Using Regex from command line to extract number

Question

I am using a tool to calculate cylomatic complexity of a javascript file.

Example:

jsc --minimal test.js

This command will give the following output.

┌─────────────────────┬─────┬────────────┬─────────────────────┐
│ File                │ LOC │ Cyclomatic │ Halstead difficulty │
├─────────────────────┼─────┼────────────┼─────────────────────┤
│ /home/shray/test.js │ 23  │ 4          │ 10                  │
└─────────────────────┴─────┴────────────┴─────────────────────┘
Cyclomatic: min 4 mean 4.0 max 4
Halstead: min 10 mean 10.0 max 10

Now I use

jsc --minimal test.js | grep "Cyclomatic:"

which gives me output as

Cyclomatic: min 4 mean 4.0 max 4

Now I have a regex, Cyclomatic:[\s]*min[\s]+([0-9]+) but I am not able to use it to extract the number showing minimum Cylomatic value.

Any help how can I just ouput the value of Min or Max Cyclomatic complexity value on the terminal output?

Answer 1

If you know that this line is always of the same format, you can use a simple cut:

cut -d' ' -f3

or with awk you can do the whole thing including your first grep:

awk '$1 == "Cyclomatic:" {print $3}'

If the line might change, use sed:

sed -E 's/.*( min )([0-9]+).*/\2/'

or grep -P if available:

grep -Po ' min \K[0-9]+'

or normal grep:

grep -o 'min [0-9]\+'

This returns min 4, which you can easily filter adding another grep or cut

grep -o '[0-9]\+$'
# or
cut -d' ' -f2

Answer 2

In Linux, using a look-behind (-P is only in Gnu grep):

jsc --minimal test.js | grep -Po '(?<=Cyclomatic: min )\d+'

(this extracts the first group of digits that follows "Cyclomatic: min ").