Migrating a Web Server to IIS 5.0

Converting UNIX File Names and Pathnames

When you migrate files and directories from a UNIX-based Web server, you need to edit file names and pathnames (called paths in Windows) so that they adhere to Windows conventions. Otherwise, hyperlink references to the files might malfunction, and in some cases files could be overwritten.

Conventions are slightly different between the Windows FAT and NTFS file systems. While for security reasons you will want to use NTFS as much as possible, there are cases where you might need to use FAT for cross-platform file sharing. The issues involved with both systems are described below.

File Name and Extension Length UNIX supports long file names and four-character file name extensions. When you copy a UNIX file or directory to a Windows FAT or NTFS file system, Windows truncates a long file name or extension using the Microsoft® MS-DOS 8.3 convention. It reduces the file name to its first six letters followed by a tilde (~) and a unique number, for a total of eight characters, not including the extension. It truncates the extension to its first two letters, followed by a tilde (~). For example, the long name ProjectOverview.esti would change to Projec~1.es~.
Case Sensitivity UNIX file names are case sensitive. Windows FAT file names are not case sensitive, and Windows NTFS file names are “case aware.” NTFS preserves case, but doesn’t use it in some functions, such as indexing. In UNIX you can have two different files in the same directory with names distinguished only by case. For example, in any given directory you could have two different files, one named MyFile and the other named myfile. If you copy these files to a FAT directory, the Windows operating system preserves the letter cases of UNIX file names, but interprets the two names as being identical as long as they use the same characters in the same order, regardless of case. In this situation, the Windows operating system reads the first name correctly, and then reads the second as a duplicate name, and appends a number to it to make it unique. Using the previous example, this would result in the file name, MyFile(2). Because this is a different file name, any hyperlinks or applications that reference the file will malfunction. Therefore, before you copy UNIX files to a FAT directory make sure they have unique Windows-style file names, and also remember to correct any hyperlink or application references to the file name.
Illegal Characters The following characters are supported in UNIX-style file names, but are not permitted in Windows file or directory names:
/ \ : * ? " < > |
The Windows operating system will replace each occurrence of one of these characters with the letter “X.”
Directory Separators UNIX pathnames use the forward slash (/) as a directory separator, and Windows paths use the backslash (\). Windows produces an error when it encounters UNIX-style pathnames.
Directory Hierarchy The UNIX file system appears to be a single directory hierarchy whereas Windows storage is divided into one or more physical or virtual disk drives with a directory hierarchy on each. To access a file on Windows, you must know what disk drive the file is on and specify the drive letter (C:, D:, E:, and so forth ) as part of the pathname for the file.

You also need to edit any file names and pathnames within UNIX application files in order to run them on Windows 2000 Server. This subject is discussed in Migrating Web Applications.