Could have a binary setting, Using the following example data:

  • Tree ( code, sauce)
  • Nice ( project, sauce) With the following tree:

--code |--sauce

--sauce

with one binary option ∝ we would get:

--code |--sauce Tree --sauce Nice

and in the other ⊚ we would get:

--code |--sauce Tree --sauce Nice Tree

i a m confusinon Why So theres more ways we can do this

So for a directory, present children under it that match the list of filters all down to that point For a directory, present children that match the list of filters down to that point and dont match any filters beyond it. Maybe implement this and see if it can match the current code dir

Currently all data that includes all of the path parts is displayed Maybe data is displayed when all its tags are completely met So have an all components filter Nope I think those methods are both functionally the same, I want it to display data that has the tags of the parts and nothing else

SO ONLYHASTAGS IS WORKING NICE, THERES STILL MORE TO DO ON THAT FRONT, DONT THINK ITS EXCLUSIVE. iN ADDITION TO THAT, IMPLEMENTING THAT EXCLUSIVITY DOESNT SPELL WELL FOR THIS BEING A SOLUTION THAT CAN APPLY TO ARBITRARY DIRECTORYS HAVING CUSTOM FILTERS. THE SOLUTION SHOULD BE APPLICABLE TO ANY FILTERS. CANT REALLY THINK TOO WELL RN BUT YHHHH THIS SHIT, KEEP IT UP X ^2fd21f

We can place a member variable on the directories of a filter list during setup of the mount and then connect the signals so that the invalidation structure automatically updates the children as to changes, then when reading the directories, we can just read from that filter list

So do I treat all directories as filters?

∝ could produce different results for the location of data which could be in multiple locations, there are different ways to do this so the method chosen should be consistent. If not then one cannot rely on this hierarchy for consistent location of the data but one could rely on the same tags being present on the data and always could rely on its unique id. Systems such as python may be built around a reliable hierarchy structure Also systems such as python can have autogenerated hierarchy like in gi.repository I dont know if this offers any new functionality over our hierarchical organisation

Not reliable

Always reliable

This technique is as reliable as filesystem methods
Changing the uniquely specified directory will still break the reference as always
  • First come first serve # Not reliable If the algorithm was based on some alphabetical sorting then this could change the location of data.
  • Time based # Always reliable The edit time of each hierarchical edit could be tracked and this can then be used to either state that the data will always be in the first match in time
  • Order based # Always reliable The order stated in the hierarchy definition can be used to always pick the first match, I think this is as reliable as a filesystem

How do i implement an order based exlusivity

Mounting as a directory from arbitrary hierarchy

ldd $( which mount) lists libraries used to mount devices on the filesystem, these libraries might be a good investigative starting point