29 releases
0.3.1 | Mar 11, 2023 |
---|---|
0.3.0 | Feb 11, 2021 |
0.2.1 | Jan 25, 2021 |
0.1.1 | Jan 27, 2020 |
0.1.0-alpha.5 | Oct 28, 2019 |
#791 in Procedural macros
415 downloads per month
Used in 3 crates
(2 directly)
205KB
4K
SLoC
Implementation of the join!
macro.
join!
Macros which provide useful shortcut combinators, combine sync/async chains, support single and multi thread (sync/async) step by step execution of branches, transform tuple of results to result of tuple.
join!
macros will just return final values. Use it if you are working with iterators/streams etc.try_join!
macros will transpose tuple ofOption
s/Result
s inOption
/Result
of tuple. Use it when you are dealing with options or results. If one of branches producesNone
/Err
at the end of step, next steps execution will be aborted. In case ofasync
macro you can only provideResult
s because::futures::try_join
doesn't supportOption
s.
Use these docs for development, they are more convenient.
- Features
- Macros
- Combinators
- Nested combinators
- Handler
- Let pattern
- Block captures
- Custom configuration
- Demos
- Single thread examples
- Multi thread examples
- Detailed steps example
Features
- Performance. Macros produce well-optimized code (it doesn't use inactive branches during steps, doesn't clone results/options or any other values, doesn't allocate any memory on heap [except wrapping futures into
Box::pin
]) - you can check it withcargo expand
. - Steps allow to write code which depends on results of branches in previous iteration.
- One-line chains which can't be created using pure
Rust
without macros. - Briefness. Less code to express the same flow. Shortcut combinators = less parentheses.
async
macros produce futures, so they can be used in non-async
functions.- Configurability. There're many options which can be configured independently to fully change macro behaviour.
Macros
try_join!
- combinesResult
s/Option
s, transposes tuple ofResult
s/Option
s intoResult
/Option
of tuple.
assert_eq!(
try_join!(Ok::<_,()>(1), Ok::<_,()>("2"), Ok::<_,()>(3.0)),
Ok::<_,()>((1, "2", 3.0))
);
try_join_async!
- combines futures, transposes tuple ofResult
s intoResult
of tuple.
assert_eq!(
try_join_async!(ok::<_,()>(1), ok::<_,()>("2"), ok::<_,()>(3.0)).await,
Ok::<_,()>((1, "2", 3.0))
);
try_join_spawn!
- spawnsstd::thread
per each branch and joins results, transposes tuple ofResult
s/Option
s intoResult
/Option
of tuple.
assert_eq!(
try_join_spawn!(Ok::<_,()>(1), Ok::<_,()>("2"), Ok::<_,()>(3.0)),
Ok::<_,()>((1, "2", 3.0))
);
try_spawn!
- alias fortry_join_spawn!
.try_join_async_spawn!
- spawns tokio task usingtokio::spawn
per each branch, transposes tuple ofResult
s intoResult
of tuple.
assert_eq!(
try_join_async_spawn!(ok::<_,()>(1), ok::<_,()>("2"), ok::<_,()>(3.0)).await,
Ok::<_,()>((1, "2", 3.0))
);
try_async_spawn!
- alias fortry_join_async_spawn!
.join!
- combines values.
assert_eq!(
join!(1, "2", 3.0), (1, "2", 3.0)
);
join_async!
- combines futures.
assert_eq!(
join_async!(ready(1), ready("2"), ready(3.0)).await, (1, "2", 3.0)
);
join_spawn!
- spawnsstd::thread
per each branch.
assert_eq!(
join_spawn!(1, "2", 3.0), (1, "2", 3.0)
);
spawn!
- alias forjoin_spawn!
.join_async_spawn!
- spawns tokio task usingtokio::spawn
per each branch.
assert_eq!(
join_async_spawn!(ready(1), ready("2"), ready(3.0)).await, (1, "2", 3.0)
);
async_spawn!
- alias forjoin_async_spawn!
