Messy data != Messy code
summary
Consumes messy third‑party APIs for a major Swiss retailer and runs an API kept up for over five years, upgraded to Symfony 4. Architecture stores raw provider payloads in MySQL, maps into Elasticsearch, and uses queues (RabbitMQ then Symfony messenger) with workers. Introduces a compiler‑pass driven mapper system to resolve dependencies and per‑language mapping. Replaces visitor serializer with generated code and sees a 55% performance gain, cutting responses from ~390ms to ~275ms and halving CPU and I/O.
AI-written summary · about
chapters (8)
- 0:00Intro
- 0:46Michelle Sanver Colour & Code addict
- 11:43Importing Data With import commands
- 20:093rd party data providers
- 28:53Mapper Dependencies With a compiler pass!
- 32:27Ensuring Quality Tests, Logging, Monitoring
- 33:11Acceptance Tests Test for critical data, often
- 34:56Serializing, Versioning & Groups
related talks
34mSymfony and Elasticsearch: debugging and optimizations
34mWhy PHP and Symfony are still in the game
52mRefactoring to Symfony Components
50mSymfony Messenger: the sharpest tool in your PHP toolbox
57mEating spaghetti with Symfony
40mDeep dive into Symfony 4 internals
57mBuilding first-class REST APIs with Symfony
01h 02mRewriting 12-Year-Old Code