From 16f7c8f6d372c4a5fe209cafdcec0def95d69169 Mon Sep 17 00:00:00 2001
From: Vladimir 'phcoder' Serbinenko <phcoder@gmail.com>
Date: Wed, 8 May 2013 11:09:08 +0200
Subject: [PATCH] 	* grub-core/commands/cat.c: Show UTF-8 characters.

---
 ChangeLog                |  4 ++
 grub-core/commands/cat.c | 85 ++++++++++++++++++++++++++++++++++------
 include/grub/charset.h   | 16 ++++++++
 3 files changed, 94 insertions(+), 11 deletions(-)

diff --git a/ChangeLog b/ChangeLog
index 0da7177af..a57d103b0 100644
--- a/ChangeLog
+++ b/ChangeLog
@@ -1,3 +1,7 @@
+2013-05-08  Vladimir Serbinenko  <phcoder@gmail.com>
+
+	* grub-core/commands/cat.c: Show UTF-8 characters.
+
 2013-05-08  Vladimir Serbinenko  <phcoder@gmail.com>
 
 	* conf/Makefile.common: Poison float and double on non-emu.
diff --git a/grub-core/commands/cat.c b/grub-core/commands/cat.c
index cd2d76099..8a7f1af78 100644
--- a/grub-core/commands/cat.c
+++ b/grub-core/commands/cat.c
@@ -24,6 +24,7 @@
 #include <grub/misc.h>
 #include <grub/extcmd.h>
 #include <grub/i18n.h>
+#include <grub/charset.h>
 
 GRUB_MOD_LICENSE ("GPLv3+");
 
@@ -39,9 +40,15 @@ grub_cmd_cat (grub_extcmd_context_t ctxt, int argc, char **args)
   struct grub_arg_list *state = ctxt->state;
   int dos = 0;
   grub_file_t file;
-  char buf[GRUB_DISK_SECTOR_SIZE];
+  unsigned char buf[GRUB_DISK_SECTOR_SIZE];
   grub_ssize_t size;
   int key = GRUB_TERM_NO_KEY;
+  grub_uint32_t code = 0;
+  int count = 0;
+  unsigned char utbuf[GRUB_MAX_UTF8_PER_CODEPOINT + 1];
+  int utcount = 0;
+  int is_0d = 0;
+  int j;
 
   if (state[0].set)
     dos = 1;
@@ -60,21 +67,65 @@ grub_cmd_cat (grub_extcmd_context_t ctxt, int argc, char **args)
 
       for (i = 0; i < size; i++)
 	{
-	  unsigned char c = buf[i];
+	  utbuf[utcount++] = buf[i];
 
-	  if ((grub_isprint (c) || grub_isspace (c)) && c != '\r')
-	    grub_printf ("%c", c);
-	  else if (dos && c == '\r' && i + 1 < size && buf[i + 1] == '\n')
-	    {
-	      grub_printf ("\n");
-	      i++;
-	    }
-	  else
+	  if (is_0d && buf[i] != '\n')
 	    {
 	      grub_setcolorstate (GRUB_TERM_COLOR_HIGHLIGHT);
-	      grub_printf ("<%x>", (int) c);
+	      grub_printf ("<%x>", (int) '\r');
 	      grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
 	    }
+
+	  is_0d = 0;
+
+	  if (!grub_utf8_process (buf[i], &code, &count))
+	    {
+	      grub_setcolorstate (GRUB_TERM_COLOR_HIGHLIGHT);
+	      for (j = 0; j < utcount - 1; j++)
+		grub_printf ("<%x>", (unsigned int) utbuf[j]);
+	      code = 0;
+	      count = 0;
+	      if (utcount == 1 || !grub_utf8_process (buf[i], &code, &count))
+		{
+		  grub_printf ("<%x>", (unsigned int) buf[i]);
+		  code = 0;
+		  count = 0;
+		  utcount = 0;
+		  grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
+		  continue;
+		}
+	      grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
+	      utcount = 1;
+	    }
+	  if (count)
+	    continue;
+
+	  if ((code >= 0xa1 || grub_isprint (code)
+	       || grub_isspace (code)) && code != '\r')
+	    {
+	      grub_printf ("%C", code);
+	      count = 0; 
+	      code = 0;
+	      utcount = 0;
+	      continue;
+	    }
+
+	  if (dos && code == '\r')
+	    {
+	      is_0d = 1;
+	      count = 0; 
+	      code = 0;
+	      utcount = 0;
+	      continue;
+	    }
+
+	  grub_setcolorstate (GRUB_TERM_COLOR_HIGHLIGHT);
+	  for (j = 0; j < utcount; j++)
+	    grub_printf ("<%x>", (unsigned int) utbuf[j]);
+	  grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
+	  count = 0; 
+	  code = 0;
+	  utcount = 0;
 	}
 
       do
@@ -82,6 +133,18 @@ grub_cmd_cat (grub_extcmd_context_t ctxt, int argc, char **args)
       while (key != GRUB_TERM_ESC && key != GRUB_TERM_NO_KEY);
     }
 
+  if (is_0d)
+    {
+      grub_setcolorstate (GRUB_TERM_COLOR_HIGHLIGHT);
+      grub_printf ("<%x>", (unsigned int) '\r');
+      grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
+    }
+
+  grub_setcolorstate (GRUB_TERM_COLOR_HIGHLIGHT);
+  for (j = 0; j < utcount; j++)
+    grub_printf ("<%x>", (unsigned int) utbuf[j]);
+  grub_setcolorstate (GRUB_TERM_COLOR_STANDARD);
+
   grub_xputs ("\n");
   grub_refresh ();
   grub_file_close (file);
diff --git a/include/grub/charset.h b/include/grub/charset.h
index ab9202884..d14faea32 100644
--- a/include/grub/charset.h
+++ b/include/grub/charset.h
@@ -41,6 +41,7 @@
    You need at least three UTF-8 bytes to have 2 UTF-16 words (surrogate pairs).
  */
 #define GRUB_MAX_UTF16_PER_UTF8 1
+#define GRUB_MAX_UTF8_PER_CODEPOINT 4
 
 #define GRUB_UCS2_LIMIT 0x10000
 #define GRUB_UTF16_UPPER_SURROGATE(code) \
@@ -67,6 +68,14 @@ grub_utf8_process (grub_uint8_t c, grub_uint32_t *code, int *count)
 	  *code <<= 6;
 	  *code |= (c & GRUB_UINT8_6_TRAILINGBITS);
 	  (*count)--;
+	  /* Overlong.  */
+	  if ((*count == 1 && *code <= 0x1f)
+	      || (*count == 2 && *code <= 0xf))
+	    {
+	      *code = 0;
+	      *count = 0;
+	      return 0;
+	    }
 	  return 1;
 	}
     }
@@ -80,6 +89,13 @@ grub_utf8_process (grub_uint8_t c, grub_uint32_t *code, int *count)
     {
       *count = 1;
       *code = c & GRUB_UINT8_5_TRAILINGBITS;
+      /* Overlong */
+      if (*code <= 1)
+	{
+	  *count = 0;
+	  *code = 0;
+	  return 0;
+	}
       return 1;
     }
   if ((c & GRUB_UINT8_4_LEADINGBITS) == GRUB_UINT8_3_LEADINGBITS)