feat(server): return ListeningServer from Nickel::listen
#334
Conversation
| /// use nickel::Nickel; | ||
| /// | ||
| /// let server = Nickel::new(); | ||
| /// server.listen("127.0.0.1:6767"); | ||
| /// let listening = server.listen("127.0.0.1:6767").expect("Failed to launch server"); | ||
| /// println!("Listening on: {:?}", listening.socket()); |
There was a problem hiding this comment.
Should this print be here?
There was a problem hiding this comment.
It was to show the use of socket() to get the address, but I guess this example will print twice (since the options were not passed to disable printing). So I think it's fine, but I should make the intention more explicit perhaps?
There was a problem hiding this comment.
Okay, I see. As you wish, I think we can just leave it like this.
|
Exciting! This looks nice to me, but I'm not known to do a lot of testing so let's see if we can get some user feedback. |
|
☔ The latest upstream changes (presumably 2d91a0a) made this pull request unmergeable. Please resolve the merge conflicts. |
|
Rebased on master |
|
☔ The latest upstream changes (presumably f7658fa) made this pull request unmergeable. Please resolve the merge conflicts. |
|
Is this still in flight? |
…erver is listening on BREAKING CHANGE: This changes the return type of Nickel::listen so it may break some code. To fix broken code while maintaining the previous semantics, you can add `unwrap` to the `listen` call.
|
@euclio just rebased. Do you have any suggested edits / concerns? If not I will try to merge this tomorrow. |
|
Nope, looks good to me. Thanks for your work! On Wed, Jun 22, 2016 at 10:25 PM Ryman [email protected] wrote:
|
…ion tests When this test is included in the integration suite, the import is no longer in the root of the crate so the macro_use attribute is disallowed.
|
@Ryman friendly ping :) |
|
@homu r+ |
|
📌 Commit c9abad7 has been approved by |
…erver is listening on BREAKING CHANGE: This changes the return type of Nickel::listen so it may break some code. To fix broken code while maintaining the previous semantics, you can add `unwrap` to the `listen` call. Pull request: #334 Approved by: Ryman
Pull request: #334 Approved by: Ryman
…tabase Pull request: #334 Approved by: Ryman
…ion tests When this test is included in the integration suite, the import is no longer in the root of the crate so the macro_use attribute is disallowed. Pull request: #334 Approved by: Ryman
|
☀️ Test successful - status |
…erver is listening on BREAKING CHANGE: This changes the return type of Nickel::listen so it may break some code. To fix broken code while maintaining the previous semantics, you can add `unwrap` to the `listen` call. Pull request: nickel-org#334 Approved by: Ryman
Pull request: nickel-org#334 Approved by: Ryman
…tabase Pull request: nickel-org#334 Approved by: Ryman
…ion tests When this test is included in the integration suite, the import is no longer in the root of the crate so the macro_use attribute is disallowed. Pull request: nickel-org#334 Approved by: Ryman
Related to #328.
The tests within this repo are a bit of a hack (require shelling out to subprocesses involving cargo), with this change we can instead make it easier to integration test a server (or multiple instances of one) all within the same process which should be easier and faster.
The code within the integration test example would get tested using a normal
cargo test, but since it's in our example files I had to use a workaround to get it compiled directly into our integration test runner.BREAKING CHANGE: This changes the return type of
Nickel::listen, you can add anunwrap()to the call to maintain previous semantics.