Debug: expansions: refactor ascii-art/UTF8; mark up space & nl

[exim.git] / src / src / string.c

diff --git a/src/src/string.c b/src/src/string.c
index c0a1aad78dfb4dd3b3823afec8d40dddb66c312b..af187c1992e13384795091e03946117bc2c2a584 100644 (file)
--- a/src/src/string.c
+++ b/src/src/string.c
@@ -2,7 +2,7 @@
 *     Exim - an Internet mail transport agent    *
 *************************************************/
 
-/* Copyright (c) The Exim Maintainers 2020 - 2022 */
+/* Copyright (c) The Exim Maintainers 2020 - 2023 */
 /* Copyright (c) University of Cambridge 1995 - 2018 */
 /* See the file NOTICE for conditions of use and distribution. */
 /* SPDX-License-Identifier: GPL-2.0-or-later */
@@ -44,7 +44,6 @@ The legacy string_is_ip_address() function follows below.
 int
 string_is_ip_addressX(const uschar * ip_addr, int * maskptr, const uschar ** errp)
 {
-struct addrinfo hints, * res;
 uschar * slash, * percent, * endp = NULL;
 long int mask = 0;
 const uschar * addr = NULL;
@@ -58,6 +57,7 @@ union { /* we do not need this, but inet_pton() needs a place for storage */
 we return failure, as we do if the mask isn't a pure numerical value,
 or if it is negative. The actual length is checked later, once we know
 the address family. */
+
 if (slash = Ustrchr(ip_addr, '/'))
   {
   uschar * rest;
@@ -75,10 +75,11 @@ if (slash = Ustrchr(ip_addr, '/'))
     return 0;
     }
 
-  *maskptr = slash - ip_addr;     /* offset of the slash */
+  *maskptr = slash - ip_addr;  /* offset of the slash */
   endp = slash;
   }
-else if (maskptr) *maskptr = 0; /* no slash found */
+else if (maskptr)
+  *maskptr = 0;                        /* no slash found */
 
 /* The interface-ID suffix (%<id>) is optional (for IPv6). If it
 exists, we check it syntactically. Later, if we know the address
@@ -161,7 +162,7 @@ switch (af)
 int
 string_is_ip_address(const uschar * ip_addr, int * maskptr)
 {
-return string_is_ip_addressX(ip_addr, maskptr, 0);
+return string_is_ip_addressX(ip_addr, maskptr, NULL);
 }
 
 #endif  /* COMPILE_UTILITY */
@@ -1347,7 +1348,7 @@ The return value can be NULL to signify overflow.
 Field width:           decimal digits, or *
 Precision:             dot, followed by decimal digits or *
 Length modifiers:      h  L  l  ll  z
-Conversion specifiers: n d o u x X p f e E g G % c s S T Y D M
+Conversion specifiers: n d o u x X p f e E g G % c s S T W V Y D M
 
 Returns the possibly-new (if copy for growth or taint-handling was needed)
 string, not nul-terminated.
@@ -1596,11 +1597,77 @@ while (*fp)
     case 'Y':                  /* gstring pointer */
       {
       gstring * zg = va_arg(ap, gstring *);
-      if (zg) { s = CS zg->s; slen = zg->ptr;    }
+      if (zg) { s = CS zg->s; slen = gstring_length(zg); }
       else    { s = null;     slen = Ustrlen(s); }
       goto INSERT_GSTRING;
       }
+#ifndef COMPILE_UTILITY
+    case 'V':                  /* Maybe convert ascii-art to UTF-8 chars */
+      {
+      gstring * zg = NULL;
+      s = va_arg(ap, char *);
+      if (IS_DEBUG(D_noutf8))
+       for ( ; *s; s++)
+         zg = string_catn(zg, CUS (*s == 'K' ? "|" : s), 1);
+      else
+       for ( ; *s; s++) switch (*s)
+         {
+         case '\\': zg = string_catn(zg, US UTF8_UP_RIGHT,       3); break;
+         case '/':  zg = string_catn(zg, US UTF8_DOWN_RIGHT,     3); break;
+         case '-':
+         case '_':  zg = string_catn(zg, US UTF8_HORIZ,          3); break;
+         case '|':  zg = string_catn(zg, US UTF8_VERT,           3); break;
+         case 'K':  zg = string_catn(zg, US UTF8_VERT_RIGHT,     3); break;
+         case '<':  zg = string_catn(zg, US UTF8_LEFT_TRIANGLE,  3); break;
+         case '>':  zg = string_catn(zg, US UTF8_RIGHT_TRIANGLE, 3); break;
+         default:   zg = string_catn(zg, CUS s, 1);                  break;
+         }
+
+      if (!zg)
+       break;
+      s = CS zg->s;
+      slen = gstring_length(zg);
+      goto INSERT_GSTRING;
+      }
+
+    case 'W':                  /* Maybe mark up spaces & newlines */
+      s = va_arg(ap, char *);
+      if (Ustrpbrk(s, " \n") && !IS_DEBUG(D_noutf8))
+       {
+       gstring * zg = NULL;
+       int p = precision;
+       for ( ; *s; s++)
+         {
+         /* Take a given precision as applying to the input; expand
+         it for the transformed result */
+
+         if (p >= 0 && --p < 0) break;
+         switch (*s)
+           {
+           case ' ':
+             zg = string_catn(zg, CUS UTF8_LIGHT_SHADE, 3);
+             if (precision >= 0) precision += 2;
+             break;
+           case '\n':
+             zg = string_catn(zg, CUS UTF8_L_ARROW_HOOK "\n", 4);
+             if (precision >= 0) precision += 3;
+             break;
+           default:
+             zg = string_catn(zg, CUS s, 1);
+             break;
+           }
+         }
+       if (zg) { s = CS zg->s; slen = gstring_length(zg); }
+       else    { s = null;     slen = Ustrlen(s); }
+       }
+      else
+       {
+       if (!s) s = null;
+       slen = Ustrlen(s);
+       }
+      goto INSERT_GSTRING;
 
+#endif
     case 's':
     case 'S':                   /* Forces *lower* case */
     case 'T':                   /* Forces *upper* case */
@@ -1609,7 +1676,7 @@ while (*fp)
       if (!s) s = null;
       slen = Ustrlen(s);
 
-    INSERT_GSTRING:            /* Coome to from %Y above */
+    INSERT_GSTRING:            /* Come to from %Y above */
 
       if (!(flags & SVFMT_TAINT_NOCHK) && is_incompatible(g->s, s))
        if (flags & SVFMT_REBUFFER)
@@ -1908,3 +1975,5 @@ return 0;
 #endif
 
 /* End of string.c */
+/* vi: aw ai sw=2
+*/