我有list of dict
个:
print (L)
[{0: 'x==1', 1: 'y==2', 2: 'z!=1'}, {0: 'x==1', 1: 'y<=3', 2: 'z>1'}]
我要创建元组,其值在运算符之前,运算符在之后:
#first step
wanted = [[('x', '==', '1'), ('y', '==', '2'), ('z', '!=', '1')],
[('x', '==', '1'), ('y', '<=', '3'), ('z', '>', '1')]]
然后由运算符映射第二个值:
import operator
ops = {'>': operator.gt,
'<': operator.lt,
'>=': operator.ge,
'<=': operator.le,
'==': operator.eq,
'!=': operator.ne}
#expected final output
wanted = [[('x', <built-in function eq>, '1'),
('y', <built-in function eq>, '2'),
('z', <built-in function ne>, '1')],
[('x', <built-in function eq>, '1'),
('y', <built-in function le>, '3'),
('z', <built-in function gt>, '1')]]
我尝试:
L = [[re.findall(r'(.*)([<>=!]+)(.*)', v)[0] for k, v in x.items()] for x in L]
print (L)
[[('x=', '=', '1'), ('y=', '=', '2'), ('z!', '=', '1')],
[('x=', '=', '1'), ('y<', '=', '3'), ('z', '>', '1')]]
L = [[ops[y[1]] for y in x] for x in L]
但是问题是匹配的中间子字符串-运算符错误,然后运算符的匹配值错误。
正确匹配的正确正则表达式是什么?或者这是另一种可能的解决方案。例如由string.partition
创建?我打开所有可能的解决方案。
答案 0 :(得分:2)
如果您的输入确实如此简单,我认为最直接的方法是分割操作符:
In [1]: import re
In [2]: data = [{0: 'x==1', 1: 'y==2', 2: 'z!=1'}, {0: 'x==1', 1: 'y<=3', 2: 'z>1'}]
In [3]: rgx = re.compile(r'([<>=!]+)')
In [4]: [[rgx.split(v) for v in d.values()] for d in data]
Out[4]:
[[['x', '==', '1'], ['y', '==', '2'], ['z', '!=', '1']],
[['x', '==', '1'], ['y', '<=', '3'], ['z', '>', '1']]]
请注意,如果您将捕获组添加到拆分器正则表达式中,则会将其包括在内!
然后,完成它:
In [11]: ops = {'>': operator.gt,
...: '<': operator.lt,
...: '>=': operator.ge,
...: '<=': operator.le,
...: '==': operator.eq,
...: '!=': operator.ne}
...:
In [12]: parsed = [[rgx.split(v) for v in d.values()] for d in data]
In [13]: [[(x, ops[op], y) for x,op,y in ps] for ps in parsed]
Out[13]:
[[('x', <function _operator.eq>, '1'),
('y', <function _operator.eq>, '2'),
('z', <function _operator.ne>, '1')],
[('x', <function _operator.eq>, '1'),
('y', <function _operator.le>, '3'),
('z', <function _operator.gt>, '1')]]
答案 1 :(得分:2)
将贪婪方法1st子字符串正则表达式更改为唯一的单词字符:
L = [{0: 'x==1', 1: 'y==2', 2: 'z!=1'}, {0: 'x==1', 1: 'y<=3', 2: 'z>1'}]
L = [[re.findall(r'(\w)([<>=!]+)(.*)', v)[0] for k, v in x.items()] for x in L]
[[(y[0],ops[y[1]],y[2]) for y in x] for x in L]
[[('x', <function _operator.eq>, '1'),
('y', <function _operator.eq>, '2'),
('z', <function _operator.ne>, '1')],
[('x', <function _operator.eq>, '1'),
('y', <function _operator.le>, '3'),
('z', <function _operator.gt>, '1')]]
或根据评论中的jezrael
建议(1行列表的内容):
L = [[[(z[0], ops[z[1]], z[2]) for z in re.findall(r'(\w)([<>=!]+)(.*)', v)][0] for k, v in x.items()] for x in L]
或者我们不需要键,因此可以直接使用值:
L = [[[(z[0], ops[z[1]], z[2]) for z in re.findall(r'(\w)([<>=!]+)(.*)', v)][0] for v in x.values()] for x in L]
答案 2 :(得分:0)
问题在于*
是一个贪婪的字符匹配器。因此,在x==1
中,如果*
可以匹配多个字符,它将在仍用单个([<>=!]+)
字符满足第二组=
的情况下
解决方案:
假定非运营商组将永远不会包含<
,>
,=
或!
,而不是使用*
,请使用负数字符集:
re.findall(r'([^<>=!]+)([<>=!]+)([^<>=!]+)', v)
与竖线交替使用以捕获操作员:
re.findall(r'(.*)((?:>|<|<=|>=|==|!=))(.*)', v)