From 3778869567f60509115fbce935c3bc6cd88a66e7 Mon Sep 17 00:00:00 2001
From: Aadhavan Srinivasan <aadhavan@twomorecents.org>
Date: Mon, 28 Oct 2024 17:38:43 -0400
Subject: [PATCH] Use stateContents type to allow a state to store multiple
 characters

---
 nfa.go | 52 +++++++++++++++++++++++++++++++++++-----------------
 1 file changed, 35 insertions(+), 17 deletions(-)

diff --git a/nfa.go b/nfa.go
index 233a5cb..a3977f5 100644
--- a/nfa.go
+++ b/nfa.go
@@ -3,7 +3,7 @@ package main
 const EPSILON int = 0
 
 type State struct {
-	content     int              // Contents of current state
+	content     stateContents    // Contents of current state
 	isEmpty     bool             // If it is empty - Union operator and Kleene star states will be empty
 	isLast      bool             // If it is the last state (acept state)
 	output      []*State         // The outputs of the current state ie. the 'outward arrows'. A union operator state will have more than one of these.
@@ -22,17 +22,25 @@ func verifyLastStatesHelper(state *State, visited map[*State]bool) {
 		state.isLast = true
 		return
 	}
-	if len(state.transitions) == 1 && len(state.transitions[state.content]) == 1 && state.transitions[state.content][0] == state { // Eg. a*
-		state.isLast = true
-		return
+	//	if len(state.transitions) == 1 && len(state.transitions[state.content]) == 1 && state.transitions[state.content][0] == state { // Eg. a*
+	if len(state.transitions) == 1 { // Eg. a*
+		var moreThanOneTrans bool // Dummy variable, check if all the transitions for the current's state's contents have a length of one
+		for _, c := range state.content {
+			if len(state.transitions[c]) != 1 || state.transitions[c][0] != state {
+				moreThanOneTrans = true
+			}
+		}
+		state.isLast = !moreThanOneTrans
 	}
 
-	if len(state.transitions) == 1 && state.isKleene { // A State representing a Kleene Star has a transition going out, which loops back to it. If that is the only transition (and it contains only one state), then it must be a last-state
-		for _, v := range state.transitions { // Should only loop once
-			if len(v) == 1 {
-				state.isLast = true
-				return
-			}
+	if state.isKleene { // A State representing a Kleene Star has transitions going out, which loop back to it. If all those transitions point to the same (single) state, then it must be a last state
+		transitionDests := make([]*State, 0)
+		for _, v := range state.transitions {
+			transitionDests = append(transitionDests, v...)
+		}
+		if allEqual(transitionDests...) {
+			state.isLast = true
+			return
 		}
 	}
 	if visited[state] == true {
@@ -55,7 +63,9 @@ func verifyLastStates(start []*State) {
 
 func concatenate(s1 *State, s2 *State) *State {
 	for i := range s1.output {
-		s1.output[i].transitions[s2.content], _ = unique_append(s1.output[i].transitions[s2.content], s2)
+		for _, c := range s2.content { // Create transitions for every element in s2's content to s2'
+			s1.output[i].transitions[c], _ = unique_append(s1.output[i].transitions[c], s2)
+		}
 	}
 	s1.output = s2.output
 	return s1
@@ -64,14 +74,18 @@ func concatenate(s1 *State, s2 *State) *State {
 func kleene(s1 State) *State {
 	toReturn := &State{}
 	toReturn.transitions = make(map[int][]*State)
-	toReturn.content = EPSILON
+	toReturn.content = newContents(EPSILON)
 	toReturn.isEmpty = true
 	toReturn.isKleene = true
 	toReturn.output = append(toReturn.output, toReturn)
 	for i := range s1.output {
-		s1.output[i].transitions[toReturn.content], _ = unique_append(s1.output[i].transitions[toReturn.content], toReturn)
+		for _, c := range toReturn.content {
+			s1.output[i].transitions[c], _ = unique_append(s1.output[i].transitions[c], toReturn)
+		}
+	}
+	for _, c := range s1.content {
+		toReturn.transitions[c], _ = unique_append(toReturn.transitions[c], &s1)
 	}
-	toReturn.transitions[s1.content], _ = unique_append(toReturn.transitions[s1.content], &s1)
 	return toReturn
 }
 
@@ -85,9 +99,13 @@ func alternate(s1 *State, s2 *State) *State {
 	// For example, given the transition 'a', the state 's1' can only be mentioned once.
 	// This would lead to multiple instances of the same set of match indices, since both
 	// 's1' states would be considered to match.
-	toReturn.transitions[s1.content], _ = unique_append(toReturn.transitions[s1.content], s1)
-	toReturn.transitions[s2.content], _ = unique_append(toReturn.transitions[s2.content], s2)
-	toReturn.content = EPSILON
+	for _, c := range s1.content {
+		toReturn.transitions[c], _ = unique_append(toReturn.transitions[c], s1)
+	}
+	for _, c := range s2.content {
+		toReturn.transitions[c], _ = unique_append(toReturn.transitions[c], s2)
+	}
+	toReturn.content = newContents(EPSILON)
 	toReturn.isEmpty = true
 
 	return toReturn