readline: keypress event trigger for escape character #7382
Conversation
nodejs-github-bot
added
the
readline
| @@ -149,3 +154,8 @@ addTest('\x1b[31ma\x1b[39ma', [ | |||
| { name: 'undefined', sequence: '\x1b[39m', code: '[39m' }, | |||
| { name: 'a', sequence: 'a' }, | |||
| ]); | |||
|
|
|||
| // escape character | |||
| addTest('\x1b', [ | |||
There was a problem hiding this comment.
can you add a test for multiple escape characters one after another?
princejwesley
force-pushed
the
repl-keypress-escape
branch
2 times, most recently
from
ad7bf8f to
9509910
Compare
| function onData(b) { | ||
| if (stream.listenerCount('keypress') > 0) { | ||
| var r = stream[KEYPRESS_DECODER].write(b); | ||
| if (r) { | ||
| if (timeoutRef) { | ||
| timeoutRef = clearTimeout(timeoutRef); |
There was a problem hiding this comment.
This does not return anything. See https://github.com/nodejs/node/blob/ede98a77677afcc5e75a126043a11a720f7ae28b/lib/timers.js -- you can use just use clearTimeout(timeoutRef) without the if or re-assigning.
There was a problem hiding this comment.
@Fishrock123 Yes, I just assigned the undefined return value to timeout object. I'll remove that
princejwesley
force-pushed
the
repl-keypress-escape
branch
4 times, most recently
from
435a870 to
bcb95b7
Compare
|
Looking good on the timeout value part from my side. By the way: changes over time are much easier to review if you don't force push after every change. |
| // that can be called to run tests in sequence | ||
| const addKeyIntervalTest = (sequences, expectedKeys, interval = 550, | ||
| assertDelay = 550) => { | ||
| return (next) => () => { |
There was a problem hiding this comment.
Maybe just add a comment stating which functions these are mocking?
|
Local testing shows this doing ok. LGTM with green CI but would like others to review also. |
|
Gave it a try. While it seems to work on single escape presses, 2 quick presses seem to get swallowed, 3 quick presses gives me: |
|
@silverwind Do we need to reduce timeout value below 500ms? |
|
I don't think the timeout value is the issue, it's what happens when another escape arrives during the timeout. |
|
@nodejs/collaborators Is it good to merge? |
|
I'm fine with this as is but would like @silverwind to be comfortable with it also. |
|
Not really happy that it prints that garbage |
|
@silverwind I'll fix it |
princejwesley
force-pushed
the
repl-keypress-escape
branch
from
e2476c7 to
9327a39
Compare
|
@silverwind Please try now |
|
Looking good. Only nit I have a are ESC+ESC and ESC+ESC+ESC entered during the timeout which are given as these keypress events: Is ESC+ESC a valid escape sequence? If not, maybe we should emit individual events per keypress. Above issue isn't really something that should prevent this from landing. LGTM. |
|
@silverwind As per the existing implementation, CI: https://ci.nodejs.org/job/node-test-pull-request/3659/ For the invalid escape sequence, we may need a new PR to emit individual events per keypress as its a breaking change. |
|
CI is green. |
Fixes: #7379 PR-URL: #7382 Reviewed-By: jasnell - James M Snell <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
|
Landed in |
I'd say leave it as is. I can't find any reference that ESC+ESC is not a escape sequence of its own. |
Fixes: #7379 PR-URL: #7382 Reviewed-By: jasnell - James M Snell <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
Fixes: #7379 PR-URL: #7382 Reviewed-By: jasnell - James M Snell <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
Fixes: #7379 PR-URL: #7382 Reviewed-By: jasnell - James M Snell <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
Fixes: #7379 PR-URL: #7382 Reviewed-By: jasnell - James M Snell <[email protected]> Reviewed-By: Roman Reiss <[email protected]>
Checklist
make -j4 test(UNIX) orvcbuild test nosign(Windows) passesAffected core subsystem(s)
readline
Description of change
Fixes: #7379