) column using OpenESQL in .NET Core VS COBOL

  • December 18, 2025
  • 5 replies
  • 59 views
Peter Restorick
Forum|alt.badge.img+1

We currently have a native COBOL application that uses dynamic SQL (via OpenESQL) to fetch, insert and update into Microsoft SQL.

This includes varbinary(max) columns which we have working and I suspect it was based on the demo which was available from this link https://community.rocketsoftware.com/rocket-visual-cobol-28/demo-using-the-varbinary-max-column-type-from-cobol-4560. Unfortunately the link to the actual demo is no longer available.

I now have the fun and games of trying to convert this logic to .NET core Visual COBOL and I am really struggling to work out how to ‘stitch’ the 65k blocks back together into the correct format and would really appreciate some guidance on this (the same goes for splitting it into the 65k blocks to insert/update).

Is there any chance for the demo mentioned to be converted into .NET Core VC so that I can refer to it and work out what I need to do?

Thanks in advance

 

Best answer by Chris Glazier

Hi Peter, I understand now that you are using native OO COBOL.

Although there are .NET data types that would support what you are doing, the example below, using a dynamic field would handle this as well. Here is a scaled down example:

       program-id. Program1 as "testcharappend.Program1".
       data division.
       working-storage section.
       01 dynamic-field  pic x dynamic length.
       01 buffer         pic x(5) value spaces.
       01 sub-1          pic 9(3) value zeroes.
       procedure division.

           move "ABCDE" to buffer
           move buffer to dynamic-field
           move "FGHIJ" to buffer
           perform 100-append-dynamic-field
           move "KLMNO" to buffer
           perform 100-append-dynamic-field

           move spaces to buffer
           perform varying sub-1 from 1 by length of buffer
              until sub-1 > length of dynamic-field
              move dynamic-field(sub-1:length of buffer) to buffer
              display buffer
           end-perform
           goback.

       100-append-dynamic-field.

           string dynamic-field buffer into dynamic-field
           display dynamic-field
           display length of dynamic-field.

       end program Program1.

 

5 replies

Chris Glazier
Forum|alt.badge.img+3

Hi Peter,

I am not sure why you would need to convert the logic of the program as you should be able to recompile it for .NET as is and use the SQL Server ADO provider instead of ODBC. Are you having problems with this?

According to the OpenESQL docs, the max size of a column that can be used with a varbinary(max) column is 450MB.

I just tested this here, using static SQL instead of dynamic and it worked just fine and inserted a 400MB data item into my varbinary(max) column.

The following is from the docs:

SQL [TYPE] [IS] BLOB(lob-length)

Where lob-length is a value between 1 and 1073741823 expressed either as a number or a number followed by K (kilobytes), M (megabytes), or G (gigabytes).

Note: Although this SQL TYPE has a theoretical size limitation of 2G, for all practical purposes, the actual limitation is approximately 450M, which is the data size actually allocated to the application program.

Peter Restorick
Forum|alt.badge.img+1

Hi Chris,

Thanks for getting back to me. Yes I have changed my connection to use ADO instead of ODBC.

In the Native COBOL application we read the column in 65k blocks (the limit in OpenESQL) and store each of these blocks in a field pic x(65535). These are then output/appended to a character array using the withByteLengthValue option. This loop continues, appending to the character array until the read is complete.

When inserting/updating the information is extracted, in blocks, from a character array using “getValueWithSizeAtPosition”

I guess what I am looking at is what type of .NET object I can use to replace the native character array which I can interact with allowing me to continue using OpenESQL

I have tried the static SQL statements fetching the result into a byte array and that does seem to work correctly however we rely on the dynamic SQL quite heavily, via a generic program that gets used by a number of programs and to change this logic to use static statements is, unfortunately, not feasible.

Peter

Kim Hoskin
Forum|alt.badge.img+2
  • Kim HoskinModerator
  • Moderator
  • December 19, 2025

For information, Peter your colleague Neil contacted me via email about this post, we have located the missing sample files and will provide that to Neil directly.

    Chris Glazier
    Forum|alt.badge.img+3
    • Chris GlazierModerator
    • Moderator
    • Answer
    • December 19, 2025

    Hi Peter, I understand now that you are using native OO COBOL.

    Although there are .NET data types that would support what you are doing, the example below, using a dynamic field would handle this as well. Here is a scaled down example:

           program-id. Program1 as "testcharappend.Program1".
           data division.
           working-storage section.
           01 dynamic-field  pic x dynamic length.
           01 buffer         pic x(5) value spaces.
           01 sub-1          pic 9(3) value zeroes.
           procedure division.

               move "ABCDE" to buffer
               move buffer to dynamic-field
               move "FGHIJ" to buffer
               perform 100-append-dynamic-field
               move "KLMNO" to buffer
               perform 100-append-dynamic-field

               move spaces to buffer
               perform varying sub-1 from 1 by length of buffer
                  until sub-1 > length of dynamic-field
                  move dynamic-field(sub-1:length of buffer) to buffer
                  display buffer
               end-perform
               goback.

           100-append-dynamic-field.

               string dynamic-field buffer into dynamic-field
               display dynamic-field
               display length of dynamic-field.

           end program Program1.

     

    Peter Restorick
    Forum|alt.badge.img+1

    Hi Chris, 

    This looks promising thanks very much let me give it a try - just to confirm is a pic x dynamic length the equivanent of a char array in .NET terms (char[])? 

    I have just noticed in the documentation that the dynamic length clause is only supported for native COBOL?

    Peter

    Recent badge winners

    Explore all badges
    Terms & ConditionsAccessibility statement