Started working on lazy quantifier support

master
Aadhavan Srinivasan 4 weeks ago
parent 3924502b72
commit 277cbc0fc5

@ -64,7 +64,7 @@ const (
)
func isOperator(c rune) bool {
if c == '+' || c == '?' || c == '*' || c == '|' || c == concatRune {
if c == '+' || c == '?' || c == '*' || c == '|' || c == concatRune || c == lazyPlusRune || c == lazyKleeneRune || c == lazyQuestionRune {
return true
}
return false
@ -72,7 +72,7 @@ func isOperator(c rune) bool {
/* priority returns the priority of the given operator */
func priority(op rune) int {
precedence := []rune{'|', concatRune, '+', '*', '?'}
precedence := []rune{'|', concatRune, '+', lazyPlusRune, '*', lazyKleeneRune, '?', lazyQuestionRune}
return slices.Index(precedence, op)
}
@ -208,9 +208,6 @@ func shuntingYard(re string, flags ...ReFlag) ([]postfixNode, error) {
// metacharacter. Later, in thompson(), these will be converted back. This avoids
// confusion in detecting whether a character is escaped eg. detecting
// whether '\\[a]' has an escaped opening bracket (it doesn't).
//
// 5. Check for non-greedy operators. These are not supported at the moment, so an error
// must be thrown if the user attempts to use a non-greedy operator.
for i := 0; i < len(re_runes_orig); i++ {
c := re_runes_orig[i]
if c == '<' && (i == 0 || (re_runes_orig[i-1] != '\\' && re_runes_orig[i-1] != '?')) {
@ -257,8 +254,16 @@ func shuntingYard(re string, flags ...ReFlag) ([]postfixNode, error) {
} else if c == ']' && (i == 0 || re_runes[len(re_runes)-1] != '\\') {
re_runes = append(re_runes, rbracketRune)
continue
} else if slices.Contains([]rune{'+', '*', '?'}, c) && (i < len(re_runes_orig)-1 && re_runes_orig[i+1] == '?') {
return nil, fmt.Errorf("non-greedy operators are not supported")
} else if slices.Contains([]rune{'+', '*', '?'}, c) && (i > 0 && re_runes_orig[i-1] != '\\') && (i < len(re_runes_orig)-1 && re_runes_orig[i+1] == '?') {
switch c {
case '+':
re_runes = append(re_runes, lazyPlusRune)
case '*':
re_runes = append(re_runes, lazyKleeneRune)
case '?':
re_runes = append(re_runes, lazyQuestionRune)
}
i++
} else {
re_runes = append(re_runes, c)
}
@ -421,7 +426,7 @@ func shuntingYard(re string, flags ...ReFlag) ([]postfixNode, error) {
}
if i < len(re_runes) && (re_runes[i] != '(' && re_runes[i] != nonCapLparenRune && re_runes[i] != '|' && re_runes[i] != '\\') || (i > 0 && re_runes[i-1] == '\\') { // Every character should be concatenated if it is escaped
if i < len(re_runes)-1 {
if re_runes[i+1] != '|' && re_runes[i+1] != '*' && re_runes[i+1] != '+' && re_runes[i+1] != '?' && re_runes[i+1] != ')' && re_runes[i+1] != '{' {
if re_runes[i+1] != '|' && re_runes[i+1] != '*' && re_runes[i+1] != lazyKleeneRune && re_runes[i+1] != '+' && re_runes[i+1] != lazyPlusRune && re_runes[i+1] != '?' && re_runes[i+1] != lazyQuestionRune && re_runes[i+1] != ')' && re_runes[i+1] != '{' {
re_postfix = append(re_postfix, concatRune)
}
}

Loading…
Cancel
Save