topbanner_forum
  *

avatar image

Welcome, Guest. Please login or register.
Did you miss your activation email?

Login with username, password and session length
  • Saturday December 14, 2024, 12:10 pm
  • Proudly celebrating 15+ years online.
  • Donate now to become a lifetime supporting member of the site and get a non-expiring license key for all of our programs.
  • donate

Author Topic: IDEA:Intelligently gather files & folders with same words in names to clusters  (Read 4587 times)

tmpusr

  • Member
  • Joined in 2005
  • **
  • Posts: 154
  • Instantiation stuck in meatspace with no backup
    • View Profile
    • Donate to Member
When you search stuff, download it, it just stays in the download folders, which you may have on several hard drives. Despite all this computation power, it doesn't go where it's supposed to by itself. You need to move it where it belongs.

I want a program that does that. How do you code a program that does that?

Just model what you do when you do that.

In a nutshell: Big clusters of similarly named files and folders act like magnets that suck similar named files to their vicinity.

You look at the filenames. When you see similar words, you know they should go into a folder.
Which folder? You know where there are files and folders with similar names. Those clusters of files and folders are your destinations. For example, you already have several albums by some artist in a folder. All the spread-around files should go there.  program would have to locate every file, perhaps utilizing Locate32's database, count how many times words occur in all filenames, and determine matches when those words are in any order, the more similar, the higher the match.

something A something B something
something B something A something
A B
B A

A B C
C B A
B C A
A C B

etc.

Option: Move only files that have two or more matching words in their filenames.
Option: Only move files or folders or files with some extension.
Option: Limit search folders.
Option: Simulate to see what would happen.

Just ONE run of it would seriously clean up your hard drives. It should check whether there's enough space on the drive. If it notices that you're going to run out of space, it suggests moving the cluster, i.e. the whole subfolder where similar named files are, to another drive with enough space under the same tree structure.

After it has gathered files into big piles, it would sort them further into subfolders if tighter matches are found.

Build as much intelligence into it as you can. Make it work exactly like you would.

Oh, and why not throw in automatic archive extraction into subfolders and clean up unnecessary folders. Just like you would.

And file & folder name cleanup with a simple txt file of words and sentences to kill. It could determine that if you have two words, like

Artist Name - Song

And you have lots of names like

Name, Artist - Song

it would think, "perhaps I need to rename all 'Artist Name - Song' to  'Name, Artist - Song'" while I'm at it. The user seems to prefer it that way. It might ask you.
« Last Edit: November 13, 2008, 09:09 PM by tmpusr »

tmpusr

  • Member
  • Joined in 2005
  • **
  • Posts: 154
  • Instantiation stuck in meatspace with no backup
    • View Profile
    • Donate to Member
See also

IDEA: DeEntropizer - File and Folder Auto Organizer by Similarity and Patterns
https://www.donation...96.msg43589#msg43589

Not much enthusiasm... no messy file systems, eh?

We need to start a Manhattan Project for seriously intelligent file management automation... The goal: You drop your files on the desktop or some Inbox folder and they find their way to the right place, exactly where you would have put them, creating subfolders if necessary and renamed like you would have renamed them. A shortcut list of the locations of the new stuff gets updated.
« Last Edit: November 19, 2008, 09:22 PM by tmpusr »

tmpusr

  • Member
  • Joined in 2005
  • **
  • Posts: 154
  • Instantiation stuck in meatspace with no backup
    • View Profile
    • Donate to Member
Locate32's database would be perfectly suited for implementing this sort of an application because it deals with searching or "knowing" the locations, names, and other attributes of files and folders. You would simply create a list of files and folders that have, in their names, groups of words appearing together in any order (even with some random words in between) and then move each found member of such group to a common folder, resulting in the clustering of files and folders with similar names.

Seems doable and rather simple, and UberUseful - I mean, who would do such tedious file and folder grouping ever again if it was handled automagically? You could restrict to one folder or its subfolders or a set of folders anywhere (also network shares) or do it on all folders (not so useful because most of the stuff you want organized is probably already under a few folders, which may be on different drives).

Perhaps some of these features could already be created simply as somewhat complex searches, or if not yet possible, perhaps Locate32's search terms and capabilities could be extended to make such smart searches possible. It would automatically create searches based on groups of words that exist in names, like: if it finds files and folders with names such as "Word1 Word2" and "Word2 Word1" and "random words Word1 random words Word2 random words" it would create a search folder (search preset, for now, until search folders in Explorer are implemented) that searches all those files, naming it "Word1 Word2 cluster" or something. From there you could move everything to a common folder manually, until full automatic functionality is in place.
« Last Edit: November 30, 2008, 03:35 PM by tmpusr »