Improper addition of NaN/Infinity recognition to float8in()
authorTom Lane
Sun, 24 Jan 1999 00:12:59 +0000 (00:12 +0000)
committerTom Lane
Sun, 24 Jan 1999 00:12:59 +0000 (00:12 +0000)
commit157349e3afdb5282cc177e53f4c435ab343b113f
treee6063c9ee448fe063224d09dac40aa14f4a1d061
parent598a4e15dd469e41d69d6588bac1c45bb893ecee
Improper addition of NaN/Infinity recognition to float8in()
was causing it not to detect out-of-range float values, as evidenced by
failure of float8 regression test.  I corrected that logic and also
modified expected float8 results to account for new error message
generated for out-of-range inputs.
src/backend/utils/adt/float.c
src/test/regress/expected/float8.out