The Book of Ninja

THE SAP KILLER
The SAP Killer
Build or Migrate
any DBMS or MIS
Up to a full blown ERP
By Redhuan D. Oon (RED1)
Dated January 1-16, 2019

(Book in progress, 37% done)
What Can This Book Do?
This book shows you, without doubt, how to do anything related to creating a DBMS or MIS
from scratch or migrate from a legacy to a brand new modern powerful open source suite
called iDempiere.
The author has more than 36 years struggle in the industry with much travel and work
experience from 40 countries facing formidable challenges which culminate in a set of best
practice and his own novel approach that cuts through the hype and myth of software
engineering.
Now, you can build your MIS or ERP in the quickest time imaginable compared to many
options out there. You are guided in easy steps from practical real life scenarios.
Using the Ninja toolset, it simplifies the use of the best free heavy weight ERP, iDempiere,
which has replaced many large ERP systems in many countries around the world. Now you
can use this as a real SAP killer.
The first part of the book assumes you have setup of iDempiere from www.idempiere.com.
The second or appendix part of the book will give you all the basics from setting up iDempiere
either in-house or on the cloud or use our free cloud for the back-end, up to the fundamentals
of iDempiere structure, features and functionality.
Further paid consulting support is available by contacting the author at red1@red1.org.

Meanwhile, you may continue your DIY journey by accessing freely the technical notes at
www.red1.org/adempiere.
Copyright 2019 Redhuan D. Oon Page 1

THE SAP KILLER
Who The Heck Is SAP?
For those who doesn’t know can actually Google exactly the above and there is a blog on it.
SAP is actually the most expensive software in the world not due to its code but the know how
and consulting or advisory services that goes with it. This is because it is usually used by the
Deutsche Banks, Red Bulls and Petronases of the world. Its cost is thus sky high running into
hundreds of millions of dollars and that is in licensing alone.
Does that means Open Source is free? No, it is free as in Freedom, not Free Lunch. Then how
do you make money? Read here https://www.gnu.org/philosophy/selling.en.html.
So can I switch from SAP to iDempiere with Ninja? Not until you think about it by reading the
19th October, 2011 article down here http://www.red1.org/pos.html.
You may still be deeply confused, that is why I make life easy with philosophy.
RED1’s THREE LINE MANTRA
INFORMATION IS FREE, YOU HAVE TO KNOW
PEOPLE ARE NOT, YOU HAVE TO PAY
CONTRIBUTORS ARE PRICELESS, YOU HAVE TO BE
https://www.youtube.com/user/redhuanoon/videos

THE SAP KILLER
Before You Continue
Have you read my 19th October 2011 essays? If no, then you may use this series of questions
to determine if you need to continue with this or look elsewhere. This is to save everyone’s
time.
1. Are you looking for freedom to have sovereignty over your system, the freedom to have it fit
into your needs rather than the other way around? If Yes, proceed.
2. Are you hoping this will replace your SAP? If Yes, then be careful and tick all the advice
given here.
3. Are you looking for more security, independence, inter-operability, and sustainability of your
ERP system? If Yes, then this is certainly the direction to pursue.
4. Are you trying to have a free as in free lunch ERP system? If Yes, then this is not for you. It
involves much more costs than you think. Not just your time, but the emotional burden and
attitude towards this project may hurt. I have seen people from as far as Slovakia and Japan
cursing this project for faults of their own making. So stop here, if you think this will save you
money outright without homework.
5. Are you ready to be well prepared with a clearly defined budget and resources and
expectations of time schedule of deliveries. At least someone like me has giving you a transfer
of know how course? If Yes, then proceed.
6. Is your team ready to go through the sharp and short learning curve for 3 months? If yes,
proceed.
7. Are you willing to accept best practice advice that maybe in contravention with your own
best practice? If Yes, you may proceed.
Under part two there will be a chapter for you to understand why I put these in and the rest of
my best practice that you can read through and decide if you can proceed on your own or you
need me to come in to explain further usually in a paid Skype session or on site workshop.
Not For You
This iDempiere is not for you if you are looking for a standalone ERP system, handling a single
POS. It will be overkill. It is like driving a Ferrari to go next door. It is better you use simple free
download-able apps out there that probably cost below USD1,000. Or Excel, also below
USD1,000.
In fact there is no such thing as a standalone ERP. The word enterprise in there means it is for
a starship enterprise effort.
However the Ninja is making this ERP into a better faster easier MIS. So if you are looking for
an MIS, which actually is bigger than a standalone. It has to be a multi-user, multi-organisation,
multi-role, multi-user and multi module, then this will be worth your while.

THE SAP KILLER
What is iDempiere?
iDempiere is a fork of a fork of Compiere, the first Open Source ERP project to hit the world in
1999. In mid of 2006 Compiere was sold off. The community under the leadership of the author
forked as ADempiere during the first week of September, 2006. Later the project evolved
further migrating onto the Eclipse-like OSGi as iDempiere.
Compiere is originally meant for heavy duty serious ERP that covers accounting, inventory,
supply chain, CRM, and dcoument management. Like any other ERP software, it will try to be
one size fits all, which no one can. There shall always be gaps where new modules or verticals
have to be built.
What is powerful is it has within its design the Application Dictionary or AD which is actually an
application within the ERP application. The AD uses metadata to quickly and easily configure
Menus, Windows, Tabs, Tables and Columns as well as Print Formats and Proceses and a
host of many other things without the developer having to modify code. Thus it is not error
prone, but versatile and stretchable. Even so, grappling with the AD is still always a challenge
and this book lays out how best to meet this challenge.
iDempiere since its first fork of ADempiere from Compiere, has incorporated amazing features
such as Composite Web-Services API, FitNesse testing engine, and Self-Help screens. You
can refer to many of them here http://wiki.idempiere.org/en/Category:New_Features
Of course, the OSGi framework is what separated iDempiere from ADempiere. It allows
separation of concern where all developers may add on without touching the core thus
ensuring maintainability, always synching with upgrades of the core and easier testing. See
the many plugins done by the world wide community here:
http://wiki.idempiere.org/en/Category:Available_Plugins.
Now it is the Application Dictionary or AD that I want everyone to pay attention to. I realised
that this AD is under-marketed as the focus and lingo surrounding this project has been mostly
on the ERP which is a complex subject matter. This lesser affair of simple module or new
application building is overlooked and not realiased by non ERP users. This lesser realm
actually solves the needs of the world of MIS or DBMS (Management Information System or
Database Management System).
So I intend to promote that useful nature and I even come out with a wizard suite called Ninja
which focus on letting the AD be unleashed via script strings which generate the metada AD
on the fly.
I also have a tool call the Wukong Menu Controller which can hide all the ERP sub menus from
the main menu tree giving it the appearance of a simple empty shell to be sculptured quickly to
any MIS or DBMS system. The Wukong tool can then reactivate back any ERP module and
this gives the appearance that iDempiere can start as an MIS all the way to a full fledge ERP.
Now, another overlooked killer app outside all this, is the spreadsheet. In my case I used to
use MacBook Numbers but now switched to Linux based operating system for speed, using
Libre Calc. Now it is this wondrous spreadsheet that I build my scripts from. You shall see in
later chapters how I progressively exploit the functions within the cells to produce script
formats in a more visible and easy form, to avoid redundant typo and human errors.

