this repo has no description
evaluator: regression test for order independent eval fixed in evalv3
The old evaluator does not correctly handle declarations in a different
order, despite CUE being order independent.
The new evaluator correctly handles this situation.
Add a regression test to lock in the behaviour of the new evaluator.
Fixes #3376
Signed-off-by: Paul Jolly <paul@myitcv.io>
Change-Id: I1c020846663e265b6c244eabd1d678650e8f1fe2
Reviewed-on: https://review.gerrithub.io/c/cue-lang/cue/+/1199680
Reviewed-by: Daniel Martí <mvdan@mvdan.cc>
TryBot-Result: CUEcueckoo <cueckoo@cuelang.org>
Unity-Result: CUE porcuepine <cue.porcuepine@gmail.com>
+114
+114
cmd/cue/cmd/testdata/script/eval_issue3376.txtar
+114
cmd/cue/cmd/testdata/script/eval_issue3376.txtar
···
1
+
# https://cuelang.org/issue/3376
2
+
#
3
+
# The old evaluator does not correctly handle the difference in order of
4
+
# declarations. To the extent that scenario2 gives an error:
5
+
#
6
+
# items.a: cannot add to field _:
7
+
# ./scenario2.cue:1:32
8
+
#
9
+
# and scenario4 results in a SIGSEV.
10
+
#
11
+
# The new evaluator correctly results in no errors in these cases.
12
+
13
+
# evalv2 - scenario 1
14
+
exec cue eval ./scenario1.cue
15
+
16
+
# evalv2 - scenario 2
17
+
#
18
+
# This should not fail but it does. Fixed in evalv3
19
+
! exec cue eval ./scenario2.cue
20
+
cmp stderr scenario2.evalv2.stderr
21
+
22
+
# evalv2 - scenario 3
23
+
exec cue eval ./scenario3.cue
24
+
25
+
# evalv2 - scenario 4
26
+
#
27
+
# This should not fail but it does with a SIGSEV. Fixed in evalv3
28
+
! exec cue eval ./scenario4.cue
29
+
stderr 'panic: runtime error: invalid memory address or nil pointer dereference'
30
+
31
+
env CUE_EXPERIMENT=evalv3
32
+
33
+
# evalv3 - scenario 1
34
+
exec cue eval ./scenario1.cue
35
+
36
+
# evalv3 - scenario 2
37
+
exec cue eval ./scenario2.cue
38
+
39
+
# evalv3 - scenario 3
40
+
exec cue eval ./scenario3.cue
41
+
42
+
# evalv3 - scenario 4
43
+
exec cue eval ./scenario4.cue
44
+
45
+
-- scenario1.cue --
46
+
#Item: {
47
+
id!: string
48
+
deps: [...or(known_ids)] | or(known_ids)
49
+
}
50
+
51
+
known_ids: [for k,v in items { v.id } ]
52
+
53
+
items: [string]: #Item
54
+
55
+
items: [Key=_]: { id: (Key) + "-id" }
56
+
57
+
items: a: deps: items.b.id
58
+
items: b: deps: []
59
+
items: c: deps: [items.a.id, items.b.id]
60
+
61
+
-- scenario2.cue --
62
+
known_ids: [for k,v in items { v.id } ]
63
+
64
+
#Item: {
65
+
id!: string
66
+
deps: [...or(known_ids)] | or(known_ids)
67
+
}
68
+
69
+
70
+
items: [string]: #Item
71
+
72
+
items: [Key=_]: { id: (Key) + "-id" }
73
+
74
+
items: a: deps: items.b.id
75
+
items: b: deps: []
76
+
items: c: deps: [items.a.id, items.b.id]
77
+
78
+
-- scenario3.cue --
79
+
#Item: {
80
+
id!: string
81
+
deps?: [...or(known_ids)] | or(known_ids)
82
+
}
83
+
84
+
known_ids: [for k,v in items { v.id } ]
85
+
86
+
items: [string]: #Item
87
+
88
+
items: [Key=_]: { id: (Key) + "-id" }
89
+
90
+
items: a: deps: items.b.id
91
+
items: b: deps: []
92
+
items: c: deps: [items.a.id, items.b.id]
93
+
items: d: _
94
+
95
+
-- scenario4.cue --
96
+
known_ids: [for k,v in items { v.id } ]
97
+
98
+
#Item: {
99
+
id!: string
100
+
deps?: [...or(known_ids)] | or(known_ids)
101
+
}
102
+
103
+
items: [string]: #Item
104
+
105
+
items: [Key=_]: { id: (Key) + "-id" }
106
+
107
+
items: a: deps: items.b.id
108
+
items: b: deps: []
109
+
items: c: deps: [items.a.id, items.b.id]
110
+
items: d: _
111
+
112
+
-- scenario2.evalv2.stderr --
113
+
items.a: cannot add to field _:
114
+
./scenario2.cue:1:32