Need help reading damaged file

[PoulsenL at capanalysis.com]

I have about 100+ files that are a dump of old tape from a database.  Most
of the data is good, but it is interspersed with damage that contains
backspace characters and I _believe_ EOF characters.  When we try to import
the data it only imports 1/3 or 1/10, etc of the data depending on the file.
I can pull it up in WinEdt and see that it contains far more lines.  I
created a script that reads through the file and counts the lines (not the
most efficient script in the world, I'm sure).  The problem is that the
script suffers from the same problem as the import utility.  It stops far
short of the the end of the file.  Any help would be appreciated.

Here is the script:

import glob, os

def countlines(a,b,c):
    for file in c:
        if os.path.isfile(b + '\\' + file):
            input = open(b + '\\' + file)
            print b + '\\' + file
            x = 0
            for y in input:
                x += 1 
            print x
        
os.path.walk('\\\\Server\\Dir\\',countlines, None)