The following documentation comment has been logged on the website:
Page: https://www.postgresql.org/docs/18/datatype-character.html
Description:
The description of BPCHAR in section 8.3. Character Types is misleading and
incomplete.
The first problem is that, contrary to table 8.4, BPCHAR is not actually
blank-trimmed. The wording "as-if-blank-trimmed" or "blank-ignoring" may be
better suited here. The following query explains the problem:
SELECT bpchar_val, length(bpchar_val) as bplen, concat('[', bpchar_val, ']')
as bpbrack
, varchar_val, length(varchar_val) as vclen, concat('[',
varchar_val, ']') as vcbrack
FROM (VALUES
('abc '::bpchar, 'abc '::varchar),
('abc '::bpchar, 'abc'::varchar),
('abc'::bpchar, 'abc '::varchar),
('abc'::bpchar, 'abc'::varchar))
AS bpchar_test (bpchar_val, varchar_val)
WHERE bpchar_val = varchar_val;
As can be seen, it returns four rows (so, it indeed treats values that
differ only in number of trailing spaces as equal, regardless of the type on
the other side, and the length of bpchar_val is always 3 in all 4 rows -
trailing spaces are ignored in comparison. But the bpbrack column shows that
actual value is NOT trimmed: two of four rows have '[abc ]' value (with 3
spaces before the closing bracket).
The example above also illustrate the need of more detailed explanation of
how BPCHAR actually behaves in different situation, particularly, in which
contexts trailing blanks are ignored (comparison, length, what else?) and in
which they are still significant (displaying, client interfaces,
concatenation, what else?).
The second problem is that 'Tip' before the example 8.1 mentions only three
types, also in a misleading way: 'There is no performance difference among
these three types' - as if there were only 3, not 4 distinct types.