.
Combinators
- Then:
->
join! { value -> expr }; // => expr(value)
- Map:
|>
join! { value |> expr }; // => value.map(expr)
- AndThen:
=>
join! { value => expr }; // => value.and_then(expr)
- Filter:
?>
join! { value ?> expr }; // => value.filter(expr)
- Dot:
..
or>.
join! { value .. expr }; // => value.expr
join! { value >. expr }; // => value.expr
- Or:
<|
join! { value <| expr }; // => value.or(expr)
- OrElse:
<=
join! { value <= expr }; // => value.or_else(expr)
- MapErr:
!>
join! { value !> expr }; // => value.map_err(expr)
- Collect:
=>[]
(type is optional)
join! { value =>[] T }; // => value.collect::<T>()
join! { value =>[] }; // => value.collect()
- Chain:
>@>
join! { value >@> expr }; // => value.chain(expr)
- FindMap:
?|>@
join! { value ?|>@ expr }; // => value.find_map(expr)
- FilterMap:
?|>
join! { value ?|> expr }; // => value.filter_map(expr)
- Enumerate:
|n>
join! { value |n> }; // => value.enumerate()
- Partition:
?&!>
join! { value ?&!> expr }; // => value.partition(expr)
- Flatten:
^^>
join! { value ^^> }; // => value.flatten()
- Fold:
^@
join! { value ^@ init_expr, fn_expr }; // => value.fold(init_expr, fn_expr)
- TryFold:
?^@
join! { value ?^@ init_expr, fn_expr }; // => value.try_fold(init_expr, fn_expr)
- Find:
?@
join! { value ?@ expr }; // => value.find(expr)
- Zip:
>^>
join! { value >^> expr }; // => value.zip(expr)
- Unzip:
<->
(types are optional)
join! { value <-> A, B, FromA, FromB }; // => value.unzip::<A, B, FromA, FromB>()
join! { value <-> }; // => value.unzip()
- Inspect:
??
join! { value ?? expr }; // => (|value| { (expr)(&value); value })(value) // for sync
join_async! { value ?? expr }; // => value.inspect(expr) // for async
where value
is the previous value.
Every combinator prefixed by ~
will act as deferred action (all actions will wait until completion in every step and only after move to the next one).
Nested combinators
- Wrap:
combinator
>>>
combinator
(s)...
try_join! { value => >>> |> |v| v + 2 } // => value.and_then(|value| value.map(|v| v + 2))
Use to enter to nested constructions like
a.and_then(
// >>>
|b| b.and_then(
// >>>
|c| c.and_then(
|v| Ok(v + 2)
)
)
)
- Unwrap:
<<<
try_join! {
value
=> >>>
|> |v| v + 2
<<<
|> |v| Some(v + 4)
} // => value.and_then(|value| value.map(|v| v + 2)).map(|v| Some(v + 4))
Use to move out of nested constructions
a.and_then(
// >>>
|b| b.and_then(
// >>>
|c| c.and_then(
|v| Ok(v + 2)
)
// <<<
)
// <<<
).map(
|v| v + 1
)
Handler
might be one of
map
=> Only valid fortry
macros. Will act asresults.map(|(result0, result1, ..)| handler(result0, result1, ..))
assert_eq!(
try_join! {
Some(1),
Some(2),
Some(3),
map => |a, b, c| a + b + c
},
Some(6)
);
and_then
=> Only valid fortry
macros. Will act asresults.and_then(|(result0, result1, ..)| handler(result0, result1, ..))
assert_eq!(
try_join! {
Some(1),
Some(2),
Some(3),
and_then => |a, b, c| Some(a + b + c)
},
Some(6)
);
then
=> Only valid for nottry
macros. Will be executed in any case, act ashandler(result0, result1, ..)
assert_eq!(
join! {
Some(1),
Some(2),
Some(3),
then => |a: Option<u8>, b: Option<u8>, c: Option<u8>|
Some(a.unwrap() + b.unwrap() + c.unwrap())
},
Some(6)
);
or not specified - then Result<(result0, result1, ..), Error>
or Option<(result0, result1, ..)>
will be returned for try
macros and (result0, result1, ..)
for not try
macros.
