def merge_status(*args):
"""Merge two binary entries and return a FHIR CarePlan.status"""
for inact, val in args:
if is_empty(val) or is_empty(inact):
return "unknown"
if inact == 1:
return "revoked"
else:
if val == 1:
return "active"
if val == 0:
return "on hold"
def clean_time(raw_input): # noqa: C901
if not isinstance(raw_input, str):
raw_input = str(raw_input)
if utils.is_empty(raw_input):
return ""
time = None
formats = [
"%H:%M:%S",
"%H:%M:%S.%f",
"%Y%m%d%H%M%S",
"%Y%m%d%H%M",
"%Y-%m-%d %H:%M:%S",
"%Y-%m-%dT%H:%M:%S",
"%Y-%m-%dT%H:%M:%S%z",
"%a, %d %b %Y %H:%M:%S %Z",
"%H:%M:%S%z",
"%H::%M::%S",
"%H %M %S",
"%H%M%S",
"%H%M",
]
for fmt in formats:
try:
time = datetime.datetime.strptime(raw_input, fmt).time()
except ValueError:
pass
if not time:
return "00:00:00"
return time.strftime("%H:%M:%S")
def merge_insee(value1, value2):
"""Custom script (x, y) -> x if x is not None else y"""
if not utils.is_empty(value1):
insee = value1
else:
insee = value2
return insee.strip()
def if_valid_func(value):
if not utils.is_empty(process(value)):
if callable(callback):
return callback(value)
else:
return callback
else:
return ""
def map_intent(raw_input):
"""Map (0,1,NULL) to (plan, order, proposal)"""
mapping = {0: intent.PLAN.value, 1: intent.ORDER.value}
if is_empty(input):
return intent.PROPOSAL.value
elif raw_input in mapping.keys():
return mapping[raw_input]
else:
return None
def map_permission(raw_input):
"""Map UMLS codes (Yes, No) to (permit, deny)"""
if utils.is_empty(raw_input):
return None
mapping = {
"UMLS:C1298907": Authorization.PERMIT.value,
"UMLS:C1298908": Authorization.DENY.value,
}
if raw_input in mapping.keys():
return mapping[raw_input]
def clean_codes(raw_input):
"""Remove terminology system from code ("HL7:male") to ("male")"""
if utils.is_empty(raw_input):
return None
code = re.match(r"([A-z0-9\-]*)( *: *)(.*)", raw_input)
if not code or code.group(1) not in codeTypes:
return raw_input
else:
return code.group(3)
def clean_instant(raw_input): # noqa: C901
if not isinstance(raw_input, str):
raw_input = str(raw_input)
if utils.is_empty(raw_input):
return ""
date = None
formats = [
"%Y",
"%Y-%m",
"%Y%m",
"%Y-%m-%d",
"%Y%m%d",
"%Y%m%d%H%M",
"%Y-%m-%d %H:%M:%S",
"%Y-%m-%dT%H:%M:%S",
]
for fmt in formats:
try:
date = datetime.datetime.strptime(raw_input, fmt)
