(swift) Revamped keywords and added support for operators

CHANGES.md

@@ -8,8 +8,14 @@ Language improvements:
 
 - enh(cpp): Support C++ pack expansion in function arguments [Martin Dørum][]
 - enh(makefile): Add `make` as an alias (#2883) [tripleee][]
-- fix(http) avoid recursive sublanguage and tighten rules (#2893) [Josh Goebel][]
 - enh(swift) Improved grammar for strings (#2819) [Steven Van Impe][]
+- enh(swift) Grammar improvements (#2908) [Steven Van Impe][]
+  - New grammar for keywords and built-ins
+  - Added support for operator highlighting
+  - New grammar for attributes
+  - Added support for quoted identifiers, implicit parameters, and property wrapper projections
+  - Support for more complex expressions in string interpolation
+- fix(http) avoid recursive sublanguage and tighten rules (#2893) [Josh Goebel][]
 - fix(asciidoc): Handle section titles level 5 (#2868) [Vaibhav Chanana][]
 
 Grammar improvements:

src/languages/lib/kws_swift.js

@@ -0,0 +1,307 @@
+import {
+  concat,
+  either
+} from '../../lib/regex.js';
+
+export const keywordWrapper = keyword => concat(
+  /\b/,
+  keyword,
+  /\w$/.test(keyword) ? /\b/ : /\B/
+);
+
+// Keywords that require a leading dot.
+export const dotKeywords = [
+  'Protocol', // contextual
+  'Type' // contextual
+].map(keywordWrapper);
+
+// Keywords that may have a leading dot.
+export const optionalDotKeywords = [
+  'init',
+  'self'
+].map(keywordWrapper);
+
+// should register as keyword, not type
+export const keywordTypes = [
+  'Any',
+  'Self'
+];
+
+// Regular keywords and literals.
+export const keywords = [
+  // strings below will be fed into the regular `keywords` engine while regex
+  // will result in additional modes being created to scan for those keywords to
+  // avoid conflicts with other rules
+  'associatedtype',
+  /as\?/, // operator
+  /as!/, // operator
+  'as', // operator
+  'break',
+  'case',
+  'catch',
+  'class',
+  'continue',
+  'convenience', // contextual
+  'default',
+  'defer',
+  'deinit',
+  'didSet', // contextual
+  'do',
+  'dynamic', // contextual
+  'else',
+  'enum',
+  'extension',
+  'fallthrough',
+  'fileprivate(set)',
+  'fileprivate',
+  'final', // contextual
+  'for',
+  'func',
+  'get', // contextual
+  'guard',
+  'if',
+  'import',
+  'indirect', // contextual
+  'infix', // contextual
+  /init\?/,
+  /init!/,
+  'inout',
+  'internal(set)',
+  'internal',
+  'in',
+  'is', // operator
+  'lazy', // contextual
+  'let',
+  'mutating', // contextual
+  'nonmutating', // contextual
+  'open(set)', // contextual
+  'open', // contextual
+  'operator',
+  'optional', // contextual
+  'override', // contextual
+  'postfix', // contextual
+  'precedencegroup',
+  'prefix', // contextual
+  'private(set)',
+  'private',
+  'protocol',
+  'public(set)',
+  'public',
+  'repeat',
+  'required', // contextual
+  'rethrows',
+  'return',
+  'set', // contextual
+  'some', // contextual
+  'static',
+  'struct',
+  'subscript',
+  'super',
+  'switch',
+  'throws',
+  'throw',
+  /try\?/, // operator
+  /try!/, // operator
+  'try', // operator
+  'typealias',
+  'unowned(safe)', // contextual
+  'unowned(unsafe)', // contextual
+  'unowned', // contextual
+  'var',
+  'weak', // contextual
+  'where',
+  'while',
+  'willSet' // contextual
+];
+
+// NOTE: Contextual keywords are reserved only in specific contexts.
+// Ideally, these should be matched using modes to avoid false positives.
+
+// TODO: Create a PRECEDENCE_GROUP mode to match the remaining contextual keywords:
+// assignment associativity higherThan left lowerThan none right
+// These aren't included in the list because they result in mostly false positives.
+
+// Literals.
+export const literals = [
+  'false',
+  'nil',
+  'true'
+];
+
+// Keywords that start with a number sign (#).
+// #available is handled separately.
+export const numberSignKeywords = [
+  '#colorLiteral',
+  '#column',
+  '#dsohandle',
+  '#else',
+  '#elseif',
+  '#endif',
+  '#error',
+  '#file',
+  '#fileID',
+  '#fileLiteral',
+  '#filePath',
+  '#function',
+  '#if',
