Merge pull request #16090 from hashemi/non-breaking-spaces

[Parser] Detect non-breaking space (U+00A0) and offer a fix-it

Author: rintaro

include/swift/AST/DiagnosticsParse.def

@@ -121,6 +121,8 @@ ERROR(lex_invalid_curly_quote,none,
 NOTE(lex_confusable_character,none,
       "unicode character '%0' looks similar to '%1'; did you mean to use '%1'?",
       (StringRef, StringRef))
+WARNING(lex_nonbreaking_space,none,
+        "non-breaking space (U+00A0) used instead of regular space", ())
 
 ERROR(lex_unterminated_block_comment,none,
       "unterminated '/*' comment", ())

lib/Parse/Lexer.cpp

@@ -682,6 +682,12 @@ static bool isLeftBound(const char *tokBegin, const char *bufferBegin) {
     else
       return true;
 
+  case '\xA0':
+    if (tokBegin - 1 != bufferBegin && tokBegin[-2] == '\xC2')
+      return false; // Non-breaking whitespace (U+00A0)
+    else
+      return true;
+
   default:
     return true;
   }
@@ -716,6 +722,12 @@ static bool isRightBound(const char *tokEnd, bool isLeftBound,
     else
       return true;
 
+  case '\xC2':
+    if (tokEnd[1] == '\xA0')
+      return false; // Non-breaking whitespace (U+00A0)
+    else
+      return true;
+
   default:
     return true;
   }
@@ -1894,6 +1906,12 @@ bool Lexer::lexUnknown(bool EmitDiagnosticsIfToken) {
         .fixItReplaceChars(getSourceLoc(CurPtr - 1), getSourceLoc(Tmp), " ");
     CurPtr = Tmp;
     return false; // Skip presumed whitespace.
+  } else if (Codepoint == 0x000000A0) {
+      // Non-breaking whitespace (U+00A0)
+      diagnose(CurPtr - 1, diag::lex_nonbreaking_space)
+          .fixItReplaceChars(getSourceLoc(CurPtr - 1), getSourceLoc(Tmp), " ");
+      CurPtr = Tmp;
+      return false;
   } else if (Codepoint == 0x0000201D) {
     // If this is an end curly quote, just diagnose it with a fixit hint.
     if (EmitDiagnosticsIfToken) {

test/Parse/nonbreaking_space.swift

@@ -0,0 +1,6 @@
+// RUN: %target-typecheck-verify-swift
+
+let nonBreakingSpace1 = 3 // expected-warning {{non-breaking space (U+00A0) used instead of regular space}} {{22-24= }}
+
+let nonBreakingSpace2 = 3 // expected-warning {{non-breaking space (U+00A0) used instead of regular space}} {{24-26= }}
+

test/Syntax/tokens_nonbreaking_space.swift

@@ -0,0 +1,44 @@
+// RUN: cat %s | sed -e 's/'$(echo -ne "\x5a")'/'$(echo -ne "\xc2\xa0")'/g' > %t.tmp
+// RUN: cp -f %t.tmp %t
+// RUN: %swift-syntax-test -input-source-filename %t -dump-full-tokens 2>&1 | %FileCheck %t
+let a =Z3Z // nbsp(Z)
+let bZ= 3Z
+let cZ=Z3
+
+// CHECK: 4:8: warning: non-breaking space (U+00A0) used instead of regular space
+// CHECK: 4:11: warning: non-breaking space (U+00A0) used instead of regular space
+// CHECK: 5:6: warning: non-breaking space (U+00A0) used instead of regular space
+// CHECK: 5:11: warning: non-breaking space (U+00A0) used instead of regular space
+// CHECK: 6:6: warning: non-breaking space (U+00A0) used instead of regular space
+// CHECK: 6:9: warning: non-breaking space (U+00A0) used instead of regular space
+
+// CHECK-LABEL: 4:7
+// CHECK-NEXT:(Token equal
+// CHECK-NEXT: (text="=")
+// CHECK-NEXT: (trivia garbageText \302\240))
+
+// CHECK-LABEL: 4:10
+// CHECK-NEXT:(Token integer_literal
+// CHECK-NEXT: (text="3")
+// CHECK-NEXT: (trivia garbageText \302\240)
+// CHECK-NEXT: (trivia space 1))
+
+// CHECK-LABEL: 5:5
+// CHECK-NEXT:(Token identifier
+// CHECK-NEXT: (text="b")
+// CHECK-NEXT: (trivia garbageText \302\240))
+
+// CHECK-LABEL: 5:10
+// CHECK-NEXT:(Token integer_literal
+// CHECK-NEXT: (text="3")
+// CHECK-NEXT: (trivia garbageText \302\240)
+
+// CHECK-LABEL: 6:5
+// CHECK-NEXT:(Token identifier
+// CHECK-NEXT: (text="c")
+// CHECK-NEXT: (trivia garbageText \302\240))
+
+// CHECK-LABEL: 6:8
+// CHECK-NEXT:(Token equal
+// CHECK-NEXT: (text="=")
+// CHECK-NEXT: (trivia garbageText \302\240))