def into_semgrep_error(self) -> SemgrepError:
if self._check_id == "Timeout":
return MatchTimeoutError(self._path, self._rule_id)
elif self._check_id == "OutOfMemory":
return OutOfMemoryError(self._path, self._rule_id)
elif self._check_id == "LexicalError":
return LexicalError(self._path, self._rule_id)
else:
try:
with open(self._path, errors="replace") as f:
file_hash = SourceTracker.add_source(f.read())
except IOError as e:
return SemgrepError(f"Could not open '{self._path}': {e}")
error_span = Span(
start=self._start,
end=self._end,
source_hash=file_hash,
file=str(self._path),
)
return SourceParseError(
short_msg="parse error",
long_msg=f"Could not parse {self._path.name} as {self._language}",
spans=[error_span],
help="If the code appears to be valid, this may be a semgrep bug.",
)
def into_semgrep_error(self) -> SemgrepError:
with open(self._path) as f:
file_hash = SourceTracker.add_source(f.read())
error_span = Span(
start=self._start,
end=self._end,
source_hash=file_hash,
file=self._path.name,
)
return SourceParseError(
short_msg="parse error",
long_msg=f"Could not parse {self._path.name} as {self._language}",
spans=[error_span],
help="If the code appears to be valid, this may be a semgrep bug.",
)
def test_same_hash():
# SemgrepErrors with all same fields have the same hash
error_1 = SourceParseError(
short_msg="1",
long_msg="2",
spans=[],
help="4",
)
error_2 = SourceParseError(
short_msg="1",
long_msg="2",
spans=[],
help="4",
)
assert error_1.__hash__() == error_2.__hash__()
errors = set()
errors.add(error_1)
assert error_2 in errors
def test_different_hash():
# SemgrepErrors with differing fields have different hash
error_1 = SourceParseError(
short_msg="1",
long_msg="2",
spans=[],
help="4",
)
error_2 = SourceParseError(
short_msg="not 1",
long_msg="2",
spans=[],
help="4",
)
assert error_1.__hash__() != error_2.__hash__()
errors = set()
errors.add(error_1)
assert error_2 not in errors
