(Finsert_file_contents): Avoid detecting a code twice.

diff --git a/src/fileio.c b/src/fileio.c
index cb983cc..c5089cb 100644 (file)
--- a/src/fileio.c
+++ b/src/fileio.c
@@ -4415,11 +4415,10 @@ actually used.  */)
          /* Since we are sure that the current buffer was empty
             before the insertion, we can toggle
             enable-multibyte-characters directly here without taking
-            care of marker adjustment and byte combining problem.  By
-            this way, we can run Lisp program safely before decoding
-            the inserted text.  */
+            care of marker adjustment.  By this way, we can run Lisp
+            program safely before decoding the inserted text.  */
          Lisp_Object unwind_data;
-             int count = SPECPDL_INDEX ();
+         int count = SPECPDL_INDEX ();
 
          unwind_data = Fcons (current_buffer->enable_multibyte_characters,
                               Fcons (current_buffer->undo_list,
@@ -4468,15 +4467,7 @@ actually used.  */)
   if (!NILP (visit))
     {
       /* When we visit a file by raw-text, we change the buffer to
-        unibyte.  If we have not yet decided how to decode a text,
-        decide it at first by detecting the file's encoding.  */
-      if (CODING_REQUIRE_DETECTION (&coding))
-       {
-         coding_system = detect_coding_system (PT_ADDR, inserted, inserted,
-                                               1, 0, coding_system);
-         setup_coding_system (coding_system, &coding);
-       }
-
+        unibyte.  */
       if (CODING_FOR_UNIBYTE (&coding)
          /* Can't do this if part of the buffer might be preserved.  */
          && NILP (replace))