Heirtags are a better way to tag, classify and categorize entities in your system. Contemporary tagging makes use of a many-to-many relationship to allow developers to express multiple attributes on an entity. However these tags are often simple strings that can quicky become unmanageable (if you've ever stared at a tag cloud to try generate insight you'll know what I mean).
Heirtags make use of tags that are heirarchical in nature. Each tag looks something like:
/root/child group/something/else
Kind of like a file system path would look like. From this we can explore the tags far easier,
as they have built-in heirachy. It's kind of like tagging your entities with file paths, and you can
do anything that you can with a file path. Such as show me all the items under /root/child_group/*
Take a library of books as an example. Previously you would have worked to guess all the attributes up-front and add them accordingly. With heirtags you could have a list of books (we'll ID them with integers, 1, 2, 3 etc) and then tags to describe them. For example in our library we could have:
/core/authors/Fitzgerald
/core/authors/Wilson
/core/authors/Poe
/core/authors/Churchill
/core/year/2000
/core/year/2001
/core/year/2002
/core/type/Fiction/genre/Mystery
/core/type/Fiction/genre/Sci-fi
/core/type/Non-Fiction/genre/Biography
/core/type/Non-Fiction/genre/Historical
To find all the fiction books you can query
/core/type/Fiction/
while to get more specific and find all books by Edgar Allen Poe you could do
/core/authors/Poe
Every book (entity) could have multiple tags accordingly, and it's easy to add new tags!
This is a python implementation of heirtags. It's an in-memory data structure that supports adding entities referenced by ID's to the heirachy tree, finding entities via tags and adding new tags to entities.
See main.py for an example.
- Support deleting entities/tags
- Testing absolute/relative look ups
- Printing none at end of dump?