1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
|
#include <cmath>
#include <pqxx/transaction>
#include "../test_helpers.hxx"
namespace
{
/// Test conversions for some floating-point type.
template<typename T> void infinity_test()
{
T inf{std::numeric_limits<T>::infinity()};
std::string inf_string;
T back_conversion;
inf_string = pqxx::to_string(inf);
pqxx::from_string(inf_string, back_conversion);
PQXX_CHECK_LESS(
T(999999999), back_conversion,
"Infinity doesn't convert back to something huge.");
inf_string = pqxx::to_string(-inf);
pqxx::from_string(inf_string, back_conversion);
PQXX_CHECK_LESS(
back_conversion, -T(999999999), "Negative infinity is broken");
}
void test_infinities()
{
infinity_test<float>();
infinity_test<double>();
infinity_test<long double>();
}
/// Reproduce bug #262: repeated float conversions break without charconv.
template<typename T> void bug_262()
{
pqxx::connection cx;
cx.prepare("stmt", "select cast($1 as float)");
pqxx::work tr{cx};
// We must use the same float type both for passing the value to the
// statement and for retrieving result of the statement execution. This is
// due to an internal stringstream being instantiated as a a parameterized
// thread-local singleton. So, there are separate stream<float>,
// stream<double>, stream<long double>, but every such instance is a
// singleton. We should use only one of them for this test.
pqxx::row row;
// Nothing bad here, select a float value.
// The stream is clear, so just fill it with the value and extract str().
row = tr.exec("SELECT 1.0").one_row();
// This works properly, but as we parse the value from the stream, the
// seeking cursor moves towards the EOF. When the inevitable EOF happens
// 'eof' flag is set in the stream and 'good' flag is unset.
row[0].as<T>();
// The second try. Select a float value again. The stream is not clean, so
// we need to put an empty string into its buffer {stream.str("");}. This
// resets the seeking cursor to 0. Then we will put the value using
// operator<<().
// ...
// ...
// OOPS. stream.str("") does not reset 'eof' flag and 'good' flag! We are
// trying to read from EOF! This is no good.
// Throws on unpatched pqxx v6.4.5
row = tr.exec("SELECT 2.0").one_row();
// We won't get here without patch. The following statements are just for
// demonstration of how are intended to work. If we
// simply just reset the stream flags properly, this would work fine.
// The most obvious patch is just explicitly stream.seekg(0).
row[0].as<T>();
row = tr.exec("SELECT 3.0").one_row();
row[0].as<T>();
}
/// Test for bug #262.
void test_bug_262()
{
bug_262<float>();
bug_262<double>();
bug_262<long double>();
}
/// Test conversion of malformed floating-point values.
void test_bad_float()
{
float x [[maybe_unused]];
PQXX_CHECK_THROWS(
x = pqxx::from_string<float>(""), pqxx::conversion_error,
"Conversion of empty string to float was not caught.");
PQXX_CHECK_THROWS(
x = pqxx::from_string<float>("Infancy"), pqxx::conversion_error,
"Misleading infinity was not caught.");
PQXX_CHECK_THROWS(
x = pqxx::from_string<float>("-Infighting"), pqxx::conversion_error,
"Misleading negative infinity was not caught.");
PQXX_CHECK_THROWS(
x = pqxx::from_string<float>("Nanny"), pqxx::conversion_error,
"Conversion of misleading NaN was not caught.");
}
template<typename T> void test_float_length(T value)
{
auto const text{pqxx::to_string(value)};
PQXX_CHECK_GREATER_EQUAL(
pqxx::size_buffer(value), std::size(text) + 1,
"Not enough buffer space for " + text + ".");
}
/// Test conversion of long float values to strings.
void test_long_float()
{
test_float_length(0.1f);
test_float_length(0.1);
test_float_length(std::numeric_limits<float>::denorm_min());
test_float_length(-std::numeric_limits<float>::denorm_min());
test_float_length(std::numeric_limits<float>::min());
test_float_length(-std::numeric_limits<float>::min());
test_float_length(std::numeric_limits<float>::max());
test_float_length(-std::numeric_limits<float>::max());
test_float_length(-std::nextafter(1.0f, 2.0f));
test_float_length(std::numeric_limits<double>::denorm_min());
test_float_length(-std::numeric_limits<double>::denorm_min());
test_float_length(std::numeric_limits<double>::min());
test_float_length(-std::numeric_limits<double>::min());
test_float_length(std::numeric_limits<double>::max());
test_float_length(-std::numeric_limits<double>::max());
test_float_length(-std::nextafter(1.0, 2.0));
test_float_length(std::numeric_limits<long double>::denorm_min());
test_float_length(-std::numeric_limits<long double>::denorm_min());
test_float_length(std::numeric_limits<long double>::min());
test_float_length(-std::numeric_limits<long double>::min());
test_float_length(std::numeric_limits<long double>::max());
test_float_length(-std::numeric_limits<long double>::max());
test_float_length(-std::nextafter(1.0L, 2.0L));
// Ahem. I'm not proud of this. We really can't assume much about the
// floating-point types, but I'd really like to try a few things to see that
// buffer sizes are in the right ballpark. So, if "double" is at least 64
// bits, check for some examples of long conversions.
if constexpr (sizeof(double) >= 8)
{
auto constexpr awkward{-2.2250738585072014e-308};
auto const text{pqxx::to_string(awkward)};
PQXX_CHECK_LESS_EQUAL(
std::size(text), 25u, text + " converted to too long a string.");
}
if constexpr (sizeof(double) <= 8)
{
auto const text{pqxx::to_string(0.99)};
PQXX_CHECK_LESS_EQUAL(
pqxx::size_buffer(0.99), 25u, text + " converted to too long a string.");
}
}
PQXX_REGISTER_TEST(test_infinities);
PQXX_REGISTER_TEST(test_bug_262);
PQXX_REGISTER_TEST(test_bad_float);
PQXX_REGISTER_TEST(test_long_float);
} // namespace
|
