dataformatting in nodeJS SQL interface with IBM db2a

One of the tjing I have noticed is that when yo run a SQL SELECT * mytable
the interface returns the resultset in a javascript object.

That is nice because one dosn't have to JSON.parse the resultset if it was
in
a plain JSON string.

However all fields (also numeric) has the typeof string. No big deal, one
just make
a loop through the resultset and change the "typeof" for each property node
(field)

The second problem is more general. DB's are seldom constructed to support a
browser-based UI which means that all alpha-fields in the DB are defined as
fixed
length fields where browser based UI's typical requires trimmed fields so
the user
is able to insert characters in the string without deleting the trailing
blanks first.

This is actually a performance killer.

First of all sending all these trailing blanks will in most case double the
transmission
in size that eventually will end up as GB of nothing passed between
different layors
but it will eventually also means processing (RegEX) in removing the
trailing blanks.

So the question is if there are a generic SQL instruction that is able to
generally remove
all the trailing blanks? Or du one have to redefine all DB tables to make
specific columns
varying?

I have tried SELECT RTRIM(*) FROM XXX and seached the net but can't find a
simple
solution.