THE SAP KILLER
What Is Red1’s Ninja?
Ninja is meant to be a killer wizard that I took the time since early 2016 to craft which helps you
overcome the complexity of building any form of database application and implementation,
working alongside the best free ERP software in the world, iDempiere.
The Ninja takes an Excel spreadsheet template to generate most of an iDempiere module from
the ground up instantly on the fly. It can also do so in tandem with the migration process of
legacy data. The artifacts it can generate or handle within a single blow are:
1. Each data model and its columns with data type settings
2. 7 mandatory columns automatically inserted
3. Master-Detail relationship with choice of parent key column
4. Menu item on Menu tree with Window and Process
5. Window Tab with attached Process stub
6. Process java stub with defined parameters
7. Workflow linking a series of windows as a ready user guide
8. Info-Window creation from View of multiple models and link on main panel
9. Info-Window additional displayed columns
10. Info-Window Report output from the selected rows.
11. Info-Window with attached Process stub
12. Translation of menu and window tabs and fields into Translation tab
13. Master-Detail Print-Format reflecting displayed layout, with Corporate image and
window description as headline.
14. Document Type with Action/Status Model
15. Kanban Board with 12 column status for cards
16. ORM Classes generation
17. Model Factory classes
18. OSGi XML definitions
19. Callout JSR 223 generation
20. PackOut format
21. New blank plugin for the newly generated module
22. Migration of master data in a single zip, generating any new data models.
All the above are handled from a single spreadsheet which any dumb user can understand
instead of confronting the complex bulky parts of any database configuration framework. The
idea is to give high level single reference visibility.
What is left for a the programmer is to just code the logic of:
A. Process
B. Model class events
Sample Excel template and blank OSGi plugin are provided. Also great data model design
ideas from me for some common user cases. No haggling with the setup, scaffolding, and
boiler plate code. All this is contributed freely by me to make iDempiere more popular and easy
to use. Now, lets get going to build a brand new ERP module or domain vertical in record time.

THE SAP KILLER
CHAPTER 0 - A Crash Course in MIS
My first job in 1982 was a COBOL Programmer for a HR Payroll system. My second job in
1985 was given the title EDP Supervisor Programmer. EDP stands for Electronic Data
Processing. Today it has evolved into DBMS (Database Management System), MIS
(Management Information System), and ERP (Enterprise Resource Planning).
The programmer label is now developer, supposedly doing everything from requirements
analysis through programming through testing. In 2010 I received a new title CTFL (Certified
Testing Foundation Level). So now, I am going to debunk many myths by going back to only 2
things you need to know.
1. GIGO - Garbage In Garbage Out. Whatever goes into the system, remains as such and
you won’t find anything else. That also means if data has gone in, do not worry. It is there.
a) This leads to the sub law called non-redundancy. It means once recorded, it is there.
You need not repeat it. It may not be well formatted, or raw, but again, do not worry.
What is most important is that it has been captured. It is somewhere. We shall get it.
b) This thinking is important because later and in fact quite right away in the next
chapter you see my use of the Excel spreadsheet where i capture basic data model,
and like magic it becomes a full fledged MIS or ERP system. How? It is because of
this sub law of non redundancy, which its mother law is GIGO. Read it 15 times.
Every day. Until the day you attain Nirvana.
2. Input, Process, Output. This simple three word is your only mantra to good systems design.
There is only and only three parts to any system. Again, note how it follows the first law of
GIGO. Input is what goes in. Output is what comes out. But there is something in between
to rearrange or set in order the data within. That is where programming supposed to be.
a) Both Input and Output is the same. GIGO. Thus that is why a single spreadsheet can
represent both. Once you design well this stage of data model, you need not worry
where to fetch the data for the output.
b) Input are all formats used to capture data. They can be input screen, mobile input,
sensors, keyboard, mouse or voice. They are usually found in the form of input
formats. That can be put into a spreadsheet. As your data design.
c) Output are all required data demanded of a system, mostly reporting. Even charts
and analysis is a mesh of such output. Again, once input is in, output is derivable.
Now, even migration, which was hell, is now a breeze. Once I got the data, I got it.
d) Process is what to do with this data bit or that data bit. Usually it is taking a piece of
data, looking up another related to it, applying a calculation if needed, and writing the
result back or in a new model. Now this seems to break the first GIGO law, that
suddenly a new model exists that maybe non-garbage. No. It is its non-redundancy
derived data. It didn’t come from somewhere else. It came from what has gone in.
Thus, you have to read all these and meditate on it. Daily.
What I actually did, is using another modern rule called, Separation of Concern. You can call
this the 3rd law but it is a modern software design principle. I am a traditional fundamentalist. I
do not wish your mind to be clouded. But this 3rd law is actually in the first 2 just that it is made
more subtle. By using the easy and universal Excel, I am separating the hard work and the
Input/Output Data Model part. I am also putting the grunt work of Process, which are the code
stubs scaffolding there too. What is left is just pure coding work of business logic. No grappling
with redundant repetitive code preparation, as you shall see later in advance chapters.
Now, the journey shall be painful. It is a matter of degree of pain, depending on how smart, or
experienced or trained you are. I maybe a guru but that is not important. There are no bad
students but only bad gurus. My task is to guide you in a way that is most effective and at times
it maybe non-conventional. I know many things, but they are part of the problem. I am putting
finally the solution.

THE SAP KILLER
CHAPTER 1 - The Excel Template
This is a clean module setup template. Later we shall look at the legacy data migration
template. You can fetch this first template from
https://sourceforge.net/projects/red1/files/ModuleCreator/Ninja_Excel.xlsx/download
Open it and see in the first sheet called 1_RO_ModelHeader
This defines the Menu Tree item. In this case it is NewMenu. That shall appear as the new
item in the iDempiere’s menu tree. Next we move to 2_RO_ModelMaker.
Here a lot of actions happens. So we start from the left. Look at the header row. The
RO_ModelHeader_ID means this is a lookup field. The Excel already set that its value is
referencing the first sheet’s value. So you need not refill such.
This is the power of Excel and the main reason it is used here. I could easily setup a sample
template with many rules and formulas already set. The lay user also is more familiar with
Excel than anything else.
Whenever you encounter a header value with a suffix _ID, it means the content below it is the
Name column of the record referred.
Now the next header. It is Seq. No. This has to be a running number as Ninja will sort the
records and execute them according to their sequence numbering. In this case we only has
one record. You may proceed to try more after this. For now, we just proceed very slowly so as
to get all the concepts grasped tightly first.

