[SE-0163] Add Latin1 to Unicode codecs

Author: benrimmington

stdlib/public/core/CMakeLists.txt

@@ -2,7 +2,7 @@
 #
 # This source file is part of the Swift.org open source project
 #
-# Copyright (c) 2014 - 2019 Apple Inc. and the Swift project authors
+# Copyright (c) 2014 - 2020 Apple Inc. and the Swift project authors
 # Licensed under Apache License v2.0 with Runtime Library Exception
 #
 # See https://swift.org/LICENSE.txt for license information
@@ -219,6 +219,7 @@ set(SWIFTLIB_SOURCES
   SliceBuffer.swift
   SIMDVector.swift
   UnfoldSequence.swift
+  UnicodeLatin1.swift
   VarArgs.swift
   Zip.swift
   "${SWIFT_SOURCE_DIR}/stdlib/linker-support/magic-symbols-for-install-name.c"

stdlib/public/core/GroupInfo.json

@@ -43,6 +43,7 @@
     "Unicode.swift",
     "UnicodeEncoding.swift",
     "UnicodeHelpers.swift",
+    "UnicodeLatin1.swift",
     "UnicodeParser.swift",
     "UnicodeScalar.swift",
     "UnicodeScalarProperties.swift",

stdlib/public/core/UnicodeLatin1.swift

@@ -0,0 +1,66 @@
+//===--- UnicodeLatin1.swift ----------------------------------------------===//
+//
+// This source file is part of the Swift.org open source project
+//
+// Copyright (c) 2020 Apple Inc. and the Swift project authors
+// Licensed under Apache License v2.0 with Runtime Library Exception
+//
+// See https://swift.org/LICENSE.txt for license information
+// See https://swift.org/CONTRIBUTORS.txt for the list of Swift project authors
+//
+//===----------------------------------------------------------------------===//
+
+@available(macOS 9999, iOS 9999, tvOS 9999, watchOS 9999, *)
+extension Unicode {
+
+  @frozen
+  public enum Latin1 {}
+}
+
+@available(macOS 9999, iOS 9999, tvOS 9999, watchOS 9999, *)
+extension Unicode.Latin1: Unicode.Encoding {
+
+  @frozen
+  public struct Parser {
+
+    @inlinable
+    public init() {}
+  }
+
+  public typealias CodeUnit = UInt8
+  public typealias EncodedScalar = CollectionOfOne<CodeUnit>
+  public typealias ForwardParser = Parser
+  public typealias ReverseParser = Parser
+
+  @inlinable
+  public static var encodedReplacementCharacter: EncodedScalar {
+    EncodedScalar(0x1A) // U+001A SUBSTITUTE
+  }
+
+  @inlinable
+  public static func decode(_ encodedScalar: EncodedScalar) -> Unicode.Scalar {
+    Unicode.Scalar(encodedScalar[0])
+  }
+
+  @inlinable
+  public static func encode(_ unicodeScalar: Unicode.Scalar) -> EncodedScalar? {
+    CodeUnit(exactly: unicodeScalar.value).map { EncodedScalar($0) }
+  }
+
+  @inlinable
+  public static func _isScalar(_: CodeUnit) -> Bool { true }
+}
+
+@available(macOS 9999, iOS 9999, tvOS 9999, watchOS 9999, *)
+extension Unicode.Latin1.Parser: Unicode.Parser {
+
+  public typealias Encoding = Unicode.Latin1
+
+  @inlinable
+  public mutating func parseScalar<I: IteratorProtocol>(
+    from codeUnits: inout I
+  ) -> Unicode.ParseResult<Encoding.EncodedScalar>
+  where I.Element == Encoding.CodeUnit {
+    codeUnits.next().map { .valid(Encoding.EncodedScalar($0)) } ?? .emptyInput
+  }
+}

test/stdlib/Unicode.swift

@@ -2,7 +2,7 @@
 //
 // This source file is part of the Swift.org open source project
 //
-// Copyright (c) 2014 - 2017 Apple Inc. and the Swift project authors
+// Copyright (c) 2014 - 2020 Apple Inc. and the Swift project authors
 // Licensed under Apache License v2.0 with Runtime Library Exception
 //
 // See https://swift.org/LICENSE.txt for license information
@@ -110,4 +110,81 @@ UnicodeAPIs.test("UTF-8 and UTF-16 queries") {
   }
 }
 
+if #available(macOS 9999, iOS 9999, tvOS 9999, watchOS 9999, *) {
+  var UnicodeLatin1 = TestSuite("UnicodeLatin1")
+
+  UnicodeLatin1.test("Encoding") {
+    let unicodeScalars = (UInt8.min ... UInt8.max).map { Unicode.Scalar($0) }
+    let encodedScalars = unicodeScalars.compactMap { Unicode.Latin1.encode($0) }
+    let decodedScalars = encodedScalars.map { Unicode.Latin1.decode($0) }
+    expectEqualSequence(unicodeScalars, decodedScalars)
+    expectNil(Unicode.Latin1.encode("\u{0100}"))
+    expectNil(Unicode.Latin1.encode("\u{10FFFF}"))
+  }
+
+  UnicodeLatin1.test("Parser") {
+    let codeUnits = UInt8.min ... UInt8.max
+    var codeUnitsIterator = codeUnits.makeIterator()
+    var encodedScalars: [Unicode.Latin1.EncodedScalar] = []
+    var forwardParser = Unicode.Latin1.ForwardParser()
+    loop: while true {
+      switch forwardParser.parseScalar(from: &codeUnitsIterator) {
+      case .valid(let encodedScalar):
+        encodedScalars.append(encodedScalar)
+      case .emptyInput:
+        expectEqualSequence(codeUnits, encodedScalars.joined())
+        break loop
+      case .error:
+        expectUnreachable()
+        break loop
+      }
+    }
+  }
+
+  UnicodeLatin1.test("Transcode") {
+    let codeUnitsAndText: [ClosedRange<UInt8>: String] = [
+      UInt8(0x20) ... UInt8(0x7E):
+        """
+        \u{20}!"#$%&'()*+,-./0123456789:;<=>?\
+        @ABCDEFGHIJKLMNOPQRSTUVWXYZ[\\]^_\
+        `abcdefghijklmnopqrstuvwxyz{|}~
+        """,
+      UInt8(0xA0) ... UInt8(0xFF):
+        """
+        \u{A0}¡¢£¤¥¦§¨©ª«¬\u{AD}®¯°±²³´µ¶·¸¹º»¼½¾¿\
+        ÀÁÂÃÄÅÆÇÈÉÊËÌÍÎÏÐÑÒÓÔÕÖ×ØÙÚÛÜÝÞß\
+        àáâãäåæçèéêëìíîïðñòóôõö÷øùúûüýþÿ
+        """,
+    ]
+    for (codeUnits, expectedText) in codeUnitsAndText {
+      let actualText = String(decoding: codeUnits, as: Unicode.Latin1.self)
+      expectEqual(expectedText, actualText)
+    }
+    for (expectedCodeUnits, text) in codeUnitsAndText {
+      var actualCodeUnits: [UInt8] = []
+      let hadError = transcode(
+        text.utf8.makeIterator(),
+        from: Unicode.UTF8.self,
+        to: Unicode.Latin1.self,
+        stoppingOnError: false,
+        into: { actualCodeUnits.append($0) }
+      )
+      expectEqualSequence(expectedCodeUnits, actualCodeUnits)
+      expectFalse(hadError)
+    }
+    do {
+      var actualCodeUnits: [UInt8] = []
+      let hadError = transcode(
+        "A\u{0100}B\u{10FFFF}C".utf8.makeIterator(),
+        from: Unicode.UTF8.self,
+        to: Unicode.Latin1.self,
+        stoppingOnError: false,
+        into: { actualCodeUnits.append($0) }
+      )
+      expectEqualSequence([0x41, 0x1A, 0x42, 0x1A, 0x43], actualCodeUnits)
+      expectFalse(hadError)
+    }
+  }
+}
+
 runAllTests()