Remove empty literals when concatenating regexes

nfa.py

@@ -0,0 +1,10 @@
+from collections import namedtuple
+
+NFA = namedtuple('NFA', ['start', 'accept', 'transitions'])
+
+def copy_nfa(nfa):
+	transitions_copy = {}
+	for from_state in nfa.transitions:
+		transitions_copy[from_state] = nfa.transitions[from_state].copy()
+
+	return NFA(nfa.start, nfa.accept, transitions_copy)

nfa_to_regex.py

@@ -1,16 +1,7 @@
 import enum
-from collections import namedtuple
 
 from regex import lit, concat, bar, star
-
-NFA = namedtuple('NFA', ['start', 'accept', 'transitions'])
-
-def copy_nfa(nfa):
-	transitions_copy = {}
-	for from_state in nfa.transitions:
-		transitions_copy[from_state] = nfa.transitions[from_state].copy()
-
-	return NFA(nfa.start, nfa.accept, transitions_copy)
+from nfa import NFA, copy_nfa
 
 def remove_states(nfa):
 	start, accept, transitions = nfa
@@ -150,16 +141,19 @@ def prettyprint(nfa):
 		print(process_state(from_state) + '\t' + '\t'.join(t))
 
 def main():
-	nfa = NFA('start', ['end'], {
-		'start': {'0': lit('s')},
-		'0': {'0': lit('0'), '1': lit('1'), 'end': lit('e'), 'start': lit('r')},
-		'1': {'0': lit('1'), '1': lit('0'), 'start': lit('r')},
-		'end': {'end': lit('e'), 'start': lit('n')}
+	nfa = NFA('start', ['0'], {
+		'start': {'1': lit('i'), '2': lit('d')},
+		'0': {'1': lit('i'), '2': lit('d')},
+		'1': {'0': lit('d'), '2': lit('i')},
+		'2': {'0': lit('i'), '1': lit('d')}
 	})
 
 	prettyprint(nfa)
 
-	print(to_regex(nfa))
+	regex = to_regex(nfa)
+
+	print(repr(regex))
+	print(regex)
 
 if __name__ == '__main__':
 	main()

regex.py

@@ -81,7 +81,11 @@ def concat(*elements):
 
 	combined = []
 	for element in flattened:
-		if len(combined) > 0 and type(combined[-1]) == Literal and type(element) == Literal:
+		if type(element) == Literal and element.text == '':
+			# Drop empty literals
+			continue
+
+		elif len(combined) > 0 and type(combined[-1]) == Literal and type(element) == Literal:
 			# Combine two literals next to each other
 			# into one literal
 			previous = combined.pop()
@@ -90,7 +94,11 @@ def concat(*elements):
 		else:
 			combined.append(element)
 
-	if len(combined) == 1:
+	if len(combined) == 0:
+		# Empty regex, represent with empty literal
+		return lit('')
+
+	elif len(combined) == 1:
 		element, = combined
 		return element
 	else: