def test_mct_utf8():
'''Test that utf8 content can be read and written correctly in
make_clean_tmpfile.
'''
filepath = os.path.join(os.path.dirname(os.path.abspath(__file__)),
"utf.f90")
_ = make_clean_tmpfile(filepath)
def test_mct_invalid_filename_arg():
'''Test that the expected exception occurs if an invalid input file
is provided.
'''
with pytest.raises(IOError) as excinfo:
_ = make_clean_tmpfile("nonexistantfile")
assert "No such file or directory" in str(excinfo.value)
def test_mct_invalid_encode_arg(tmpdir):
'''Test that an appropriate exception is raised if the encode argument
is invalid.
'''
input_filepath = create_tmp_file("", tmpdir)
with pytest.raises(InternalError) as excinfo:
_ = make_clean_tmpfile(input_filepath, encoding="invalid")
assert "unknown encoding: invalid'." in str(excinfo.value)
def test_mct_works(tmpdir):
'''Test that if there are no errors in the input file then we get an
exact copy of the input file in the temporary file.'''
content = ("program valid\n" "end program valid\n")
input_filepath = create_tmp_file(content, tmpdir)
output_filepath = make_clean_tmpfile(input_filepath)
with open(output_filepath, "r") as cfile:
output = cfile.read()
assert output == content
def test_mct_invalid_skip_arg(tmpdir):
'''Test that the expected exception occurs if an invalid
skip_bad_input argument is provided.
'''
input_filepath = create_tmp_file("", tmpdir)
with pytest.raises(InternalError) as excinfo:
_ = make_clean_tmpfile(input_filepath, skip_bad_input="INVALID")
assert ("skip_bad_input argument should be False or True but "
"found 'INVALID'.") in str(excinfo.value)
def test_mct_parse_error(tmpdir):
'''Test that an appropriate exception is raised if there is an invalid
character in the input file and skip_bad_argument is set to
False.
'''
invalid_content = u"\xca"
input_filepath = create_tmp_file(invalid_content, tmpdir)
with pytest.raises(ParseError) as excinfo:
_ = make_clean_tmpfile(input_filepath, skip_bad_input=False,
encoding="ascii")
assert ("Bad character in input file. Error returned was 'ascii' "
"codec can't decode byte ") in str(excinfo.value)
# Can't check the actual value as some versions of Python3 return
# a different value to the one above.
assert "in position 0: ordinal not in range(128)." in str(excinfo.value)
def test_mct_skip_error(tmpdir, caplog):
'''Test that invalid characters are skipped in an input file by
default and that logging messages are created.
'''
content = "HELLO"
invalid_content = u"\xca".join(content)
input_filepath = create_tmp_file(invalid_content, tmpdir)
output_filepath = make_clean_tmpfile(input_filepath, encoding="ascii")
with open(output_filepath, "r") as cfile:
output = cfile.read()
assert output == content
for record in caplog.records:
assert record.levelname != 'CRITICAL'
assert ("Skipped bad character in input file. Error returned was 'ascii' "
"codec can't decode byte ") in caplog.text
# Can't check the actual value as some versions of Python3 return
# a different value to the one above.
assert "in position 1: ordinal not in range(128)." in caplog.text
def get_source_info(file_candidate):
'''
Determines the format of Fortran source held in a file.
:param file_candidate: a filename or a file object
:type file_candidate: str or (file (py2) or _io.TextIOWrapper (py3))
:returns: the Fortran format encoded as a string.
:rtype: str
'''
if hasattr(file_candidate, 'name') and hasattr(file_candidate, 'read'):
filename = file_candidate.name
# The behaviour of file.name when associated with a file without a
# file name has changed between Python 2 and 3.
#
# Under Python 3 file.name holds an integer file handle.
if isinstance(filename, int):
filename = None
# Under Python 2 file.name holds a string of the form "<..>".
elif filename.startswith('<') and filename.endswith('>'):
filename = None
elif isinstance(file_candidate, six.string_types):
# The preferred method for identifying strings changed between Python2
# and Python3.
filename = file_candidate
else:
message = 'Argument must be a filename or file-like object.'
raise ValueError(message)
if filename:
_, ext = os.path.splitext(filename)
if ext == '.pyf':
return FortranFormat(True, True)
if hasattr(file_candidate, 'read'):
# If the candidate object has a "read" method we assume it's a file
# object.
#
# If it is a file object then it may be in the process of being read.
# As such we need to take a note of the current state of the file
# pointer so we can restore it when we've finished what we're doing.
#
pointer = file_candidate.tell()
file_candidate.seek(0)
source_info = get_source_info_str(file_candidate.read())
file_candidate.seek(pointer)
return source_info
else:
# It isn't a file and it passed the type check above so it must be
# a string.
#
# If it's a string we assume it is a filename. In which case we need
# to open the named file so we can read it.
#
# It is closed on completion so as to return it to the state it was
# found in.
#
from fparser.common.utils import make_clean_tmpfile
tmpfile = make_clean_tmpfile(file_candidate)
with io.open(tmpfile, 'r', encoding='utf8') as file_object:
string = get_source_info_str(file_object.read())
os.remove(tmpfile)
return string