如何在Python2中计算字符串中的不等式

我有一个文本文件（另一个进程的输出，我无法更改），其中包含逻辑比较（只有这三个：

>，经过几个小时的工作，我找到了一种获取
ast.literal\u eval（）的方法
开始工作！我的逻辑是查看
x>2
的两个方面，即
x
和
2
，通过
literal\u eval
评估这两个方面，确保它们是安全的，然后通过我的
check（）
函数运行它，该函数进行评估。对于（'def'，abc'）中的
z也一样
：首先确保
z
和
（'def'，'abc'）
都是安全的，然后使用
check（）
函数执行实际的布尔检查

因为我完全信任我的输入，所以我本可以用更简单的
eval（）
方法，但我只想加倍小心。并想为所有存在安全问题（用户输入等）且需要安全评估逻辑的人构建一些代码。希望它能帮助某些人

请参阅下面我的完整代码，欢迎提出任何意见/建议

import re
import ast

myStr = "x>2 and y<=30 and z in ('def', 'abc')"
categoricalVars = ('z')
x, y, z = '5', '6', 'abc'
delim = "(\>|\<=|\ in )" # Put in group to find in the func check() which delimiter is used

def pyRules(s):
    """
    Place {} around variable names so that we can str.format() in the func check()
    """
    varName = re.split(delim, s)[0]
    rest = "".join(re.split(delim, s)[1:])
    return "'{" + varName + "}'" + rest

def check(x):
    """
    If operation is > or <= then it is a numeric var, use double literal_eval to
    parse floats e.g. "'5'" (dual quotes) to 5.0. This is equivalent to:
    float(ast.literal_eval(first)). Else it is categorical, just literal_eval once
    """
    first, operation, second = re.split(delim, x)
    if operation == ">":
        return ast.literal_eval(ast.literal_eval(first)) > ast.literal_eval(second)
    elif operation == "<=":
        return ast.literal_eval(ast.literal_eval(first)) <= ast.literal_eval(second)
    elif operation == " in ":
        return ast.literal_eval(first) in ast.literal_eval(second)

# These are my raw rules:
print [pyRules(e) for e in myStr.split(' and ')] 

# These are my processed rules:
print [pyRules(e).format(x='5',y='6',z='abc') for e in myStr.split(' and ')] 

# And these are my final results of logical evaluation:
print [check(pyRules(e).format(x='5',y='6',z='abc')) for e in myStr.split(' and ')] 

重新导入
导入ast
myStr=“x>2和y |\or”：
返回ast.literal_eval（ast.literal_eval（第一））>ast.literal_eval（第二）
elif操作==“2”、“{y}'2”、“'6'
eval（）
对于不受信任的字符串是有问题的。否则，它只是在运行代码。请看一看，它可能有一些用处。谢谢@stephernauch！如果没有什么好消息，我将使用
eval（）
。@cᴏʟᴅsᴘᴇᴇᴅ 谢谢！乍一看似乎很有希望，我现在正在检查。如果您完全信任您的代码，我认为使用
eval
就可以了。
categoricalVars = ('z')

x, y, z = '5', '6', 'abc'

import re
delim = "(\>|\<=|\ in )" # Put in group to find later which delimiter is used

def pyRules(s):
    varName = re.split(delim, s)[0]
    rest = "".join(re.split(delim, s)[1:])
    if varName in categoricalVars:
        return varName + rest
    else:
        return "float(" + varName + ")" + rest
# Call:
[pyRules(e) for e in myStr.split(' and ')] 
# Result:
['float(x)>2', 'float(y)<=30', "z in ('def', 'abc')"]

[eval(pyRules(e)) for e in myStr.split(' and ')]
# Result:
[True, True, True]

import ast
[ast.literal_eval(pyRules(e)) for e in myStr.split(' and ')]
# Result:
Traceback (most recent call last):

  File "<ipython-input-556-dae16951de03>", line 1, in <module>
    ast.literal_eval(ast.parse(conds[0]))

  File "C:\ProgramData\Anaconda2\lib\ast.py", line 80, in literal_eval
    return _convert(node_or_string)

  File "C:\ProgramData\Anaconda2\lib\ast.py", line 79, in _convert
    raise ValueError('malformed string')

ValueError: malformed string

def pyRules(s):
    varName = re.split(delim, s)[0]
    operation = "".join(re.split(delim, s)[1:])
    if varName in categoricalVars:
        return "'{" + varName + "}'" + operation
    else:
        return "float({" + varName + "})" + operation

rules = [pyRules(e).format(x='5',y='6',z='abc') for e in myStr.split(' and ')] 
# rules is:
['float(5)>2', 'float(6)<=30', "'abc' in ('def', 'abc')"]

def check(x):
    first, operation, second = re.split(delim, x)
    if operation == ">":
        return first > second
    elif operation == "<=":
        return first <= second
    elif operation == " in ":
        return first in second
# Call:
[check(pyRules(e).format(x='5',y='6',z='abc')) for e in myStr.split(' and ')] 
# Result:
[True, False, True]

import re
import ast

myStr = "x>2 and y<=30 and z in ('def', 'abc')"
categoricalVars = ('z')
x, y, z = '5', '6', 'abc'
delim = "(\>|\<=|\ in )" # Put in group to find in the func check() which delimiter is used

def pyRules(s):
    """
    Place {} around variable names so that we can str.format() in the func check()
    """
    varName = re.split(delim, s)[0]
    rest = "".join(re.split(delim, s)[1:])
    return "'{" + varName + "}'" + rest

def check(x):
    """
    If operation is > or <= then it is a numeric var, use double literal_eval to
    parse floats e.g. "'5'" (dual quotes) to 5.0. This is equivalent to:
    float(ast.literal_eval(first)). Else it is categorical, just literal_eval once
    """
    first, operation, second = re.split(delim, x)
    if operation == ">":
        return ast.literal_eval(ast.literal_eval(first)) > ast.literal_eval(second)
    elif operation == "<=":
        return ast.literal_eval(ast.literal_eval(first)) <= ast.literal_eval(second)
    elif operation == " in ":
        return ast.literal_eval(first) in ast.literal_eval(second)

# These are my raw rules:
print [pyRules(e) for e in myStr.split(' and ')] 

# These are my processed rules:
print [pyRules(e).format(x='5',y='6',z='abc') for e in myStr.split(' and ')] 

# And these are my final results of logical evaluation:
print [check(pyRules(e).format(x='5',y='6',z='abc')) for e in myStr.split(' and ')] 

["'{x}'>2", "'{y}'<=30", "'{z}' in ('def', 'abc')"]
["'5'>2", "'6'<=30", "'abc' in ('def', 'abc')"]
[True, True, True]