PSR-7 / PSR-15 support

[spotman]

Hi there!

I hope you're doing well! What about implementing support for PSR-7 HTTP messages and PSR-15 middleware? I have some ideas about refactoring the core to support this standards right from the box, in the same manner like PSR-16 cache that we implemented last year. This will allow to use page-cache with modern applications and frameworks, including async processing in ReactPHP. The PHP is moving formward and we need to do so :)

Have a nice day!

[spotman]

@mmamedov ping

[mmamedov]

Hi Denis, That’s a great idea, we could definetely try that. We have another issue at the moment, for some reason on a very few occasions I was getting a blank page. I’d like to focus on that first. It’d be great if you also had time to go over the code to stabilize cache serving. Thanks!

…

On Tue 9. Jan 2018 at 6:56 AM, Denis Terekhov ***@***.***> wrote: Hi there! I hope you're doing well! What about implementing support for PSR-7 HTTP messages and PSR-15 middleware? I have some ideas about refactoring the core to support this standards right from the box, in the same manner like PSR-16 cache that we implemented last year. This will allow to use page-cache with modern applications and frameworks, including async processing in ReactPHP. The PHP is moving formward and we need to do so :) Have a nice day! — You are receiving this because you are subscribed to this thread. Reply to this email directly, view it on GitHub <#9>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADnp5YN87aInkCN9xLiTv8QtuJbKwYPLks5tIv-EgaJpZM4RXWVQ> .

[spotman]

@mmamedov OK, I have some spare time now and will try to debug this. Unfortunately, I had not this issue in my env so your PageCache initialization code is highly appreciated. I may emulate high load on this code and to check if everything is going well. Maybe we can create unit test for this case?

[mmamedov]

Yeah, makes sense. We could try to cover that logic with more tests. I used to run brianium parallel tests, but I think we disabled them because of Travis issue.

…

On Thu 25. Jan 2018 at 3:16 AM, Denis Terekhov ***@***.***> wrote: @mmamedov <https://github.com/mmamedov> OK, I have some spare time now and will try to debug this. Unfortunately, I had not this issue in my env so your PageCache initialization code is highly appreciated. I may emulate high load on this code and to check if everything is going well. Maybe we can create unit test for this case? — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub <#9 (comment)>, or mute the thread <https://github.com/notifications/unsubscribe-auth/ADnp5c8Iu9QBPbKsOerlza8fFo9Pkmipks5tN-P2gaJpZM4RXWVQ> .

[spotman]

We may write one test script based on symfony/process and run multiple instances of the PageCache with the same configuration. This will emulate concurrency and/or cache stampede. I think that capturing PSR response would be much cleaner than raw echo and header so we may test your test case and PSR messages implementation in one test suite. Also we may run these tests on a multiple set of configuration options so more use cases will be covered.