sed awk substring challenge 2

[GROKING]

Hello,
I have an input file like this
1, "1,123", 10, 5

I need to remove the comma , which is between the double quotes so the data appears like "123", and newline looks like this
1, "123", 10, 5

any ideas??
thanks for the help!

 

[Annihilannic]

This should do the trick:

echo '1, "1,123", 10, 5, "3,123,234"' | sed '
        :a
        s/\("[0-9]*\),\([0-9,]*"\)/\1\2/g
        ta
'

ta causes it to branch back to label "a" (defined by :a) if the search and replace was successful. This is to ensure that numbers in the millions (i.e. multiple commas) are handled properly by repeating the replacement.

Annihilannic.