Testsuite: munging for Alpine Linux

[user/henk/code/exim.git] / src / src / expand.c

diff --git a/src/src/expand.c b/src/src/expand.c
index fea6501fe881d140aebf5dc096394701ef5ca5ef..590b40383dbbe41254df3822ed2fa1d3545a9c58 100644 (file)
--- a/src/src/expand.c
+++ b/src/src/expand.c
@@ -475,6 +475,7 @@ typedef struct {
 
 typedef uschar * stringptr_fn_t(void);
 static uschar * fn_recipients(void);
+static uschar * fn_recipients_list(void);
 static uschar * fn_queue_size(void);
 
 /* This table must be kept in alphabetical order. */
@@ -694,6 +695,7 @@ static var_entry var_table[] = {
   { "recipient_verify_failure",vtype_stringptr,&recipient_verify_failure },
   { "recipients",          vtype_string_func, (void *) &fn_recipients },
   { "recipients_count",    vtype_int,         &recipients_count },
+  { "recipients_list",     vtype_string_func, (void *) &fn_recipients_list },
   { "regex_cachesize",     vtype_int,         &regex_cachesize },/* undocumented; devel observability */
 #ifdef WITH_CONTENT_SCAN
   { "regex_match_string",  vtype_stringptr,   &regex_match_string },
@@ -839,6 +841,7 @@ uschar * fn_arc_domains(void) {return NULL;}
 uschar * fn_hdrs_added(void) {return NULL;}
 uschar * fn_queue_size(void) {return NULL;}
 uschar * fn_recipients(void) {return NULL;}
+uschar * fn_recipients_list(void) {return NULL;}
 uschar * sender_helo_verified_boolstr(void) {return NULL;}
 uschar * smtp_cmd_hist(void) {return NULL;}
 
@@ -1800,7 +1803,9 @@ return g;
 *************************************************/
 /* A recipients list is available only during system message filtering,
 during ACL processing after DATA, and while expanding pipe commands
-generated from a system filter, but not elsewhere. */
+generated from a system filter, but not elsewhere.  Note that this does
+not check for comman in the elements, and uses comma-space as seperator -
+so cannot be used as an exim list as-is. */
 
 static uschar *
 fn_recipients(void)
@@ -1815,6 +1820,23 @@ for (int i = 0; i < recipients_count; i++)
   s = recipients_list[i].address;
   g = string_append2_listele_n(g, US", ", s, Ustrlen(s));
   }
+gstring_release_unused(g);
+return string_from_gstring(g);
+}
+
+/* Similar, but as a properly-quoted exim list */
+
+
+static uschar *
+fn_recipients_list(void)
+{
+gstring * g = NULL;
+
+if (!f.enable_dollar_recipients) return NULL;
+
+for (int i = 0; i < recipients_count; i++)
+  g = string_append_listele(g, ':', recipients_list[i].address);
+gstring_release_unused(g);
 return string_from_gstring(g);
 }
 
@@ -2119,7 +2141,7 @@ switch (vp->type)
   case vtype_string_func:
     {
     stringptr_fn_t * fn = (stringptr_fn_t *) val;
-    uschar* s = fn();
+    uschar * s = fn();
     return s ? s : US"";
     }
 
@@ -5623,7 +5645,7 @@ while (*s)
       {
       FILE * f;
       const uschar * arg, ** argv;
-      BOOL late_expand = TRUE;
+      unsigned late_expand = TSUC_EXPAND_ARGS | TSUC_ALLOW_TAINTED_ARGS | TSUC_ALLOW_RECIPIENTS;
 
       if (expand_forbid & RDO_RUN)
         {
@@ -5635,7 +5657,7 @@ while (*s)
 
       while (*s == ',')
        if (Ustrncmp(++s, "preexpand", 9) == 0)
-         { late_expand = FALSE; s += 9; }
+         { late_expand = 0; s += 9; }
        else
          {
          const uschar * t = s;
@@ -5695,7 +5717,6 @@ while (*s)
            late_expand,                /* expand args if not already done */
             0,                          /* not relevant when... */
             NULL,                       /* no transporting address */
-           late_expand,                /* allow tainted args, when expand-after-split */
             US"${run} expansion",       /* for error messages */
             &expand_string_message))    /* where to put error message */
           goto EXPAND_FAILED;
@@ -7862,7 +7883,7 @@ NOT_ITEM: ;
        case EOP_UTF8CLEAN:
          {
          int seq_len = 0, index = 0, bytes_left = 0, complete;
-         long codepoint = -1;
+         u_long codepoint = (u_long)-1;
          uschar seq_buff[4];                   /* accumulate utf-8 here */
 
          /* Manually track tainting, as we deal in individual chars below */
@@ -7896,6 +7917,15 @@ NOT_ITEM: ;
                if (--bytes_left == 0)          /* codepoint complete */
                  if(codepoint > 0x10FFFF)      /* is it too large? */
                    complete = -1;      /* error (RFC3629 limit) */
+                 else if ( (codepoint & 0x1FF800 ) == 0xD800 ) /* surrogate */
+                   /* A UTF-16 surrogate (which should be one of a pair that
+                   encode a Unicode codepoint that is outside the Basic
+                   Multilingual Plane).  Error, not UTF8.
+                   RFC2279.2 is slightly unclear on this, but 
+                   https://unicodebook.readthedocs.io/issues.html#strict-utf8-decoder
+                   says "Surrogates characters are also invalid in UTF-8:
+                   characters in U+D800—U+DFFF have to be rejected." */
+                   complete = -1;
                  else
                    {           /* finished; output utf-8 sequence */
                    yield = string_catn(yield, seq_buff, seq_len);
@@ -7905,27 +7935,25 @@ NOT_ITEM: ;
              }
            else        /* no bytes left: new sequence */
              {
-             if(!(c & 0x80))   /* 1-byte sequence, US-ASCII, keep it */
+             if (!(c & 0x80))  /* 1-byte sequence, US-ASCII, keep it */
                {
                yield = string_catn(yield, &c, 1);
                continue;
                }
-             if((c & 0xe0) == 0xc0)            /* 2-byte sequence */
-               {
-               if(c == 0xc0 || c == 0xc1)      /* 0xc0 and 0xc1 are illegal */
+             if ((c & 0xe0) == 0xc0)           /* 2-byte sequence */
+               if (c == 0xc0 || c == 0xc1)     /* 0xc0 and 0xc1 are illegal */
                  complete = -1;
                else
                  {
-                   bytes_left = 1;
-                   codepoint = c & 0x1f;
+                 bytes_left = 1;
+                 codepoint = c & 0x1f;
                  }
-               }
-             else if((c & 0xf0) == 0xe0)               /* 3-byte sequence */
+             else if ((c & 0xf0) == 0xe0)              /* 3-byte sequence */
                {
                bytes_left = 2;
                codepoint = c & 0x0f;
                }
-             else if((c & 0xf8) == 0xf0)               /* 4-byte sequence */
+             else if ((c & 0xf8) == 0xf0)              /* 4-byte sequence */
                {
                bytes_left = 3;
                codepoint = c & 0x07;