Add Rails Runner client #250
Conversation
|
|
||
| sig { params(name: String).returns(T.nilable(T::Hash[Symbol, T.untyped])) } | ||
| def model(name) | ||
| make_request("model", name: name) |
There was a problem hiding this comment.
The param was called models in the Rack app, which matched up with the RESTful route name, but here I think model is more appropriate.
andyw8
force-pushed
the
andyw8/add-runner-client
branch
2 times, most recently
from
54cca6f to
72f4c60
Compare
|
This still needs a little clean-up but feedback is welcome. |
andyw8
force-pushed
the
andyw8/add-runner-client
branch
2 times, most recently
from
56f75b8 to
afc4f28
Compare
andyw8
force-pushed
the
andyw8/add-runner-client
branch
from
afc4f28 to
a306d81
Compare
| end | ||
|
|
||
| sig { returns(T::Boolean) } | ||
| def stopped? |
There was a problem hiding this comment.
To avoid the test having to poke inside with instance_var_get.
| end | ||
| end | ||
|
|
||
| RubyLsp::Rails::Server.new.start if ARGV.first == "start" |
There was a problem hiding this comment.
Do we need this argument? This script is always invoked from the runner, so would there be a case where we wouldn't want to start it?
There was a problem hiding this comment.
We can revisit this, but currently without it the check_docs task will hang because it requires each file.
There was a problem hiding this comment.
I see. We can probably ignore this file in check_docs.
| } | ||
| end | ||
|
|
||
| schema_file = ActiveRecord::Tasks::DatabaseTasks.schema_dump_path(const.connection.pool.db_config) |
There was a problem hiding this comment.
This line break the compatibility with ActiveRecord 6
schema_dump_path is not present in ActiveRecord 6.1: https://github.com/rails/rails/blob/v6.1.7.6/activerecord/lib/active_record/tasks/database_tasks.rb
schema_dump_path is available from 7.0: https://github.com/rails/rails/blob/v7.0.0/activerecord/lib/active_record/tasks/database_tasks.rb#L448
[EDIT: open an issue #260]
This PR is part of a migration towards communicating with the Ruby LSP server over a pipe, rather than HTTP. It was copied out of the branch
andyw8/implement-lsp-runnerfor easier review.I have tophatted the above branch against Code DB and Core. On Core, it correctly jumps the correct one of the two SQL structure files.
Previous discussion on prototype.