THE SAP KILLER
Next is Workflow Structure. This when set to Y will generate the model as a DocAction aware
type. It will have extra fields to handle Document workflow actions and status..
Next is KanbanBoard. It will create a Kanban Board display with 12 status columns
automatically for the DocAction aware model. But it must be set to N first and allow Workflow
Structure to be Y first. After the first run of the Ninja Generate Module, the Kanban Board flag
can be set to Y.
Next is Master. This is usually blank. If it has an entry, it means there is going to be a
Master-Detail model relationship. However, whatever is stated here must have been defined
prior. It has to be created first before been placed as a Master. So in this case C_Partner is a
core table which already existed.
Now, the main item, Name. This is simply the name of the Table or Window Tab that shall be
created. As stated, if the Master is blank, this will be a single Table. Don’t worry if you are
puzzled by anything above. After going through this round, we shall examine a more elaborate
example ‘Hospital’ already imported into iDempiere Ninja to follow and learn quicker.
Next is the Help column. Whatever is put here shall appear in the Help section of the Tab
created. By having this, will make life easier for all, as the designer, user and implementer has
a common reference spreadsheet to refer to and understand the purpose of the design.
Finally, the ColumnSet, containing 2 fields only. These two, Description and IsActive shall be
appearing in the intended Tab panel.
So in conclusion there shall be a sub tab to the Business Partner tab, called MY_Model.
Actually the MY_ prefix would not appear in the sub window tab, and only be used in the
underlying database table creation. This sub tab shall have two fields Description and Active.
The ‘Is’ part is also hidden and it shall appear as a checkbox.
At the time of this writing, but before updating this spreadsheet, I just improved the
ModelMaker with a new column checkbox, PrintFormat, which works in conjunction with
Master-Detail definition to generate a Print Format with Sub Print Format and attach itself to
the new document. It has a special way to do it, which is amazing as you can rearrange its
layout from the Tab Editor of the Window Tab. This checkbox is in play only after you have
imported your master-detail model and gone thru its layout in the Tab Editor.
Now we go to the 3_RO_CodeMaker sheet.
This is used to generate OSGi XMLs and code stubs and the model ORM classes
automatically for the defined model in the previous sheet.
PluginLocation refers to the location of your new plugin. You may take my free sample plugin
from https://bitbucket.org/red1/org.red1.blank/. However if you are not coding further logic to
your new module, you need not have the CodeMaker involved. The ModelMaker should be in
theory be sufficient to let you fly off with a simple database system, accepting data input, and
exporting data to external parties.

THE SAP KILLER
GenerateModel and GenerateModelFactory are usually checked Y together to generate

your new model I_ and X_ classes, as well as the Factory and Model classes.
GenerateProcess will specify a new process java class stub be generated and attached to the
model’s window tab at the same time. If you specify this on a new line and leave the
ModelMaker_ID blank, then it will be attached under the new menu item tree.
The advantage of been attached with the model is that you can fetch its active record and
process it.
With the above done, all we have to do now, is just to export the sheets as CSVs and zipped
them up. Then attach them to the Ninja to GenerateModule and get the results within few
seconds.
Store each sheet as its namesake as shown below. In MacBook, this can be done at one shot.
But in Libre Office, you have to do it each sheet at a time. The numbering prefix is vital as it will
ensure they are imported later sequentially.
Compress the 3 files into a zip and you can give it any name. Here i gave ANinja.zip. What
matters is the individual CSV file names with the numerical sequence prefix.

THE SAP KILLER
Chapter 2 - Ninja Generate Module
This assumes you have setup your iDempiere instance. You may run this either in your binary
instance or Eclipse development environment. And you should have got the Ninja plugin
working. You may fetch them as follows:
1. Binary - https://sourceforge.net/projects/red1/files/ModuleCreator/
2. Source - https://bitbucket.org/red1/org.red1.ninja
I am using the source on Eclipse. I also have a blank plugin (as a new plugin to house the
generated code and attached a new 2Pack of the data models) fetched from
3. https://bitbucket.org/red1/org.red1.blank
From the Menu tree, call up the Module Designer
Inside the Module Designer, make a new record, and just attach the zipped CSVs to the
ModelMaker window. No need to put in any details. Now, you can either run the attached
process Import Model or at the Header, Generate Module.

THE SAP KILLER
After Import Model you will get a positive result at the Description box:
Go to the Model Header main tab and refresh.
You will see a new record New Menu and sub detail record for Model Maker. It corresponds to
the imported CSVs. You will also notice that there are more columns that were not included in
the Excel. They, particularly Callout and Workflow Model, can be included whenever needed.
Later, during the second round of explanation, I shall address these columns usage.
You can also go to the Code Maker tab and see that it has also been populated accordingly.

THE SAP KILLER
Now, we can execute them with the header tab process icon, Generate Module.
Click OK, without checking anything. You will receive the result within 5 seconds and it is all
done. Notice the message on the top of the panel.
COMPLETED - HEADERS:1 MODELS:1(COLUMNS:3) CODES:1

This means one menu header is created with a data model that has 3 significant columns, and
its code stub set.
Notice that the Active box in the details are unchecked and they are grayed out. For both
Model Maker and Code Maker. This means that they have executed properly. If they are still
active, then something went wrong, and you have to check the console logs.
Now we can restart and login into the Garden Admin role to examine the new model.
A new menu item appears under the main menu tree with a window item called Model and a
Process item called MY_Process. So congratulations, you basically passed Ninja Basics. But
wait there is more. Remember our original Model Maker entry was a Master-Detail setting for it

THE SAP KILLER
to be under the C_Partner model or Business Partner window. So call that up and you can see
there is a new tab:
You can see it is correctly linked to individual parent records. Lets try producing some sub
detail data and see if it works.
It does work and without needing its code plugin to be active. That means the meta-data
framework, thanks to Compiere and improved iDempiere holds very well. If you check the left
hand lower panel, you will see under recent items, it is also tracked.
Now, let the AD framework show more of its ready made goodness and glory. Click on the
report icon button on the top menu bar. OK it through and see an automated report is
generated:
Just imagine, a single Excel detail can translate quite easily and instantly into your ERP
system. From now on, your imagination is the only limit. But wait, there is more. Look for the
gear icon on the top right and click on it.

THE SAP KILLER
You can see the MY_Process which was defined

in the CodeMaker tab. Try to test it. You will get an
error message.
This is because the code side has not come into play. We shall now leave this data side and
go to the code side by going to the Eclipse IDE. Expand the blank plugin.
Under the src folder, under

org.idempiere.component are the DocEvent and
ModelFactory classes for the MY_Model item.
These saves the developer from coding or creating

them from scratch. Even copy/paste is not needed.
You may delete the other java classes that are just
place holders.
Under the org.my.model structure, are the ORM

classes. The I_ and X_ and Model classes of
MY_Model.
Again you may delete the Acme place holder class

after this.
Under org.my.process is the new MY_Process

class.
Under the OSGI-INF folder are two new XMLs.

THey both relate to the DocEvent and
ModelFactory java classes respectively. Again,
delete the other placeholders.
There is one task to do here before we can activate

this new plugin to be used for coding further any
logic.

