Update comment explaining why we read the bytes ourselves

* The encoding is no longer an issue, TruffleFile.FileTypeDetector solves that.

Author: eregon

src/main/java/org/truffleruby/language/loader/FileLoader.java

@@ -64,9 +64,9 @@ public RubySource loadFile(Env env, String path) throws IOException {
         }
 
         /*
-         * We must read the file bytes ourselves - otherwise Truffle will read them, assume they're UTF-8, and we will
-         * not be able to re-interpret the encoding later without the risk of the values being corrupted by being
-         * passed through UTF-8.
+         * We read the file's bytes ourselves because the lexer works on bytes and Truffle only gives us a CharSequence.
+         * We could convert the CharSequence back to bytes, but that's more expensive than just reading the bytes once
+         * and pass them down to the lexer and to the Source.
          */
 
         final byte[] sourceBytes = file.readAllBytes();

src/main/java/org/truffleruby/language/loader/MainLoader.java

@@ -85,9 +85,9 @@ public RubySource loadFromFile(Env env, RubyNode currentNode, String path) throw
         fileLoader.ensureReadable(file);
 
         /*
-         * We must read the file bytes ourselves - otherwise Truffle will read them, assume they're UTF-8, and we will
-         * not be able to re-interpret the encoding later without the risk of the values being corrupted by being
-         * passed through UTF-8.
+         * We read the file's bytes ourselves because the lexer works on bytes and Truffle only gives us a CharSequence.
+         * We could convert the CharSequence back to bytes, but that's more expensive than just reading the bytes once
+         * and pass them down to the lexer and to the Source.
          */
 
         byte[] sourceBytes = file.readAllBytes();