Find Duplicate Files

This problem is based on LeetCode 609 (Find Duplicate File in System). If you haven’t solved that one yet, it’s recommended to do so first.

In a simulated file system rooted at "/", your task is to identify all groups of files that share exactly the same binary content.

Two files are considered duplicates only when their complete binary contents are byte-for-byte identical; file names and directory paths are irrelevant. The content may be large, and you can access it only as a binary stream—direct string access is not available. You need to design an efficient approach that can handle a large number of files and directories.

Implement the DuplicateFileFinder class:

• DuplicateFileFinder(FileSystem fs)
  Constructs an instance that will operate on the given file system fs.

• List<List<String>> findDuplicateFiles()
  Explores the file system starting from the root ("/") and returns a list of duplicate groups:

  • Every group must contain two or more file paths whose contents are identical.
  • The order of the groups, as well as the order of paths within a group, does not matter.

The following FileSystem interface is provided; do not modify it. You can use its methods to list directory contents, determine whether a path is a directory, obtain file sizes, and open binary input streams:

class FileSystem {
    /* List all files in the directory */
    List<String> listFiles(String path){...}

    /* Check if the path is a directory */
    boolean isDirectory(String path){...}

    /* Get the size of the file in bytes */
    int getFileSize(String path){...}

    /* Open the file and return a binary stream */
    InputStream openStream(String path){...}
}

Constraints:

• All files are under the root directory "/".
• No two files with different sizes have identical content.
• 1 ≤ number of files ≤ 10⁶
• 1 ≤ file size ≤ 10⁹ (in bytes)
• Only files with identical content and size must be grouped as duplicates.
• You may assume standard hash utilities are collision-free for this problem.

Example: