Posts Tagged ‘Individual Standards’


May - Oct 2006 Calendar



Finding the difference between two dates isn’t as straightforward as subtracting the two dates and dividing the result by /*(24 * 60 * 60 * 1000). Infact, its erroneous!

Going the ‘milliseconds way’ will lead to rounding off errors and they become most evident once you have a little thing like “Daylight Savings Time” come into the picture.

The Correct Way:


/** Using Calendar – THE CORRECT WAY**/
//assert: startDate must be before endDate
public static long daysBetween(Calendar startDate, Calendar endDate) {
Calendar date = (Calendar) startDate.clone();
long daysBetween = 0;
while (date.before(endDate)) {
date.add(Calendar.DAY_OF_MONTH, 1);
return daysBetween;






or more efficiently,   if you’re using the Gregorian Calendar:


/** Using Calendar – THE CORRECT (& Faster) WAY**/
/****Needs testing …… Anyone?****/
//assert: startDate must be before endDate
public static long daysBetween(final Calendar startDate, final Calendar endDate) {
int MILLIS_IN_DAY = 1000 * 60 * 60 * 24;
long endInstant = endDate.getTimeInMillis();
int presumedDays = (int) ((endInstant – startDate.getTimeInMillis()) / MILLIS_IN_DAY);
Calendar cursor = (Calendar) startDate.clone();
cursor.add(Calendar.DAY_OF_YEAR, presumedDays);
long instant = cursor.getTimeInMillis();
if (instant == endInstant)
return presumedDays;
final int step = instant < endInstant ? 1 : -1;
do {
cursor.add(Calendar.DAY_OF_MONTH, step);
presumedDays += step;
} while (cursor.getTimeInMillis() != endInstant);
return presumedDays;

Lets take, for example, the difference between the two dates 03/24/2007 and 03/25/2007 should be 1 day;
However, using the millisecond route, you’ll get 0 days,


import java.text.SimpleDateFormat;
import java.util.Calendar;
import java.util.Date;
import java.util.TimeZone;

public class DateTest {

public class DateTest {

static SimpleDateFormat sdf = new SimpleDateFormat(“dd-MMM-yyyy”);

public static void main(String[] args) {


//diff between these 2 dates should be 1
Date d1 = new Date(“01/01/2007 12:00:00”);
Date d2 = new Date(“01/02/2007 12:00:00”);

//diff between these 2 dates should be 1
Date d3 = new Date(“03/24/2007 12:00:00”);
Date d4 = new Date(“03/25/2007 12:00:00”);

Calendar cal1 = Calendar.getInstance();cal1.setTime(d1);
Calendar cal2 = Calendar.getInstance();cal2.setTime(d2);
Calendar cal3 = Calendar.getInstance();cal3.setTime(d3);
Calendar cal4 = Calendar.getInstance();cal4.setTime(d4);

printOutput(“Manual   “, d1, d2, calculateDays(d1, d2));
printOutput(“Calendar “, d1, d2, daysBetween(cal1, cal2));
printOutput(“Manual   “, d3, d4, calculateDays(d3, d4));
printOutput(“Calendar “, d3, d4, daysBetween(cal3, cal4));

private static void printOutput(String type, Date d1, Date d2, long result) {
System.out.println(type+ “- Days between: ” + sdf.format(d1)
+ ” and ” + sdf.format(d2) + ” is: ” + result);

/* This method is used to find the no of days between the given dates */
public static long calculateDays(Date dateEarly, Date dateLater) {
return (dateLater.getTime() – dateEarly.getTime()) / (24 * 60 * 60 * 1000);

/** Using Calendar – THE CORRECT WAY**/
public static long daysBetween(Calendar startDate, Calendar endDate) {
Calendar date = (Calendar) startDate.clone();
long daysBetween = 0;
while (date.before(endDate)) {
date.add(Calendar.DAY_OF_MONTH, 1);
return daysBetween;




Manual – Days between: 01-Jan-2007 and 02-Jan-2007 is: 1
Calendar – Days between: 01-Jan-2007 and 02-Jan-2007 is: 1

Manual – Days between: 24-Mar-2007 and 25-Mar-2007 is: 0
Calendar – Days between: 24-Mar-2007 and 25-Mar-2007 is: 1




Financial Transaction Tax campaign


reference Jimmy Blog

All (or most?) financial transaction is using ISO 8583 standard, which at first I thought is a complicated standard. But after learn and see how it implemented it’s not as complex as I thought.

