Category talk:Wren-str: Difference between revisions

}

// Converts a pattern into a list of tokens for processing by the 'isMatch' method.

// Characters within the pattern are represented as follows:

// Non-wildcard characters as themselves (i.e. single character strings);

// * (or **) by the number 0;

// ? (or *?) by the number 1;

// [set] by a list of the tokens within the set:

// single characters by themselves;

// a range of characters, a-b, by a Range of codepoints from 'a' to 'b'.

// If the first character of the set is '!' then the number -1 is inserted

// as a separate token immediately before the list.

static tokenize(pattern) {

var tokens = []

var i = 0

var j

while (i < pattern.count) {

var c = pattern[i]

if (c == "*") {

if (i == 0 || tokens[-1] != 0) tokens.add(0)

} else if (c == "?") {

if (i > 0 && tokens[-1] == 0) tokens[-1] = 1 else tokens.add(1)

} else if (c == "[") {

if (i == pattern.count - 1) {

tokens.add(c)

} else if ((j = indexOf(pattern, "]", i + 1)) == -1) {

tokens.add(c)

} else {

var l = []

var s = sub(pattern, i+1...j)

var k = 0

while (k < s.count) {

var d = s[k]

if (d == "!") {

if (k == 0) tokens.add(-1) else l.add(d)

} else if (k < s.count - 2 && s[k+1] == "-") {

l.add(d.codePoints[0]..s[k+2].codePoints[0])

k = k + 2

} else {

l.add(d)

}

k = k + 1

}

if (l.count == 0) Fiber.abort("set cannot be empty.")

tokens.add(l)

i = i + s.count + 1

}

} else {

tokens.add(c)

}

i = i + 1

}

return tokens

}

// Returns whether a string 's' matches a 'pattern' which may already be tokenized

// if many strings are to be matched. Matching is case sensitive.

// Patterns may contain the following wildcards:

// * (or **) matches zero or more characters until the next token (if any) matches

// and doesn't backtrack in the event of subsequent failure;

// ? (or *?) matches exactly one character;

// [set] matches a single character from the set within the brackets e.g. [aeiou].

// The set can also contain ranges of characters separated by '-' e.g. [a-zA-Z].

// If the first character of the set is '!' then only characters NOT within the rest

// of the set are matched e.g. [!0-9] matches any character other than a digit.

static isMatch(s, pattern) {

var tokens = pattern

if (tokens is String) tokens = tokenize(tokens)

if (!((tokens is List) && tokens.count > 0)) {

Fiber.abort("'pattern' must be a non-empty string or list of tokens.")

}

var i = 0

var j = 0

var star = false

var neg = false

while (i < s.count && j < tokens.count) {

var c = s[i]

var t = tokens[j]

if (t is Num) {

if (t == 0) {

star = true

} else if (t == 1) {

i = i + 1

star = false

} else if (t == -1) {

neg = true

} else {

Fiber.abort("'%(t)' is not a recognized token.")

}

j = j + 1

} else if (t is String) {

if (!star && c != t) return false

if (star && c == t) star = false

i = i + 1

if (!star) j = j + 1

} else if (t is List) {

var matched = false

for (e in t) {

if (e is String) {

if (e == c) {

matched = true

break

}

} else if (e is Range){

var cp = c.codePoints[0]

if (cp >= e.from && cp <= e.to) {

matched = true

break

}

} else {

Fiber.abort("'%(e)' is not a recognized token within a set.")

}

}

if (!star && !neg && !matched) return false

if (!star && neg && matched) return false

if (star && matched) star = false

i = i + 1

neg = false

if (!star) j = j + 1

} else {

Fiber.abort("'%(t)' is not a recognized token.")

}

}

if (i == s.count && j == tokens.count) return true

if (j == tokens.count && tokens[-1] == 0) return true

if (j == tokens.count - 1 && tokens[-1] == 0) return true

return false