Sorry for my late answer. I copy and paste from the software help file:
The components line has a more technical purpose, and is more abbreviated, than the Explanation section. Inclusion of 女 (nǚ ‘woman’) and 子 (zǐ ‘child’) on the components line of 好 (hǎo ‘good’) causes the dictionary entry for 好 to be automatically indexed to both the components 女 and 子, so that when you list all the character containing 女 (or 子) as a component, then 好 will be included in the list. This is useful as a method of looking up a character when you don’t know the pronunciation (similar to, but more flexible than, the traditional system of dictionary radicals). It is also useful as an aid to memorization, for making meaningful associations (and avoiding confusion) between characters with similar shapes.
The “components” on this line do not always have any “true” (historical) relation to the character they are listed under. For example, the character 克 may have 兄 listed as one of its components, simply because the modern form of 克 seems to have 兄 at the bottom; but historically these two characters have nothing to do with each other (the historical analysis belongs in the Explanation section). The point of the components line is simply to provide access to any character by way of anything that seems to be a graphical component of it; and also to allow for comparisons between characters with similar shapes. The criteria used in deciding the analysis into components are flexible. Chapter 9 explains how to edit a character dictionary entry, including changing the components line.
.............................................................
What to include on the components line, is sometimes a matter of choice. The main idea is to help anyone who might try to look up the character by choosing Characters Containing Components from the List menu. You can follow this procedure:
• Look at the whole character
• Try to divide it in two
Most characters can be divided in two from left-to-right, and many others from top-to-bottom. If the there is more than one way of dividing it in two – if it isn’t obvious which way is historically relevant – then try to do it in all ways that one might imagine.
For example, there are two obvious way to divide 章 (zhāng ‘badge’) in two:
1: 音 (yīn ‘sound’) over 十 (shí ‘ten’)
2: 立 (lì ‘stand’) over 早 (zǎo ‘early’)
音 over 十 is the historically accurate analysis. On the other hand, if one didn’t know better, one might imagine it to be composed of 立 and 早. Including 音 and 十 and 立 and 早 on the components line provides a way for anyone to locate 章 if they recognize any of its apparent major components.
Although 章 could be divided in three, with 日 (rì ‘sun’) in the middle, we try to divide it in two; so we don’t regard 日 as a major or essential component. Nevertheless, you could optionally include 日 on the components line as well.
3: 立 over 日 over 十
This approach is more flexible than the traditional classification by radicals, where you are forced to guess what the radical is (and waste a lot of time if you guess wrong). In Wenlin, any reasonable guess based on the above procedure can work. The goal is to help people find a character, not to hide it.
The components line allows for an efficient dictionary look-up method, and some useful lists; but it’s not a theoretical statement. Remember, the historical analysis of a character belongs in the explanation section, not on the components line.
One might reasonably ask, “If you want to help people look up characters, why stop with a division by two; why not keep diving down until you get all the components?” Consider the horizontal line, which is the character 一 yī ‘one’; and consider 口 kǒu ‘mouth’. Horizontal lines, and boxes, appear almost everywhere in Chinese characters. If 一 and 口 were included on the components line every time they appeared in a character, it would make the list of characters containing 一 and 口 very long indeed. This might be self-defeating, and nothing helpful would be likely to come of it (for many common purposes).
