refactor: id prefixes #1546
refactor: id prefixes #1546
Conversation
This change makes sure that going forward all IDs generated will have ID prefixes, causing easier identification of the entity in question. The old data is not modified.
turip
added
release-note/breaking-change
| @@ -169,3 +196,11 @@ func (c CustomerAddressMixin) Fields() []ent.Field { | |||
| field.String(fmt.Sprintf("%s_address_phone_number", c.FieldPrefix)).Optional().Nillable(), | |||
| } | |||
| } | |||
|
|
|||
| func Must(m ent.Mixin, err error) ent.Mixin { | |||
There was a problem hiding this comment.
if we're already using it then i suggest maybe lo.Must
| // Table of ID prefixes for each entity type. | ||
| // Please make sure you are picking a unique prefix for each entity type. | ||
| // Recommendation: | ||
| // - at least 3 characters of the entity type name. | ||
| // - max length is 8 characters (field size limit) |
There was a problem hiding this comment.
I suppose this is the hard part to get right. I'm not 100% scoping is the way to go, I'd rather define what to do in case of name collisions and use as short names as possible... wdyt?
There was a problem hiding this comment.
Yes, scoping is 100% TBD. Initially I was doing stuff like
- BillingInvoiveItem => bii_
- Customer => c_
But then I didn't like the one-character names, so that's why I ended up with this scoping thing. I am not against dropping that part.
I would just mandate that there are no collisions, as seven characters should be enough.
| } | ||
|
|
||
| func (i idMixin) ULIDWithPrefix() string { | ||
| return fmt.Sprintf("%s_%s", i.IDPrefix, ulid.Make().String()) |
There was a problem hiding this comment.
TODO: With the current IDs the _ is a duplicate.
| -- modify "billing_invoice_items" table | ||
| ALTER TABLE "billing_invoice_items" ALTER COLUMN "id" TYPE character varying(34), ALTER COLUMN "invoice_id" TYPE character varying(34); | ||
| -- modify "billing_invoices" table | ||
| ALTER TABLE "billing_invoices" ALTER COLUMN "id" TYPE character varying(34), ALTER COLUMN "billing_profile_id" TYPE character varying(34), ALTER COLUMN "workflow_config_id" TYPE character varying(34); |
There was a problem hiding this comment.
TODO: we need to change the schema override too (or remove in case of edge fields as it seems like ent autogenerates the types)
Overview
This change makes sure that going forward all IDs generated will have ID prefixes, causing easier identification of the entity in question.
The old data is not modified.
Notes for reviewer