sqlalchemy update a specific specific column element where value is?

Question

I have a sqlite db, created as below:

class VPN_routers(Base):
    __tablename__ = 'vpnrouters'
    __table_args__ = TABLE_ARGS
    sp_id = Column(String(), primary_key=True)
    router_id = Column(String())
    external_gw = Column(String())
    subnet_attached = Column(String())
    vpn_connectivity = Column(String()

and the contents of the table are as below:

sp-10.1.10.148|66c49f9b-11c4-4|172.24.4.6|cb3a722e-2ec4-4caf-a34b-93004d885f28|NO
sp-10.1.10.147|282608ba-8870-l|172.24.4.7|Not Attached|NO
sp-10.1.10.149|44cc77fe-3b2e-4|172.24.4.7|50853e0b-918f-4c5f-8188-3acbb23a829a|NO

I wanted to update the row where sp_id="sp-10.1.10.147"(identifying using the sp_id) and update the subnet_id and vpn_connectivity to new values. after updating the table could be like this.

sp-10.1.10.148|66c49f9b-11c4-4|172.24.4.6|cb3a722e-2ec4-4caf-a34b-93004d885f28|NO
sp-10.1.10.147|282608ba-8870-l|172.24.4.7|cbcbeiowj-d34f5tgygt-6g7ggrydwqrer3d|YES
sp-10.1.10.149|44cc77fe-3b2e-4|172.24.4.7|50853e0b-918f-4c5f-8188-3acbb23a829a|NO

I am trying to write this update as a function where I can have the feasibility to give anyone value as an identifier(not necessarily sp_id always) of the row, and its corresponding columns new elements as inputs to the function.

ex:

update_DB(VPN_routers, identifier={'sp_id':"sp-10.1.10.147"}, changes={'subnet_attached': "cbcbeiowj-d34f5tgygt-6g7ggrydwqrer3d", 'vpn_connectivity':"YES"})

*Note: identifier and changes are not same always, the identifier will be of fixed length 1 but changes could be of variable length.

I have tried many methods, but I could not make it, completely relying on this post for an answer.

Answer 1

The "identifier" is easy to handle: just use Query.filter_by() and unpack the given dictionary as keyword arguments .

Since Query.update() accepts a dictionary with attribute names as keys, handling the changes is as easy as passing the dictionary forward:

def update_DB(cls, identifier, changes):
    session.query(cls).\
        filter_by(**identifier).\
        update(changes, synchronize_session=False)

As long as both dictionaries use attribute names as keys, it'll work, but your example includes "subnet_id" as key, which is not an attribute of the mapped class. Perhaps it should've been "subnet_attached".