Tree Traversal

I'd suggest you apply as Support Engineer @ InterSystems
and over the years you might learn all the secrets of Caché.
I doubt what you ask for is subject for a public forum.

Hi Ponnumani,

I agree with Robert and Eduard, it's very difficult to answer such a broad question in any detail without duplicating in a single post (at length and time) what has already been written many times over.

However, tree traversal and its methods are very pertinent to mastering the fundamentals of Caché and worthy of a reference answer to help anyone else starting out on this path.

It's important to understand that Caché is essentially a modern implementation of the M programming language / database.

M was first developed back in the late 60's. It was mainly implemented on PDP servers which at that time where limited to a few hundred K of memory. These limitations greatly influenced the design and efficiency of M's data globals which would go on to be its core strength.

Caché was an evolutionary step in this technology time line which added a new high level object oriented programming language as well as a comprehensive set of libraries and tools.

Despite being an evolutionary step, Caché is still considered an implementation of M, adhering to much of the 1995 ANSI language standards.

Whilst some people distance themselves from this heritage, many of us are embrace these roots and like to combine both the modern aspects of Caché with much of what we mastered as M developers, the combination of which makes for a very powerful platform that you won't find in any other single technology.

It's important to understand this history as much of the information and teachings to master the raw fundamentals of Caché can be found by using information and teachings that still exist for the M programming language.

To help on-ramp any new Caché developers, I always buy them "M Programming: A comprehensive Guide" by Richard Walters as a short staple learning tool. There are also many other good books out there and a simple search on Mumps books will bring up a good selection.

All of these books will dive into globals, the hierarchical tree data storage which is very relevant to your question. They will also explain the $order and $query functions which are key to traversing global's.

Most of the time you will have shallow global data and traversing this data will be implemented with an equal number of nested $order functions. When the data is deeper, or the traversal needs to be more generic then you can either use the $query function for a brute force traversal, or implement a combination of $order and a recursive loop.

So the short answer to your question would be to seek out everything that you can find on Caché / M globals and the $order and $query functions. Take the time to read and practice using globals directly and you will master tree traversal. In addition, if you haven't already done so, I would also seek out generic books and teachings on programming algorithms that cover topics such as recursion and trees.

I hope this helps, and we are more than happy to help when you get stuck with more specific (and hopefully more detailed) questions.