+  '#imageLiteral',
+  '#keyPath',
+  '#line',
+  '#selector',
+  '#sourceLocation',
+  '#warn_unqualified_access',
+  '#warning'
+];
+
+// Global functions in the Standard Library.
+export const builtIns = [
+  'abs',
+  'all',
+  'any',
+  'assert',
+  'assertionFailure',
+  'debugPrint',
+  'dump',
+  'fatalError',
+  'getVaList',
+  'isKnownUniquelyReferenced',
+  'max',
+  'min',
+  'numericCast',
+  'pointwiseMax',
+  'pointwiseMin',
+  'precondition',
+  'preconditionFailure',
+  'print',
+  'readLine',
+  'repeatElement',
+  'sequence',
+  'stride',
+  'swap',
+  'swift_unboxFromSwiftValueWithType',
+  'transcode',
+  'type',
+  'unsafeBitCast',
+  'unsafeDowncast',
+  'withExtendedLifetime',
+  'withUnsafeMutablePointer',
+  'withUnsafePointer',
+  'withVaList',
+  'withoutActuallyEscaping',
+  'zip'
+];
+
+// Valid first characters for operators.
+export const operatorHead = either(
+  /[/=\-+!*%<>&|^~?]/,
+  /[\u00A1-\u00A7]/,
+  /[\u00A9\u00AB]/,
+  /[\u00AC\u00AE]/,
+  /[\u00B0\u00B1]/,
+  /[\u00B6\u00BB\u00BF\u00D7\u00F7]/,
+  /[\u2016-\u2017]/,
+  /[\u2020-\u2027]/,
+  /[\u2030-\u203E]/,
+  /[\u2041-\u2053]/,
+  /[\u2055-\u205E]/,
+  /[\u2190-\u23FF]/,
+  /[\u2500-\u2775]/,
+  /[\u2794-\u2BFF]/,
+  /[\u2E00-\u2E7F]/,
+  /[\u3001-\u3003]/,
+  /[\u3008-\u3020]/,
+  /[\u3030]/
+);
+
+// Valid characters for operators.
+export const operatorCharacter = either(
+  operatorHead,
+  /[\u0300-\u036F]/,
+  /[\u1DC0-\u1DFF]/,
+  /[\u20D0-\u20FF]/,
+  /[\uFE00-\uFE0F]/,
+  /[\uFE20-\uFE2F]/
+  // TODO: The following characters are also allowed, but the regex isn't supported yet.
+  // /[\u{E0100}-\u{E01EF}]/u
+);
+
+// Valid operator.
+export const operator = concat(operatorHead, operatorCharacter, '*');
+
+// Valid first characters for identifiers.
+export const identifierHead = either(
+  /[a-zA-Z_]/,
+  /[\u00A8\u00AA\u00AD\u00AF\u00B2-\u00B5\u00B7-\u00BA]/,
+  /[\u00BC-\u00BE\u00C0-\u00D6\u00D8-\u00F6\u00F8-\u00FF]/,
+  /[\u0100-\u02FF\u0370-\u167F\u1681-\u180D\u180F-\u1DBF]/,
+  /[\u1E00-\u1FFF]/,
+  /[\u200B-\u200D\u202A-\u202E\u203F-\u2040\u2054\u2060-\u206F]/,
+  /[\u2070-\u20CF\u2100-\u218F\u2460-\u24FF\u2776-\u2793]/,
+  /[\u2C00-\u2DFF\u2E80-\u2FFF]/,
+  /[\u3004-\u3007\u3021-\u302F\u3031-\u303F\u3040-\uD7FF]/,
+  /[\uF900-\uFD3D\uFD40-\uFDCF\uFDF0-\uFE1F\uFE30-\uFE44]/,
+  /[\uFE47-\uFFFD]/
+  // The following characters are also allowed, but the regexes aren't supported yet.
+  // /[\u{10000}-\u{1FFFD}\u{20000-\u{2FFFD}\u{30000}-\u{3FFFD}\u{40000}-\u{4FFFD}]/u,
+  // /[\u{50000}-\u{5FFFD}\u{60000-\u{6FFFD}\u{70000}-\u{7FFFD}\u{80000}-\u{8FFFD}]/u,
+  // /[\u{90000}-\u{9FFFD}\u{A0000-\u{AFFFD}\u{B0000}-\u{BFFFD}\u{C0000}-\u{CFFFD}]/u,
+  // /[\u{D0000}-\u{DFFFD}\u{E0000-\u{EFFFD}]/u
+);
+
+// Valid characters for identifiers.
+export const identifierCharacter = either(
+  identifierHead,
+  /\d/,
+  /[\u0300-\u036F\u1DC0-\u1DFF\u20D0-\u20FF\uFE20-\uFE2F]/
+);
+
+// Valid identifier.
+export const identifier = concat(identifierHead, identifierCharacter, '*');
+
+// Built-in attributes, which are highlighted as keywords.
+// @available is handled separately.
+export const keywordAttributes = [
+  'autoclosure',
+  concat(/convention\(/, either('swift', 'block', 'c'), /\)/),
+  'discardableResult',
+  'dynamicCallable',
+  'dynamicMemberLookup',
+  'escaping',
+  'frozen',
+  'GKInspectable',
+  'IBAction',
+  'IBDesignable',
+  'IBInspectable',
+  'IBOutlet',
+  'IBSegueAction',
+  'inlinable',
+  'main',
+  'nonobjc',
+  'NSApplicationMain',
+  'NSCopying',
+  'NSManaged',
+  concat(/objc\(/, identifier, /\)/),
+  'objc',
+  'objcMembers',
+  'propertyWrapper',
+  'requires_stored_property_inits',
+  'testable',
+  'UIApplicationMain',
+  'unknown',
+  'usableFromInline'
+];
+
+// Contextual keywords used in @available and #available.
+export const availabilityKeywords = [
+  'iOS',
+  'iOSApplicationExtension',
+  'macOS',
+  'macOSApplicationExtension',
+  'macCatalyst',
+  'macCatalystApplicationExtension',
+  'watchOS',
+  'watchOSApplicationExtension',
+  'tvOS',
+  'tvOSApplicationExtension',
+  'swift'
+];