Thanks for making this! Insta is great.

Suppose I am testing the old and new versions of an API - I may want two flavors of every snapshot. Naively I'd write:

assert snapshot("v2.json") == (await client.get_new_api()).json()
assert snapshot("v1.json") == (await client.get_old_api()).json()

That would work fine as long as I called this only once. But what happens when my test wants to call each API multiple times? If I run with --insta update, I end up with the last response saved. If I run without --insta update, the test obviously fails.

There are two problems here:

• there's no way to ask for a snapshot to have a name suffix (and the "name" argument is overloaded to mean both "type" and "exact name").
• the framework does not detect the same snapshot being overwritten multiple times in the same update run.

I'd be happy to send a pull request! Let me know how you feel about this.

Insta has a neat feature called redactions which let you omit certain values from the snapshot (e.g. a timestamp field which will be different on every run).

Is it possible to do something similar with pytest-insta?

Hi @vberlier

First off, what a wonderful library! Thanks for the time and effort you've put into this, it really scratches an itch I've experienced when it comes to testing in the Python world.

I wanted to discuss if there is interest in extending the functionality (which I'd be more than happy to contribute the work towards) to allow for setting "partial" expectations in a snapshot.

Here's a use case to hopefully illustrate that sort of functionality.

I often find myself writing tests like this....

def test_posts_detail(api_client):
    post = PostFactory(title="Amazing deal!", body="Don't miss out", published_at=date(2021, 1, 1))

    response = api_client.get(f"/posts/{post.id}/")

    assert response.json() == {
        "id": str(post.id), 
        "title": "Amazing deal!",
        "body": "Don't miss out", 
        "published_at": "2021-1-1"
    }

It's super useful to be able this against a snapshot. Especially as project go on, and API change over time, the review process really help keep things manageable.
However, there is one snag, most of all the fields are "frozen" but in this case the id is dynamic. I can't control what it will be (i.e depends on how many tests, what operation happened before, etc).

I would love the ability to provide the snapshot with a few selected expectations, i.e something like this....

def test_posts_detail(api_client, snapshot):
    post = PostFactory(title="Amazing deal!", body="Don't miss out", published_at=date(2021, 1, 1))

    response = api_client.get(f"/posts/{post.id}/")

    assert snapshot("json", {"id": str(post.id)}) == response.json()

In this case, the id will always be updated and set at runtime in the test, but the rest of are snapshotted.

This looks like a really nice project! I was looking for a snapshot testing library to automate tests that compare some JSON responses. One feature I would really like is to store the snapshots in code. For example, something like this:

def test_foo():
    expected_value = snapshot(
        {...}   # some large dict
    )
    actual_value = foo()
    assert expected_value == actual_value

The reason for this is to keep the expected/snapshotted value close to where it is being used, which makes it a lot easier to inspect the expected value, e.g. in code review.

Although I've never used it, it seems that insta (Rust version) supports this, albeit only for strings. I think native Python objects would be ideal, but storing inline as JSON might be easier.

Steps to reproduce

• Create a snapshot like snapshot('audit.txt')
• Run test
• Rename snapshot like like snapshot('audit.json')
• Run test
• Now run test with --insta review

Expected behavior

Things work just fie.
Review works as expected or alternatively I am instructed to do some kind of cleanup.

Actual behavior

As audit.txt is still existing in the cache folder the last pytest call fails with an error like this:

FileNotFoundError: [Errno 2] No such file or directory: 'flaskapp/snapshots/views__cm_views__audit.txt'

Running --insta clear fixes the issue.

For new users, this behavior can be quite unexpected. Especially as one might not even remember anymore that a snapshot with the unexpected name once existed.

I'm having trouble getting the snapshot tests to pass in my CI as the snapshot location is different in CI than locally. For example the github runner fails with:

assert <not found:'ast_parsing__vyper_cfgir_home_runner_work_slither_slither_tests_e2e_vyper_parsing_test_data_precedence_vy_foo__0.txt'> 

Locally, it is named ast_parsing__vyper_cfgir_builtins_test_builtins__0.txt and the test passes

Firstly, thank you for creating this package. It seems really well made and I have been liking it so far! ❤️

Overview

When I first create a snapshot test with snapshot() and run pytest --insta review, I see this in the output:

========================================================================================== SNAPSHOTS ===========================================================================================
CREATE tests/snapshots/<snapshot name>.json

But I don't get to accept/reject this first snapshot. I then have to manually cat the file and review it.

Why?

IIRC Armin's insta (the rust one), asks you to review the first snapshot too. It would be nice if pytest-insta did the same too.