validater can validate and convert value to python object by schema, support python 2.7.x and python 3.3+
validater 可以依据 schema 校验数据并将数据转换成相应的 python 对象，支持 python 2.7.x 和 python 3.3+
pip install validater
py.test
or
tox
>>> from validater import parse, validate
>>> validate('5',parse('int(0,9)&required&default=0'))
([], 5)
>>> validate('-1',parse('int(0,9)&required&default=0'))
([(u'', u"must be 'int(0, 9)'")], None)
>>> validate(None,parse('int(0,9)&required&default=0'))
([], 0)
>>>
>>> snippet = 'int(0,9)&required&default=0','input a int value, >=0 and <= 9'
>>> schema = parse({'key':snippet,'list':[snippet]})
>>> obj={"key":None, "list":["1","23","asd"]}
>>> error,value=validate(obj,schema)
>>> error
[(u'list[1]', u"must be 'int(0, 9)': input a int value, >=0 and <= 9"),
(u'list[2]', u"must be 'int(0, 9)': input a int value, >=0 and <= 9")]
>>> value
{'list': [1, None, None], 'key': 0}
>>> validaters is a dict contains all validaters, it will be default_validaters by default
schema has 5 styles:
-
string snippet
'int(0,9)&required&default=0' -
tuple snippet
'int(0,9)&required&default=0', 'desc' -
dict snippet
{ 'validater':'int', 'args':(0,9), 'required':True, 'default':0, 'desc':'desc' } -
list schema
[snippet] -
dict schema
{ 'key':snippet, 'list':[snippet], 'inner':{ 'key':snippet } }
string snippet has 3 part
------------------------------------------
|validater |args |kwargs |
------------------------------------------
|int |(0,9) |&required&default=0 |
------------------------------------------
validater: the name of validater
args: optional, the args passed to validater
kwargs: optional, the kwargs passed to validater, it's value is determine by eval()
built-in kwargs
required: is required or not, empty string is treated as missing
default: default value, it can be callable without args(used in dict snippet). default value will also be validated
desc: desc of the snippet
obj can be dict, list or just a value, it should has the same struct as schema
schema is the return value of parse
return tuple(error,value)
error: a list of tuple(key, msg)
value: a dict, it's struct is the same as schema
| name | valid value |
|---|---|
| any | anything |
| str | six.string_types(basestring on py2, str on py3) |
| unicode | six.text_type(unicode on py2, str on py3) |
| bool | bool |
| int(start,end) | int value >=start and <= end |
| +int | int value >=0 |
| float(start,end) | float value >=start and <= end |
| datetime(format,output,input) | format is %Y-%m-%dT%H:%M:%S.%fZ(ISO 8601) by default, output means convert value to string, input means convert value to datetime, if both output and input is False, convert to string if value is datetime, else convert string to datetime |
| date(format,output,input) | format is %Y-%m-%d by default, output means convert value to string, input means convert value to date, if both output and input is False, convert to string if value is date/datetime, else convert string/datetime to date |
| ipv4 | ipv4 |
| phone | phone number |
| idcard | idcard number of chinese |
| url | url, support urls without 'http://' |
| name(minlength,maxlength) | name combine of chars in [a-zA-Z0-9_], and start with chars in [a-zA-Z], minlength is 4 and maxlength is 16 by defeault |
| password(minlength,maxlength) | password combine of ascii chars, and don't has space chars, minlength is 6 and maxlength is 16 by defeault |
| safestr | escape unsafe string: & > < ' " |
string snippet
sche = "unicode&required&default='hello world'"tuple snippet
sche = "unicode&required&default='hello world'", "welcome words"dict snippet
from datetime import datetime
sche = {
"desc":"a iso8601 format datetime string",
"required":True,
"validate":"datetime",
"input":True,
"default":datetime.utcnow,
}list schema
sche = ["int&required"]
sche = {'userid_list': ["int&required"]}dict schema
sche = {
"page_num": ('+int&required&fefault=1', 'the page num'),
"page_size": ('int(1,50)&required&fefault=10', 'the page size')
} reuse schema
snippet = {"name": ("safestr","your name")}
schema = {
"user1": snippet,
"user2": snippet,
}validater is a callable object and return a tuple
def validater(v, args, kwargs):
return (True/False, validated_value)re_validater and type_validater
from validater import re_validater,type_validater
# build a validater by regex_object
validater = re_validater(regex_object)
# build a validater by type or types
validater = type_validater(cls, empty='')custom validaters
from validater import default_validaters
my_validaters = {}
my_validaters.update(default_validaters)
# then you can use my_validaters as params of parse,
# add_validater, remove_validateradd_validater and remove_validater
from validater import add_validater, remove_validater
add_validater('name', validater, validaters=my_validaters)
remove_validater('name', validaters=my_validaters)for example
from validater import re_validater,type_validater
from validater import default_validaters
from validater import add_validater, remove_validater
my_validaters = {}
my_validaters.update(default_validaters)
year_validater = re_validater(re.compile(r"^\d{4}$"))
add_validater("year", year_validater, my_validaters)
add_validater("list", type_validater(list, empty=[]), my_validaters)
def abs_validater(v, debug=False):
try:
return True, abs(v)
except:
if debug:
raise
return False, None
add_validater('abs', abs_validater, my_validaters)
schema = parse("abs&required&debug", my_validaters)
err,val = validate(-1, schema)ProxyDict can wrap custome type object and use it as dict
ProxyDict(obj, types)
validate custome type
class User(object):
def __init__(self, userid):
self.userid = userid
sche = parse({
'userid': "int&required",
'friends': [{'userid': "int&required"}]})
jack, f1, f2 = User(0), User(1), User(2)
jack.friends = [f1, f2]
err, val = validate(jack, sche, proxy_types=[User])MIT License