Do you guys think that storing the data in an XML file
and using the XPATH search will be faster than the text file search?
I would make a guess that a widely-used XML parser and XPath
search engine will have been more carefully optimized than your
home-grown equivalent, but that's only a guess, since you haven't provided any