feat: pass in index to FactoryValueGenerator

[simplenotezy]

Description

This PR introduces a new feature in the FactoryValueGenerator class to include the current index in its operations. This enhancement allows for greater flexibility, e.g. the creation of consistent UUIDs across different seeds, which is particularly useful during development.

Key Benefits

• Consistent UUID Generation: By incorporating the current index, the generated UUIDs can remain consistent across reseeds.
• Improved Developer Productivity: Developers can then refire the same API requests without needing to reconfigure IDs, streamlining the development process.

Implementation Details

• The FactoryValueGenerator has been updated to include the current index in its UUID generation logic.
• No breaking changes are introduced; existing functionalities remain intact. I had to add ts-ignore though, which I am not particularly proud of, but seems to be the most sensible solution to maintain backwards compatability

Example

@Factory((_, __, index) =>
  uuidv5(index.toString(), '6ba7b810-9dad-11d1-80b4-00c04fd430c8'),
)
id: string;

Checklist

• Code has been tested locally
• Documentation has been updated
• All tests pass N/A

Please review the changes and provide feedback. Thanks!

[edwardanthony]

Hey Mattias,

Thank you for submitting the PR. I'm considering whether injecting the index would be beneficial.

When creating IDs, it's generally best for the database or the driver to determine the sequence. This approach minimizes logic errors since the database acts as the single source of truth for ID resolution.

For example, in MongoDB, ObjectId handles this automatically. Similarly, SQL databases use features like autoincrement or bigserial to generate sequential IDs. Even in cases where IDs are created manually, such as Twitter's Snowflake ID, an index isn't necessary.

However, if you still need to access the index, you can easily achieve this with the following code:

const users = DataFactory.createForClass(User)
  .generate(10)
  .map((user, index) => ({ ...user, id: index }));

Feel free to let me know what you think.

[simplenotezy]

Hey @edwardanthony - that's defiantly also an approach that could work, didn't think of that. This naturally wouldn't be an issue if you use auto incrementing IDs, but I have stopped using that years ago in favour of UUIDs which are a bit more discrete, espeically when it comes to revealing business metrics to the outside world.

While I have you; how do you go about seeding relationships? E.g. when using TypeORM.

I find my self having to first create the entities, and then map over each entity and assign relationships. Having the IDs prepared in advance, prior to saving them in the DB, also greatly helps this process.