Custom configuration
You can specify any params at the beginning of macro call.
futures_crate_path
- specifies custom crate path forfutures
crate, which will be used for allfutures
-related items, used byasync
join!
macros. Only valid forasync
macros.custom_joiner
- specifies custom joiner function or macro, which will join active branches in step if their count is greater than 1.transpose_results
- specifies should macro transpose tuple ofResult
s/Option
s intoResult
/Option
of tuple or not. Useful when provided joiner already returnsResult
of tuple and there's no need to transpose it.lazy_branches
- wrap every branch intomove || {}
when pass values to joiner. By defaulttrue
fortry_join_spawn!
,try_spawn!
andjoin_spawn!
,spawn!
macros because they usethread::spawn
call. Only if active branch count > 1.
use join::try_join_async;
use futures::future::ok;
macro_rules! custom_futures_joiner {
($($futures: expr),+) => {
::futures::try_join!($($futures),*);
}
}
#[tokio::main]
async fn main() {
let value = try_join_async! {
futures_crate_path(::futures)
custom_joiner(custom_futures_joiner!)
transpose_results(false)
ok::<_,()>(2u16), ok::<_,()>(3u16),
map => |a, b| a + b
}.await.unwrap();
assert_eq!(value, 5);
}
Rayon demo
use join::{try_join, join};
fn fib(num: u8) -> usize {
let mut prev = 0;
let mut cur = if num > 0 { 1 } else { 0 };
for _ in 1..num as usize {
let tmp = cur;
cur = prev + cur;
prev = tmp;
}
cur
}
fn main() {
let pool = rayon::ThreadPoolBuilder::new().build().unwrap();
let calculated = pool.install(||
try_join! {
custom_joiner(rayon::join)
|| Some(fib(50)),
|| Some(
join! {
custom_joiner(rayon::join)
lazy_branches(true)
fib(20) -> |v| v + 25,
fib(30) -> |v| vec![v; 10].into_iter() |n> |> |(index, value)| value + index ..sum::<usize>(),
then => |a, b| a + b
}
),
map => |a, b| a * b
}
);
assert_eq!(calculated.unwrap(), 104808819944395875);
}
Let pattern
You can specify let
pattern for each branch in order to share result with other branches, or in case if you need to have mut
value between steps.
assert_eq!(
try_join! {
let mut branch_0 = Ok::<_,()>(1) ~|> |v| v + 1,
let branch_1 = Ok::<_,()>(2) ~|> { let value_0 = branch_0.as_ref().unwrap(); move |v| v + value_0 },
map => |b_0, b_1| b_0 * b_1
}.unwrap(),
6
);
Block captures
In order to capture variables (for ex. values of other branches in example above) you can pass block statements instead of functions:
let mut some_value = Some("capture me");
assert_eq!(try_join! {
Some(0) |> |v| {
// assign `None` to some_value in step expr
some_value = None;
v
} |> {
// capture value before step and get str len
let captured_len = some_value.as_ref().unwrap().len();
move |v| v + captured_len
}
}.unwrap(), 10);
These blocks will be placed before actual step expressions.
Demos
Sync demo
Using this macro you can write things like
use rand::prelude::*;
use std::sync::Arc;
use join::try_join_spawn;
// Problem: generate vecs filled by random numbers in parallel, make some operations on them in parallel,
// find max of each vec in parallel and find final max of 3 vecs
// Solution:
fn main() {
// Branches will be executed in parallel, each in its own thread
let max = try_join_spawn! {
let branch_0 =
generate_random_vec(1000, 10000000u64)
.into_iter()
// .map(power2) (Multiply every element by itself)
|> power2
// .filter(is_even) (Filter even values)
?> is_even
// .collect::<Vec<_>>() (Collect values into `Vec<_>`)
=>[] Vec<_>
// Arc::new(Some(...))
