it is still long way to go

Every day i am dealing with this encoding decoding thing because i am working in a telco company and my position is a Software Architech in billing departement so what i am doing is design and develop billing system. The data that billed by billing is data that come from network elements like Mobile Switching Center (MSC) orServing GPRS Support Node (SGSN).All data or call detail record come from network element is encoded in ASN.1 format. My billing system should read this ASN.1 format.From this data i have to get information about the calling number, called number, event time, duration, volume (gprs only), trunk (for MSC), cgi, MSC id (for msc), sgsn address (for sgsn).
The challenge of ASN.1 is that this format is in hexadecimal format. If we open this ASN.1 file we will get hexadecimal number like this:
a0 82 01 30 a4 82 01 2c  9f 21 01 03 8d 03 00 00 36 8a 03 0a 01 04 94 0f  49 44 30 32 4d 53 4d 44 33 20 42 2f 31 33 36 8e  03 00 00 00 82 03 b5 44 e9 92 02 00 01 93 01 00  8b 03 17 30 2e 8c 03 17 31 28 9f 66 01 04 9f 42  02 24 22 86 08 15 10 01 05 72 00 85 f9 9b 07 15  f0 11 56 d0 77 1b 9f 74 08 53 57 89 00 56 87 20  60 87 08 53 57 89 00 56 87 20 f0 9a 03 17 30 28  88 07 11 26 18 08 41 58 84 9f 22 01 00 9d 01 11  9f 44 01 06 95 07 11 26 18 48 54 47 f9 99 08 11  26 18 48 54 47 29 f8 9f 49 05 30 06 f0 23 c3 9f  41 07 11 26 18 48 54 08 f9 96 06 4f 55 48 42 39  31 85 08 11 26 78 68 58 35 73 f2 9f 75 07 00 22  24 7d d5 5e 84 84 07 41 78 68 58 38 25 f7 97 06  49 41 48 4e 4d 31 9f 39 01 30 9f 26 01 00 81 03  75 92 4f 83 01 01 9f 3a 01 00 80 03 01 02 0e 9f  3f 01 00 91 01 00 90 01 00 8f 03 00 00 0b 9f 23  02 03 03 9f 20 01 01 9f 3d 01 05 9f 3e 05 02 01  00 05 03 9c 07 15 f0 11 56 d0 77 1b 9f 4e 06 4f  55 48 42 39 31 9f 4a 01 02 89 01 00

a0 81 a2 a5  81 9f 87 03 0a 01 04 8b 0f 49 44 30 32 4d 53 4d  44 33 20 42 2f 31 33 36 82 03 b5 44 ea 88 03 17  31 29 9f 66 01 04 9f 1f 02 24 22 81 03 75 e3 f2  89 01 00 85 08 15 10 11 12 96 13 07 f3 8e 07 15  f0 11 56 cf 77 11 8f 01 22 80 03 03 02 09 83 01  01 86 08 53 28 96 30 19 83 01 f0 91 08 53 28 96  30 19 83 01 03 8d 06 49 55 48 42 4d 39 8a 01 00  84 08 11 26 78 68 09 37 19 f6 99 01 06 8c 07 11  26 18 48 54 47 f9 90 07 11 26 18 48 54 00 f9 9f  21 01 42 9f 22 01 04 9f 20 07 21 80 31 57 42 70  96

Above is a sample of call detail record from msc.Ofcourse all encoded format have rule to decode it. ASN.1 have a grammar that describe the rule to encode and decode it. Every hexadecimal number have meaning that described in asn.1 grammar.
My company have a application that can decode the CDR into text file. But the decoder is very slow to decode the cdr. It takes about 15 to 30 minutes just to decode one cdr files with size 16 MB. I need a fast decoder because the billing must decode at least 14.000 cdr files a day andthe billing system should be near real time. Then i am googling to search faster ASN.1 decoder. Because the limited budget my company didn’twant to spend a lot for this decoder so i search for open source ASN.1 decoder. I found http://lionet.info/asn1c/blog/ but it is not working when asn1c compile my ASN.1 grammar so i decide to write my own decoder.
My decoder doesn’t use the grammar at all. But it solve the TLV pattern of ASN.1. I see that ASN.1 constructed in TLV (tag length value). Tag is the tag number of record, length is the length of value and the value is the value of tag number. The value may contains another TLV that it called nested TLV
T L VT L [ T L V ]T L [ T L [ T L V ] ]
My method is just simply read each byte of ASN.1 file then decide what tag that will be decoded and write the value into appropriate type.
below is my code to get each CDR types like mobile originating call, mobile teminating call, roaming call forwarding call, call forwarding call, transit call, sms originating and sms terminating.

#include <stdio.h>

#include <stdlib.h>

int main (int argc, const char * argv[]) {

// insert code here…

FILE *erifile;

erifile = fopen(“/Users/gentur/Documents/eridecoder/EB_MDN3_100104017257.DAT”, “rb”);

if( erifile == NULL ) {

exit(0);

}

int byte_read;

int y = 0;

while( feof(erifile) == 0 ) {

byte_read = getc(erifile);

if( byte_read != 0×00 ) {

if( byte_read == 0xA0 ) {

printf(“<%02d>\n”, y);

int length = getc(erifile);

int length_flag = (length & 0×80 ) >> 7;

if( length_flag == 1 ) {

int num_len_byte = (length & 0x7F );

int i;

length = 0;

for( i=0; i<num_len_byte ; i++ ) {

if( i>0 ) {

length <<= 8;

}

length = (getc(erifile) | length);

}

}

char *cdr_value = malloc(length);

fread(cdr_value, sizeof(char), length, erifile);

int j;

for( j=0; j<length; j++ ) {

if((j%16)==0) printf(“\n”);

printf(” %02X “, cdr_value[j]&0xFF);

}

printf(“\n===================================================\n”);

free(cdr_value);

y++;

}

}

}

fclose(erifile);

return 0;

}

i think this is enough for to day