repost: [PATCH] SMPP: data_coding (deliver_sm) votes?!

? charset.patch
? data_coding-patch
? data_coding.patch
? migrate.sh
? smsc_smpp.c-charset
? smsc_smpp.c-diff
? smsc_smpp.c-patch
? smsc_smpp.c-tcl
? smsc_smpp.patch
? doc/arch/arch.aux
? doc/arch/arch.dvi
? doc/arch/arch.log
? doc/arch/arch.tex
? doc/userguide/userguide.xml-patch
? gw/diff
? gw/smsc_at2.c.diff
Index: config.h.in
===================================================================
RCS file: /home/cvs/gateway/config.h.in,v
retrieving revision 1.42
diff -u -r1.42 config.h.in
--- config.h.in	9 Sep 2002 10:49:28 -0000	1.42
+++ config.h.in	6 Nov 2002 07:35:37 -0000
@@ -113,6 +113,9 @@
 /* Define if you have <syslog.h>.  */
 #undef HAVE_SYSLOG_H
 
+/* Define if you have <iconv.h>.  */
+#undef HAVE_ICONV_H
+
 /* Define if you have and want to use the ssl library (-lssl) */
 #undef HAVE_LIBSSL
 
Index: configure
===================================================================
RCS file: /home/cvs/gateway/configure,v
retrieving revision 1.105
diff -u -r1.105 configure
--- configure	9 Sep 2002 10:49:28 -0000	1.105
+++ configure	6 Nov 2002 07:35:38 -0000
@@ -1781,7 +1781,7 @@
 fi
 done
 
-for ac_hdr in pthread.h getopt.h syslog.h
+for ac_hdr in pthread.h getopt.h syslog.h iconv.h
 do
 ac_safe=`echo "$ac_hdr" | sed 'y%./+-%__p_%'`
 echo $ac_n "checking for $ac_hdr""... $ac_c" 1>&6
Index: configure.in
===================================================================
RCS file: /home/cvs/gateway/configure.in,v
retrieving revision 1.107
diff -u -r1.107 configure.in
--- configure.in	9 Sep 2002 10:49:28 -0000	1.107
+++ configure.in	6 Nov 2002 07:35:39 -0000
@@ -95,7 +95,7 @@
 
 AC_HEADER_STDC
 AC_CHECK_HEADERS(sys/ioctl.h sys/time.h sys/types.h unistd.h sys/poll.h)
-AC_CHECK_HEADERS(pthread.h getopt.h syslog.h)
+AC_CHECK_HEADERS(pthread.h getopt.h syslog.h iconv.h)
 
 dnl Checks for typedefs, structures, and compiler characteristics.
 
Index: gw/smsc/smsc_smpp.c
===================================================================
RCS file: /home/cvs/gateway/gw/smsc/smsc_smpp.c,v
retrieving revision 1.14
diff -u -r1.14 smsc_smpp.c
--- gw/smsc/smsc_smpp.c	17 Oct 2002 08:28:10 -0000	1.14
+++ gw/smsc/smsc_smpp.c	6 Nov 2002 07:35:39 -0000
@@ -102,6 +102,7 @@
     time_t throttling_err_time;
     int smpp_msg_id_type;  /* msg id in C string, hex or decimal */
     int autodetect_addr;
+    Octstr* alt_charset;
     SMSCConn *conn; 
 } SMPP; 
  
@@ -115,7 +116,8 @@
                          int alt_dcs, int enquire_link_interval, 
                          int max_pending_submits, int reconnect_delay,
                          int version, int priority, Octstr *my_number,
-                         int smpp_msg_id_type, int autodetect_addr) 
+                         int smpp_msg_id_type, int autodetect_addr, 
+			 Octstr* alt_charset)
 { 
     SMPP *smpp; 
      
@@ -147,6 +149,7 @@
     smpp->quitting = 0; 
     smpp->version = version;
     smpp->priority = priority;
+    smpp->alt_charset = octstr_duplicate(alt_charset);
     smpp->conn = conn; 
     smpp->throttling_err_time = 0; 
     smpp->smpp_msg_id_type = smpp_msg_id_type;    
@@ -169,6 +172,8 @@
 	octstr_destroy(smpp->system_type); 
 	octstr_destroy(smpp->address_range); 
 	octstr_destroy(smpp->our_host); 
+        octstr_destroy(smpp->my_number);
+        octstr_destroy(smpp->alt_charset);
 	gw_free(smpp); 
     } 
 } 
@@ -223,7 +228,7 @@
 } 
  
