Explicitly set UTF-8 charset in unicode test cases

so that they don't fail on systems where UTF-8 is not properly set on the terminal #1

src/unicode.c

@@ -265,6 +265,7 @@ const char *check_encoding(const char *manual_encoding, const char *system_encod
         }
         fprintf(stderr, "%s: Invalid character encoding: %s - falling back to %s\n",
                 PROJECT, manual_encoding, system_encoding);
+        fflush(stderr);
     }
     return system_encoding;
 }

test/106_unicode_align_center.txt

@@ -1,5 +1,8 @@
+:DESC
+Test center alignment of unicode text without escape sequences mixed in, but with some chars in double-width
+(chinese). We always set the charset manually because else the build would fail on systems which have it set wrong.
 :ARGS
--ac
+-n UTF-8 -ac
 :INPUT
          a
      ab

test/107_unicode_align_left_lolcat.txt

@@ -1,5 +1,5 @@
 :ARGS
--al
+-n UTF-8 -al
 :INPUT
          a
      ab

test/108_unicode_align_center_lolcat.txt

@@ -1,5 +1,5 @@
 :ARGS
--ac
+-n UTF-8 -ac
 :INPUT
          a
      ab

test/109_unicode_align_right_lolcat.txt

@@ -1,5 +1,5 @@
 :ARGS
--ar
+-n UTF-8 -ar
 :INPUT
          a
      ab

test/110_unicode_align_complex_lolcat.txt

@@ -1,5 +1,5 @@
 :ARGS
--ahrvbjr -ph3 -sx26
+-n UTF-8 -ahrvbjr -ph3 -sx26
 :INPUT
          a
      ab

test/117_unicode_ansi_mending.txt

@@ -1,7 +1,7 @@
 :DESC
 This test case checks that a broken box with unicode and ansi sequences can be properly mended.
 :ARGS
--d diamonds -ac -m
+-n UTF-8 -d diamonds -ac -m
 :INPUT
        /\          /\          /\          /\
     /\//\\/\    /\//\\/\    /\//\\/\    /\//\\/\

test/118_unicode_ansi_align_center.txt

@@ -1,7 +1,7 @@
 :DESC
 Center align colored unicode text.
 :ARGS
--d unicornsay -ac
+-n UTF-8 -d unicornsay -ac
 :INPUT
          試
      常盤

test/119_unicode_ansi_align_right.txt

@@ -1,7 +1,7 @@
 :DESC
 Right align colored unicode text.
 :ARGS
--d unicornsay -ar
+-n UTF-8 -d unicornsay -ar
 :INPUT
          試
      常盤