THE SAP KILLER
Further down the blank plugin structure (hidden from the screenshot view), you will find
PLUGIN.txt. Open it up, copy the content, and open the plugin.xml and insert it in as an added
extension. This is to allow the Process to be linked when called.
Now, restart your Eclipse Run/Debug As Configuration, and you will notice the blank plugin
referred to as org.my.plugin.
You need to check that to include it. You now have full programmatic control over your model,
and soon you can readily use as well as introduce business code logic easily.
You can test again the process and see. But first let’s try something to make you jump. Go to
the src folder and open the MY_Process.java

THE SAP KILLER
Put a break at line 32 and add some comments within the quotes. Now press the MY_Process
button and see.
First you will get the break button to

pause the execution. This means that
the plugin is active and you can now
modify that java class. Press continue
and you shall receive the success with
that comment displayed.

THE SAP KILLER
Chapter 3 - Ninja Starter Cheat Sheet
Here i provide a cheat-sheet that summarizes the above steps taken. You may call this the
Ninja Starter cheat-sheet.
1. Install iDempiere source code in Eclipse IDE
2. Install Ninja Plugin in Eclipse IDE
3. Install Blank Plugin in Eclipse IDE
(Estimated newbie time - few days. Estimated expert time - half day)
4. Export Excel Template first 3 sheets as CSVs in same sheet names
5. Zipped up as any name
6. Launch iDempiere in Eclipse with Ninja plugin checked
7. Attach to Model Maker tab of Module Creator
8. Run Import Model or Generate Module
9. Refresh to new record at Header and Generate Module
10. In Eclipse, insert PLUGIN.txt into plugin.xml
11. Restart iDempiere in Eclipse with My Plugin checked
12. Login Client of GardenWorld and start entering data and exporting report
(Estimated newbie time - > half day. Estimated expert time - 30 mins).
Next, we shall go through the the other powerful actions of the Ninja.

THE SAP KILLER
Chapter 4 - Data Types and Callout
Now we examine the Model Maker again, and up close with sample imported content already
present when you have the Ninja plugin running. Look at the first header record ‘Hospital’.
Look at the lower detail tab, Model Maker and you can see many entries. Some of them are
Master-Detail and some are not but standalone models. We shall examine the first one up
close. The Bed Model. Click on its Edit Record to bring it up as main panel display.
The Name means it will be create a Window Tab and Table model named ‘Bed’. It shall have
properties or Field-Columns as stated in the ColumnSet, separated by commas:
1. Name as String data type.
2. Bed No as list of values “1,2,3,4,5,6”
3. Floor as list of values “1,2,3,4”
4. Wing as list of values “North,South,East,West
5. In Service as a Y/N checkbox.
Here we encountered 3 type of data types: String, Reference of fixed Data List, and Checkbox.

THE SAP KILLER
We go to the next model, BedRegistration. Take note, that we remove spaces and during
generation, the leading caps will space them out. So in this case, it shall appear as Bed
Registration.
There are _IDs and now I explain what they means. An _ID means it is a lookup to a foreign
table. The looked up table is the same name as the ID without the _ID suffix. Thus
C_BPartner_ID is looking up the C_BPartner table. Unlike the fixed list stipulated by L#, this
will be a dynamic list.
The next data type is D# which is obviously the Date data type. Ninja also automatically
recognise a Date value as a D#. Thus D#DateOut can be written as DateOut and it shall
reproduce the same result. As long the Date is the first part of the value.
Now we go down the list of Model Maker records to examine new data types and script. Take a
look at record 6, PatientLog. It is a master-detail under the C_BPartner model.
It has Q# which is a Quantity data type. Note the Callout field has a value:
QtyDosage>Percentage=QtyDosage / MaxDosage * 100
This will be a JSR223 meta-data callout rule. It means the Qty Dosage field when changed
shall update the Percentage field with the value of field, QtyDosage divided by the value of
field, MaxDosage multiply by 100. Basically this is a percentage calculation.
The JSR script generated will be more complex but that is the work of Ninja. You just use such
simpler syntax. After generation, you can see under the Column Callout pointing to a Rule that
has the following script.

THE SAP KILLER
In the next record, we can see something interesting. It is the same PatientLog model, but it
has another Callout. This means it shall create just as describe -
Name>Description=C_BPartner.C_BP_Group.Name.
The field Name when changed, shall have the Description field updated with the C_BPartner
contained property C_BP_Group which will be looked up and return its value for Name. In
other words, it will just pass the BPartner Group’s Name and store in the Description field.
Again, this simple script will be converted to a more complex JSR223 rule set, which is as
follows:

THE SAP KILLER
Chapter 5 - Workflow Model and Info-Window
Now we look at the next record that has a Workflow Model value:
NewWorkFlow=Bed Registration,Patient Activity,Treatment.

Note that the equals sign has no spaces before and after, and the
commas are tight too. The names of the Windows are exact names
as appear in the windows with spaces included.
A new Workflow Model will be named New Work Flow which hall
appear in the Menu Tree. Double clicking it and it shall show open to
show a linked icon graph of 3 windows as below.
Next record has an interesting method.
Its Name field has multiple table models:

PatientLog, C_BPartner, BedRegistration, Bed
This shall create an Info-Window that is based on an SQL View of these tables. Make sure
these tables are already created prior to this. Their sequence must be earlier. And make sure
their spelling is exact.

THE SAP KILLER
The Column Set field contains columns that shall be added Info-Columns. The result shall be
as below.
When you run Generate Module for this Hospital set, you should get an all grayed success.
You can unchecked the Active columns of the Code Maker side if you just wish to see the
model generation success as well the Callout Rules and Info-Window as explained earlier.
Then you can relogin as Garden Admin and view the new Hospital module.

THE SAP KILLER
Chapter 6 - Spreadsheet Power
Now I show my most recent improvement which was done as I am writing this guide. Take a
look at this Ninja Excel sample:
Notice how unwieldy the ColumnSet is. I am making two improvements. First, I am batching all
long form hash datatype such as L#, Q#, Y# into a preliminary ModelMaker record so that they
need not be repeatedly defined with the hash, as it will have been created and Ninja shall
reuse what is done.
Second I create another source for this ColumnSet. Notice the formula behind it (above left of
the screenshot. It is:
=TEXTJOIN(",",,$DataTypeMaker.B2:R2)
It is referring to another sheet and joining the contents of the columns in the range selected.
So we go to that sheet and it is below:
Here, it is very obvious. I am setting out the string making even further separated into logical
sets so that it is now easy to enter just the values without wielding other signage and symbols.
For instance, the first L# list you only need to enter the values in yellow highlighted columns.
Qualification and then the list values below - Phd, Masters, Bachelor, Diploma, and Certificate.
It shall automatically result in the top row as
L#Qualification=Phd/Masters/Bachelor/Diploma/Certificate
There is now less human redundant typo activity. And the layout is more elegant and cleaner
for the data model designer to work with. Everything is in its own place and consultants and
users won’t be hunting high and low for any documentation or definition. Changes or
corrections will be much faster and even lay users won’t be confused at all.
Now, I didn’t stop there. I converted quite all mundane ColumnSet sets to be hands-off and
refer to another sheet where all model fields or properties are nicely laid out.
See the following screenshot:

