"unlink" has a specific meaning, and it doesn't refer to the way a file is assembled out of scattered blocks. Sorry if this is already familiar to you, but from

"unlink" has a specific meaning, and it doesn't refer to the way a file is assembled out of scattered blocks. Sorry if this is already familiar to you, but from what you say it may not be. On Unix/Linux filesystems, a link is a directory entry pointing at a file, and there may be more than one such link. Unlinking a file looks like deleting the file from the user's perspective, but if another directory has a link to it, the file is not affected - not deleted, not split in to scattered blocks. Hence a normal copy of the filesystem will copy it over - a block copy is not required.