# By default, we set the timezone to UTC+2 (Paris)... Until we expand worldwide!
date_with_tz = date.replace(tzinfo=datetime.timezone(datetime.timedelta(hours=2)))
result = date_with_tz.isoformat()
break
except ValueError:
pass
# Handle YYYY-MM-DDTH:M:S+zz:zz
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%dT%H:%M:%S+%z")
result = date.isoformat()
except ValueError:
pass
# Handle YYYY-MM-DDTH:M:S-zz:zz
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%dT%H:%M:%S-%z")
result = date.isoformat()
except ValueError:
pass
# Handle RFC 1123 format
try:
date = datetime.datetime.strptime(raw_input, "%a, %d %b %Y %H:%M:%S GMT")
date_with_tz = date.replace(tzinfo=datetime.timezone(datetime.timedelta(hours=0)))
result = date_with_tz.isoformat()
except ValueError:
pass
if date is None:
return raw_input
return result
def clean_date(raw_input):
"""Convert date to ISO format"""
if utils.is_empty(raw_input):
return ""
# Handle YYYYMMDD
try:
date = datetime.datetime.strptime(raw_input, "%Y%m%d")
iso_date = date.isoformat()
return iso_date
except ValueError:
pass
return raw_input
def map_gender(raw_input):
"""Map gender from (M,F) or (HL7:M, HL7:F) to (male,female)"""
if utils.is_empty(raw_input):
return None
mapping = {
"M": Gender.MALE.value,
"F": Gender.FEMALE.value,
"HL7:M": Gender.MALE.value,
"HL7:F": Gender.FEMALE.value,
}
if raw_input in mapping.keys():
return mapping[raw_input]
else:
return Gender.UNKNOWN.value
def clean_date(raw_input): # noqa: C901
if not isinstance(raw_input, str):
raw_input = str(raw_input)
if utils.is_empty(raw_input):
return ""
date = None
# Correct format
try:
pattern = re.compile(
r"([0-9]([0-9]([0-9][1-9]|[1-9]0)|[1-9]00)|[1-9]000)"
r"(-(0[1-9]|1[0-2])(-(0[1-9]|[1-2][0-9]|3[0-1])(T([01][0-9]|2[0-3])"
r":[0-5][0-9]:([0-5][0-9]|60)(\.[0-9]+)?(Z|(\+|-)((0[0-9]|1[0-3]):"
r"[0-5][0-9]|14:00)))?)?)?")
full_match = re.fullmatch(raw_input, pattern)
date = datetime.datetime.strptime(
full_match.group(0)[0:10], "%Y-%m-%d")
except Exception:
pass
formats = [
"%Y%m%d%H%M",
"%Y-%m-%d %H:%M:%S",
"%Y-%m-%dT%H:%M:%S",
"%Y-%m-%dT%H:%M:%S%z",
"%a, %d %b %Y %H:%M:%S %Z",
"%Y-%m-%d",
"%Y%m%d",
"%Y-%m",
"%Y%m",
"%Y",
]
for fmt in formats:
try:
date = datetime.datetime.strptime(raw_input, fmt)
except ValueError:
pass
if date is None:
return raw_input
# We only want the date
return date.isoformat().split("T")[0]
def clean_quantity(raw_input):
"""
[deprecated: river parses types automatically]
Removes input not conform to FHIR quantity type
"""
if isinstance(raw_input, (float, int)):
return raw_input
if utils.is_empty(raw_input):
return None
number = re.search(r"-?(0|[1-9][0-9]*)(\.[0-9]+)?([eE][+-]?[0-9]+)?",
raw_input)
if not number or number.group(0) != raw_input:
logging.warning(
"The quantity cleaning-script has removed input {}, not of decimal type"
.format(raw_input))
return None
else:
return float(raw_input)
def map_marital_status(code):
"""Map MIMIC marital status"""
status = MaritalStatus
mapping = {
"MARRIED": status.Married.value,
"SINGLE": status.Unmarried.value,
"WIDOWED": status.Widowed.value,
"SEPARATED": status.LegallySeparated.value,
"DIVORCED": status.Divorced.value,
"UNKNOWN": status.Unknown.value,
}
if code in mapping.keys():
return mapping[code]
elif utils.is_empty(code):
return status.Unknown.value
else:
logging.warning("In {}, args {} not recognised".format(
"marital_status", code))
return status.Unknown.value
def map_2_true(raw_input):
"""Map some code from (2->True) and None otherwise"""
if utils.is_empty(raw_input) or raw_input != 2:
return None
else:
return True
def merge_insee(value1, value2):
if not utils.is_empty(value1):
insee = value1
else:
insee = value2
return insee.strip()
def test_is_empty():
assert utils.is_empty(None)
assert utils.is_empty("NaN")
assert utils.is_empty("")
assert utils.is_empty(" ")
assert not utils.is_empty(" holà ")
def select_first_not_empty(*args):
"""Merging script which select the first input not empty"""
for arg in args:
if not is_empty(arg):
return arg
def code_to_empty(raw_input):
"""Return None when entry is some code: -1 or (sans)"""
if utils.is_empty(raw_input) or raw_input == "-1" or raw_input == "(sans)":
return None
else:
return raw_input
def clean_dateTime(raw_input): # noqa: C901
if not isinstance(raw_input, str):
raw_input = str(raw_input)
if utils.is_empty(raw_input):
return ""
date = None
# Correct format
try:
pattern = re.compile(
r"([0-9]([0-9]([0-9][1-9]|[1-9]0)|[1-9]00)|[1-9]000)"
r"(-(0[1-9]|1[0-2])(-(0[1-9]|[1-2][0-9]|3[0-1])(T([01][0-9]|2[0-3])"
r":[0-5][0-9]:([0-5][0-9]|60)(\.[0-9]+)?(Z|(\+|-)((0[0-9]|1[0-3]):"
r"[0-5][0-9]|14:00)))?)?)?")
