feat/handle_hex_audiodata #145
Conversation
companion to OpenVoiceOS/ovos-audio#34
companion to OpenVoiceOS/ovos-audio#34
Codecov ReportAttention:
Additional details and impacted files@@ Coverage Diff @@
## dev #145 +/- ##
==========================================
- Coverage 60.23% 59.71% -0.52%
==========================================
Files 34 34
Lines 3347 3366 +19
==========================================
- Hits 2016 2010 -6
- Misses 1331 1356 +25
☔ View full report in Codecov by Sentry. |
companion to OpenVoiceOS/ovos-audio#34
| # if sessions is not default we also need to do it since | ||
| # it likely is a remote client such as hivemind | ||
| send_binary = os.environ.get("IS_OVOS_CONTAINER") or \ | ||
| SessionManager.get(message).session_id != "default" |
There was a problem hiding this comment.
Is the default session ID guaranteed to always be 'default', or should this reference SessionManager.default_session?
There was a problem hiding this comment.
it is guaranteed to be "default"
| mtype = "mycroft.audio.queue" | ||
|
|
||
| if not send_binary: | ||
| data = {"uri": filename} |
There was a problem hiding this comment.
Should this be excluded from binary messages? The URI could be useful for non-default sessions running with a shared file system
There was a problem hiding this comment.
it is not used and further reinforces (to humans debugging) that a different code path is used
i'd rather not send a path that is supposed to be invalid, i'd argue if its running on same device then a non-default session should not be used, but i can imagine a case where a different device still has a shared file system so no strong feelings either way
could also argue this is leaking information about host system (unwanted in some scenarios) and increasing message size/bandwidth usage for no reason
companion to OpenVoiceOS/ovos-audio#34