Slow replace 'x.y' to 'x,y' for large Quark document

Hi,
I’m trying to write script that will replace all appearance of ‘x.y’ to ‘x,y’ where x and y are digits. I made to loop from 0 to 9 for x and y, but this script takes 30 min for 100 pages text document.

tell application “QuarkXPress Passport™”
tell document 1
repeat with x from 0 to 9
repeat with y from 0 to 9
try
set (every text of every story where it = (x as string) & “.” & (y as string)) to (x as string) & “,” & (y as string)
end try
end repeat
end repeat
end tell
end tell

Anyone have any idea how to speed up this replace?
How can I find ‘.’ and check if previous and next characters are digits?
Thanks for help