  
-static Msg *pdu_to_msg(SMPP_PDU *pdu) 
+static Msg *pdu_to_msg(SMPP *smpp, SMPP_PDU *pdu) 
 { 
     Msg *msg; 
  
@@ -236,9 +241,44 @@
     pdu->u.deliver_sm.destination_addr = NULL; 
     msg->sms.msgdata = pdu->u.deliver_sm.short_message; 
     pdu->u.deliver_sm.short_message = NULL; 
-    charset_gsm_to_latin1(msg->sms.msgdata); 
+    dcs_to_fields(&msg, pdu->u.deliver_sm.data_coding);
+
+    switch (pdu->u.deliver_sm.data_coding) {// handle default data coding
+        case 0x00: // default SMSC alphabet
+            if (smpp->alt_charset) { // try to convert from something interesting if specified so
+                if (octstr_compare(smpp->alt_charset, octstr_imm("binary")) != 0) // unless it was specified binary
+                    if (charset_convert(msg->sms.msgdata, octstr_get_cstr(smpp->alt_charset), "ISO-8859-1") != 0)
+                        error(0, "Failed to convert msgdata from %s to %s, will leave as is",
+                             octstr_get_cstr(smpp->alt_charset), "ISO-8859-1");
+                msg->sms.coding = DC_8BIT;
+
+            } else { // assume GSM 03.38 7 bit alphabet
+                charset_gsm_to_latin1(msg->sms.msgdata); 
+                msg->sms.coding = DC_7BIT;
+            }
+            break;
+        case 0x01: // ASCII or IA5 - not sure if I need to do anything
+        case 0x02: // 8 bit binary - do nothing
+        case 0x04: // 8 bit binary - do nothing
+                break;
+        case 0x03: // ISO-8859-1 - do nothing
+                msg->sms.coding = DC_8BIT; break;
+        case 0x05: // JIS - what do I do with that ?
+                break;
+        case 0x06: // Cyrllic - iso-8859-5, I'll convert to unicode
+            if (charset_convert(msg->sms.msgdata, "ISO-8859-5", "UCS-2BE") != 0)
+                error(0, "Failed to convert msgdata from cyrllic to UCS-2, will leave as is");
+            msg->sms.coding = DC_UCS2; break;
+        case 0x07: // Hebrew iso-8859-8, I'll convert to unicode
+            if (charset_convert(msg->sms.msgdata, "ISO-8859-8", "UCS-2BE") != 0)
+                error(0, "Failed to convert msgdata from hebrew to UCS-2, will leave as is");
+            msg->sms.coding = DC_UCS2; break;
+        case 0x08: // unicode UCS-2, yey
+            msg->sms.coding = DC_UCS2; break;
+        default: // don't much care about the others - you implement them if you feel like it
+            msg->sms.coding = DC_8BIT;
+    }
     msg->sms.pid = pdu->u.deliver_sm.protocol_id; 
-    dcs_to_fields(&msg, pdu->u.deliver_sm.data_coding); 
  
     return msg; 
 } 
@@ -359,8 +399,16 @@
 	       octstr_format("%S%S", msg->sms.udhdata, msg->sms.msgdata); 
     } else { 
         pdu->u.submit_sm.short_message = octstr_duplicate(msg->sms.msgdata); 
-        if (pdu->u.submit_sm.data_coding == 0 ) /* no reencoding for unicode! */ 
-            charset_latin1_to_gsm(pdu->u.submit_sm.short_message);		 
+        if (pdu->u.submit_sm.data_coding == 0 ) /* no reencoding for unicode! */ {
+	    if (smpp->alt_charset && octstr_compare(smpp->alt_charset, octstr_imm("binary")) != 0) {
+		if (charset_convert(pdu->u.submit_sm.short_message, "ISO-8859-1",
+                    octstr_get_cstr(smpp->alt_charset)) != 0)
+        	    error(0, "Failed to convert msgdata from %s to %s will send as is", 
+	    		"ISO-8859-1", octstr_get_cstr(smpp->alt_charset));
+		debug("smsc_smpp.c", 0, "alt_charset!=0");
+	    } else
+                charset_latin1_to_gsm(pdu->u.submit_sm.short_message);		 
+        } 
     } 
 
     /*
@@ -724,7 +772,7 @@
             } else /* MO-SMS */
             {
                 /* ensure the smsc-id is set */ 
-                msg = pdu_to_msg(pdu); 
+                msg = pdu_to_msg(smpp, pdu); 
  
                 /* Replace MO destination number with my-number */ 
                 if (octstr_len(smpp->my_number)) { 
@@ -1126,6 +1174,7 @@
     long max_pending_submits;
     long reconnect_delay;
     long version;
+    Octstr* alt_charset;
     long priority;
     long smpp_msg_id_type;
     int autodetect_addr;
@@ -1234,13 +1283,17 @@
             panic(0,"SMPP: Invlid value for msg-id-type directive in configuraton"); 
     }
 
+    
+    /* check for charset specific */
+    alt_charset = cfg_get(grp, octstr_imm("alt-charset"));
+
     smpp = smpp_create(conn, host, port, receive_port, system_type,  
     	    	       username, password, address_range, our_host, 
                        source_addr_ton, source_addr_npi, dest_addr_ton,  
                        dest_addr_npi, alt_dcs, enquire_link_interval, 
                        max_pending_submits, reconnect_delay, 
                        version, priority, my_number, smpp_msg_id_type,
-                       autodetect_addr); 
+                       autodetect_addr, alt_charset); 
  
     conn->data = smpp; 
     conn->name = octstr_format("SMPP:%S:%d/%d:%S:%S",  
@@ -1261,6 +1314,7 @@
     octstr_destroy(our_host); 
     octstr_destroy(my_number); 
     octstr_destroy(smsc_id);
+    octstr_destroy(alt_charset);
  
     conn->status = SMSCCONN_CONNECTING; 
        
Index: gwlib/charset.c
===================================================================
RCS file: /home/cvs/gateway/gwlib/charset.c,v
retrieving revision 1.11
diff -u -r1.11 charset.c
--- gwlib/charset.c	25 Jan 2002 11:49:05 -0000	1.11
+++ gwlib/charset.c	6 Nov 2002 07:35:39 -0000
@@ -7,6 +7,11 @@
  */
 
 #include "gwlib/gwlib.h"
+#include <errno.h>
+
+#if HAVE_ICONV_H
+#include <iconv.h>
+#endif
 
 /* Map GSM default alphabet characters to ISO-Latin-1 characters.
  * The greek characters at positions 16 and 18 through 26 are not
@@ -400,4 +405,50 @@
     xmlBufferFree(frombuffer);
 
     return ret;
+}
+
+int charset_convert(Octstr* string, char* charset_from, char* charset_to)
+{
+#if HAVE_ICONV_H
+    char *from_buf, *to_buf, *pointer;
+    size_t inbytes, outbytes;
+    int ret;
+    iconv_t cd;
+     
+    if (!charset_from || !charset_to || !string) /* sanity check */
+         return -1;
+         
+    cd = iconv_open(charset_to, charset_from);
+    /* Did I succeed in getting a conversion descriptor ? */
+    if (cd == (iconv_t)(-1)) {
+        /* I guess not */
+        error(0,"Failed to convert string from %s to %s - probably broken type names.", 
+            charset_from, charset_to);
+        return -1; 
+    }
+    from_buf = octstr_get_cstr(string);
+    /* allocate max sized buffer, assuming target encoding may be 4 byte unicode */
+    inbytes = octstr_len(string);
+    outbytes = sizeof(char) * octstr_len(string) * 4;
+    pointer = to_buf = gw_malloc(outbytes + 1);
+    memset(to_buf,0,outbytes+1);
+    ret = iconv(cd, (const char**)&from_buf, &inbytes, &pointer, &outbytes);
+    iconv_close(cd);
+    if (ret != -1) {
+         /* conversion succeeded */
+         octstr_delete(string,0,octstr_len(string));
+         octstr_append_cstr(string, to_buf);
+	 if (ret)
+	    debug("charset",0,"charset_convert did %d non-reversible conversions",ret);
+	ret = 0;
+    } else
+         error(0,"Failed to convert string from %s to %s, errno: %d",charset_from, charset_to, errno);
+    if (errno == EILSEQ)
+    {
+        debug("charset_convert",0,"found an invalid multibyte sequence at position %d",from_buf - octstr_get_cstr(string));     
+    }
+    gw_free(to_buf);
+    return ret;
+#endif
+    return 0;
 }
Index: gwlib/charset.h
===================================================================
RCS file: /home/cvs/gateway/gwlib/charset.h,v
retrieving revision 1.6
diff -u -r1.6 charset.h
--- gwlib/charset.h	7 Dec 2001 13:48:57 -0000	1.6
+++ gwlib/charset.h	6 Nov 2002 07:35:39 -0000
@@ -72,4 +72,9 @@
  */
 int charset_from_utf8(Octstr *utf8, Octstr **to, Octstr *charset_to);
 
+/* use iconv library to convert an Octstr in place, from source character set to
+ * destination character set
+ */
+int charset_convert(Octstr* string, char* charset_from, char* charset_to);
+
 #endif