I have tested both of the commands.

ls -l | awk 'NR%2 != 0'
ls -l | awk 'NR%2'

I saw the difference in the output, but, I do not know how can I explain it when I am teaching it to someone else.

Could you explain to me what != 0 means exactly?

CodePudding user response：

In awk, the variable NR is preset with "the total number of input records seen so far". Thus, when processing the first line it is set to 1, for the second line to 2, etc.

The % operator calculates the remainder of a numerical division. Thus, with NR%2 denoting the remainder when dividing the records number by 2, for the first line this computes to 1, for the second to 0, for the third to 1 again, and so on.

To determine whether a record is an odd row or an even row, just check whether NR%2 evaluates to 1 or to 0, i.e. NR%2 == 1 or NR%2 == 0. As 1 and 0 are the only possible values, one could also check for the negation of the opposite: NR%2 != 0 and NR%2 != 1, respectively.

In awk, a typical instruction consists of a pattern or condition (evaluating to a boolean value) and an action statement. In your case, the action statements are missing, which means the current record should be printed if the condition evaluates to true. For convenience, if the condition evaluates to a number instead (as in your second case), the number is automatically cast to a boolean value, with 0 becomng false, and any other value becoming true.

Therefore, awk 'NR%2 != 0' which evaluates to true on every other line starting with the first, will print odd rows and skip even rows. awk 'NR%2 == 1' will do the same. And as 1 is cast to true, just awk 'NR%2' will also do the same.

CodePudding user response：

My guess is that you need to teach that someone else about modular arithmetic first. Then it should be easier to explain the AWK part.