THE SAP KILLER
Now the ColumnSet sets below the

Terminology are also been factored out.
Notice the cell function on the top left of the
screen. It is another TEXTJOIN referring to
a new sheet, ModelMakerSource as shown
on the right.
Now each ColumnSet refers to this

separate sheet from rows 2 downwards
column-wise.Row 1 is for Master-Detail
relationship.
The heading names are also referenced

into the ModelMaker. Thus, all changes or
creation of data-models can be done here
without worrying of the actual ColumnSet
typo such as the commas. Extra commas are taken care by Ninja and truncated during actual
processing.
Again, the data types are also not and issue as seen in some of these field names. They also
reference the previous DataTypeMaker sheet to avoid typos. Thus we always make changes
at the same source, at one place rather than all over.
The magic is that, whatever is done here shall translate to exact and actual appearance in the
software for your MIS or DBMS or ERP later.
So now i easily complete my data model design within a single day interspersed with long
walks and leisure, and when i am done and unleash my Ninja, it is voila! Look at the following.
I upload the sample HRMIS (Human Resource Management Information System) that i am
building from scratch at the moment as a brand new module (free) for iDempiere!
Fetch the Ninja_HRMIS.xlsx here - https://sourceforge.net/projects/red1/files/HRMIS/

THE SAP KILLER
So all I have to do is just zipped up the Excel

exported 2_RO_ModelMaker.csv with its
1_RO_ModelHeader.csv and attach to an
empty line in ModelMaker of the Module
Designer and then run the Generate Module
process. I get the new Model Maker called
HRMIS (see below), and run Generate Module
again on this new set.
Relogin as the Client and you can see at the Main Menu
Tree pull down at the top mid of the panel, a brand new
menu called HRMIS. Upon opening it up, you see all the
models created in its entirety. Relationship, master-details,
lookups, lists, and data types.
Simply magic.
On the following screenshot is the important Employment

window with its sub-tabs. I quickly arrange its main fields by
using the Tab Editor from the Window/Tab Process.
Now, who won’t be impressed? You can quickly then go

through with your prospect or client or user on the look and
feel but more importantly the data model design whether it
has taken into account all required fields and data
relationship.
And the beauty is that you can do all this changes right at the
original source Excel spreadsheet, and then export the
CSVs and generate again, all within less than a minute, once
you get a hang of it.

THE SAP KILLER
All the data models as designed in the Excel are now transferred into the iDempiere System
and ready to use. It can now accept data, stored automatically within the database. The data
relationship is all taken care of. Master-detail association correctly managed. There shall be no
errors of any sort. All data type referencing be they check-boxes, pull down list, look up
dynamic values done.
The most powerful part is as shown before this, is the reporting function. Row based reports
are easily made on the fly. Later we shall see Master-Detail PrintFormat creation also done
with a Ninja tick box. I also have nice patches to modify the underlying code to make
JasperReport add-on (which was integrated into iDempiere to beautify print format) redundant
90% of the time.

THE SAP KILLER
Chapter 7 - The Module Design Approach
I have my preferred way of designing modules. Let’s look at the HRMIS spreadsheet again:
Remember that in any new module, we want to integrate some of the above to core tables
such AD_User and C_BPartner. (In appendix we explain such tables in detail.)
There are two ways to put in new tables that needs reference to core tables. Either as a
sub-table to the intended main table, or a separate table altogether with just a field within it
looking up the Business Partner. In the above screenshot, you see both. The left most is the
standalone table. The rest are Master-Detail.
However in this HRMIS example, none of them, both standalone and master-details are
children of any core table. By right, HR is part of the C_BPartner (Employee) model. They can
be, just that in my HRMIS, I do not wish them to be bundled to the core making them bulky. I
wish the HRMIS to be a more loosely coupled module and only use the 2nd approach which is
to look up instead. My HRMIS sample module makes all the tables children of either
HR_HumanResourceManagemant and HR_Employment (to be the stand-in for C_BPartner).
So in the first leftmost table of HR_HumanResourceManagement, I look up C_BPartner by

pointing to its main primary key, C_BPartner_ID. In iDempiere (again this is an inherited
feature from the original project, Compiere), it automatically refers to data records in the table
referenced without the _ID suffix, which is C_BPartner. And the Name column values shall
appear as a pull down dynamic list in the HR_HumanResourceManagement table.
Let me give another reason why in this case, I look up instead of make a child of. By looking up,
you have the choice of leaving that field blank. Meaning it is not mandatory to have a core
record in the Financials ERP core. This is both good or bad depending what you wish to
achieve. I wish to have an HRMIS that can exists without the ERP. It can then be minimalist
and standalone small MIS catering to a small client who wishes only for a HR Payroll system.
It is not so bad, for being separated from the Business Partner (Employee) main window,
because with the Info-Window (another magic step from Ninja) you can easily join
HR_Employment to C_BPartner or other models for information joining and reporting.
Now what do we do after defining the above in our spreadsheet? How do we setup the
meta-data or configuration or any coding? It is all meta-data within the Application Dictionary
and normally from Compiere days you have to do quite a number of steps to do this. But now
with my Ninja, you won’t notice it. Ninja code solves both for you. All you need is just that
Ninja_HRMIS template.
You will notice in the HR_Trainer table, there is another reference to C_BPartner_ID. That is
meant for the trainer record ID. The label on the window tab for this table will show ‘Trainer’.

THE SAP KILLER
How is the label changed? It is done via a smart use of a language translation file. This will be
in later chapters.
In the HRMIS spreadsheet, we scroll further to the
right and see HR_Employment as a new standalone
and next to it is a sub-table to it. This is to put in past
job experience or employment history of employees.
Such data will be created anew. What we shall do is
to prepare the data as Excel format also.
This new table is HR_EmploymentHistory. The prefix

HR_ is to help identify the table easily as there are
about 500 tables already in the core iDempiere. All
are prefixed and we shall explain them in the
appendix for basics.
The first line belongs to HR_Employment. This

means It is a sub tab of HR_Employment table as
can be seen below.
See how the properties in the spreadsheet appear on

the actual window panel. If any changes wish to be
made, the implementer just go back to this sheet and
make the changes there. S/he can also add and
delete fields and RUN_ImportIdempiere to clean
back the database and ImportModel again. This
allows fast prototyping without running around
correcting the AD manually and losing sight of what
goes into what.
Hopefully by now, you can begin to grasp the whole idea behind Ninja and the smart
integration it has with the spreadsheet to create new models. Knowing such tool doesn’t mean
we can be all powerful. The GIGO principle still applies. You can use this tool to design a
fragmented or elegant module. A tight or loosely coupled design.