date = re.fullmatch(pattern, raw_input).group(0)
result = date
except AttributeError:
pass
# Handle YYYY
try:
date = datetime.datetime.strptime(raw_input, "%Y")
result = date.isoformat().split("-")[0]
except ValueError:
pass
# Handle YYYY-MM
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m")
result = date.isoformat()[:7]
except ValueError:
pass
# Handle YYYYMM
try:
date = datetime.datetime.strptime(raw_input, "%Y%m")
result = date.isoformat()[:7]
except ValueError:
pass
# Handle YYYY-MM-DD
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%d")
result = date.isoformat().split("T")[0]
except ValueError:
pass
# Handle YYYYMMDD
try:
date = datetime.datetime.strptime(raw_input, "%Y%m%d")
result = date.isoformat().split("T")[0]
except ValueError:
pass
# Handle YYYYMMDDHHMM
try:
date = datetime.datetime.strptime(raw_input, "%Y%m%d%H%M")
# By default, we set the timezone to UTC+2 (Paris)... Until we expand worldwide!
date_with_tz = date.replace(
tzinfo=datetime.timezone(datetime.timedelta(hours=2)))
result = date_with_tz.isoformat()
except ValueError:
pass
# Handle YYYY-MM-DD H:M:S
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%d %H:%M:%S")
# By default, we set the timezone to UTC+2 (Paris)... Until we expand worldwide!
date_with_tz = date.replace(
tzinfo=datetime.timezone(datetime.timedelta(hours=2)))
result = date_with_tz.isoformat()
except ValueError:
pass
# Handle YYYY-MM-DDTH:M:S
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%dT%H:%M:%S")
# By default, we set the timezone to UTC+2 (Paris)... Until we expand worldwide!
date_with_tz = date.replace(
tzinfo=datetime.timezone(datetime.timedelta(hours=2)))
result = date_with_tz.isoformat()
except ValueError:
pass
# Handle YYYY-MM-DDTH:M:S+zz:zz
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%dT%H:%M:%S+%z")
result = date.isoformat()
except ValueError:
pass
# Handle YYYY-MM-DDTH:M:S-zz:zz
try:
date = datetime.datetime.strptime(raw_input, "%Y-%m-%dT%H:%M:%S-%z")
result = date.isoformat()
except ValueError:
pass
# Handle RFC 1123 format
try:
date = datetime.datetime.strptime(raw_input,
"%a, %d %b %Y %H:%M:%S GMT")
date_with_tz = date.replace(
tzinfo=datetime.timezone(datetime.timedelta(hours=0)))
result = date_with_tz.isoformat()
except ValueError:
pass
if date is None:
return raw_input
return result
def zero_to_empty(raw_input):
"""Return None when entry is 0"""
if utils.is_empty(raw_input) or raw_input == "0":
return None
else:
return raw_input
