ANSI styling sequences within the `valueHelp` parameter of addOption() causes incorrect usage display

[timmaffett]

If you include ANSI styling sequences in valueHelp parameters of addOption() is causes the usage display to display incorrect formatting because it is including the hidden ANSI escape sequences in the string length calculations.

Example

The correct display would be:

[osa1]

If anyone wants to contribute, this should just be a matter of adding https://github.com/flutter/flutter/blob/92879427fc87c8fb705c7d1ffeead5900d1eda67/packages/flutter_tools/lib/src/base/logger.dart#L705-L710

to

core/pkgs/args/lib/src/utils.dart

Lines 85 to 143 in dc97530

	List<String> wrapTextAsLines(String text, {int start = 0, int? length}) {
	assert(start >= 0);
	/// Returns true if the code unit at [index] in [text] is a whitespace
	/// character.
	///
	/// Based on: https://en.wikipedia.org/wiki/Whitespace_character#Unicode
	bool isWhitespace(String text, int index) {
	var rune = text.codeUnitAt(index);
	return rune >= 0x0009 && rune <= 0x000D ||
	rune == 0x0020 ||
	rune == 0x0085 ||
	rune == 0x1680 ||
	rune == 0x180E ||
	rune >= 0x2000 && rune <= 0x200A ||
	rune == 0x2028 ||
	rune == 0x2029 ||
	rune == 0x202F ||
	rune == 0x205F ||
	rune == 0x3000 ||
	rune == 0xFEFF;
	}
	if (length == null) return text.split('\n');
	var result = <String>[];
	var effectiveLength = math.max(length - start, 10);
	for (var line in text.split('\n')) {
	line = line.trim();
	if (line.length <= effectiveLength) {
	result.add(line);
	continue;
	}
	var currentLineStart = 0;
	int? lastWhitespace;
	for (var i = 0; i < line.length; ++i) {
	if (isWhitespace(line, i)) lastWhitespace = i;
	if (i - currentLineStart >= effectiveLength) {
	// Back up to the last whitespace, unless there wasn't any, in which
	// case we just split where we are.
	if (lastWhitespace != null) i = lastWhitespace;
	result.add(line.substring(currentLineStart, i).trim());
	// Skip any intervening whitespace.
	while (isWhitespace(line, i) && i < line.length) {
	i++;
	}
	currentLineStart = i;
	lastWhitespace = null;
	}
	}
	result.add(line.substring(currentLineStart).trim());
	}
	return result;
	}

[lrhn]

That RegExp is overly permissive. The only codes that it makes sense to ignore are the color-selecting ones.
Anything that moves the cursor, or has a global effect that is not just about the text being printed, can't just be ignored (and probably shouldn't be printed to begin with).

A RegExp of RegExp(r'(?:\x1b\[\d+(?:;\d+)m)+') would likely be sufficient and preferable.

(May want to check for other escapes, to disallow them.)