THE SAP KILLER
Chapter 8 - The Migration Approach
Now, I describe how you can migrate whole legacy data the easiest way into iDempiere. I have
simplified and revolutionise the approach to be dead simple which is crucial as migration is
one of the scariest and deadliest part of a large MIS or ERP effort. Many projects failed even
before it began due to the migration part.
Why does that happen? Firstly, data needs to be scrubbed or verified. There maybe redundant
repetitive data, some or lots of it outdated. Mostly will be in multiple sheets of soft copy or
worse still have to be input again from scratch and human errors abound.
In iDempiere as inherited from Compiere, there is an Import Loader regime to migrate data into
the system. But it has to be properly formatted first. And the way it looks up and populate is
hard coded. Thus it can break as time goes by due to changes to data models or you need to
code each time you have a new model.
My approach dispense with all that. The principle is to match exactly the data model property
headers to the underlying database design. Column for column. It is WYSIWYG - What you
see is what you get.
But how do I solve the raw jumbled up formats? I again use the Excel spreadsheet as the killer
app. As you will see further in the same HRMIS spreadsheet, there are another three parts of
the spreadsheet, to the right of it.
The left or first part is where raw data is taken from the legacy system as is basis. Then the 2nd
or middle part is the master sheet where the raw data is combined. Then the right side is
where the data is transposed into core iDempiere database models.
I shall explain more clearly by taking the actual data in it.
Now before that, let me explain abit about the two types of actual user data that may exist in
the database. One is the core tables. Such as C_BPartner, which is the Business Partner
master data that is used for Vendors, Customers and Employees alike. So what we are to use
shall be the exact format of this table to inject new data or migrated data from the outside. The
point to remember is that to import data into such, requires all field labels or headers to be
exact and its prerequisite or mandatory fields to be available.
Here is a list of core master tables that any data to get into it, must fit exactly:
1. C_BPartner - Business Partners
2. M_Product - Products
3. C_Order - Orders
4. C_OrderLine - Order details
Let us try a real life case which I already prepared. Fetch ProductMigration.xlsx from
https://sourceforge.net/projects/red1/files/UnicentaPOS/. Open it up and go through each
sheet.
Notice there are a number of sheets beginning from Master, 1_M_Product_Category,

2_M_Product and so on. Right away you can deduce that those with prefixes of 1_ until 6_ are
main core tables in iDempiere. You can look through each of them and see that they are
matching the actual database tables. Before we open up the Master sheet, let me explain what
this spreadsheet is all about. It is to prepare data for BOM or Build Of Materials for a chain of
cafe restaurants. Each product BOM has children or formula of ingredients at the precise or
estimated quantities so that when a parent product such as Spaghetti or Coffee is ordered, its
exact amounts of ingredients used are calculated. This is so that the central store will be
notified of each branch’s usage automatically. This is one brilliant idea I myself did.

THE SAP KILLER
Like mentioned in passing earlier, there are three sides to such a spreadsheet to migrate
actual data. The left side is the raw data from the legacy. In this case we do not need it as this
is fresh data. The middle is the Master data and it is as shown above. On the right of it, are all
core tables data.
When you look through each sheet of the core table data, in each column data, you will see
that they are referring to this master data. If you wish to change anything, you just change it
here in the Master. It is that simple. The idea is to let you focus on your own legacy data and
not worried where the exact fitting table layout is.
Let us look through one by one to see this concept

in practice. First we take the first sheet tab to the
right labelled 1_M_Product_Category.
Look at cell C3. It refers to =$Master.A3. So this

way we just edit data in the Master and the rest of
the sheets shall change accordingly.
This table data is a pre-requisite

to the M_Product data to be
grouped under these new product
categories. The row 2 header is
the exact match of the table
column-names. During Ninja’s
ImportModel, the rows of data
below them will be inserted
accordingly.
We now look at the next sheet tab,

M_Product. We see again how
the cells are referring to the
Master sheet for non-redundancy
convenience and accuracy.
Here can proceed and examine

each tab and see how the new
data is been laid out. It is intuitive,
and amazingly easy for the data
migration process. There can be no error because these sheets are tested in my real user
case. I am now sharing for reuse to the public and community.

THE SAP KILLER
As I said before, this to me is more superior and handy than the current Import Loader as data
is more visible and easily managed in a single master sheet. Also, the new iDempiere window
tab direct Import isn’t that easy when you have many different table sets to handle. Each table
requires you go to that Window Tab and attach the data set individually. That is cumbersome
and doesn’t allow you single spot visibility and reference. So, for a multi-table data-set such as
for BOM sets, this is no longer a hindrance.
This way is more quicker and zipping them (only the exact core right-hand side tables) into a
single zip and using Import Model process at Ninja’s ModelMaker lets you do the importing
with a single act.
So, this is the way of migrating to existing core tables. But what about non-core? The concept
is still the same. But I added one more magic. You may skip the Table-Column creation and
jump. No extra steps if your data type references are already created such as L#, D# and Y#.
Just put your data in defined headers, and Ninja goes further by creating new tables at the
same time while migrating! You can return back to the HRMIS spreadsheet tabs sheets’
design to see the same principle. Take those data sheets and just zip them and import them.
You only need to import first the HR_Terminology under the Model Maker.
However, if you want the option to code logic, where you have to create the process and
model code stubs, you have to include the rest under the ModelMaker column.
So now, for example, to import data directly into one such model, HR_EmploymentHistory, we
just need to prepare a sheet containing data that fits the properties in the table. Just remember
that since it is a sub table, it needs to have a value at the HR_Employment_ID that already is
present in the Name property of the parent HR_Employment table. Just go through those
sample data in the HRMIS Excel to get the idea.
(To be cont’d)

THE SAP KILLER
PART II
Chapter 1 - My Best Practice
What is the best way to learn programming? Why, it is bugs, of course. Make lots of them.
Actually you do not have to. Just strike the keyboard for the first time. 110% you hit a bug. It is
natural part of life. First time you try to feed yourself. First time you go on a bicycle. That
learning process cannot be stopped. Or else you never learn. Up till adulthood that is our
pattern and best practice of learning and living. Trial and error.
What is the best way to learn iDempiere and now with Ninja? Read above or listen carefully to
me, someone who has been around the IT world or IoT world if you prefer the latest rehash of
old terminology, for since 1980. Right after I left high-school. At a time when there were only
mini mainframes. No Internet, no mobile phones and certainly no Google to fact-check
anything we say.
But experience is something that Google often miss. It is something best not said in words as it
can be misleading or becoming a ten commandments cast in stone which gets broken
everytime it is tested.
My best practice is something i went through uniquely and it cannot be cast in stone. But i be
the happiest person if any of them gets broken. I love learning something new. And i will
always shout it back out to the world. That is why this project is so made for me. Openness,
transparency, no information hiding. So this is my first best practice rule. If you do not publish,
you perish. ‘Publish or Perish’. Wait, it was already said. So here i am going to rehash that to,
And guess what? You be blamed later anyway. In fact, the original creator of Linux is a
university professor teaching software, Dr Andrew Tanenbaum. He called it Minix as a
sandbox of Unix, the prevailing server side operating system that time. Linus called it Freax,
an acronym of Free and X. Someone else called it Linux and it stuck. Why I put Linus Torvalds
name in there? To prove a point more clearly. Prof Dr. Andrew Tanenbaum could have
unleash it first, but instead he was too high up in his cathedral. He even criticized Linus for
releasing it.
But Linus real mantra isn’t that. He is attributed to the phrase,
When he released early, mistakes were abound, and other techies rushed in to correct them.
Thus his time spent is actually a very small fraction of the total or actual cost of releasing Linux.
This i found to be true in the iDempiere project and in my own. Others gave me bits and parts
that eventually helps me more in the long run. I have no worry of anyone stealing my trade
secrets because A: I am not the smartest guy in the room, and B: It is still too difficult to make
what you steal work the way it supposed to be, no matter how much you read up or open it up.
It is much better and cleaner to let your mechanic change your engine oil for you.
In a way, choose your mistakes well. Do not try to learn everything at one time. It takes a long
time to arrive at where I am. And i am moving my goal posts every day. Sometimes every hour
as i sit to study and improve my work.
In essence, what i write is the culmination of many mistakes and Aha moments.

