Expansions: New ${ipv6denorm:<string>} and ${ipv6norm:<string>} operators. Bug 1650

7 files changed, 148 insertions, 8 deletions

diff --git a/doc/doc-docbook/spec.xfpt b/doc/doc-docbook/spec.xfpt
index 101df6b90..54480ee8d 100644
--- a/doc/doc-docbook/spec.xfpt
+++ b/doc/doc-docbook/spec.xfpt
@@ -10107,6 +10107,27 @@ as is, and other byte values are converted to &`\xNN`&, for example a
 byte value 127 is converted to &`\x7f`&.
 
 
+.new
+.vitem &*${ipv6denorm:*&<&'string'&>&*}*&
+.cindex "&%ipv6denorm%& expansion item"
+.cindex "IP address" normalisation
+This expands an IPv6 address to a full eight-element colon-separated set
+of hex digits including leading zeroes.
+A trailing ipv4-style dotted-decimal set is converted to hex.
+Pure IPv4 addresses are converted to IPv4-mapped IPv6.
+
+.vitem &*${ipv6norm:*&<&'string'&>&*}*&
+.cindex "&%ipv6norm%& expansion item"
+.cindex "IP address" normalisation
+.cindex "IP address" "canonical form"
+This converts an IPv6 address to canonical form.
+Leading zeroes of groups are omitted, and the longest
+set of zero-valued groups is replaced with a double colon.
+A trailing ipv4-style dotted-decimal set is converted to hex.
+Pure IPv4 addresses are converted to IPv4-mapped IPv6.
+.wen
+
+
 .vitem &*${lc:*&<&'string'&>&*}*&
 .cindex "case forcing in strings"
 .cindex "string" "case forcing"
diff --git a/doc/doc-txt/NewStuff b/doc/doc-txt/NewStuff
index a7185a4e2..ca65c98ea 100644
--- a/doc/doc-txt/NewStuff
+++ b/doc/doc-txt/NewStuff
@@ -17,6 +17,13 @@ Version 4.87
 
  3. Transports now take a "max_parallel" option, to limit concurrency.
 
+ 4. Expansion operators ${ipv6norm:<string>} and ${ipv6denorm:<string>}.
+    The latter expands to a 8-element colon-sep set of hex digits including
+    leading zeroes. A trailing ipv4-style dotted-decimal set is converted
+    to hex.  Pure ipv4 addresses are converted to IPv4-mapped IPv6.
+    The former operator strips leading zeroes and collapses the longest
+    set of 0-groups to a double-colon.
+
 
 Version 4.86
 ------------
diff --git a/src/src/expand.c b/src/src/expand.c
index fbbc68154..90ffe78c0 100644
--- a/src/src/expand.c
+++ b/src/src/expand.c
@@ -210,6 +210,8 @@ static uschar *op_table_main[] = {
   US"hash",
   US"hex2b64",
   US"hexquote",
+  US"ipv6denorm",
+  US"ipv6norm",
   US"l",
   US"lc",
   US"length",
@@ -248,6 +250,8 @@ enum {
   EOP_HASH,
   EOP_HEX2B64,
   EOP_HEXQUOTE,
+  EOP_IPV6DENORM,
+  EOP_IPV6NORM,
   EOP_L,
   EOP_LC,
   EOP_LENGTH,
@@ -6406,6 +6410,39 @@ while (*s != 0)
         continue;
         }
 
+      case EOP_IPV6NORM:
+      case EOP_IPV6DENORM:
+	{
+        int type = string_is_ip_address(sub, NULL);
+	int binary[4];
+	uschar buffer[44];
+
+	switch (type)
+	  {
+	  case 6:
+	    (void) host_aton(sub, binary);
+	    break;
+
+	  case 4:	/* convert to IPv4-mapped IPv6 */
+	    binary[0] = binary[1] = 0;
+	    binary[2] = 0x0000ffff;
+	    (void) host_aton(sub, binary+3);
+	    break;
+
+	  case 0:
+	    expand_string_message =
+	      string_sprintf("\"%s\" is not an IP address", sub);
+	    goto EXPAND_FAILED;
+	  }
+
+	yield = string_cat(yield, &size, &ptr, buffer,
+		  c == EOP_IPV6NORM
+		    ? ipv6_nmtoa(binary, buffer)
+		    : host_nmtoa(4, binary, -1, buffer, ':')
+		  );
+	continue;
+	}
+
       case EOP_ADDRESS:
       case EOP_LOCAL_PART:
       case EOP_DOMAIN:
diff --git a/src/src/functions.h b/src/src/functions.h
index 2fe681394..9df8030f5 100644
--- a/src/src/functions.h
+++ b/src/src/functions.h
@@ -233,6 +233,8 @@ extern int     ip_tcpsocket(const uschar *, uschar **, int);
 extern int     ip_unixsocket(const uschar *, uschar **);
 extern int     ip_streamsocket(const uschar *, uschar **, int);
 
