Skip to content

Commit 53ac49f

Browse files
Mic92Mic92
authored
Merge pull request #13741 from xokdvium/toml-timestamps
libexpr: Canonicalize TOML timestamps for toml11 > 4.0
2 parents 8e5ca78 + 320b2c7 commit 53ac49f

5 files changed

Lines changed: 158 additions & 25 deletions

File tree

‎packaging/dependencies.nix‎

Lines changed: 10 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -64,6 +64,16 @@ scope: {
6464
NIX_CFLAGS_COMPILE = "-DINITIAL_MARK_STACK_SIZE=1048576";
6565
});
6666

67+
toml11 = pkgs.toml11.overrideAttrs rec {
68+
version = "4.4.0";
69+
src = pkgs.fetchFromGitHub {
70+
owner = "ToruNiina";
71+
repo = "toml11";
72+
tag = "v${version}";
73+
hash = "sha256-sgWKYxNT22nw376ttGsTdg0AMzOwp8QH3E8mx0BZJTQ=";
74+
};
75+
};
76+
6777
# TODO Hack until https://github.com/NixOS/nixpkgs/issues/45462 is fixed.
6878
boost =
6979
(pkgs.boost.override {

‎src/libexpr/meson.build‎

Lines changed: 6 additions & 0 deletions
Original file line numberDiff line numberDiff line change
@@ -71,6 +71,12 @@ toml11 = dependency(
7171
method : 'cmake',
7272
include_type : 'system',
7373
)
74+
75+
configdata_priv.set(
76+
'HAVE_TOML11_4',
77+
toml11.version().version_compare('>= 4.0.0').to_int(),
78+
)
79+
7480
deps_other += toml11
7581

7682
config_priv_h = configure_file(

‎src/libexpr/primops/fromTOML.cc‎

Lines changed: 97 additions & 22 deletions
Original file line numberDiff line numberDiff line change
@@ -1,73 +1,140 @@
11
#include "nix/expr/primops.hh"
22
#include "nix/expr/eval-inline.hh"
33

4+
#include "expr-config-private.hh"
5+
46
#include <sstream>
57

68
#include <toml.hpp>
79

810
namespace nix {
911

12+
#if HAVE_TOML11_4
13+
14+
/**
15+
* This is what toml11 < 4.0 did when choosing the subsecond precision.
16+
* TOML 1.0.0 spec doesn't define how sub-millisecond ranges should be handled and calls it
17+
* implementation defined behavior. For a lack of a better choice we stick with what older versions
18+
* of toml11 did [1].
19+
*
20+
* [1]: https://github.com/ToruNiina/toml11/blob/dcfe39a783a94e8d52c885e5883a6fbb21529019/toml/datetime.hpp#L282
21+
*/
22+
static size_t normalizeSubsecondPrecision(toml::local_time lt)
23+
{
24+
auto millis = lt.millisecond;
25+
auto micros = lt.microsecond;
26+
auto nanos = lt.nanosecond;
27+
if (millis != 0 || micros != 0 || nanos != 0) {
28+
if (micros != 0 || nanos != 0) {
29+
if (nanos != 0)
30+
return 9;
31+
return 6;
32+
}
33+
return 3;
34+
}
35+
return 0;
36+
}
37+
38+
/**
39+
* Normalize date/time formats to serialize to the same strings as versions prior to toml11 4.0.
40+
*
41+
* Several things to consider:
42+
*
43+
* 1. Sub-millisecond range is represented the same way as in toml11 versions prior to 4.0. Precisioun is rounded
44+
* towards the next multiple of 3 or capped at 9 digits.
45+
* 2. Seconds must be specified. This may become optional in (yet unreleased) TOML 1.1.0, but 1.0.0 defined local time
46+
* in terms of RFC3339 [1].
47+
* 3. date-time separator (`t`, `T` or space ` `) is canonicalized to an upper T. This is compliant with RFC3339
48+
* [1] 5.6:
49+
* > Applications that generate this format SHOULD use upper case letters.
50+
*
51+
* [1]: https://datatracker.ietf.org/doc/html/rfc3339#section-5.6
52+
*/
53+
static void normalizeDatetimeFormat(toml::value & t)
54+
{
55+
if (t.is_local_datetime()) {
56+
auto & ldt = t.as_local_datetime();
57+
t.as_local_datetime_fmt() = {
58+
.delimiter = toml::datetime_delimiter_kind::upper_T,
59+
// https://datatracker.ietf.org/doc/html/rfc3339#section-5.6
60+
.has_seconds = true, // Mandated by TOML 1.0.0
61+
.subsecond_precision = normalizeSubsecondPrecision(ldt.time),
62+
};
63+
return;
64+
}
65+
66+
if (t.is_offset_datetime()) {
67+
auto & odt = t.as_offset_datetime();
68+
t.as_offset_datetime_fmt() = {
69+
.delimiter = toml::datetime_delimiter_kind::upper_T,
70+
// https://datatracker.ietf.org/doc/html/rfc3339#section-5.6
71+
.has_seconds = true, // Mandated by TOML 1.0.0
72+
.subsecond_precision = normalizeSubsecondPrecision(odt.time),
73+
};
74+
return;
75+
}
76+
77+
if (t.is_local_time()) {
78+
auto & lt = t.as_local_time();
79+
t.as_local_time_fmt() = {
80+
.has_seconds = true, // Mandated by TOML 1.0.0
81+
.subsecond_precision = normalizeSubsecondPrecision(lt),
82+
};
83+
return;
84+
}
85+
}
86+
87+
#endif
88+
1089
static void prim_fromTOML(EvalState & state, const PosIdx pos, Value ** args, Value & val)
1190
{
1291
auto toml = state.forceStringNoCtx(*args[0], pos, "while evaluating the argument passed to builtins.fromTOML");
1392

1493
std::istringstream tomlStream(std::string{toml});
1594

16-
std::function<void(Value &, toml::value)> visit;
17-
18-
visit = [&](Value & v, toml::value t) {
95+
auto visit = [&](auto & self, Value & v, toml::value t) -> void {
1996
switch (t.type()) {
2097
case toml::value_t::table: {
2198
auto table = toml::get<toml::table>(t);
22-
23-
size_t size = 0;
24-
for (auto & i : table) {
25-
(void) i;
26-
size++;
27-
}
28-
29-
auto attrs = state.buildBindings(size);
99+
auto attrs = state.buildBindings(table.size());
30100

31101
for (auto & elem : table) {
32102
forceNoNullByte(elem.first);
33-
visit(attrs.alloc(elem.first), elem.second);
103+
self(self, attrs.alloc(elem.first), elem.second);
34104
}
35105

36106
v.mkAttrs(attrs);
37107
} break;
38-
;
39108
case toml::value_t::array: {
40109
auto array = toml::get<std::vector<toml::value>>(t);
41110

42111
auto list = state.buildList(array.size());
43112
for (const auto & [n, v] : enumerate(list))
44-
visit(*(v = state.allocValue()), array[n]);
113+
self(self, *(v = state.allocValue()), array[n]);
45114
v.mkList(list);
46115
} break;
47-
;
48116
case toml::value_t::boolean:
49117
v.mkBool(toml::get<bool>(t));
50118
break;
51-
;
52119
case toml::value_t::integer:
53120
v.mkInt(toml::get<int64_t>(t));
54121
break;
55-
;
56122
case toml::value_t::floating:
57123
v.mkFloat(toml::get<NixFloat>(t));
58124
break;
59-
;
60125
case toml::value_t::string: {
61126
auto s = toml::get<std::string_view>(t);
62127
forceNoNullByte(s);
63128
v.mkString(s);
64129
} break;
65-
;
66130
case toml::value_t::local_datetime:
67131
case toml::value_t::offset_datetime:
68132
case toml::value_t::local_date:
69133
case toml::value_t::local_time: {
70134
if (experimentalFeatureSettings.isEnabled(Xp::ParseTomlTimestamps)) {
135+
#if HAVE_TOML11_4
136+
normalizeDatetimeFormat(t);
137+
#endif
71138
auto attrs = state.buildBindings(2);
72139
attrs.alloc("_type").mkString("timestamp");
73140
std::ostringstream s;
@@ -80,16 +147,24 @@ static void prim_fromTOML(EvalState & state, const PosIdx pos, Value ** args, Va
80147
throw std::runtime_error("Dates and times are not supported");
81148
}
82149
} break;
83-
;
84150
case toml::value_t::empty:
85151
v.mkNull();
86152
break;
87-
;
88153
}
89154
};
90155

