Simple diff tool for comparing directories written in Java.

There is nothing to use in this project yet!

TODO:

• Next steps:
  • Support "exclude directories" (and handle exception "access denied")
  • Compare two files with the lists.
• Compare directory files including contents (md5)
  • Done: Create implementation, which goes through the directory files, calculates their MD5 sums and stores them into text file.
  • Done: Create sub-command of the application, which uses the implementation from above and create those files.
  • Create unit tests for the algorithm, which writes hashes.
  • Create implementation, which would read the text files with relative paths and MD5 sums and compare them with the actual file system.
  • Create sub-command, which would use it.
  • Create unit tests for algorithms, which checks hashes.
• Single dir analysis:
  • Add some profiling code. Measure, what takes time - listing of the files, or comparison etc. Experiment with it.
  • For the files get file size
  • For the files calculate some check-sum/hash.
  • For large files estimate access speed (how to do it intelligently)?
  • Report following things (create object to query this information, then also create the reports and corresponding command line options):
    • Name of dir being examined
    • Number of files, dirs, links, other objects
    • Total size of files, max file size
    • List all elements
      • Sort by type, by name, by size
    • Calculate dir sizes
    • Show largest dirs
    • Show number of files inside dirs (direct children/nested)
    • Sort by number of files
  • Think about how to print the report about content of the directory.
  • Advanced: Improve the results data structure, so that one can group files inside subdirectories, i.e. that one can take a subdirectory and list it's children
  • Advanced: implement it using "walk" and "walkTree" and see if any of the two have any benefits.
• Dump information about some root dir as a file. Later use this file as a "side" for comparison.
• write function to calculate the diff.
  • Improve the output of comparison:
    • Sort the output
    • If directory is missing, don't print the whole content, but onsly some status (like how many files are missing).
  • Add progress and some simple profiling print-outs - to see, what takes long time.
  • Write test-cases and unit tests for this method
  • Think, how to use it in for real thing.
• Think how to make the API nice to be usable in interactive Java or Groovy session.

Target volume: 3M-5M fs items. Memory (very roughly: 0.5K*5M = 2.5G - should be OK. Adjust Java options?)