In this post I will try to explain (based on my experience :) before I go deeper on the programming side.

Although it’s a ’standard’ I see that’s many variant for the detail implementation of ISO 8583.

Financial transaction is communication between 2 system through socket connection.
After connection established, each system can send message in ISO 8583 format which commonly will be request and the other system will response.
From my experience the communication will be start with sign on and then the financial transactions.
Periodically an echo message is send to make sure the other system is still alive.

If I break down the flow:

  1. System A open connection to System B (through specific IP and Port).
  2. Connection established.
  3. System A send Sign-On request message.
  4. System B send Sign-on response message.
  5. System A will start send Echo request message periodically (e.g every 1 minute).
  6. System B will send Echo response message when receive Echo request message.
  7. When financial transaction happen, System A will send Transaction request message.
  8. Then System B will send Transaction response message.
  9. If something wrong happen (usually timeout so System A didn’t get the response), System A will send Reversal request message to cancel the previous transaction.
  10. System B will send Reversal response message.

I hope my explaination give clear basic understanding about the flow.
Next the question will be what’s this ISO 8583 message looks like.
We can separate the message into 3 parts:

ISO8583 - structure

ISO8583 – structure

Message Type Identifier

Message Type Identifier or MTI is 4 digits numeric that describe the message type. It will explain the message function.
Commonly used:
– 02xx : Financial Message (e.g 0200 for request, 0210 for response)
– 04xx : Reversal Message (e.g 0400 for request, 0410 for response)
– 08xx : Network Management Message (e.g 0800 for request, 0810 for response)
* I found more detail & complete list at wikipedia


Bitmaps is field that contain information about which data element is presence or absence. Based on the variant it could be 16 hexadecimal characters.
An example will make it clear.

The bitmap is:


If we break down to binary (I hope you understand how to convert hexadecimal to binary:) :


Since I’m nice ;) I create graphical illustration.

Bitmap sample

Bitmap sample

You can see that in this bitmap it explain that data element in the message is 1, 3, 4, 7, 11.
There’s special meaning of first bit of bitmap, if it has value of 1 that mean there’s secondary bitmap.
And what the hell is this secondary bitmap?
Since 16 hexadecimal characters will can only contain info of 64 data element, some transactions contain data element number 64 – 128. That mean the 16 hexadecimal characters is not sufficient.
With set the first bit to ‘1′ that will inform there’s additional bitmap, which is another 16 hexadecimal characters.
So in this case the full bitmap example should be:
Convert to binary:
From the bitmap, data element presence in the message : 3, 4, 7, 11, 105.

Sample Bitmap complete

Sample Bitmap complete

Data Elements

Data Elements is the essense of the whole ISO message, contain information about the transaction (transaction type, amount, customer id, etc).
Each data element have their on format, attribute and length.
Each data element number also have standard purpose, for example DE #4 is transaction amount.
I will explain based on the example above, since this is only introduction I don’t want to confuse beginner reader. In wikipedia there’s full list.
For our examples, the data element list:

  • #3 – Processing code – n 6
  • #4 – Transaction amount – n 12
  • #7 – Transmission date & time – n 10
  • #11 – Systems trace audit number – n 6
  • #44 – Additional response data – an ..25
  • #105 – Reserved for ISO use – ans …999

To make it practical let’s just use example.
We want to send :

  • DE #3 : 201234
  • DE #4 : 10000
  • DE #7 : 1107221830
  • DE #11 : 123456
  • DE #44 : A5DFGR
  • DE #105 : ABCDEFGHIJ 1234567890

Using above value, the data in data elements :

ISO8583 - de

ISO8583 – de

Based on the type each value will be formatted.

  • a : alpha (including blanks)
  • n : numeric value
  • s : special characters
  • x (no dot) : length is x (fixed)
  • .x (one dot) :  max length is x (1 digit in front as length info)
  • ..xx (two dot) :  max length is xx (2 digit in front as length info)
  • …xxx (three dot) :  max length is xxx (3 digit in front as length info)

ISO Message

From above examples the whole ISO message will be :

0200B2200000001000000000000000800000201234000000010000110722183012345606A5DFGR021ABCDEFGHIJ 1234567890

ISO8583 - full


Ok. This article already long enough :)
Actually I always prefer short article, but on this case it’s cannot be to short.
Anyway, I hope this article is clear enough for you.

Like always, just drop question in the comment if you confused.