Question

I have an array of Python objects, and I would like to get all objects that evaluate to true for a given logical formula.

Let’s say the objects are tasks in a task tracking system. Each task has a status, an assignee and a due date. Now I’d like to get all objects that evaluate to true for this formula:

(assignee=cdecker OR due=tomorrow) AND status=open

For things like the object fields (assignee and status) it would be a simple matter of checking equality, whereas due=tomorrow would be a function to be called on the object.

Is there a simple parser that I could use to parse the expression and then evaluate it for all the objects in my starting set (basically wire all literals to function calls)?

Answer 1

Try pyparsing. Here’s a parser for a grammar that supports equality, boolean ops and nested brackets:

from pyparsing import *

expr   = Forward()
factor = Forward()
term   = Forward()

iand   = Literal("AND").suppress()
ior    = Literal("OR").suppress()
ieq    = Literal("=").suppress()
lpar   = Literal("(").suppress()
rpar   = Literal(")").suppress()
ident  = Word(alphas)
atom   = ident | lpar + expr + rpar

factor << Group(atom + ZeroOrMore(ieq + atom))("equal")
term   << Group(factor + ZeroOrMore(iand + term))("and")
expr   << Group(term  + ZeroOrMore(ior + expr))("or")

You can apply this to your formulas to obtain tree-alike structures, which you can evaluate at some point later, for example:

formula = "(assignee=cdecker OR due=tomorrow) AND status=open"
tree = expr.parseString(formula, parseAll=True)
print tree.asXML()