Re: OCESQL and Data Types

[James K. Lowden]

On Fri, 17 Feb 2023 20:39:04 -0600
Kevin Monceaux <Kevin@RawFedDogs.net> wrote:

> The one in particular I'm wondering about at the moment is how to
> handle PostgreSQL unlimited character varying fields in COBOL.

Hi Kevin, 

Can you give me a sense of the kind of information stored in that
column?  If it's what it should be -- an atomic datum -- then it's
misdefined in the database, and the question of how to handle it is a
management problem, not a technical one (because the display will
truncate the data arbitrarily).  

If it's JSON or XML, that's really a 1NF problem: a database design
error resulting in a technical parsing hurdle.  If it's an inseparable
BLOB, like a PDF, then it has to be treated as a string of bits.  

Very large fields are unusual in databases, and present unusual
difficulties when they crop up.  In a lower-level interface like ODBC,
there are cursor-like operations that read/write partial columns.  I
don't know of similar functionality in any ESQL system.  

--jkl