THE SAP KILLER
This is often heard among my project members or students. It is wide spread in quite all my
writings. You probably hear the same among car mechanics not to touch the engine to ensure
nothing is broken further. Instead check the fusebox. It could be just a matter of a fuse, instead
of pulling a long wire to rewire the headlamps. It maybe a quick fix. But it can be a fatal one.
Something else may be broken. What is worse, when you change mechanics, the new one
has no idea what that wire is for.
OSGi is the technology that allows the core engine to be decoupled from your module or
vertical functionality. It is used by Eclipse when IBM acquired Equinox to be contributed
together with the IDE (Integrated Development Environment).
What this means is not to over-write or even append to any master table. Instead append as a
sub-table, or as I recommended in my HRMIS sample, a separate table that has a field that
looks up the core table reference.
I give a good example. A classic one as explained before in my BlackPaper here:

http://www.red1.org/BlackPaper.pdf
In fact my Black Paper tells everything you need to know what Not to do. So read it together
with another PDF below of more best practice that you must do.
If you read my FitNesseTesting pdf here, it says to do it from outside your organisation!
https://sourceforge.net/projects/red1/files/Testing/iDempiereFitNesse.pdf/download
Most people test their software from the inside, by the very programmer who programs it. That
is an anti-pattern of Software Engineering. Yes, you may test it, but that test is not official nor
acceptable that you have passed the testing stage. This is not to be confused with UAT, nor to
be passed on to the users during UAT, as some will inadvertently do.
UAT or User Acceptance Trial is as the name suggests. It is for the user to do their own final
assessment. But for them to find bugs that are deep is not fair. There are bugs that crop up at
year end, or Xmas holidays, when no one is working. Including the software.
And continuously test, test, test, as the mantra goes. This can be achieved by a best practice
tool called Jenkins, a fork of Hudson. Here is the iDempiere testing site -
http://ci.idempiere.org/.

THE SAP KILLER
It is not to self-promote, but I created the Ninja wizard is to implement the best practice
forcefully. It facilitates that all modules are built as they should be - as modules, separate,
decoupled and talking to the core via clearly defined interfaces.
With nothing changed in the core, you enjoy:
1. No breakage during upward and forward migration.
2. Less maintenance as the ones that break are only your own modules. And you already
know where to look. Debugging can be tricky as you work across an enterprise. So narrowing
your sight allows you to focus and be up and running to the maximum.
3. Non-repetitive redundant work in setting up new AD models in your module or creating new
Java code that is always boilerplate.
4. The Kanban Board document event handling also helps in making your own module less
clumsy and spaghetti strewn as event changes to the Document status are handled in a single
spot, namely the DocEvent java already generated automatically for you to code just the logic.
5. It avoids you from making human mistakes in the AD changes to setup menus, windows,
tables and columns, as well as processes parameters, info-windows, data types, and print
formats and thus saves you a lot time even days on end.
6. It gives a simple point of reference as Ninja picks off from a single spreadsheet.

THE SAP KILLER
Chapter 2 - Just Enough Java
Java is perhaps one of the hardest thing to learn, even for me as i was a Cobol programmer in
1982. Object Oriented concepts are really alien, and it took me years , from 1996 when it was
at version 1.0 till 2003, perhaps Java 1.6. Today at time of my writing it is at Java 11 and
change owners from Sun Microsystems to Oracle Corporation. But with the Compiere project
where I could intercept the code during execution and see what is actually happening, i began
to learn it for the first time.
Today, I still consider myself in the learning process, but I have learned enough to at least craft
the basic data model’s business logic and leave the IoT style framework handicraft to the top
gunners of our project, Carlos Ruiz and Low Heng Sin and Hiep Le Quy. They make up the
Circle of Trust that reviews the final code versions in the iDempiere project.
To learn to their level is redundant or too difficult for most people, me included. However to use
iDempiere to make specific functionality not covered by the horizontal ERP or MIS, we have to
sooner or later touch code. Or at least read the code to understand how things work or go
wrong. It is also the hallmark of been open source to have the ability to access code and
modify it and eventually improve it and share the improvements on a global scale.
I have instituted the Ninja template to allow such data model coding to create processes, and
data model events. The logic will be more focus to real life business functionality and not be
burdened by any boilerplate code. Of course the most perfect world is the use of a Business
Rules Engine, but that is another big learning curve and so far no one in the top circle has
made any mention of it. Other than me. And I cannot get my head around the best way to do it
yet. Whether i be using Drools or a Ninja style approach. But I am still thinking about it now
and then.
Meanwhile, I feel the Java that is needed is minimal and sufficient to fulfil quite all the needs of
any vertical modules that may exist. I have work on numerous module creations such as
Budgeting, Manufacturing, and Warehousing and know the scope of Java skills needed. The
Ninja itself as my creation was done using the same Java knowledge. Nothing further.
In my other project work such as Android Mobile integration to WebServices, and Unicenta
POS integration via Active Message Queue, I encounter more Java frameworks and those are
particular and specific but not needed here where most business logic resides.
With this sufficient set of Java knowledge, coupled with the Ninja wizard, one should be able to
build any module needed in the world.
However for the front end or the world of IoT, Ninja or iDempiere is not needed. MIS or ERP is
essentially the back-end of a system. The front end it has is sufficient but not as modern as
some front-ends that are emerging as we speak. Thus, let them be, as they usually talk to the
back end via WebServices and Message Queue technology. I might cover them in this guide
at a later version. For now, i shall go to the basic set of Java code that is intended.
Below is a sample already shown in the early chapter of what is left for the coder to do in the
case of a process. First, let us understand what is a process and what does it do. A process is
to do something further within a standard data model. The generated data model already has
just simple CRUD stuff. CRUD stands for Create, Read, Update and Delete. A process is a
more advanced CRUD realm, where data from the table is now used to access further data
outside and subject to business rules before updating back to itself or other tables.
An iDempiere process is first defined in the AD, and thus is a meta-data configuration, not
needing any hard code but to point to a record or instance, to retrieve the intended full details
to be iterated and acted upon. The Ninja setups all the steps in the AD for a process and its
parameters, so that the developer only needs to code the pure business logic.

