add AudioContextOptions to AudioContext #36
Conversation
| id: AudioNodeId, | ||
| } | ||
|
|
||
| impl AudioContextRegistration { | ||
| pub fn id(&self) -> &AudioNodeId { |
There was a problem hiding this comment.
There's no point in marking these functions const as the registration creation is not const. Marking functions const is a liability for the future, so I'd rather not
src/io.rs
Outdated
| config.sample_rate = CpalSampleRate(opts.sample_rate.unwrap_or(config.sample_rate.0)); | ||
| config.channels = opts.channels.unwrap_or(config.channels); | ||
|
|
||
| match opts.latency_hint { |
There was a problem hiding this comment.
Can you move this block into a helper function, makes it a bit more readable and then we can re-use if for the InputBuilder
| @@ -134,7 +219,8 @@ pub(crate) fn build_output( | |||
| (stream, config, sender) | |||
| } | |||
|
|
|||
| pub(crate) fn build_input() -> (Stream, StreamConfig, Receiver<AudioBuffer>) { | |||
| /// Builds the input | |||
| pub fn build_input() -> (Stream, StreamConfig, Receiver<AudioBuffer>) { | |||
There was a problem hiding this comment.
You should port the latency hint to the input builder too (can be done in another PR if it's too much work). They should be symmetrical in the end
src/context.rs
Outdated
| self.inner.render_channel.send(message).unwrap(); | ||
| self.inner | ||
| .render_channel | ||
| .send_timeout(message, Duration::from_millis(10)) |
There was a problem hiding this comment.
Why opt for timeout? If the requested latency > 10millis, this will fail for no good reason
| } | ||
| } | ||
|
|
||
| /// `OfflineAudioContext` doesn't start rendering automatically | ||
| /// You need to call this function to start the audio rendering | ||
| pub fn start_rendering(&mut self) -> AudioBuffer { | ||
| // make buffer_size always a multiple of BUFFER_SIZE, so we can still render piecewise with |
There was a problem hiding this comment.
I think actually this is no longer needed, now we can handle varying buffer lengths.
We can remove this code in another PR
src/context.rs
Outdated
| pub fn length(&self) -> usize { | ||
| /// get the length of rendering audio buffer | ||
| #[must_use] | ||
| pub const fn length(&self) -> usize { |
| /// | ||
| /// * `options` - options contains latency hint information | ||
| /// | ||
| /// # Warning |
There was a problem hiding this comment.
As mentioned here the correctness of the API is not enforce by the type system. I could make it more strict but at the cost of more complexity. What do you think ?
is it Ok with you, if i open a good first issue about this topic? |
|
Go ahead! |
The user can now specify:
when building its
AudioContext.If the intented options are not met by the audio hardware capabilities, the
AudioContextis built withdefault_configoptions, and a warning is printed to the user.If the user options are missing, the
AudioContextis built withprimary_configoptions. If theprimary_configoptions are not met by the audio harware capabilities, theAudioContextis built withdefault_configoptions, and a warning is printed to the user.This PR modify
DestinationNode:max_channels_countmethodChannelCountConstraintsonDestinationNodeThis PR adds merger.rs example to demonstrate new capabilities.
Clippy passes on
context.rsandio.rsrelated to #35
related to #28