def JSON2object(json_string, params=Null, flexible=False):
try:
#REMOVE """COMMENTS""", #COMMENTS, //COMMENTS, AND \n \r
if flexible: json_string=re.sub(r"\"\"\".*?\"\"\"|^\s*//\n|#.*?\n|\n|\r", r" ", json_string) #DERIVED FROM https://github.com/jeads/datasource/blob/master/datasource/bases/BaseHub.py#L58
if params != Null:
params=dict([(k,CNV.value2quote(v)) for k,v in params.items()])
json_string=expand_template(json_string, params)
obj=json_decoder.decode(json_string)
if isinstance(obj, list): return StructList(obj)
return struct.wrap(obj)
except Exception, e:
# Log.error("Can not decode JSON:\n\t"+json_string, e)
print e
pass
def expand_template(template, values):
if values == Null: values={}
values=struct.wrap(values)
def replacer(found):
var=found.group(1)
try:
val=values[var[2:-2]]
val=toString(val)
return val
except Exception, e:
try:
if e.message.find(u"is not JSON serializable"):
#WORK HARDER
val=json_scrub(val)
val=toString(val)
return val
except Exception:
raise Exception(u"Can not find "+var[2:-2]+u" in template:\n"+indent(template))
from json2sql.util.logs import Log
from json2sql.util.struct import Struct
from numpy.core import records
OBJECT="object" # USED IN property.type TO INDICATE FURTHER STRUCTURE
NESTED="nested" # USED TO INDICATE column.type IS AN OBJECT
MULTI="multi" # True IF multi-valued column, "ordered" if multivalued and ordered (has index)
ORDERED="ordered"
INDEX="index" # True, "yes", "not_analyzed", <PARSER NAME> to indicate the for of text parsing used
NOT_ANALYZED="not_analyzed"
PRIMITIVES=struct.wrap({
"string":{"sql_type":"VARCHAR(100)"},
"integer":{"sql_type":"LONG"},
"float":{"sql_type":"DOUBLE"},
"boolean":{"sql_type":"DECIMAL(1)"}
})
TYPES=[OBJECT, "nested"]
PARSERS = {
True: simple_words,
"yes": simple_words,
NOT_ANALYZED: no_parse,
}
def simple_words(value):
return value.split(" ")
