[flang][runtime] Handle READ of non-UTF-8 data into multi-byte CHARACTER

[klausler]

Summary

When a READ statement reads into a CHARACTER(2 or 4) variable from a
unit whose encoding is not UTF-8, don't copy bytes directly; they must
each be zero-extended.

flang/runtime/edit-input.cpp

if (connection.isUTF8) {
       } else if (auto ucs{DecodeUTF8(input)}) {
         *x++ = *ucs;
         --length;
       } else if (chunk == 0) {
         // error recovery: skip bad encoding
         chunk = 1;
       }
       --remaining;
+    } else if constexpr (sizeof *x > 1) {
+      // Read single byte with expansion into multi-byte CHARACTER
+      chunk = 1;
+      if (skipping) {
+        --skip;
       } else {
+        *x++ = static_cast<unsigned char>(*input);
+        --length;
+      }
+      --remaining;
+    } else { // single bytes -> default CHARACTER
       if (skipping) {
         chunk = std::min<std::size_t>(skip, ready);
         skip -= chunk;
       } else {
         chunk = std::min<std::size_t>(remaining, ready);
         std::memcpy(x, input, chunk);
         x += chunk;
         length -= chunk;