Yes, i guess you have a point. The whole purpose of storing the meta data in the text file was for optimising the speed. But unfortunately handling of large amount of data is slow.
I thought that by using the XML XPATH, I need not go through all the other data (by not doing a file.readline) if I wanted to search for documents with a certain name thus saving some time. But perhaps XPATH has to go through the whole document anyway.
Or am I cycling out in the forrest?