Parse Unstructured Data in Shell script/Linux

Question

Example log : "some data7.575=tf.some data2.0=tf.some data1.23=tf.some data.."

I want to get each TF value as output. 
i.e. output should be 
TF=7.575
TF=2.0
TF=1.23

How can I parse this? Using Shell script (preferrable). Unix command or using Java.

Answer 1

You can use this grep -P command:

grep -oP '[\d.]+(?==tf\.some)' file
7.575
2.0
1.23

Or using this awk:

awk -F 'data|=' '{for(i=2; i<=NF; i+=2) print "TF=" $i}' file
TF=7.575
TF=2.0
TF=1.23