Sentence count

Hi,

The following code counts the number of times a specific word appears in a file by separating each word onto a new line and then using grep to count how many lines contain that word.

Is there a way to change this so that it counts the number of sentences?

tr ' \t' '\n\n' < text | grep -c "$word"

Thanks.

First off, define what ends a sentence. That could be, but is not limited to: .!?" - then just look for them (and count!).

thanks for the reply. I want it to count the following '.!?'. I tried this, but I think my syntax is wrong. sorry im a newbie

tr -dc '.' < text | wc -c

Thanks again.

You could try using the fold command to seperate each character onto a seperate line. Then you could count the characters that end a sentence (.!?) using grep and wc together.

This won't, however, take into account any periods, exclamations or question marks placed mid sentence, such as an abbreviation (Mr., Ph.D., etc.)

Not at all simple, is it? A lot would depend on just how critical the accuracy is. If it is important that the value returned/printed is accurate as possible you are hoing to need to parse the input carefully, keeping track of context - are we in a quote and thus may not need to cound a '.', or was the previous character a '.' too; maybe implying we are in mid-ellipsis (the ... 'character').

Knowing *nix systems there is probably am existing base utility that will, with the flick of a switch, do exactly what you want!