Is there a way to associate data with a file in a folder hierarchy?

77 views Asked by At

Using Java, I am creating a program that indexes a folder structure and allows a user to search for files and also tag a file with keywords and then search for files based off of those tags.

I have been traversing through the folder hierarchy using the FileUtils listFiles method at the moment which is essentially this question: Recursively list files in Java

I haven't yet begun to code the tagging functionality, but thinking ahead I'm fearing that if a file is renamed or moved after I associate it with a tag then it will lose the tag. This defeats the purpose of my program, so can anybody offer suggestions as to how to store each file located in the folder hierarchy or associate the tag so that if a file is renamed or moved it will still have the tag associated with it.

3

There are 3 answers

3
VGR On BEST ANSWER

If you want to keep track of a file, even when its name and/or location changes, you should use its unique identifier, which in most file systems is called its inode. (I think NTFS/Windows calls it a "file ID.") You can read a file's inode using its BasicFileAttributes.fileKey:

Object key = Files.getAttribute(file.toPath(), "fileKey");

That key is suitable for use as a HashMap key.

0
satnam On

If the OS doesn't support file tagging, you could:

  1. maintain a mapping of file path to tags
  2. maintain a mapping of file hash to tags

Using option #2, your tags would be preserved even if a file was moved. But if someone moved AND modified the file, then the tags would be lost.

0
MiiinimalLogic On

I don't think there's a way to do this without updating your tag relationship to the newly create file since the rename/mv operation is at the disk level and there is actually a delete and 'create' file compound action happening in the background. Because of that, there's no guarantee that a file will even be in the same place on the disk. If you know for sure that the file will have the same contents, you could take an MD5 signature of the file's contents in a String object and then always compare those when a tag is queried, of course this has its down falls too when the file's content changes.

Your best bet is to use a hash map w/ the files' paths and tags and then use directory watcher to update the hash map when a file name changes. Thats the best I can think of!