91156
try {
92-
visit(val, toml::parse(tomlStream, "fromTOML" /* the "filename" */));
157+
visit(
158+
visit,
159+
val,
160+
toml::parse(
161+
tomlStream,
162+
"fromTOML" /* the "filename" */
163+
#if HAVE_TOML11_4
164+
,
165+
toml::spec::v(1, 0, 0) // Be explicit that we are parsing TOML 1.0.0 without extensions
166+
#endif
167+
));
93168
} catch (std::exception & e) { // TODO: toml::syntax_error
94169
state.error<EvalError>("while parsing TOML: %s", e.what()).atPos(pos).debugThrow();
95170
}

‎tests/functional/lang/eval-okay-fromTOML-timestamps.exp‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -1 +1 @@
1-
{ "1234" = "value"; "127.0.0.1" = "value"; a = { b = { c = { }; }; }; arr1 = [ 1 2 3 ]; arr2 = [ "red" "yellow" "green" ]; arr3 = [ [ 1 2 ] [ 3 4 5 ] ]; arr4 = [ "all" "strings" "are the same" "type" ]; arr5 = [ [ 1 2 ] [ "a" "b" "c" ] ]; arr7 = [ 1 2 3 ]; arr8 = [ 1 2 ]; bare-key = "value"; bare_key = "value"; bin1 = 214; bool1 = true; bool2 = false; "character encoding" = "value"; d = { e = { f = { }; }; }; dog = { "tater.man" = { type = { name = "pug"; }; }; }; flt1 = 1; flt2 = 3.1415; flt3 = -0.01; flt4 = 5e+22; flt5 = 1e+06; flt6 = -0.02; flt7 = 6.626e-34; flt8 = 9.22462e+06; fruit = [ { name = "apple"; physical = { color = "red"; shape = "round"; }; variety = [ { name = "red delicious"; } { name = "granny smith"; } ]; } { name = "banana"; variety = [ { name = "plantain"; } ]; } ]; g = { h = { i = { }; }; }; hex1 = 3735928559; hex2 = 3735928559; hex3 = 3735928559; int1 = 99; int2 = 42; int3 = 0; int4 = -17; int5 = 1000; int6 = 5349221; int7 = 12345; j = { "ʞ" = { l = { }; }; }; key = "value"; key2 = "value"; ld1 = { _type = "timestamp"; value = "1979-05-27"; }; ldt1 = { _type = "timestamp"; value = "1979-05-27T07:32:00"; }; ldt2 = { _type = "timestamp"; value = "1979-05-27T00:32:00.999999"; }; lt1 = { _type = "timestamp"; value = "07:32:00"; }; lt2 = { _type = "timestamp"; value = "00:32:00.999999"; }; name = "Orange"; oct1 = 342391; oct2 = 493; odt1 = { _type = "timestamp"; value = "1979-05-27T07:32:00Z"; }; odt2 = { _type = "timestamp"; value = "1979-05-27T00:32:00-07:00"; }; odt3 = { _type = "timestamp"; value = "1979-05-27T00:32:00.999999-07:00"; }; odt4 = { _type = "timestamp"; value = "1979-05-27T07:32:00Z"; }; physical = { color = "orange"; shape = "round"; }; products = [ { name = "Hammer"; sku = 738594937; } { } { color = "gray"; name = "Nail"; sku = 284758393; } ]; "quoted \"value\"" = "value"; site = { "google.com" = true; }; str = "I'm a string. \"You can quote me\". Name\tJosé\nLocation\tSF."; table-1 = { key1 = "some string"; key2 = 123; }; table-2 = { key1 = "another string"; key2 = 456; }; x = { y = { z = { w = { animal = { type = { name = "pug"; }; }; name = { first = "Tom"; last = "Preston-Werner"; }; point = { x = 1; y = 2; }; }; }; }; }; "ʎǝʞ" = "value"; }
1+
{ "1234" = "value"; "127.0.0.1" = "value"; a = { b = { c = { }; }; }; arr1 = [ 1 2 3 ]; arr2 = [ "red" "yellow" "green" ]; arr3 = [ [ 1 2 ] [ 3 4 5 ] ]; arr4 = [ "all" "strings" "are the same" "type" ]; arr5 = [ [ 1 2 ] [ "a" "b" "c" ] ]; arr7 = [ 1 2 3 ]; arr8 = [ 1 2 ]; bare-key = "value"; bare_key = "value"; bin1 = 214; bool1 = true; bool2 = false; "character encoding" = "value"; d = { e = { f = { }; }; }; dog = { "tater.man" = { type = { name = "pug"; }; }; }; flt1 = 1; flt2 = 3.1415; flt3 = -0.01; flt4 = 5e+22; flt5 = 1e+06; flt6 = -0.02; flt7 = 6.626e-34; flt8 = 9.22462e+06; fruit = [ { name = "apple"; physical = { color = "red"; shape = "round"; }; variety = [ { name = "red delicious"; } { name = "granny smith"; } ]; } { name = "banana"; variety = [ { name = "plantain"; } ]; } ]; g = { h = { i = { }; }; }; hex1 = 3735928559; hex2 = 3735928559; hex3 = 3735928559; int1 = 99; int2 = 42; int3 = 0; int4 = -17; int5 = 1000; int6 = 5349221; int7 = 12345; j = { "ʞ" = { l = { }; }; }; key = "value"; key2 = "value"; ld1 = { _type = "timestamp"; value = "1979-05-27"; }; ldt1 = { _type = "timestamp"; value = "1979-05-27T07:32:00"; }; ldt10 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123456789"; }; ldt11 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123456789"; }; ldt2 = { _type = "timestamp"; value = "1979-05-27T07:32:00.100"; }; ldt3 = { _type = "timestamp"; value = "1979-05-27T07:32:00.120"; }; ldt4 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123"; }; ldt5 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123400"; }; ldt6 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123450"; }; ldt7 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123456"; }; ldt8 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123456700"; }; ldt9 = { _type = "timestamp"; value = "1979-05-27T00:32:00.123456780"; }; lt1 = { _type = "timestamp"; value = "07:32:00"; }; lt10 = { _type = "timestamp"; value = "00:32:00.123456789"; }; lt11 = { _type = "timestamp"; value = "00:32:00.123456789"; }; lt2 = { _type = "timestamp"; value = "00:32:00.100"; }; lt3 = { _type = "timestamp"; value = "00:32:00.120"; }; lt4 = { _type = "timestamp"; value = "00:32:00.123"; }; lt5 = { _type = "timestamp"; value = "00:32:00.123400"; }; lt6 = { _type = "timestamp"; value = "00:32:00.123450"; }; lt7 = { _type = "timestamp"; value = "00:32:00.123456"; }; lt8 = { _type = "timestamp"; value = "00:32:00.123456700"; }; lt9 = { _type = "timestamp"; value = "00:32:00.123456780"; }; name = "Orange"; oct1 = 342391; oct2 = 493; odt1 = { _type = "timestamp"; value = "1979-05-27T07:32:00Z"; }; odt10 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123456Z"; }; odt11 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123456700Z"; }; odt12 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123456780Z"; }; odt13 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123456789Z"; }; odt14 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123456789Z"; }; odt2 = { _type = "timestamp"; value = "1979-05-27T00:32:00-07:00"; }; odt3 = { _type = "timestamp"; value = "1979-05-27T00:32:00.999999-07:00"; }; odt4 = { _type = "timestamp"; value = "1979-05-27T07:32:00Z"; }; odt5 = { _type = "timestamp"; value = "1979-05-27T07:32:00.100Z"; }; odt6 = { _type = "timestamp"; value = "1979-05-27T07:32:00.120Z"; }; odt7 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123Z"; }; odt8 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123400Z"; }; odt9 = { _type = "timestamp"; value = "1979-05-27T07:32:00.123450Z"; }; physical = { color = "orange"; shape = "round"; }; products = [ { name = "Hammer"; sku = 738594937; } { } { color = "gray"; name = "Nail"; sku = 284758393; } ]; "quoted \"value\"" = "value"; site = { "google.com" = true; }; str = "I'm a string. \"You can quote me\". Name\tJosé\nLocation\tSF."; table-1 = { key1 = "some string"; key2 = 123; }; table-2 = { key1 = "another string"; key2 = 456; }; x = { y = { z = { w = { animal = { type = { name = "pug"; }; }; name = { first = "Tom"; last = "Preston-Werner"; }; point = { x = 1; y = 2; }; }; }; }; }; "ʎǝʞ" = "value"; }