+extern int     ipv6_nmtoa(int *, uschar *);
+
 extern uschar *local_part_quote(uschar *);
 extern int     log_create(uschar *);
 extern int     log_create_as_exim(uschar *);
diff --git a/src/src/host.c b/src/src/host.c
index 429b6352c..67d33a9b8 100644
--- a/src/src/host.c
+++ b/src/src/host.c
@@ -1146,20 +1146,14 @@ if (count == 1)
   {
   j = binary[0];
   for (i = 24; i >= 0; i -= 8)
-    {
-    sprintf(CS tt, "%d.", (j >> i) & 255);
-    while (*tt) tt++;
-    }
+    tt += sprintf(CS tt, "%d.", (j >> i) & 255);
   }
 else
-  {
   for (i = 0; i < 4; i++)
     {
     j = binary[i];
-    sprintf(CS tt, "%04x%c%04x%c", (j >> 16) & 0xffff, sep, j & 0xffff, sep);
-    while (*tt) tt++;
+    tt += sprintf(CS tt, "%04x%c%04x%c", (j >> 16) & 0xffff, sep, j & 0xffff, sep);
     }
-  }
 
 tt--;   /* lose final separator */
 
@@ -1175,6 +1169,63 @@ return tt - buffer;
 }
 
 
+/* Like host_nmtoa() but: ipv6-only, canonical output, no mask
+
+Arguments:
+  binary      points to the ints
+  buffer      big enough to hold the result
+
+Returns:      the number of characters placed in buffer, not counting
+	      the final nul.
+*/
+
+int
+ipv6_nmtoa(int * binary, uschar * buffer)
+{
+int i, j, k;
+uschar * c = buffer;
+uschar * d;
+
+for (i = 0; i < 4; i++)
+  {			/* expand to text */
+  j = binary[i];
+  c += sprintf(CS c, "%x:%x:", (j >> 16) & 0xffff, j & 0xffff);
+  }
+
+for (c = buffer, k = -1, i = 0; i < 8; i++)
+  {			/* find longest 0-group sequence */
+  if (*c == '0')	/* must be "0:" */
+    {
+    uschar * s = c;
+    j = i;
+    while (c[2] == '0') i++, c += 2;
+    if (i-j > k)
+      {
+      k = i-j;		/* length of sequence */
+      d = s;		/* start of sequence */
+      }
+    }
+  while (*++c != ':') ;
+  c++;
+  }
+
+c[-1] = '\0';	/* drop trailing colon */
+
+/* debug_printf("%s: D k %d <%s> <%s>\n", __FUNCTION__, k, d, d + 2*(k+1)); */
+if (k >= 0)
+  {			/* collapse */
+  c = d + 2*(k+1);
+  if (d == buffer) c--;	/* need extra colon */
+  *d++ = ':';	/* 1st 0 */
+  while (*d++ = *c++) ;
+  }
+else
+  d = c;
+
+return d - buffer;
+}
+
+
 
 /*************************************************
 *        Check port for tls_on_connect           *
diff --git a/test/scripts/0000-Basic/0002 b/test/scripts/0000-Basic/0002
index 9afb556a3..bf2c1068c 100644
--- a/test/scripts/0000-Basic/0002
+++ b/test/scripts/0000-Basic/0002
@@ -208,6 +208,17 @@ mask:   ${mask:192.168.10.206/33}
 mask:   ${mask:192.168.10.206/0}
 mask:   ${mask:192.168.10.206}
 mask:   ${mask:a.b.c.d}
+ipv6denorm: ${ipv6denorm:::1}
+ipv6denorm: ${ipv6denorm:fe00::1}
+ipv6denorm: ${ipv6denorm:192.168.0.1}
+ipv6denorm: ${ipv6denorm:fe80::192.168.0.1}
+ipv6norm:   ${ipv6norm:0:0:0::1}
+ipv6norm:   ${ipv6norm:2a00::0}
+ipv6norm:   ${ipv6norm:2a00::1}
+ipv6norm:   ${ipv6norm:2a00:eadf:0000:0000:0000:0000:0001:0000}
+ipv6norm:   ${ipv6norm:2a00:eadf:0000:0001:0000:0000:0000:0000}
+ipv6norm:   ${ipv6norm:2a00:0:0:0::}
+ipv6norm:   ${ipv6norm:2a00:2:3:4:5:6:7:8}
 nhash:  ${nhash_24:monty} ${nhash_8_63:monty python}
 lc/uc:  ${lc:The Quick} ${uc: Brown Fox}
 length: ${length_10:The quick brown fox} ${l_10:abc}
diff --git a/test/stdout/0002 b/test/stdout/0002
index 9a3219d59..3018dce1f 100644
--- a/test/stdout/0002
+++ b/test/stdout/0002
@@ -189,6 +189,17 @@
 > mask:   0.0.0.0/0
 > Failed: missing mask value in "192.168.10.206"
 > Failed: "a.b.c.d" is not an IP address
+> ipv6denorm: 0000:0000:0000:0000:0000:0000:0000:0001
+> ipv6denorm: fe00:0000:0000:0000:0000:0000:0000:0001
+> ipv6denorm: 0000:0000:0000:0000:0000:ffff:c0a8:0001
+> ipv6denorm: fe80:0000:0000:0000:0000:0000:c0a8:0001
+> ipv6norm:   ::1
+> ipv6norm:   2a00::
+> ipv6norm:   2a00::1
+> ipv6norm:   2a00:eadf::1:0
+> ipv6norm:   2a00:eadf:0:1::
+> ipv6norm:   2a00::
+> ipv6norm:   2a00:2:3:4:5:6:7:8
 > nhash:  19 0/61
 > lc/uc:  the quick  BROWN FOX
 > length: The quick  abc