THE SAP KILLER
What is to be coded is at the grayed line 32 of the doIT() method.. More logic can be inserted
before that line returns any results intended. I will show a typical code sample and then explain
what it really does.
This is from the BlockChangeMenu.java under my other magic plugin, The Wukong Menu. It
basically make any part of the main menu tree appear or disappear. Like magic. The trick is
just to mark the checkbox active or not. So let’s see the doIt() method above.
After it takes the AD_Menu_ID parameter from the selection by the user, it checks if it is not
selected with, if (AD_Menu_ID<1). It will then return to the panel with the message ‘No Menu
Selected’ reminding the user to choose something.
It shall then pass the process to a sub method called iterateMenuTree(AD_Menu_ID). At that
method, the passed value is put into a Query method. Below it is copied in full.
List<X_AD_TreeNodeMM> parents = new

Query(Env.getCtx(),X_AD_TreeNodeMM.Table_Name,X_AD_TreeNodeMM.COLUMNNAME_Parent_ID+"=?",get_TrxName())
.setParameters(menu_id)
.list();

THE SAP KILLER
First let us understand what is this Query method. It is what in the Java world called an ORM
or Object Relationship Model where data from a database is queried. Previously even in
Compiere, the way it was done is using the SQL or Structured Query Language method. As
the name implies, that is a structured and not an object oriented method. What do we mean by
that?
Structured is monolithic, linear as how English sentences are written. Below is an example of
the Query method rewritten in the old SQL:
SELECT * FROM AD_TreeNodeMM WHERE Parent_ID = AD_Menu_ID;
Now that looks not too bad and simpler right? Wrong. It is actually a string and it can be
hacked during execution by what is known as SQL Injection. It is also bulky and unwieldy. The
full code to execute the string is as follows:
public <T extends PO> List<T> list() throws DBException

{
List<T> list = new ArrayList<T>();
String sql = buildSQL(null, true);
PreparedStatement pstmt = null;

ResultSet rs = null;
try
{
pstmt = DB.prepareStatement (sql, trxName);
rs = createResultSet(pstmt);
while (rs.next ())
{
T po = (T)table.getPO(rs, trxName);
list.add(po);
}
}
catch (SQLException e)
{
log.log(Level.SEVERE, sql, e);
throw new DBException(e, sql);
} finally {
DB.close(rs, pstmt);
rs = null; pstmt = null;
}
return list;
}
Now you understand? It is not only bulky but error prone because as a string typed script, the
developer can inadvertently make typo errors and string content mistakes are not noticed
during compile time but during run-time. Containing this bulky method in a Java pattern inside
the ORM framework controls the SQL injection not to be possible by the calling method and
strictly housed inside the Query method which has measures to prevent SQL Injection. For
example the buildSQL sub method it to build the SQL string is doing it this way:
private final String buildSQL(StringBuilder selectClause
Notice it is declaring a final String object which does not let the object be mutated during its
lifetime after it has been created here. Then the DB method is finally terminated at the end of
the Query method with
finally {
DB.close(rs, pstmt);
rs = null; pstmt = null;

THE SAP KILLER
This is coded once here and thus guarantees no lazy programmer forgot to do so, as there is
no need to repeat the whole DB method each time you wish to handle an SQL. Using a Query
method ensures no typo error as you be using object arguments in the script such as
TableName, COLUMNNAME, and .list. Any mistake while typing those will not compile but
highlighted for you to correct time immediately rather than unknown and during run-time which
is fatal.
The shorter Query method immediately returns this bulky work as an object called ‘list’ that is
cast back into the calling method object, in this case, AD_TreeNodeMM. The developer can
focus more easily with better visibility. I will touch about the TableName and COLUMNAME
objects after going through the rest of the iterateMenu logic.
This ORM Query framework was introduced quite early after the ADempiere fork by Teo Sarca
and Victor Perez. It immediately got adopted as our best practice.
(To be cont’d. Comments are welcome. Tell me what you do not understand, or can be written
more clearly, and i will edit the content)

The Book of Ninja

Uploaded by

Copyright:

Available Formats

The Book of Ninja

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

The Book of Ninja

Uploaded by

Copyright:

Available Formats

THE SAP KILLER

The SAP Killer

Dated January 1-16, 2019

What Can This Book Do?

Further paid consulting support is available by contacting the author at red1@red1.org.

Copyright 2019 Redhuan D. Oon Page 1

Who The Heck Is SAP?

RED1’s THREE LINE MANTRA

INFORMATION IS FREE, YOU HAVE TO KNOW

PEOPLE ARE NOT, YOU HAVE TO PAY

CONTRIBUTORS ARE PRICELESS, YOU HAVE TO BE

Copyright 2019 Redhuan D. Oon Page 2

Before You Continue

Not For You

Copyright 2019 Redhuan D. Oon Page 3

Copyright 2019 Redhuan D. Oon Page 4

What Is Red1’s Ninja?

Copyright 2019 Redhuan D. Oon Page 5

CHAPTER 0 - A Crash Course in MIS

Copyright 2019 Redhuan D. Oon Page 6

CHAPTER 1 - The Excel Template

Open it and see in the first sheet called 1_RO_ModelHeader

Copyright 2019 Redhuan D. Oon Page 7

Now we go to the 3_RO_CodeMaker sheet.

Copyright 2019 Redhuan D. Oon Page 8

GenerateModel and GenerateModelFactory are usually checked Y together to generate

Copyright 2019 Redhuan D. Oon Page 9

Chapter 2 - Ninja Generate Module

From the Menu tree, call up the Module Designer

Copyright 2019 Redhuan D. Oon Page 10

Go to the Model Header main tab and refresh.

Copyright 2019 Redhuan D. Oon Page 11

COMPLETED - HEADERS:1 MODELS:1(COLUMNS:3) CODES:1

Copyright 2019 Redhuan D. Oon Page 12

Copyright 2019 Redhuan D. Oon Page 13

You can see the MY_Process which was defined

Under the src folder, under

These saves the developer from coding or creating

Under the org.my.model structure, are the ORM

Again you may delete the Acme place holder class

Under org.my.process is the new MY_Process

Under the OSGI-INF folder are two new XMLs.

There is one task to do here before we can activate

Copyright 2019 Redhuan D. Oon Page 14

Copyright 2019 Redhuan D. Oon Page 15

First you will get the break button to

Copyright 2019 Redhuan D. Oon Page 16

Chapter 3 - Ninja Starter Cheat Sheet

1. Install iDempiere source code in Eclipse IDE

2. Install Ninja Plugin in Eclipse IDE

3. Install Blank Plugin in Eclipse IDE

4. Export Excel Template first 3 sheets as CSVs in same sheet names

5. Zipped up as any name

6. Launch iDempiere in Eclipse with Ninja plugin checked

7. Attach to Model Maker tab of Module Creator

8. Run Import Model or Generate Module

9. Refresh to new record at Header and Generate Module

10. In Eclipse, insert PLUGIN.txt into plugin.xml

11. Restart iDempiere in Eclipse with My Plugin checked

Copyright 2019 Redhuan D. Oon Page 17