‎tests/functional/lang/eval-okay-fromTOML-timestamps.nix‎

Lines changed: 44 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -55,11 +55,53 @@ builtins.fromTOML ''
5555
odt2 = 1979-05-27T00:32:00-07:00
5656
odt3 = 1979-05-27T00:32:00.999999-07:00
5757
odt4 = 1979-05-27 07:32:00Z
58+
# milliseconds
59+
odt5 = 1979-05-27 07:32:00.1Z
60+
odt6 = 1979-05-27 07:32:00.12Z
61+
odt7 = 1979-05-27 07:32:00.123Z
62+
# microseconds
63+
odt8 = 1979-05-27t07:32:00.1234Z
64+
odt9 = 1979-05-27t07:32:00.12345Z
65+
odt10 = 1979-05-27t07:32:00.123456Z
66+
# nanoseconds
67+
odt11 = 1979-05-27 07:32:00.1234567Z
68+
odt12 = 1979-05-27 07:32:00.12345678Z
69+
odt13 = 1979-05-27 07:32:00.123456789Z
70+
# no more precision after nanoseconds
71+
odt14 = 1979-05-27t07:32:00.1234567891Z
72+
5873
ldt1 = 1979-05-27T07:32:00
59-
ldt2 = 1979-05-27T00:32:00.999999
74+
# milliseconds
75+
ldt2 = 1979-05-27T07:32:00.1
76+
ldt3 = 1979-05-27T07:32:00.12
77+
ldt4 = 1979-05-27T07:32:00.123
78+
# microseconds
79+
ldt5 = 1979-05-27t00:32:00.1234
80+
ldt6 = 1979-05-27t00:32:00.12345
81+
ldt7 = 1979-05-27t00:32:00.123456
82+
# nanoseconds
83+
ldt8 = 1979-05-27 00:32:00.1234567
84+
ldt9 = 1979-05-27 00:32:00.12345678
85+
ldt10 = 1979-05-27 00:32:00.123456789
86+
# no more precision after nanoseconds
87+
ldt11 = 1979-05-27t00:32:00.1234567891
88+
6089
ld1 = 1979-05-27
6190
lt1 = 07:32:00
62-
lt2 = 00:32:00.999999
91+
# milliseconds
92+
lt2 = 00:32:00.1
93+
lt3 = 00:32:00.12
94+
lt4 = 00:32:00.123
95+
# microseconds
96+
lt5 = 00:32:00.1234
97+
lt6 = 00:32:00.12345
98+
lt7 = 00:32:00.123456
99+
# nanoseconds
100+
lt8 = 00:32:00.1234567
101+
lt9 = 00:32:00.12345678
102+
lt10 = 00:32:00.123456789
103+
# no more precision after nanoseconds
104+
lt11 = 00:32:00.1234567891
63105
64106
arr1 = [ 1, 2, 3 ]
65107
arr2 = [ "red", "yellow", "green" ]

0 commit comments

Comments
 (0)