qmail-smtputf8.patch

This patch adds EAI support to qmail; EAI allows UTF8 almost everywhere in
email.

Thanks for CNNIC for sponsoring this work.

--- /dev/null
+++ netqmail-1.06/Makefile
@@ -1446,7 +1446,7 @@ substdio.a error.a str.a fs.a auto_qmail.o dns.lib socket.lib
 	timeoutwrite.o timeoutconn.o tcpto.o now.o dns.o ip.o \
 	ipalloc.o ipme.o quote.o ndelay.a case.a sig.a open.a \
 	lock.a seek.a getln.a stralloc.a alloc.a substdio.a error.a \
-	str.a fs.a auto_qmail.o  `cat dns.lib` `cat socket.lib`
+	str.a fs.a auto_qmail.o  `cat dns.lib` `cat socket.lib` -lidn2
 
 qmail-remote.0: \
 qmail-remote.8
--- /dev/null
+++ netqmail-1.06/qmail-remote.c
@@ -2,6 +2,7 @@
 #include <sys/socket.h>
 #include <netinet/in.h>
 #include <arpa/inet.h>
+#include <idn2.h>
 #include "sig.h"
 #include "stralloc.h"
 #include "substdio.h"
@@ -42,6 +43,7 @@
 stralloc routes = {0};
 struct constmap maproutes;
 stralloc host = {0};
+stralloc asciihost = {0};
 stralloc sender = {0};
 
 saa reciplist = {0};
@@ -53,12 +55,13 @@
 # include "tls.h"
 # include "ssl_timeoutio.h"
 # include <openssl/x509v3.h>
-# define EHLO 1
 
 int tls_init();
 const char *ssl_err_str = 0;
 #endif 
 
+# define EHLO 1
+
 void out(s) char *s; { if (substdio_puts(subfdoutsmall,s) == -1) _exit(0); }
 void zero() { if (substdio_put(subfdoutsmall,"\0",1) == -1) _exit(0); }
 void zerodie() { zero(); substdio_flush(subfdoutsmall); _exit(0); }
@@ -156,6 +159,7 @@
 substdio smtpfrom = SUBSTDIO_FDBUF(saferead,-1,smtpfrombuf,sizeof smtpfrombuf);
 
 stralloc smtptext = {0};
+stralloc firstpart = {0};
 
 void get(ch)
 char *ch;
@@ -308,6 +312,8 @@
   int r;
   char ch;
 
+  substdio_put(&smtpto,firstpart.s,firstpart.len);
+
   for (;;) {
     r = substdio_get(&ssin,&ch,1);
     if (r == 0) break;
@@ -518,6 +524,88 @@
 
 stralloc recip = {0};
 
+int containsutf8(p, l) unsigned char * p; int l;
+{
+  int i = 0;
+  while (i<l)
+    if(p[i++] > 127) return 1;
+  return 0;
+}
+
+int utf8message;
+
+void checkutf8message()
+{
+  int pos;
+  int i;
+  int r;
+  char ch;
+  int state;
+
+  if (containsutf8(sender.s, sender.len)) { utf8message = 1; return; }
+  for (i = 0;i < reciplist.len;++i)
+    if (containsutf8(reciplist.sa[i].s, reciplist.sa[i].len)) {
+      utf8message = 1;
+      return;
+    }
+
+  state = 0;
+  pos = 0;
+  for (;;) {
+    r = substdio_get(&ssin,&ch,1);
+    if (r == 0) break;
+    if (r == -1) temp_read();
+
+    if (!stralloc_append(&firstpart,&ch)) temp_nomem();
+
+    if (ch == '\r')
+      continue;
+    if (ch == '\t')
+      ch = ' ';
+
+    switch (state) {
+    case 6: /* in Received, at LF but before WITH clause */
+      if (ch == ' ') { state = 3; pos = 1; continue; }
+      state = 0;
+      /* FALL THROUGH */
+
+    case 0: /* start of header field */
+      if (ch == '\n') return;
+      state = 1;
+      pos = 0;
+      /* FALL THROUGH */
+
+    case 1: /* partway through "Received:" */
+      if (ch != "RECEIVED:"[pos] && ch != "received:"[pos]) { state = 2; continue; }
+      if (++pos == 9) { state = 3; pos = 0; }
+      continue;
+
+    case 2: /* other header field */
+      if (ch == '\n') state = 0;
+      continue;
+
+    case 3: /* in Received, before WITH clause or partway though " with " */
+      if (ch == '\n') { state = 6; continue; }
+      if (ch != " WITH "[pos] && ch != " with "[pos]) { pos = 0; continue; }
+      if (++pos == 6) { state = 4; pos = 0; }
+      continue;
+
+    case 4: /* in Received, having seen with, before the argument */
+      if (pos == 0 && (ch == ' ' || ch == '\t')) continue;
+      if (ch != "UTF8"[pos] && ch != "utf8"[pos]) { state = 5; continue; }
+      if(++pos == 4) { utf8message = 1; state = 5; continue; }
+      continue;
+
+    case 5: /* after the RECEIVED WITH argument */
+      /* blast() assumes that it copies whole lines */
+      if (ch == '\n') return;
+      state = 1;
+      pos = 0;
+      continue;
+    }
+  }
+}
+
 void smtp()
 {
   unsigned long code;
@@ -571,9 +659,12 @@
   }
 #endif
  