// Use `Arc` to share data with branch 1
-> Arc::new -> Some
// Find max and clone its value
// .and_then(|v| v.iter().max().cloned())
~=> >>> ..iter().max() |> Clone::clone,
generate_random_vec(10000, 100000000000000f64)
.into_iter()
// .map(sqrt) (Extract sqrt from every element)
|> sqrt
// Some(...)
-> Some
// .and_then(|v| v...)
~=> >>>
// .enumerate() (Add index in order to compare with the values of branch_0)
|n>
// .map(...)
|> {
// Get data from branch 0 by cloning arc
let branch_0 = branch_0.as_ref().unwrap().clone();
let len = branch_0.len();
// Compare every element of branch 1 with element of branch_0
// with the same index and take min
move |(index, value)|
if index < len && value as u64 > branch_0[index] {
branch_0[index]
} else {
value as u64
}
}..max(),
generate_random_vec(100000, 100000u32)
.into_iter()
-> Some
// .and_then(|v| v.max())
~=> >>> ..max(),
and_then => |max0, max1, max2|
// Find final max
[max0, max1, max2 as u64].iter().max().cloned()
}
.unwrap();
println!("Max: {}", max);
}
fn generate_random_vec<T>(size: usize, max: T) -> Vec<T>
where
T: From<u8>
+ rand::distributions::uniform::SampleUniform
+ rand::distributions::uniform::SampleBorrow<T>
+ Copy,
{
let mut rng = rand::thread_rng();
(0..size)
.map(|_| rng.gen_range(T::from(0u8), max))
.collect()
}
fn is_even<T>(value: &T) -> bool
where
T: std::ops::Rem<Output = T> + std::cmp::PartialEq + From<u8> + Copy
{
*value % 2u8.into() == 0u8.into()
}
fn sqrt<T>(value: T) -> T
where
T: Into<f64>,
f64: Into<T>,
{
let value_f64: f64 = value.into();
value_f64.sqrt().into()
}
fn power2<T>(value: T) -> T
where
T: std::ops::Mul<Output = T> + Copy,
{
value * value
}
extern crate rand;
extern crate join;
use rand::prelude::*;
use join::try_join;
fn main() {
let mut rng = rand::thread_rng();
let result = try_join! {
(0..10)
// .map(|index| { let value ... })
|> |index| { let value = rng.gen_range(0, index + 5); if rng.gen_range(0f32, 2.0) > 1.0 { Ok(value) } else { Err(value) }}
// .filter(|result| ...)
?> |result| match result { Ok(_) => true, Err(value) => *value > 2 }
// .map(|v| v.map(|value| value + 1))
|> >>> |> |value| value + 1
<<<
// .try_fold(0i32, |acc, cur| {...})
?^@ 0i32, |acc, cur| {
cur.map(|cur| acc + cur).or_else(|cur| Ok(acc - cur))
}
// .and_then(|value| if ...)
