Allow UTF-8 sequences in message subjects (passing K_UTF8 to getstr)

This was the behavior for all string input in SBBS until recently
(Commit 55ccda62), with this change, an exception is being
made for message subjects, but the vast majority of other string
input is still limited to single-byte (i.e. US-ASCII or CP437)
characters.

This should fix issue #720

src/sbbs3/writemsg.cpp

@@ -511,7 +511,7 @@ bool sbbs_t::writemsg(const char *fname, const char *top, char *subj, int mode,
 			max_title_len = LEN_TITLE;
 		if(draft_restored)
 			user_get_property(&cfg, useron.number, draft_desc, "subject", subj, max_title_len);
-		if(!getstr(subj,max_title_len,mode&WM_FILE ? K_LINE|K_TRIM : K_LINE|K_EDIT|K_AUTODEL|K_TRIM)
+		if(!getstr(subj,max_title_len,mode&WM_FILE ? K_LINE|K_TRIM : K_LINE|K_EDIT|K_AUTODEL|K_TRIM|K_UTF8)
 			&& useron_level && useron.logons) {
 			free(buf);
 			return(false);
@@ -1158,7 +1158,7 @@ uint sbbs_t::msgeditor(char *buf, const char *top, char *title, uint maxlines, u
 			else if(!stricmp(strin,"/T")) { /* Edit title/subject */
 				if(title != nulstr) { // hack
 					bputs(text[SubjectPrompt]);
-					getstr(title,LEN_TITLE,K_LINE|K_EDIT|K_AUTODEL|K_TRIM);
+					getstr(title,LEN_TITLE,K_LINE|K_EDIT|K_AUTODEL|K_TRIM|K_UTF8);
 					sync();
 					CRLF;
 				}
@@ -1449,7 +1449,7 @@ bool sbbs_t::forwardmsg(smb_t* smb, smbmsg_t* orgmsg, const char* to, const char
 		else {
 			SAFEPRINTF(subj, "Fwd: %s", orgmsg->subj);
 			bputs(text[SubjectPrompt]);
-			if(!getstr(subj, sizeof(subj) - 1, K_LINE | K_EDIT | K_AUTODEL | K_TRIM))
+			if(!getstr(subj, sizeof(subj) - 1, K_LINE | K_EDIT | K_AUTODEL | K_TRIM | K_UTF8))
 				return false;
 		}
 	}