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 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226
|
"""Tests for pycddl."""
import platform
import resource
import mmap
from psutil import Process
import pytest
import cbor2
from pycddl import Schema, ValidationError
from .utils import BSTR_SCHEMA, BSTR_100M, BSTR_1K
def assert_invalid_caught(schema, data):
"""
The schema correctly identifies that data as invalid.
"""
with pytest.raises(ValidationError):
schema.validate_cbor(cbor2.dumps(data))
def test_invalid_schema_errors_out():
"""
Attempting to create a new ``CDDLSchema`` with an invalid CDDL schema
results in a ValueError.
"""
with pytest.raises(ValueError):
Schema(
"""
reputation-object = {
application: text
reputons: [* reputon]
"""
)
REPUTON_SCHEMA = """\
reputation-object = {
application: text
reputons: [* reputon]
}
reputon = {
rater: text
assertion: text
rated: text
rating: float16
? confidence: float16
? normal-rating: float16
? sample-size: uint
? generated: uint
? expires: uint
* text => any
}
"""
def test_schema_validates_good_document():
"""
A valid schema will validate a valid document (i.e. no exception is
raised).
"""
schema = Schema(REPUTON_SCHEMA)
for document in [
{"application": "blah", "reputons": []},
{
"application": "conchometry",
"reputons": [
{
"rater": "Ephthianura",
"assertion": "codding",
"rated": "sphaerolitic",
"rating": 0.34133473256800795,
"confidence": 0.9481983064298332,
"expires": 1568,
"unplaster": "grassy",
},
{
"rater": "nonchargeable",
"assertion": "raglan",
"rated": "alienage",
"rating": 0.5724646875815566,
"sample-size": 3514,
"Aldebaran": "unchurched",
"puruloid": "impersonable",
"uninfracted": "pericarpoidal",
"schorl": "Caro",
},
],
},
]:
schema.validate_cbor(cbor2.dumps(document))
def test_schema_fails_bad_documents():
"""
Bad documents cause ``validate_cbor()`` to raise a ``ValidationError``.
"""
schema = Schema(REPUTON_SCHEMA)
for bad_document in [
b"",
cbor2.dumps({"application": "blah"}), # missing reputons key
cbor2.dumps({"application": "blah", "reputons": "NOT A LIST"}),
]:
with pytest.raises(ValidationError):
schema.validate_cbor(bad_document)
def test_integer_value_enforcement():
"""
Schemas that limit minimum integer value are enforced. This is important
for security, for example.
"""
uint_schema = Schema(
"""
object = {
xint: uint
}
"""
)
for i in [0, 1, 4, 5, 500, 1000000]:
uint_schema.validate_cbor(cbor2.dumps({"xint": i}))
for i in [-1, -10000, "x", 0.3]:
assert_invalid_caught(uint_schema, i)
more_than_3_schema = Schema(
"""
object = {
xint: int .gt 3
}
"""
)
for i in [4, 5, 500, 1000000]:
more_than_3_schema.validate_cbor(cbor2.dumps({"xint": i}))
for i in [-1, -10000, "x", 0.3, 0, 1, 2, 3]:
assert_invalid_caught(more_than_3_schema, {"xint": i})
def test_schema_repr():
"""``repr(Schema)`` reflects the schema string."""
schema_text = """
object = {
xint: int .gt 3
"value": uint
}
"""
schema = Schema(schema_text)
assert repr(schema) == f'Schema("""{schema_text}""")'
def get_max_rss_mb():
# ru_maxrss is kilobytes on Linux, bytes on macOS
result = resource.getrusage(resource.RUSAGE_SELF).ru_maxrss / 1024
if platform.system() == "Darwin":
result /= 1024
return result
@pytest.mark.skipif(platform.system() == "Windows", reason="Need POSIX to check maxrss")
def test_memory_usage():
"""
Validating a large document doesn't significantly increase memory usage.
"""
process = Process()
# This returns units of bytes, so divide appropriately:
before_rss_mb = process.memory_info().rss / (1024 * 1024)
before_max_rss_mb = get_max_rss_mb()
# If this fails, this test won't be meaningful because new allocation won't
# budge max RSS. So if this fails, to fix it you should try to adjust
# runtime to ensure we don't have extra high max RSS. E.g. in utils.py we
# try to generate large CBOR docs in a subprocess.
#
# TODO Fil should really expose an API for "current in-use memory and
# current peak memory", it has the info...
assert (
before_max_rss_mb - before_rss_mb < 105
), "This is an environmental check; see code"
BSTR_SCHEMA.validate_cbor(BSTR_100M)
new_max_rss_mb = get_max_rss_mb()
# We're validating a 100MB document. The underlying parser used by the Rust
# library allocates memory linearly based on input (see
# https://github.com/anweiss/cddl/issues/167), but we at least should not
# be copying memory much beyond that. Ideally this should be more like < 10.
assert new_max_rss_mb - before_rss_mb < 120
def test_buffer_interface(tmp_path):
"""
It's possible to pass in read-only buffers, e.g. ``memoryview``.
"""
# memoryview() of bytes works fine: it's read-only.
BSTR_SCHEMA.validate_cbor(memoryview(BSTR_1K))
# But if we mmap() an existing file in read-only way, that's fine:
path = tmp_path / "out.cbor"
with path.open("wb") as f:
f.write(BSTR_1K)
with path.open("rb") as f2:
mapping = mmap.mmap(f2.fileno(), len(BSTR_1K), access=mmap.ACCESS_READ)
BSTR_SCHEMA.validate_cbor(mapping)
@pytest.mark.skipif(
platform.python_implementation() == "PyPy",
reason="PyPy is buggy, it thinks bytearray() is read-only...",
)
def test_buffer_interface_refuses_writeable():
"""
Writeable buffers won't get validated.
"""
# bytearray won't work because it's mutable:
arr = bytearray(BSTR_1K)
with pytest.raises(ValueError):
BSTR_SCHEMA.validate_cbor(arr)
# Likewise for writeable mmap():
mapping = mmap.mmap(-1, len(BSTR_1K))
mapping[:] = BSTR_1K
with pytest.raises(ValueError):
BSTR_SCHEMA.validate_cbor(mapping)
|