URNs produce the wrong result for .uri() method

[tomschr]

trafficstars

Thanks for this great project! :heart:

• Faker version: 28.4.1
• OS: openSUSE 15.5

If you run the .uri() method and pass the URN (Uniform Resource Name) scheme, it returns the wrong result.

Steps to reproduce

Consider the following snippet:

>>> from faker import Faker
>>> fake = Faker()
>>> fake.uri(schemes=["urn"])
'urn://thompson.org/tagsauthor.htm'

Expected behavior

URNs are described in RFC 1737 and later in RFC 2141, I would have expected to have something like this:

urn:isbn:0451450523
urn:mpeg:mpeg7:schema:2001
urn:mrn:iala:pub:g1143

Find more examples in the Wikipedia article about Uniform Resource Name.

Actual behavior

At the moment, the URN scheme uses slashes (/) which is wrong according to the mentioned specs. As the method name .uri() implies, it's the generic term for URNs and URLs.

Possible solution

I can think about different solutions to solve this issue:

1. Fix the .uri() method and return the correct result when a URN scheme is requested.
2. Introduce a new .urn() method.

Idea 1 would probably the more generic solution as it works for URLs and URNs. On the other hand, idea 2 would make URNs stick out.

[Abdujabbar]

@tomschr I did a little research and found this list of URI schemas: https://en.wikipedia.org/wiki/List_of_URI_schemes

Based on this list, schemas=['urn'] is not the right way because it's not in the list of URI schemas.

I think it would be better to have a new method .urn().

It's just opinion.

[tomschr]

@tomschr I did a little research and found this list of URI schemas: https://en.wikipedia.org/wiki/List_of_URI_schemes

Based on this list, schemas=['urn'] is not the right way because it's not in the list of URI schemas.

Ahh, that's interesting. I didn't know, thanks!

I think it would be better to have a new method .urn().

It's just opinion.

I'm fine with using .urn() as the way to go. 👍

My only concern is that users will find it. So one way to ensure that could be:

• If the users use .uri() for getting URNs, raise an exception and direct users to use .urn()
• Document it in the docstring and the official documentation. (I guess that's the normal approach anyway, right?)

[Abdujabbar]

@tomschr

• I think the normal it should be added some white list for schemas list, and if passed schemas will not match to whitelist schemas then raise an exception for .uri() should be good practice.
• Yes I agree this should be documented of course.

In this case, it would be great to hear @fcurella opinion also.

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.

[tomschr]

@Abdujabbar / @fcurella do we want to work on that?

[fcurella]

@Abdujabbar I think your should is the best option. If any of you want to submit a Pull Request, I can re-open this ticket :)

[jasur-py]

I am taking up this if no one is working ;)

[tomschr]

Thanks @jasur-py, please do! I've tried it, but as I'm not familiar with the code, I didn't come far.

[jasur-py]

I would love to complete this fix but my previous PR is still hanging without any review :( i am afraid it will be another stale one...

[github-actions[bot]]

This issue is stale because it has been open for 30 days with no activity.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.