Coding Conventions and Structuring

Indentation (字下げ) is the practice of adding spaces at the start of lines so that blocks of processing and nesting levels are visible on sight.

Lines inside a conditional branch or a loop are shifted one level to the right, and lines at the same level start at the same position. In most languages changing the indentation does not change the result, but some languages such as Python use indentation to decide where a block begins and ends. Either way, block boundaries are clear at a glance and misreadings become less likely.

AnalogyIt works like an outline in which sub-items are written slightly to the right under a main item, so you see at once which sub-item belongs to which main item.

How to rememberThe Japanese name literally means "lowering the characters": the act of pushing the first characters of a line to the right became the name itself.

Indentation (インデンテーション) is simply the English name for 字下げ and refers to exactly the same thing.

The leading space itself, or its width, is called an indent, and one level is made of space characters or a tab. If the width of one level is not kept the same across the whole program, the boundaries between blocks become harder to read, not easier.

How to rememberThe English word indent describes the notch where text is set back inward from the edge of the page.

Nesting depth (ネストの深さ) is how many levels of nesting are stacked, where a conditional branch or loop is placed inside another conditional branch or loop.

For example, a loop containing a branch that contains another branch is nested three levels deep. The deeper the nesting, the more levels you must trace back to see when a line runs, which invites misreadings and faulty fixes. Deep parts are kept shallow by measures such as moving them out into separate processing.

AnalogyThink of boxes inside boxes inside boxes: the innermost box takes the most opening to reach.

How to rememberNest means a bird's nest in English: just as eggs sit in a nest, processing sits inside processing.

Naming conventions (命名規則) are rules agreed in advance on how to name variables, functions and other elements.

Typical rules are to use English words that show the role, to mark word breaks with capitals or underscores, and to write constants in all capitals. Naming conventions are usually set together with rules such as the indentation width, as part of the coding standards that keep the way programs are written consistent. When names follow one scheme, people other than the author can guess a name's role from the name alone.

AnalogyIt is like naming every file in a cabinet as "year-month_client_content", so anyone looking can guess what is inside.

1 past questions on this term 命名規則 →

Module decomposition (モジュール分割) is a technique of splitting a large program into parts by function, called modules, and assembling it from them.

Each module is given a single role, and modules are combined after deciding what information they pass to one another. When a change is needed, its effect tends to stay inside that module, and parts already built can be reused in other programs. The word module is also used for the program units produced when source code is translated and linked.

AnalogyIt is like furniture built from boards, legs and drawers, where a broken drawer can be replaced on its own.

How to rememberModule means a standardized, interchangeable unit, like a spacecraft's habitation module, a part that can be swapped or added.

2 past questions on this term モジュール分割 →

A main routine (メインルーチン) is the central processing that starts first when a program runs and directs the flow of the whole program.

It leaves individual detailed tasks to lower-level processing that it calls, then moves on to the next step once that processing returns. Reading the main routine therefore tells you what the whole program does and in what order.

AnalogyIt resembles a site supervisor who sets the order of the work, hands each stage to the right tradesperson, and moves to the next stage after hearing that one is done.

A subroutine (サブルーチン) is a part that packs a specific block of processing into one unit so that other processing can call it.

← Scroll the figure sideways →

When called, it carries out its processing and then returns to where the caller left off. The same code need not be written in several places, and one fix applies everywhere it is called. Depending on the language it is also called a function or procedure, and a subroutine can itself call other subroutines.

AnalogyIt is like a recipe that says "for making stock, see the separate page", which you can turn to whenever that step comes up.

How to rememberSub is a prefix meaning below or subordinate, the same sub as in submarine, a ship that travels below the surface.