Try to support prompt_toolkit >3.0.37 #347
Conversation
FantasqueX
force-pushed
the
update-prompt_toolkit
branch
from
e44f75d
to
c4c95da
Compare
|
@kiancross Hi, could you please take a look at this? I'm not sure whether I have to do more except making all tests pass. |
FantasqueX
force-pushed
the
update-prompt_toolkit
branch
from
c4c95da
to
e3eea63
Compare
FantasqueX
force-pushed
the
update-prompt_toolkit
branch
from
e3eea63
to
ac81bf2
Compare
FantasqueX
force-pushed
the
update-prompt_toolkit
branch
from
ac81bf2
to
670653d
Compare
|
I've fixed code quality problem. Gently ping~ |
|
Any update please? |
|
@FantasqueX Please could you merge latest changes into this branch and then I can run the CI. |
Currently one test will fail with latest prompt_toolkit(3.0.43). test_blank_line_fix will throw RuntimeError: no running event loop. By bisecting commit, the culprit is prompt-toolkit/python-prompt-toolkit@a775996. This commit replaces custom `get_event_loop` with `asyncio.get_event_loop`. The former will creator a new loop if `asyncio.get_running_loop` fails while the latter won't. I mimic the changes in the examples to use `asyncio.run` and the test passes. Fixes: tmbo#344
FantasqueX
force-pushed
the
update-prompt_toolkit
branch
from
670653d
to
78df02a
Compare
@kiancross Done :) |
There was a problem hiding this comment.
The CI is currently failing because there was a breaking change in 3.0.29, which changed the return type of create_pipe_input. Prior to this version, it did not return a context manager.
prompt-toolkit/python-prompt-toolkit@97ac514
Given we support these previous versions of prompt_toolkit (for reasons I am unaware of), I suppose some conditional logic needs to be added here?
Currently one test will fail with latest prompt_toolkit(3.0.43). test_blank_line_fix will throw RuntimeError: no running event loop. By bisecting commit, the culprit is
prompt-toolkit/python-prompt-toolkit@a775996. This commit replaces custom
get_event_loopwithasyncio.get_event_loop. The former will creator a new loop ifasyncio.get_running_loopfails while the latter won't. I mimic the changes in the examples to useasyncio.runand the test passes. I'm not sure whether more changes are needed.What is the problem that this PR addresses?
Closes: #344
How did you solve it?
Use
asyncio.runto make the test pass.Checklist