Struct netsblox_vm::std_system::StdSystem
source · pub struct StdSystem<C: CustomTypes<StdSystem<C>>> { /* private fields */ }
Expand description
A type implementing the System
trait which supports all features.
StdSystem
can be configured with CustomTypes
and Config
,
which together allow for the definition of any external features (e.g., defining syscalls),
as well as overriding default behavior (e.g., rpc intercepting).
Implementations§
source§impl<C: CustomTypes<StdSystem<C>>> StdSystem<C>
impl<C: CustomTypes<StdSystem<C>>> StdSystem<C>
sourcepub fn new_sync(
base_url: CompactString,
project_name: Option<&str>,
config: Config<C, Self>,
clock: Arc<Clock>
) -> Self
pub fn new_sync( base_url: CompactString, project_name: Option<&str>, config: Config<C, Self>, clock: Arc<Clock> ) -> Self
Equivalent to StdSystem::new_async
except that it can be executed outside of async context.
Note that using this from within async context can result in a panic from, e.g., tokio
trying to create a runtime within a runtime.
Examples found in repository?
94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154
fn main() {
// read in an xml file whose path is given as a command line argument
let args = std::env::args().collect::<Vec<_>>();
if args.len() != 2 {
panic!("usage: {} [xml file]", &args[0]);
}
let mut xml = String::new();
std::fs::File::open(&args[1]).expect("failed to open file").read_to_string(&mut xml).expect("failed to read file");
// create a new shared clock and start a thread that updates it at our desired interval
let clock = Arc::new(Clock::new(UtcOffset::UTC, Some(Precision::Medium)));
let clock_clone = clock.clone();
std::thread::spawn(move || loop {
std::thread::sleep(CLOCK_INTERVAL);
clock_clone.update();
});
// create a custom config for the system - in this simple example we just implement the say/think blocks to print to stdout
let config = Config::<C, StdSystem<C>> {
request: None,
command: Some(Rc::new(|_mc, key, command, _proc| match command {
Command::Print { style: _, value } => {
if let Some(value) = value {
println!("{value:?}");
}
key.complete(Ok(())); // any request that you handle must be completed - otherwise the calling process will hang forever
CommandStatus::Handled
}
_ => CommandStatus::UseDefault { key, command }, // anything you don't handle should return the key and command to invoke the default behavior instead
})),
};
// initialize our system with all the info we've put together
let system = Rc::new(StdSystem::new_sync(CompactString::new(BASE_URL), None, config, clock.clone()));
let mut env = get_running_project(&xml, system);
// begin running the code - these are some helpers to make things more efficient in terms of memory and cpu resources
let mut idle_sleeper = IdleAction::new(YIELDS_BEFORE_SLEEP, Box::new(|| std::thread::sleep(IDLE_SLEEP_TIME)));
let mut next_collect = clock.read(Precision::Medium) + COLLECT_INTERVAL;
loop {
env.mutate(|mc, env| {
let mut proj = env.proj.borrow_mut(mc);
for _ in 0..1024 {
// step the virtual machine forward by one bytecode instruction
let res = proj.step(mc);
if let ProjectStep::Error { error, proc } = &res {
// if we get an error, we can generate an error summary including a stack trace - here we just print out the result
let trace = ErrorSummary::extract(error, proc, &env.locs);
println!("error: {error:?}\ntrace: {trace:?}");
}
// this takes care of performing thread sleep if we get a bunch of no-ops from proj.step back to back
idle_sleeper.consume(&res);
}
});
// if it's time for us to do garbage collection, do it and reset the next collection time
if clock.read(Precision::Low) >= next_collect {
env.collect_all();
next_collect = clock.read(Precision::Medium) + COLLECT_INTERVAL;
}
}
}
sourcepub async fn new_async(
base_url: CompactString,
project_name: Option<&str>,
config: Config<C, Self>,
clock: Arc<Clock>
) -> Self
pub async fn new_async( base_url: CompactString, project_name: Option<&str>, config: Config<C, Self>, clock: Arc<Clock> ) -> Self
Initializes a new instance of StdSystem
targeting the given NetsBlox server base url, e.g., https://cloud.netsblox.org
.
sourcepub async fn call_rpc_async(
&self,
service: &str,
rpc: &str,
args: &[(&str, &Json)]
) -> Result<SimpleValue, CompactString>
pub async fn call_rpc_async( &self, service: &str, rpc: &str, args: &[(&str, &Json)] ) -> Result<SimpleValue, CompactString>
Asynchronously calls an RPC and returns the result.
This function directly makes requests to NetsBlox, bypassing any RPC hook defined by Config
.
sourcepub fn get_public_id(&self) -> CompactString
pub fn get_public_id(&self) -> CompactString
Gets the public id of the running system that can be used to send messages to this client.
sourcepub fn inject_message(
&self,
msg_type: CompactString,
values: VecMap<CompactString, SimpleValue, false>
)
pub fn inject_message( &self, msg_type: CompactString, values: VecMap<CompactString, SimpleValue, false> )
Injects a message into the receiving queue as if received over the network.
