To improve our understanding of how Power Query works, let’s step back from the pieces and look at the paradigm—the methodology that M is built around. Conceptually, how does M work? How does it think? What makes it tick? The particulars, like variables, let expressions, functions and identifiers, are all important, but if we only focus on details, we might miss the big picture. We might not see the forest because we’re looking so intently at the trees. Let’s step back and learn about M’s modus operandi.
Without an awareness of M’s methodology, we might spend our days writing M code perplexed by certain behaviors, puzzled by why M seemingly doesn’t let us do certain things we might want to do and perspiring as we waste effort trying to mimic code patterns from other languages that are needless in M.
Up until now, we’ve used variables without talking much about their specifics. Since we’re operating under the assumption that you have at least a little experience with another programming or scripting language, how we’ve used variables has probably made sense.
However, there might be differences—perhaps even significant differences—between how variables work in other languages you’ve used with and how they behave in the Power Query M language. Let’s spend some time exploring variables and related concepts to clear up any lurking confusion and position you to take full advantage of what M offers.
We’ll start with a brief recap of the main unit where we define variables: the
let expression. Then, we’ll talk about how variables (and other things) are identified. Related to identifiers is scope, so we’ll cover that, too. Next time, we’ll expand our understanding of how variables work by learning about M’s paradigm.
Not only can you define and invoke functions (as we covered in part 2), you can also pass them around. The ability to pass a function around without invoking it gives lots of flexibility.
Sounds complex? Yes, in words, but not necessarily in practice. If you’ve touched the Power Query M language, you’ve probably already passed functions around—just perhaps without realizing about it.
Then there’s the mysterious
each. It shows up a lot in code generated by the query editor. What does it actually mean or do? Turns out, it’s a handy shortcut that can simplify code you write.
We have ground to cover. Let’s get going!
If you read part 1 in this series, you may have picked up a theme about expressions that produce values: A simple statement, like
1, is an expression that produces a value. let is also an expression that produces a value.
Guess what? A function is an expression that ultimately produces a value. Unlike the expressions we looked at last time, a function only produces this value when it’s invoked.
Usually, parameters are passed to a function when it’s invoked. The function can reference these inputs as it computes what it will return. Continue reading
Sooner or later, you may find yourself working directly with the Power Query M formula language. Tools like Microsoft Excel’s Get & Transform Data and Microsoft Power BI provide point-and-click interfaces allowing you to build data transformation sequences which behind-the-scenes are implemented in Power Query M. However, these query editors can’t do everything you might need. Sometimes direct editing and authoring of M is required.
Search the Internet and you’ll find many examples showing how to use this language to solve one problem or another but little is out there describing the syntax and paradigm of the language itself. A limited knowledge of these details may be insignificant when simply copying and pasting samples and editing variable and column names. Move beyond this to weaving and writing your own solutions directly in Power Query M and a solid understanding of the language’s syntax, rules and capabilities becomes most helpful.