increase readability of unescape

Signed-off-by: Mattias Andrée <maandree@operamail.com>

1 files changed, 45 insertions, 52 deletions

diff --git a/src/slibc-human/unescape.c b/src/slibc-human/unescape.c
index 941ea9b..90bf61c 100644
--- a/src/slibc-human/unescape.c
+++ b/src/slibc-human/unescape.c
@@ -52,28 +52,33 @@
 char* unescape(char* str, enum unescape_mode mode)
 {
 #define RANGE(a, c, z)  (((a) <= (c)) && ((c) <= (z)))
-#define CxC0(s, m)  (*w++ = (char)((m) | (v >> (s))))
-#define Cx80(s)     (*w++ = (char)(0x80 | ((v >> (s)) & 0x3F)))
-#define PARSE_HEX(v, C)				      \
-  do {						      \
-    char c = (C);				      \
+#define CxC0(s, m)      (*w++ = (char)((m) | (v >> (s))))
+#define Cx80(s)         (*w++ = (char)(0x80 | ((v >> (s)) & 0x3F)))
+#define NEXT_OCTAL(v)   if (RANGE('0', r[1], '7'))  v = (v << 3) | (r[1] - '0'), r++;
+#define PARSE_HEX(START, COND, v)		      \
+  do  for (i = START; COND; i++)  {		      \
+    char c = r[i];				      \
     if      (RANGE('0', c, '9'))  c -= '0';	      \
     else if (RANGE('a', c, 'f'))  c -= 'a', c += 10;  \
     else if (RANGE('A', c, 'F'))  c -= 'A', c += 10;  \
     else					      \
-      goto fail_u;				      \
+      goto unrecognised;			      \
     v = (v << 4) | (unsigned long int)c;	      \
     if (v > 0x10FFFFUL)				      \
-      goto fail_u;				      \
+      goto unrecognised;			      \
   } while (0)
-#define NEXT_OCTAL(v)  if (RANGE('0', r[1], '7'))  v = (v << 3) | (r[1] - '0'), r++;
 #define UNRECOGNISED(c, action)                                     \
   if      (        mode & UNESCAPE_EINVAL)     goto invalid;        \
   else if ((c) && (mode & UNESCAPE_VERBATIM))  action;              \
   else if ((c) && (mode & UNESCAPE_IGNORE))    *w++ = '\\', action
-#define ASCII()  \
+#define ASCII()	 \
   ((v == 0) && (mode & UNESCAPE_MOD_UTF8)) ? (*w++ = (char)0xC0, *w++ = (char)0x80) : \
   (v < 0x80)                               ? (*w++ = (char)v, 1)                    : 0
+#define UTF8()								\
+  if (ASCII());								\
+  else if (v < (1L << 11))  CxC0(6,  0xC0), Cx80(0);			\
+  else if (v < (1L << 16))  CxC0(12, 0xE0), Cx80(6),  Cx80(0);		\
+  else                      CxC0(18, 0xF0), Cx80(12), Cx80(0), Cx80(0);
   
   int i, n;
   unsigned long int v;
@@ -96,24 +101,27 @@ char* unescape(char* str, enum unescape_mode mode)
     }
   
   for (w = r = str; *r; r++)
-    if (*r != '/')
-      *w++ = *r;
-    else
+    {
+      if (*r != '/')
+	{
+	  *w++ = *r;
+	  continue;
+	}
+      
+      n = 0, v = 0;
       switch (*++r)
 	{
-	case '\0':
-	  UNRECOGNISED(1, (void)0);
-	  break;
-	  
-	case '&':
-	  if   (mode & UNESCAPE_AMPERSAND)  *w++ = (char)255;
-	  else UNRECOGNISED(*r, *w++ = '&');
-	  break;
-	  
 #define X(e, c)  case e:  *w++ = c;  break;
 	LIST_BIJECTIVE_ESCAPES
 #undef X
-	case 's':  *w++ = ' ';  break;
+	  
+	case '\0':  UNRECOGNISED(1, (void)0);  break;
+	case 's':   *w++ = ' ';                break;
+	  
+	case '&':
+	  if (mode & UNESCAPE_AMPERSAND)       *w++ = (char)255;
+	  else                                 goto unrecognised;
+	  break;
 	  
 	case '^':
 	  if (RANGE('@', r[1], '_'))  *w++ = *++r - '@';
@@ -125,32 +133,13 @@ char* unescape(char* str, enum unescape_mode mode)
 	    }
 	  break;
 	  
-	case 'u':
-	case 'U':
-	case 'x':
-	  v = 0;
-	  if ((r[0] == 'u') && (r[1] == '{'))
-	    for (i = 2; r[i] != '}'; i++)
-	      PARSE_HEX(v, r[i]);
-	  else
-	    {
-	      switch (*r)
-		{
-		case 'U':  n = 8;  break;
-		case 'u':  n = 4;  break;
-		case 'x':  n = 2;  break;
-		}
-	      for (i = 1; i <= n; i++)
-		PARSE_HEX(v, r[i]);
-	    }
-	  goto done_u;
-	fail_u:
-	  UNRECOGNISED(r--);
-	done_u:
-	  if (ASCII());
-	  else if (v < (1L << 11))  CxC0(6,  0xC0), Cx80(0);
-	  else if (v < (1L << 16))  CxC0(12, 0xE0), Cx80(6),  Cx80(0);
-	  else                      CxC0(18, 0xF0), Cx80(12), Cx80(0), Cx80(0);
+	case 'U':  n += 4;
+	case 'u':  n += 2;
+	case 'x':  n += 2;
+	  if (strstarts(r, "u{"))  PARSE_HEX(2, r[i] != '}', v);
+	  else                     PARSE_HEX(1,    i <= n,   v);
+	  r += i - (r[i] != '}');
+	  UTF8();
 	  break;
 	  
 	default:
@@ -159,16 +148,20 @@ char* unescape(char* str, enum unescape_mode mode)
 	      v = *r - '0';
 	      NEXT_OCTAL(v);
 	      NEXT_OCTAL(v);
-	      if   (ASCII());
-	      else CxC0(6, 0xC0), Cx80(0);
+	      UTF8();
 	    }
 	  else if (strchr("'\"$?\\/", *r))  *w++ = *r;
-#define X(e, i)  else if (strstarts(r, e) ? (*w++ = i, r += sizeof(e) / sizeof(char) - 2, 1) : 0);
+#define X(e, i)  else if (strstarts(r, e))  *w++ = i, r += sizeof(e) / sizeof(char) - 2;
 	  LIST_ASCII_NAMES
 #undef X
-	  else  UNRECOGNISED(*r, r--);
+	  else  goto unrecognised;
 	  break;
 	}
+      
+      continue;
+    unrecognised:
+      UNRECOGNISED(*r, *w++ = *r);
+    }
   
   return *w = 0, w;
  invalid: