def filter_pandas(df, filters):
r"""
Filter a GeoPandas DataFrame, return a new filtered DataFrame.
Currently all filters are joined by 'AND'
TODO: Support for 'OR', parentheses?
:param df: The DataFrame to filter
:param filters: An array of (attribute, operator, value) arrays\
for example [('city', 'in', ['Boston', 'New York']), ('id', '>', 10)]
:return: A filtered DataFrame
"""
for filter in filters:
attribute = filter[0]
operator = filter[1]
values = filter[2]
if operator.lower() == "in":
df = df[df[attribute].isin(values)]
elif operator.lower() == "not in":
df = df[~df[attribute].isin(values)]
elif operator.lower() in ["contains", "startswith", "endswith"]:
str_func = getattr(df[attribute].str, operator.lower())
df = df[str_func(r'{}'.format(values))]
elif operator in ops.keys():
df = df[ops[operator](df[attribute], values)]
return df
def eval_binary_expr(self,op1, operator, op2):
op1,op2 = op1, op2
if(operator == 'Contains') :
return op1.like('%'+op2+'%')
if operator.lower() == 'in':
l = [s for s in re.split("[,|;\W]+", op2)]
return op1.in_(l)
if operator.lower() == 'checked':
operator = '='
if '-1' in op2:
return None
return op1.in_(op2)
return self.get_operator_fn(operator)(op1, op2)
def _process_domain(self, result, domain):
if domain and len(domain) == 3:
field, operator, value = domain
compute_operator = {
'ilike': lambda field, value: str(value).lower() in str(field).lower(),
'like': lambda field, value: str(value) in str(field),
'not ilike': lambda field, value: str(value).lower() not in str(field).lower(),
'not like': lambda field, value: str(value) not in str(field),
'=': lambda field, value: value == field,
'!=': lambda field, value: value != field,
'>': lambda field, value: value < field,
'<': lambda field, value: value > field,
'>=': lambda field, value: value <= field,
'<=': lambda field, value: value >= field,
}
operator = operator.lower()
if field in result:
if operator in compute_operator.iterkeys():
return compute_operator.get(operator)(result.get(field), value)
raise osv.except_osv(u'错误', u'未添加的domain条件%s' % domain)
raise osv.except_osv(u'错误', u'不可识别的domain条件，请检查domain"%s"是否正确' % domain)
def eval_binary_expr(self,op1, operator, op2):
op = operator.lower()
if op in ['is', 'is not'] and isinstance(op2,str) and op2.lower() == 'null':
op2 = None
try :
return self.get_operator_fn(op)(op1, op2)
except (NameError,KeyError):
return not_(self.get_operator_fn(op.replace('not','').replace(' ',''))(op1, op2))
def eval_binary_expr(self,op1, operator, op2):
op = operator.lower()
if op in ['is null', 'is not null']:
op = op.replace('null','').replace(' ','')
op2 = None
try :
return self.get_operator_fn(op)(op1, op2)
except (NameError,KeyError):
return not_(self.get_operator_fn(op.replace('not','').replace(' ',''))(op1, op2))
def eval_binary_expr(self,op1, operator, op2):
op = operator.lower()
if op in ['is null', 'is not null']:
op = op.replace('null','').replace(' ','')
op2 = None
try :
return self.get_operator_fn(op)(op1, op2)
except (NameError,KeyError):
return not_(self.get_operator_fn(op.replace('not','').replace(' ',''))(op1, op2))
def eval_binary_expr(self, op1, operator, op2):
op = operator.lower()
if op in ['is', 'is not'] and isinstance(
op2, str) and op2.lower() == 'null':
op2 = None
try:
return self.get_operator_fn(op)(op1, op2)
except (NameError, KeyError):
return not_(
self.get_operator_fn(op.replace('not', '').replace(' ',
''))(op1,
op2))
def filter_pandas(df, filters):
r"""
Filter a GeoPandas DataFrame, return a new filtered DataFrame.
Currently all filters are joined by 'AND'
TODO: Support for 'OR', parentheses?
:param df: The DataFrame to filter
:param filters: An array of (attribute, operator, value) arrays\
for example [('city', 'in', ['Boston', 'New York']), ('id', '>', 10)]
:return: A filtered DataFrame
"""
for filter in filters:
attribute = filter[0]
operator = filter[1]
values = filter[2]
if operator.lower() == "in":
df = df[df[attribute].isin(values)]
elif operator.lower() == "not in":
df = df[~df[attribute].isin(values)]
elif operator.lower() == "contains":
df = df[df[attribute].str.contains(r'{}'.format(values))]
elif operator in ops.keys():
df = df[ops[operator](df[attribute], values)]
return df
def main():
fs = cgi.FieldStorage()
aggr=fs.getvalue("aggr")
operator=fs.getvalue("operator")
dates=fs.getvalue("dates")
if ((not aggr) or (not operator) or (not dates)):
error.errhttp("406")
return
urlaggr=getaggrurl(aggr)
if(urlaggr=="404"):
error.errhttp("404")
return
aggr=aggr.lower()
operator=operator.lower()
dates=dates.lower()
req=urllib2.Request(url=urlaggr)
req.add_header('Accept', 'application/xml, text/turtle, text/csv, application/json')
response = urllib2.urlopen(req)
restype= response.info().gettype()
resource=response.read()
response.close()
if(restype=="application/xml"):
meta=trasforma.locationfromxml(resource,loclist)
elif(restype=="text/turtle"):
meta=trasforma.locationfromturtle(resource,loclist)
elif(restype=="text/csv"):
meta=trasforma.locationfromcsv(resource,loclist)
elif(restype=="application/json"):
meta=trasforma.locationfromjson(resource,loclist)
else:
error.errhttp("406")
return
finallist=getopened(dates, operator,loclist)
if(isinstance(finallist, ( int, long ))):
error.errcode(str(finallist))
return
trasforma.formatresult(os.environ["HTTP_ACCEPT"], finallist, meta)
def oper(op):
operator, value = op.split(" ")
if operator.lower() == "inc":
return int(value)
else:
return int(value) * -1
operation_sequence = str(
input("Enter the operation sequence separated by comma"))
operators = operation_sequence.split(",")
output = None
comparisons_sum = 0
skip = False
# if len(list(set(operators)))==1:
# query_terms.sort(key=lambda x: len(inverted_index.get(x)[1]))
# print("Sorted",query_terms)
# for each operator in the given list we iterate through them to perform the opertions and obtain final result
for index, operator in enumerate(operators):
pos_lists = []
NOT = False
if "and" in operator.lower():
if index == 0:
pos_list_1 = inverted_index.get(query_terms[index])
if not pos_list_1:
print(
"Term not found: {}\nHence, operation has failed".
format(query_terms[index]))
skip = True
break
else:
pos_list_1 = pos_list_1[1]
else:
# to use the output from applying previous operation
pos_list_1 = output
pos_list_2 = inverted_index.get(query_terms[index + 1])
# covers AND NOT