+  checkutf8message();
   substdio_puts(&smtpto,"MAIL FROM:<");
   substdio_put(&smtpto,sender.s,sender.len);
-  substdio_puts(&smtpto,">\r\n");
+  substdio_puts(&smtpto,">");
+  if (utf8message) substdio_puts(&smtpto," SMTPUTF8");
+  substdio_puts(&smtpto,"\r\n");
   substdio_flush(&smtpto);
   code = smtpcode();
   if (code >= 500) quit("DConnected to "," but sender was rejected");
@@ -702,9 +793,17 @@
       relayhost[i] = 0;
     }
     if (!stralloc_copys(&host,relayhost)) temp_nomem();
+  } else {
+    char * ascii = 0;
+    host.s[host.len] = '\0';
+    switch (idn2_lookup_u8(host.s, (uint8_t**)&ascii, IDN2_NFC_INPUT)) {
+      case IDN2_OK: break;
+      case IDN2_MALLOC: temp_nomem();
+      default: perm_dns();
+    }
+    if (!stralloc_copys(&asciihost, ascii)) temp_nomem();
   }
 
-
   addrmangle(&sender,argv[2],&flagalias,0);
  
   if (!saa_readyplus(&reciplist,0)) temp_nomem();
@@ -723,7 +822,7 @@
 
  
   random = now() + (getpid() << 16);
-  switch (relayhost ? dns_ip(&ip,&host) : dns_mxip(&ip,&host,random)) {
+  switch (relayhost ? dns_ip(&ip,&host) : dns_mxip(&ip,&asciihost,random)) {
     case DNS_MEM: temp_nomem();
     case DNS_SOFT: temp_dns();
     case DNS_HARD: perm_dns();
--- /dev/null
+++ netqmail-1.06/qmail-smtpd.c
@@ -273,6 +273,7 @@
 stralloc rcptto = {0};
 stralloc fuser = {0};
 stralloc mfparms = {0};
+int smtputf8 = 0;
 
 int mailfrom_size(arg) char *arg;
 {
@@ -323,6 +324,7 @@
       while (len) {
         arg++; len--; 
         if (*arg == ' ' || *arg == '\0' ) {
+           if (case_starts(mfparms.s,"SMTPUTF8")) smtputf8 = 1;
            if (case_starts(mfparms.s,"SIZE=")) if (mailfrom_size(mfparms.s+5)) { flagsize = 1; return; }
            if (case_starts(mfparms.s,"AUTH=")) mailfrom_auth(mfparms.s+5,mfparms.len-5);  
            if (!stralloc_copys(&mfparms,"")) die_nomem;
@@ -351,7 +353,7 @@
     out("\r\n250-STARTTLS");
 #endif
   size[fmt_ulong(size,(unsigned int) databytes)] = 0;
-  out("\r\n250-PIPELINING\r\n250-8BITMIME\r\n");
+  out("\r\n250-PIPELINING\r\n250-SMTPUTF8\r\n250-8BITMIME\r\n");
   out("250-SIZE "); out(size); out("\r\n");
 #ifdef CRAM_MD5
   out("250 AUTH LOGIN PLAIN CRAM-MD5\r\n");
@@ -516,7 +518,15 @@
   if (qmail_open(&qqt) == -1) { err_qqt(); return; }
   qp = qmail_qp(&qqt);
   out("354 go ahead\r\n");
- 
+
+  if (smtputf8) {
+    stralloc utf8proto = {0};
+    if ('E' == *protocol) protocol++;
+    if (!stralloc_copys(&utf8proto, "UTF8")) die_nomem();
+    if (!stralloc_cats(&utf8proto, protocol)) die_nomem();
+    utf8proto.s[utf8proto.len] = '\0';
+    protocol = utf8proto.s;
+  }
   received(&qqt,protocol,local,remoteip,remotehost,remoteinfo,fakehelo);
   blast(&hops);
   hops = (hops >= MAXHOPS);