get elements in "AND" in logic string with Python -

i want parse logic strings , combinations of elements in "and" logic. instance, string '( , ( b or c ) )' should [[a,b],[a,c]] , string '( , b , ( c or d , f ) or f , g )' should [[a,b,c],[a,b,d,f],[f,g]].

i'm trying use pyparsing. following post here parsing complex logical expression in pyparsing in binary tree fashion manage nested list letters grouped according preferences ("and" has preference on "or", , parenthesis overrides this):

import pyparsing pp  complex_expr = pp.forward() vars = pp.word(pp.alphas, pp.alphanums + "_") | pp.regex(r"[+-]?\d+(:?\.\d*)?(:?[ee][+-]?\d+)?").setname('proteins') clause = pp.group(vars ^ (pp.suppress("(") + complex_expr + pp.suppress(")") ))  expr = pp.operatorprecedence(clause,[                             ("and", 2, pp.opassoc.left, ),                             ("or", 2, pp.opassoc.left, ),]) #print expr complex_expr << expr parseresult=complex_expr.parsestring('( , b , ( c or d , f ) or f , g )') print parseresult 

which gives:

[[[[['a'], 'and', ['b'], 'and', [[['c'], 'or', [['d'], 'and', ['f']]]]], 'or', [['f'], 'and', ['g']]]]]

now how can process result achieve desired output? greateful help. tried pyparsing i'm open other modules may better.

thanks in advance.

python libraries going little bit:

import re import itertools 

let's write required function:

def analyse(expression):     # find used symbols     symbols = set(re.findall(r"\b[a-z]\b", expression))     # find combinations of symbols , values     mappings = (dict(zip(symbols, values)) values in itertools.product([false, true], repeat=len(symbols)))     # select combinations make whole expression true     solutions = [sorted(name name in mapping if mapping[name]) mapping in mappings if eval(expression, none, mapping)]     # filter out redundant solutions     return sorted(s1 s1 in solutions if not any(set(s1) > set(s2) s2 in solutions)) 

and let's test it:

assert analyse("( , ( b or c ) )") == [["a", "b"], ["a", "c"]] assert analyse("( , b , ( c or d , f ) or f , g )") == [["a", "b", "c"], ["a", "b", "d", "f"], ["f", "g"]] 

there comments in source code. anyway, main steps are:

• the expression variables found single-character uppercase names.
• each variable can either true or false. find combinations.
• we select such combinations make whole expression true.
• we keep minimal solutions, i.e. not supersets of other solutions.

i thank nice question. python's itertools never stop surprising me. ;-)