The purge

You try to download all the photos your sister sends you but your phone produces an error message that there isn't enough space available. It seems all your constant programming and tinkering has left your phone filled with a lot of unnecessary files. For instance you see a lot of folders with names like "TEMPORARY" or "DELETE LATER".

You decide to take a few minutes to clean up the file system on your phone. Obviously you need to be careful not to delete any files your phone actually needs to operate.

Any file or folder containing the word delete or temporary are safe to delete, but just how much disk space will it free up? It's one thing to total all the files containing delete or temporary, but if a folder name contains it, then you will also be deleting all the files contained within that folder (regardless of their names), plus any sub folders within the offending folder! You could find yourself going into multiple layers of folders to total all the file sizes!

Example

Folder 0 is always the starting point of your search, and everything else is contained within it. On a Mac this might be folder /, or on Windows this could be your C:/.

Let's look at a small example...

Folder: 0
 - taskmgr.exe 5065932
 - Customer_Feedback_Compilation_2024.xlsx 2646384
 - VLC-3.0.16-win64.exe 3971817
 - ProductLaunch2024.png 3712336
 - temporary_573 5048816
 - delete_708 2054307
 - temporary_023 [FOLDER 1]
Folder: 1
 - Conference_Break_Music.mp3 5179931
 - SlackSetup-x64-4.3.2.exe 2384929
 - Strategic_Plans [FOLDER 2]
 - Client_Dinner_Mar_2024.jpeg 5364778
 - FileZilla_3.55.1_win64-setup.exe 4623628
 - Charity_Event_011524.HEIC 2134414
 - Office_Christmas_Party_2023.jpeg 687062
Folder: 2
 - Product_Video_Soundtrack.aiff 813896
 - Operations_Manuals [FOLDER 3]
 - Signed_NDA_JohnDoe_021523.pdf 3257437
 - delete_930 9940460
 - Client_Acquisition_Strategies [FOLDER 4]
 - temporary_493 1332303
 - Marketing_Brochure_Image1.PSD 5913782
Folder: 3
 - ProductLaunch2024.png 4396529
 - Motivational_Morning_Playlist.m3u 5619626
 - Network_Configuration_Settings.txt 1068226
 - CRM_Database_Export_021724.csv 5812973
 - Competitor_Analysis_0224.pdf 1088620
 - Employee_Training_Videos_Link.txt 267104
 - delete_530 7150742
Folder: 4
 - temporary_751 1051994
 - delete_208 6042521
 - Logo_Rebrand_Options.svg 3438585
 - Node-v14.17.3-x64.msi 2056068
 - Expense_Report_Jan_2024.pdf 5775782
 - user32.dll 2371618
 - delete_027 9003131

Folder 0 contains 1 file for deletion, delete_708 of size 2054307. Folder 0 also contains a folder called temporary_023. Due to the name of the folder, that means everything within it will also get deleted. To total the disk space consumed by the files in temporary_023, you need to go through the contents of that folder which we are told is [FOLDER 1].

Jump to folder 1, and total all the files within it, plus also [FOLDER 2] which is inside it.

Folder 2 contains folder 3 and folder 4, so the file sizes in those also need totalling.

Converting the folder numbers to their folder names, the above structure could be visualised with a tree of the following:

 /
 +-- temporary_023
     +-- Strategic_Plans
         +-- Operations_Manuals
         +-- Client_Acquisition_Strategies

Once you have traversed the entire folder structure in this example, the total bytes that would be freed up are 103879262, which would be the answer.

Be aware that as you are totalling files to be deleted, take care not to double-count some files. For instance the file delete_027 inside folder 4 is already getting deleted due to the folder it resides in, so the file size doesn't need counting separately.

Your task

Traverse through the files and folders in your input data, totalling the sizes of all files that should be deleted because either the filename, or one of the containing folders of a file, contains the phrase delete or temporary. The total number of bytes to be deleted is your answer value.

Hint

This problem may look like it is an ideal candidate for recursion, but you will likely hit your maximum recursion depth limits. There are other ways of solving this, in O(n) time, that do not require recursion at all. A dictionary/hashmap may prove useful.