FastStream

Супер. Тогда просто жду

Эт что-то новое

Кстати, а так должно быть, что разные прогоны тестов по разному проходят? Позавчера 2 теста падали, сегодня всё чисто

🎉 New Pull Request to ag2ai/faststream by @StepanKatin ✨ test: add syrupy JSON snapshot across NATS (#3249) 📊 +1323/-481 🌿 StepanKatin:2978-snapshot-nats → main DESCRIPTION Expand syrupy JSON snapshot testing across tests/asyncapi/nats/ for both AsyncAPI 2.6.0 and 3.0.0 . NATS schema overrides now assert the full generated document against JSONSnapshotExtension snapshots under __snapshots__/. Each snapshot is a separate .json file (one full AsyncAPI document per test) with predictable paths. Address-style tests without full-schema dict literals are left unchanged. Relates to #2978 Type of change Please delete options that are not relevant. ☑️ New feature (a non-breaking change that adds functionality) Checklist ☑️ My code adheres to the style guidelines of this project (just lint shows no errors) ☑️ I have conducted a self-review of my own code ◻️ I have made the necessary changes to the documentation ☑️ My changes do not generate any new warnings ☑️ I have added tests to validate the effectiveness of my fix or the functionality of my new feature ◻️ Both new and existing unit tests pass successfully on my local environment by running just test-coverage ◻️ I have ensured that static analysis tests are passing by running just static-analysis ◻️ I have included code examples to illustrate the modifications sent via relator

https://github.com/open-telemetry/opentelemetry.io/pull/11902 чтож, ждем мержа в OTeL Оказывается, у них ест официальный registry инструментаторов

Нихуя я репы накрутил

🎉 New Pull Request to ag2ai/faststream by @Vitaly312 ✨ Fix: exit already entered middlewares if one fails to enter (#3252) 📊 +42/-11 🌿 Vitaly312:fix/middleware-enter-leak → main DESCRIPTION Fixes #3251 Middlewares were entered by hand, and their __aexit__ was pushed to the exit stack only after the message scope had been entered — so a failing on_receive left every already entered middleware without after_processed. They cannot just be entered with stack.enter_async_context, because then they would be exited after the message/log_context scope is released. Instead, the scope now gets its own stack, reserved before the middlewares are entered, which keeps the exit order and removes both push_async_exit loops. Side effect: an error raised in on_receive now reaches CriticalLogMiddleware and is logged. Previously it disappeared entirely. Type of change ☑️ Bug fix (a non-breaking change that resolves an issue) Checklist ☑️ My code adheres to the style guidelines of this project (just lint shows no errors) ☑️ I have conducted a self-review of my own code ☑️ My changes do not generate any new warnings ☑️ I have added tests to validate the effectiveness of my fix or the functionality of my new feature ☑️ Both new and existing unit tests pass successfully on my local environment by running just test-coverage ☑️ I have ensured that static analysis tests are passing by running just static-analysis sent via relator

@diementros @ivan_kirpichnikov я там долгий тест убрал под mark slow, проверьте плз кто нибудь и слейте

Ух-ты, а вы знали о существовании такого проекта? https://faststream-outbox.modern-python.org/introduction/how-it-works/

Ага

У нас в доке есть ссылка даже

Я такого паттерна не знаю

Ты возможно путаешь распределение обязанностей 🤔

И у него 2 части: 1 пишет, другая читает

Ага, понял. Спасибо.

только немного расширил идею