> Because of this behavior, IS NULL and IS NOT NULL do not always return inverse results for row-valued expressions; in particular, a row-valued expression that contains both null and non-null fields will return false for both tests.
Using PL/SQL language, I saw a strange behavior using "EXECUTE(sql_command) INTO", then "IF rec IS NOT NULL THEN" statement.
It seems that record content infers with existence test of whole record.
You can see in attached file the possible bug in action.
Should I use "IF NOT FOUND" syntax? Is it more reliable?
tl/dr; rec IS DISTINCT FROM NULL
This does not seem like a bug.
You're reported what does happen but not what you expect to happen and why.
As Pavel points out the docs for "Obtaining the Result Status" (pl/pgsql) make an effort to point out:
"Other PL/pgSQL statements do not change the state of FOUND. Note in particular that EXECUTE changes the output of GET DIAGNOSTICS, but does not change FOUND."
You could also try:
NOT (rec IS NULL)
which is the a better way to determine whether a composite record is absent/present.
Even then that only works if at least one column of the record is guaranteed to be not null. See the docs for more details: