On Tue, Feb 2, 2021 at 12:16 PM Alan Cassidy <cfuture@xxxxxxxxxxx> wrote:
Since IBM is apparently making everything accessible using SQL...
Sometimes it seems that way, but you happened to ask for one thing
which is... not *inaccessible* but very much not straightforwardly
accessible.
Is there an IBM table or table function to get access path information
for an LF?
The link from Kevin does a pretty decent job describing the situation,
and why you have to use those obscure files to get what you're after.
Personally, I think this is one area where IBM should do *something*
to improve the situation. I am not completely sure it is too
technically challenging to deconstruct LFs into corresponding SQL
objects (chiefly indexes and views, but maybe others). If it's
feasible, then I don't see why they don't just do it, and then catalog
everything appropriately.
Now, I would not be surprised in the slightest if it turns out it's
not feasible to completely "map" LFs to SQL objects, because LFs can
be pretty complicated. But in that case, there are the workarounds
described in the article Kevin linked to. And if someone can come up
with those workarounds, then surely IBM can wrap those up in
documented, well-named table functions (or whatever).
Personally, I think it would be worth deconstructing and cataloging
the straightforward aspects of LFs, even if they have to punt on the
more esoteric aspects of LFs.
John Y.
midrange.com