Trait Implementations§
source§impl<C: CustomTypes<StdSystem<C>>> System<C> for StdSystem<C>
impl<C: CustomTypes<StdSystem<C>>> System<C> for StdSystem<C>
§type RequestKey = AsyncKey<Result<<C as CustomTypes<StdSystem<C>>>::Intermediate, CompactString>>
type RequestKey = AsyncKey<Result<<C as CustomTypes<StdSystem<C>>>::Intermediate, CompactString>>
§type CommandKey = AsyncKey<Result<(), CompactString>>
type CommandKey = AsyncKey<Result<(), CompactString>>
source§fn rand<T: SampleUniform, R: SampleRange<T>>(&self, range: R) -> T
fn rand<T: SampleUniform, R: SampleRange<T>>(&self, range: R) -> T
range
, which is assumed to be non-empty.
The input for this generic function is such that it is compatible with rand::Rng::gen_range
,
which makes it possible to implement this function with any random provider under the rand
crate standard.source§fn perform_request<'gc>(
&self,
mc: &Mutation<'gc>,
request: Request<'gc, C, Self>,
proc: &mut Process<'gc, C, Self>
) -> Result<Self::RequestKey, ErrorCause<C, Self>>
fn perform_request<'gc>( &self, mc: &Mutation<'gc>, request: Request<'gc, C, Self>, proc: &mut Process<'gc, C, Self> ) -> Result<Self::RequestKey, ErrorCause<C, Self>>
Entity
that made the request is provided for context.source§fn poll_request<'gc>(
&self,
mc: &Mutation<'gc>,
key: &Self::RequestKey,
_proc: &mut Process<'gc, C, Self>
) -> Result<AsyncResult<Result<Value<'gc, C, Self>, CompactString>>, ErrorCause<C, Self>>
fn poll_request<'gc>( &self, mc: &Mutation<'gc>, key: &Self::RequestKey, _proc: &mut Process<'gc, C, Self> ) -> Result<AsyncResult<Result<Value<'gc, C, Self>, CompactString>>, ErrorCause<C, Self>>
Entity
that made the request is provided for context.source§fn perform_command<'gc>(
&self,
mc: &Mutation<'gc>,
command: Command<'gc, '_, C, Self>,
proc: &mut Process<'gc, C, Self>
) -> Result<Self::CommandKey, ErrorCause<C, Self>>
fn perform_command<'gc>( &self, mc: &Mutation<'gc>, command: Command<'gc, '_, C, Self>, proc: &mut Process<'gc, C, Self> ) -> Result<Self::CommandKey, ErrorCause<C, Self>>
Entity
that issued the command is provided for context.source§fn poll_command<'gc>(
&self,
_mc: &Mutation<'gc>,
key: &Self::CommandKey,
_proc: &mut Process<'gc, C, Self>
) -> Result<AsyncResult<Result<(), CompactString>>, ErrorCause<C, Self>>
fn poll_command<'gc>( &self, _mc: &Mutation<'gc>, key: &Self::CommandKey, _proc: &mut Process<'gc, C, Self> ) -> Result<AsyncResult<Result<(), CompactString>>, ErrorCause<C, Self>>
Entity
that issued the command is provided for context.source§fn send_message(
&self,
msg_type: CompactString,
values: VecMap<CompactString, Json, false>,
targets: Vec<CompactString>,
expect_reply: bool
) -> Result<Option<ExternReplyKey>, ErrorCause<C, StdSystem<C>>>
fn send_message( &self, msg_type: CompactString, values: VecMap<CompactString, Json, false>, targets: Vec<CompactString>, expect_reply: bool ) -> Result<Option<ExternReplyKey>, ErrorCause<C, StdSystem<C>>>
values
to each of the specified targets
.
The expect_reply
value controls whether or not to use a reply mechanism to asynchronously receive a response from the target(s).
In the case that there are multiple targets, only the first reply (if any) should be used.source§fn poll_reply(&self, key: &ExternReplyKey) -> AsyncResult<Option<Json>>
fn poll_reply(&self, key: &ExternReplyKey) -> AsyncResult<Option<Json>>
System::send_message
.
If the client responds, a value of [Some(x)
] is returned.
The system may elect to impose a timeout for reply results, in which case None
is returned instead.source§fn send_reply(
&self,
key: InternReplyKey,
value: Json
) -> Result<(), ErrorCause<C, Self>>
fn send_reply( &self, key: InternReplyKey, value: Json ) -> Result<(), ErrorCause<C, Self>>
source§fn receive_message(&self) -> Option<IncomingMessage>
fn receive_message(&self) -> Option<IncomingMessage>
None
if there are no messages in the buffer.
If a message is received, a tuple of form (msg_type, values, reply_key)
is returned.