=> |value| if value > 0 { Ok(value as u8) } else { Err(0) }
// Wait for all branches to be successful and then calculate fib
~|> fib,
(0..6)
// .map(|index| { let value ... })
|> |index| { let value = rng.gen_range(0, index + 5); if rng.gen_range(0f32, 2.0) > 1.0 { Some(value) } else { None }}
// .filter_map(|v| v)
?|> >>>
<<<
..sum::<u16>()
// Return `Ok` only if value is less than 20
-> |value| if value < 20 { Ok(value as u8) } else { Err(0) }
// Wait for all branches to be successful and then calculate fib
~|> fib,
// In case of success, multilpy fibs
map => |v_1, v_2| v_1 * v_2
};
result.map(|value| println!("Result: {}", value)).unwrap_or_else(|err| println!("Error: {:#?}", err));
}
fn fib(num: u8) -> usize {
println!("CALLED FIB!");
let mut prev = 0;
let mut cur = if num > 0 { 1 } else { 0 };
for _ in 1..num as usize {
let tmp = cur;
cur = prev + cur;
prev = tmp;
}
cur
}
Futures demo
use join::try_join_async;
use futures::stream::{iter, Stream};
use reqwest::Client;
use futures::future::{try_join_all, ok, ready};
use failure::{format_err, Error};
#[tokio::main]
async fn main() {
println!("Hello.\nThis's is the game where winner is player, which number is closest to the max count of links (starting with `https://`) found on one of random pages.\nYou play against random generator (0-500).");
enum GameResult {
Won,
Lost,
Draw
}
let client = Client::new();
let game = try_join_async! {
// Make requests to several sites
// and calculate count of links starting from `https://`
urls_to_calculate_link_count()
|> {
// If pass block statement instead of fn, it will be placed before current step,
// so it will us allow to capture some variables from context
let ref client = client;
move |url|
// `try_join_async!` wraps its content into `Box::pin(async move { })`
try_join_async! {
client
.get(url).send()
=> |value| value.text()
=> |body| ok((url, body.matches("https://").collect::<Vec<_>>().len()))
}
}
// Collect values into `Vec<_>`
=>[] Vec<_>
|> Ok
=> try_join_all
!> |err| format_err!("Error retrieving pages to calculate links: {:#?}", err)
=> >>>
..into_iter()
.max_by_key(|(_, link_count)| *link_count)
.ok_or(format_err!("Failed to find max link count"))
-> ready
// It waits for input in stdin before log max links count
~?? >>>
..as_ref()
|> |(url, count)| {
let split = url.to_owned().split('/').collect::<Vec<_>>();
let domain_name = split.get(2).unwrap_or(&url);
println!("Max `https://` link count found on `{}`: {}", domain_name, count)
}
..unwrap_or(()),
// Concurrently it makes request to the site which generates random number
url_to_random_number()
-> ok
=> {
// If pass block statement instead of fn, it will be placed before current step,
// so it will allow us to capture some variables from context
let ref client = client;
let map_parse_error = |error, value| format_err!("Failed to parse random number: {:#?}, value: {}", error, value);
move |url|
try_join_async! {
client
.get(url)
.send()
=> |value| value.text()
!> |err| format_err!("Error retrieving random number: {:#?}", err)
=> |value| ok(value[..value.len() - 1].to_owned()) // remove \n from `154\n`
=> |value|
ready(
value
.parse::<u16>()
.map_err(|err| map_parse_error(err, value))
)
}
}
// It waits for input in stdin before log random value
~?? >>>
..as_ref()
|> |number| println!("Random: {}", number)
..unwrap_or(()),
// Concurrently it reads value from stdin
read_number_from_stdin() |> Ok,
// Finally, when we will have all results, we can decide, who is winner
map => |(_url, link_count), random_number, number_from_stdin| {
let random_diff = (link_count as i32 - random_number as i32).abs();
let stdin_diff = (link_count as i32 - number_from_stdin as i32).abs();
match () {
_ if random_diff > stdin_diff => GameResult::Won,
_ if random_diff < stdin_diff => GameResult::Lost,
_ => GameResult::Draw
}
}
};
let _ = game.await.map(
|result|
println!(
"You {}",
match result {
GameResult::Won => "won!",
GameResult::Lost => "lose...",
_ => "have the same result as random generator!"
