Added unicode support to dot metacharacter - it now matches _any_ unicode character (almost)

main.go

@@ -141,7 +141,7 @@ func shuntingYard(re string) []postfixNode {
 		}
 
 		if c == '.' { // Dot metacharacter - represents 'any' character, but I am only adding Unicode 0020-007E
-			outQueue = append(outQueue, newPostfixNode(dotChars()...))
+			outQueue = append(outQueue, newPostfixDotNode())
 			continue
 		}
 		if c == '^' { // Start-of-string assertion
@@ -282,6 +282,9 @@ func thompson(re []postfixNode) *State {
 		if c.nodetype == CHARACTER || c.nodetype == ASSERTION {
 			state := State{}
 			state.transitions = make(map[int][]*State)
+			if c.isDot {
+				state.isDot = true
+			}
 			state.content = rune2Contents(c.contents)
 			state.output = make([]*State, 0)
 			state.output = append(state.output, &state)
@@ -397,6 +400,7 @@ func main() {
 	var re string
 	re = flag.Args()[0]
 	var test_str string
+	var test_runes []rune // Rune-slice representation of test_str
 	var err error
 	// Create reader for stdin and writer for stdout // End index is one more than last index of match
 	reader := bufio.NewReader(os.Stdin)
@@ -406,7 +410,8 @@ func main() {
 	startState := thompson(re_postfix)
 	// Read every string from stdin until we encounter an error. If the error isn't EOF, panic.'
 	for test_str, err = reader.ReadString('\n'); err == nil; test_str, err = reader.ReadString('\n') {
-		matchIndices := findAllMatches(startState, []rune(test_str))
+		test_runes = []rune(test_str)
+		matchIndices := findAllMatches(startState, []rune(test_runes))
 		// Decompose the array of matchIndex structs into a flat unique array of ints - if matchIndex is {4,7}, flat array will contain 4,5,6
 		// This should make checking O(1) instead of O(n)
 		indicesToPrint := new_uniq_arr[int]()
@@ -422,7 +427,7 @@ func main() {
 			// Find all numbers from 0 to len(test_str) that are NOT in oldIndices.
 			// These are the values we want to print, now that we have inverted the match.
 			// Re-initialize indicesToPrint and add all of these values to it.
-			indicesToPrint.add(setDifference(genRange(0, len(test_str)), oldIndices)...)
+			indicesToPrint.add(setDifference(genRange(0, len(test_runes)), oldIndices)...)
 
 		}
 		// If lineFlag is enabled, we should only print something if:
@@ -435,7 +440,7 @@ func main() {
 				continue
 			}
 		}
-		for i, c := range test_str {
+		for i, c := range test_runes {
 			if indicesToPrint.contains(i) {
 				color.New(color.FgRed).Fprintf(out, "%c", c)
 				// Newline after every match - only if -o is enabled and -v is disabled.

misc.go

@@ -8,8 +8,10 @@ import (
 var whitespaceChars = []rune{' ', '\t', '\n'}
 var digitChars = []rune{'0', '1', '2', '3', '4', '5', '6', '7', '8', '9'}
 var wordChars = []rune("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789_")
+var notDotChars = []rune{'\n'}
 var LBRACKET rune = 0xF0000
 var RBRACKET rune = 0xF0001
+var ANY_CHAR rune = 0xF0002 // Represents any character - used for 'dot' metacharacter
 
 func dotChars() []rune { // Returns all possible characters represented by the dot metacharacter - this is too tedious to define as a variable, which is why it is a function
 	start := 0x0020

nfa.go

@@ -23,6 +23,7 @@ type State struct {
 	isKleene       bool             // Identifies whether current node is a 0-state representing Kleene star
 	assert         assertType       // Type of assertion of current node - NONE means that the node doesn't assert anything
 	zeroMatchFound bool             // Whether or not the state has been used for a zero-length match - only relevant for zero states
+	isDot          bool             // Whether or not the state represents a 'dot' metacharacter. A 'dot' node doesn't store any contents directly, as it would take up too much space
 }
 
 // Clones the NFA starting from the given state.
@@ -94,6 +95,9 @@ func (s State) contentContains(str []rune, idx int) bool {
 	if s.assert != NONE {
 		return s.checkAssertion(str, idx)
 	}
+	if s.isDot {
+		return !slices.Contains(notDotChars, str[idx])
+	}
 	// Default - s.assert must be NONE
 	return slices.Contains(s.content, int(str[idx]))
 }
@@ -116,7 +120,12 @@ func (s State) matchesFor(str []rune, idx int) ([]*State, int) {
 	if s.assert == NONWBOUND && isWordBoundary(str, idx) {
 		return make([]*State, 0), -1
 	}
-	return s.transitions[int(str[idx])], len(s.transitions[int(str[idx])])
+	listTransitions := s.transitions[int(str[idx])]
+	for _, dest := range s.transitions[int(ANY_CHAR)] {
+		listTransitions = append(listTransitions, dest)
+	}
+	numTransitions := len(listTransitions)
+	return listTransitions, numTransitions
 }
 
 type NFA struct {

postfixNode.go

@@ -26,6 +26,7 @@ type postfixNode struct {
 	contents  []rune // Contents of the node - the length of this would only be >1 if the node represents a character class
 	startReps int    // Minimum number of times the node should be repeated - used with numeric specifiers
 	endReps   int    // Maximum number of times the node should be repeated - used with numeric specifiers
+	isDot     bool   // Whether or not the current node represents a 'dot' metacharacter
 }
 
 // Creates a new escaped node - the given character is assumed to have been preceded by a backslash
@@ -105,9 +106,22 @@ func newPostfixNode(contents ...rune) postfixNode {
 	return to_return
 }
 
+// Creates and returns a postfixNode representing the 'dot' metacharacter.
+func newPostfixDotNode() postfixNode {
+	toReturn := postfixNode{}
+	toReturn.startReps = 1
+	toReturn.endReps = 1
+	toReturn.nodetype = CHARACTER
+	toReturn.isDot = true
+	toReturn.contents = []rune{ANY_CHAR}
+	return toReturn
+}
+
 // Creates a character node, regardless of the contents
 func newPostfixCharNode(contents ...rune) postfixNode {
 	toReturn := postfixNode{}
+	toReturn.startReps = 1
+	toReturn.endReps = 1
 	toReturn.nodetype = CHARACTER
 	toReturn.contents = append(toReturn.contents, contents...)
 	return toReturn