fix output handling for subprocesses with multi-byte chars

When using a german locale, program output may contain multi-byte chars such as "Umlaute". 'proc.stdout.read(1)' reads in only a single byte, triggering an exception when trying to decode that single byte.

By wrapping it into an TextIOWrapper, the text is read in char-by-char instead.

scripts/test/benchmark.py

@@ -128,16 +128,17 @@ class TestRun:
     def cmd_live_log(self, log_file, *cmd, cwd=None) -> bool:
         if cwd is None:
             cwd = self.main_dir
-        with open(log_file, "wb") as log:
-            with subprocess.Popen(cmd, stdout=subprocess.PIPE,
-                                  stderr=subprocess.STDOUT, cwd=cwd) as proc:
+        with open(log_file, "wb") as log, \
+            io.TextIOWrapper(log, encoding='utf8') as logger, \
+            subprocess.Popen(cmd, stdout=subprocess.PIPE, \
+                                  stderr=subprocess.STDOUT, cwd=cwd) as proc, \
+            io.TextIOWrapper(proc.stdout, encoding='utf8') as reader:
                 while True:
-                    output = proc.stdout.read(1)
+                    output = reader.read(1)
                     if not output:
                         break
-                    print(output.decode("utf-8"), end='')
-                    log.write(output)
-
+                    print(output, end='')
+                    logger.write(output)
         return proc.returncode == 0
 
     def log(self, msg, stage=None):