}
)
).unwrap_or_else(|error| eprintln!("Error: {:#?}", error));
}
fn urls_to_calculate_link_count() -> impl Stream<Item = &'static str> {
iter(
vec![
"https://en.wikipedia.org/w/api.php?format=json&action=query&generator=random&grnnamespace=0&prop=revisions|images&rvprop=content&grnlimit=100",
"https://github.com/explore",
"https://twitter.com/search?f=tweets&vertical=news&q=%23news&src=unkn"
]
)
}
fn url_to_random_number() -> &'static str {
"https://www.random.org/integers/?num=1&min=0&max=500&col=1&base=10&format=plain&rnd=new"
}
async fn read_number_from_stdin() -> u16 {
use tokio::io::{stdin, BufReader, Error, ErrorKind, AsyncBufReadExt};
let mut reader = BufReader::new(stdin()).lines();
loop {
println!("Please, enter number (`u16`)");
let next = reader.next_line();
let result = try_join_async! {
next
=> >>>
..ok_or(Error::new(ErrorKind::Other, "Failed to read value from stdin"))
=> >>>
..parse()
!> |err| Error::new(ErrorKind::Other, format!("Value from stdin isn't a correct `u16`: {:?}", err))
<<<
-> ready
}.await;
if let Ok(value) = result {
break value
}
}
}
Single thread combinations
Sync branches
Converts input in series of chained results and joins them step by step.
use std::error::Error;
use join::try_join;
type Result<T> = std::result::Result<T, Box<dyn Error>>;
fn action_1() -> Result<u16> {
Ok(1)
}
fn action_2() -> Result<u8> {
Ok(2)
}
fn main() {
let sum = try_join! {
// action_1(),
action_1(),
// action_2().map(|v| v as u16),
action_2() |> |v| v as u16,
// action_2().map(|v| v as u16 + 1).and_then(|v| Ok(v * 4)),
action_2() |> |v| v as u16 + 1 => |v| Ok(v * 4),
// action_1().and_then(|_| Err("5".into())).or(Ok(2)),
action_1() => |_| Err("5".into()) <| Ok(2),
map => |a, b, c, d| a + b + c + d
}.expect("Failed to calculate sum");
println!("Calculated: {}", sum);
}
Futures
Each branch will represent future chain. All branches will be joined using ::futures::join!
/::futures::try_join!
macro and join_async!
/try_join_async!
will return unpolled
future.
use std::error::Error;
use join::try_join_async;
use futures::future::{ok, err};
type Result<T> = std::result::Result<T, Box<dyn Error>>;
async fn action_1() -> Result<u16> {
Ok(1)
}
async fn action_2() -> Result<u8> {
Ok(2)
}
#[tokio::main]
async fn main() {
let sum = try_join_async! {
// action_1(),
action_1(),
// action_2().and_then(|v| ok(v as u16)),
action_2() => |v| ok(v as u16),
// action_2().map(|v| v.map(|v| v as u16 + 1)).and_then(|v| ok(v * 4u16)),
action_2() |> |v| v.map(|v| v as u16 + 1) => |v| ok(v * 4u16),
// action_1().and_then(|_| err("5".into())).or_else(|_| ok(2u16)),
action_1() => |_| err("5".into()) <= |_| ok(2u16),
and_then => |a, b, c, d| ok(a + b + c + d)
}.await.expect("Failed to calculate sum");
println!("Calculated: {}", sum);
}
Multi thread combinations
To execute several tasks in parallel you could use join_spawn!
(spawn!
) for sync tasks
and join_async_spawn!
(async_spawn!
) for futures. Since join_async
already provides concurrent futures execution in one thread, join_async_spawn!
spawns every branch into tokio
executor, so they will be evaluated in multi threaded executor.
Sync threads
join_spawn
spawns one ::std::thread
per each step of each branch (number of branches is the max thread count at the time).
use std::error::Error;
use join::try_join_spawn;
type Result<T> = std::result::Result<T, Box<dyn Error + Send + Sync>>;
fn action_1() -> Result<usize> {
Ok(1)
}
fn action_2() -> Result<u16> {
Ok(2)
}
fn main() {
// Branches will be executed in parallel
let sum = try_join_spawn! {
// thread::spawn(move || action_1()),
action_1(),
// thread::spawn(move || action_2().map(|v| v as usize)),
action_2() |> |v| v as usize,
// thread::spawn(move || action_2().map(|v| v as usize + 1).and_then(|v| Ok(v * 4))),
action_2() |> |v| v as usize + 1 => |v| Ok(v * 4),
// thread::spawn(move || action_1().and_then(|_| Err("5".into())).or(Ok(2))),
action_1() => |_| Err("5".into()) <| Ok(2),
map => |a, b, c, d| a + b + c + d
}.expect("Failed to calculate sum");
println!("Calculated: {}", sum);
}
Thread names
In runtime thread's name will be constructed from name of parent thread and join_%branch_index%.
