improvement: Don't use format since it seems to break at some chars

Related to scalameta/metals#7003
tgodzik · Dec 5, 2024 · 4179938 · 4179938
1 parent 4aa1e0f
commit 4179938
Show file tree

Hide file tree

Showing 2 changed files with 4 additions and 3 deletions.
diff --git a/...ameta/tokenizers/shared/src/main/scala/scala/meta/internal/tokenizers/LegacyScanner.scala b/...ameta/tokenizers/shared/src/main/scala/scala/meta/internal/tokenizers/LegacyScanner.scala
@@ -233,8 +233,8 @@ class LegacyScanner(input: Input, dialect: Dialect) {
     }
 
     @inline
-    def reportIllegalCharacter(): Unit = curr
-      .setInvalidToken(s"illegal character '\\u${"%04x".format(ch)}'")
+    def reportIllegalCharacter(): Unit =
+      curr.setInvalidToken("illegal unicode codepoint: 0x" + ch.toHexString)
 
     (ch: @switch) match {
       case ' ' =>

diff --git a/tests/shared/src/test/scala/scala/meta/tests/tokenizers/TokenizerSuite.scala b/tests/shared/src/test/scala/scala/meta/tests/tokenizers/TokenizerSuite.scala
@@ -2136,6 +2136,7 @@ class TokenizerSuite extends BaseTokenizerSuite {
          |""".stripMargin.replace("'''", "\"\"\"").nl2lf
     assertTokenizedAsStructureLines(code, struct)
   }
+
   test("unexpected-character") {
     val code = """|
                   |val 。 = 123
@@ -2145,7 +2146,7 @@ class TokenizerSuite extends BaseTokenizerSuite {
     val struct = s"""|BOF [0..0)
                      |LF [0..1)
                      |KwVal [1..4)
-                     |Invalid(illegal character '\\u3002') [5..6)
+                     |Invalid(illegal unicode codepoint: 0x3002) [5..6)
                      |MultiHS(3) [4..7)
                      |Equals [7..8)
                      |Space [8..9)