Dead-simple grammar that can read natural numbers [funcparserlib]

Funcparserlib is a parser combinator, which means it doesn't use a
text-based grammar. Instead, you combine small parsers into larger ones,
eventually arriving at a parser that can handle your whole language.

The basic unit of funcparserlib is `some`, which takes a function that
receives a character and returns a boolean, to determine whether that
character is a match. Our `digits` rule is a parser that requires its
sub-parser to match one or more times; its sub-parser is a parser that
matches if its input is a digit.

funcparserlib parsers have an __rshift__ function, which means they can
be used with the `>>` operator. However, instead of being a right
bit-shift like an int's __rshift__, this function receives a function
that takes the output of the parser and returns some munged version of
that output. In this case, we're supplying a function that takes the
parsed list-of-digit-characters, joins it into a string, and then casts
it as an int.

Author: Joaquin Lorente

ndn/funcparserlib.py

@@ -0,0 +1,9 @@
+from __future__ import absolute_import
+
+import funcparserlib.parser as p
+
+digits = p.oneplus(p.some(lambda char: char.isdigit()))
+number = digits >> (lambda ds: int(''.join(ds)))
+
+def calculate(text):
+    return number.parse(text)

requirements.txt

@@ -5,3 +5,4 @@ mock==1.0.1
 rednose==0.4.1
 ply==3.4
 parsley==1.2
+funcparserlib==0.3.6

test/test_funcparserlib.py

@@ -0,0 +1,8 @@
+from unittest import TestCase
+from nose.tools import eq_
+
+from ndn.funcparserlib import calculate
+
+class TestParsleyParse(TestCase):
+    def test_parse_a_single_number(self):
+        eq_(1, calculate('1'))