Example with several branches:
extern crate join;
use std::thread;
use join::try_join_spawn;
fn current_thread_name() -> String {
thread::current().name().unwrap().to_owned()
}
fn print_branch_thread_name(index: &Result<usize, ()>) {
println!("Branch: {}. Thread name: {}.", index.unwrap(), current_thread_name());
}
fn main() {
let _ = try_join_spawn! {
Ok(0) ?? print_branch_thread_name,
Ok(1) ?? print_branch_thread_name,
try_join_spawn! {
Ok(2) ?? print_branch_thread_name,
try_join_spawn! {
Ok(3) ?? print_branch_thread_name,
}
}
}.unwrap();
}
// Branch: 0. Thread name: main_join_0.
// Branch: 1. Thread name: main_join_1.
// Branch: 2. Thread name: main_join_2_join_0.
// Branch: 3. Thread name: main_join_2_join_1_join_0.
// Order could be different.
Future tasks
join_async_spawn!
uses ::tokio::spawn
function to spawn tasks so it should be done inside tokio
runtime
(number of branches is the max count of tokio
tasks at the time).
use std::error::Error;
use join::try_join_async_spawn;
use futures::future::{ok, err};
type Result<T> = std::result::Result<T, Box<dyn Error + Send + Sync>>;
async fn action_1() -> Result<u16> {
Ok(1)
}
async fn action_2() -> Result<u8> {
Ok(2)
}
#[tokio::main]
async fn main() {
let sum = try_join_async_spawn! {
// tokio::spawn(Box::pin(action_1()))
action_1(),
// tokio::spawn(Box::pin(action_2().and_then(|v| ok(v as u16))))
action_2() => |v| ok(v as u16),
// tokio::spawn(Box::pin(action_2().map(|v| v.map(|v| v as u16 + 1)).and_then(|v| ok(v * 4u16))))
action_2() |> |v| v.map(|v| v as u16 + 1) => |v| ok(v * 4u16),
// tokio::spawn(Box::pin(action_1().and_then(|_| err("5".into())).or_else(|_| ok(2u16))))
action_1() => |_| err("5".into()) <= |_| ok(2u16),
and_then => |a, b, c, d| ok(a + b + c + d)
}.await.expect("Failed to calculate sum");
println!("Calculated: {}", sum);
}
Detailed steps example
By separating chain in actions, you will make actions wait for completion of all of them in current step before go to the next step.
use std::error::Error;
use join::try_join;
type Result<T> = std::result::Result<T, Box<dyn Error + Send + Sync>>;
fn action_1() -> Result<u16> {
Ok(1)
}
fn action_2() -> Result<u8> {
Ok(2)
}
fn main() {
let sum = try_join! {
action_1(),
let result_1 = action_2() ~|> |v| v as u16 + 1,
action_2() ~|> {
// `result_1` now is the result of `action_2()` [Ok(1u8)]
let result_1 = result_1.as_ref().ok().cloned();
move |v| {
if result_1.is_some() {
v as u16 + 1
} else {
unreachable!()
}
}
} ~=> {
// `result_1` now is the result of `|v| v as u16 + 1` [Ok(2u16)]
let result_1 = result_1.as_ref().ok().cloned();
move |v| {
if let Some(result_1) = result_1 {
Ok(v * 4 + result_1)
} else {
unreachable!()
}
}
},
action_1() ~=> |_| Err("5".into()) <| Ok(2),
map => |a, b, c, d| a + b + c + d
}.expect("Failed to calculate sum");
println!("Calculated: {}", sum);
}
Dependencies
~1.5MB
~37K SLoC