To start the server:

Make sure you have Python3, pip, and virtualenv
source venv/bin/activate to start the virtual environment
pip install -r requirements.txt and finally, python app.py (Server should be running on localhost:5000)

Note: When using curl, make sure URLs always end in backslash

To Create a Part object:

(Must specify correct API route, header, and at least a name in data dictionary. ID will be automatically specified.) curl -X POST http://localhost:5000/create/ --header "Content-Type: application/json" -d '{"name": "phone", "color":"silver", "model": "iPhone"}'

To Update a Part Object:

curl -X PUT http://localhost:5000/update/<int:part_id>/ --header "Content-Type: application/json" -d '{"color": "orange"}'

To Delete a Part Object:

WARNING: ANYONE CAN DELETE

May want to install API key for permissions

curl -X DELETE http://localhost:5000/delete/<int:ID>/

Listing Parts:

To list all parts in BoM:

curl -X GET http://localhost:5000/bom/

To list all assemblies:

curl -X GET http://localhost:5000/assemblies/

To list all top level assemblies (assemblies that are not children of another assembly):

curl -X GET http://localhost:5000/toplevel/

To list all subassemblies (assemblies that are children of another assembly):

curl -X GET http://localhost:5000/subassems/

To list all component parts (parts that are not subassemblies, but are included in a parent assembly):

curl -X GET http://localhost:5000/components/

To list all orphan parts (parts with neither parents nor children):

curl -X GET http://localhost:5000/orphans/

To list all the first-level children of a specific assembly:

curl -X GET http://localhost:5000/assembly/children/<int: ID>

To list all parts in a specific assembly:

curl -X GET http://localhost:5000/parts/<int: ID>/

To list all assemblies that contain a specific child part, either directly or indirectly (via a subassembly):

curl -X GET http://localhost:5000/dependencies/<int: ID>/

Future Improvements:

We should store static values until we add/update/delete values, at which point only then would we need to make updates to our relational "database".

Ultimately though, if we really wanted to be efficient, using a database management system would likely abstract-away solutions to our inefficient accessors and checking methods.

If we cannot abstract-away accessors to be more efficient, then at least we can take inspiration from common database management algorithms.