1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283
use quote::{quote_spanned, ToTokens};
use super::{
DelayType, OpInstGenerics, OperatorCategory, OperatorConstraints,
OperatorInstance, OperatorWriteOutput, Persistence, WriteContextArgs, RANGE_1,
};
/// > 1 input stream of type `(K, V1)`, 1 output stream of type `(K, V2)`.
/// > The output will have one tuple for each distinct `K`, with an accumulated value of type `V2`.
///
/// If the input and output value types are the same and do not require initialization then use
/// [`reduce_keyed`](#reduce_keyed).
///
/// > Arguments: two Rust closures. The first generates an initial value per group. The second
/// > itself takes two arguments: an 'accumulator', and an element. The second closure returns the
/// > value that the accumulator should have for the next iteration.
///
/// A special case of `fold`, in the spirit of SQL's GROUP BY and aggregation constructs. The input
/// is partitioned into groups by the first field ("keys"), and for each group the values in the second
/// field are accumulated via the closures in the arguments.
///
/// > Note: The closures have access to the [`context` object](surface_flows.mdx#the-context-object).
///
/// ```dfir
/// source_iter([("toy", 1), ("toy", 2), ("shoe", 11), ("shoe", 35), ("haberdashery", 7)])
/// -> fold_keyed(|| 0, |old: &mut u32, val: u32| *old += val)
/// -> assert_eq([("toy", 3), ("shoe", 46), ("haberdashery", 7)]);
/// ```
///
/// `fold_keyed` can be provided with one generic lifetime persistence argument, either
/// `'tick` or `'static`, to specify how data persists. With `'tick`, values will only be collected
/// within the same tick. With `'static`, values will be remembered across ticks and will be
/// aggregated with pairs arriving in later ticks. When not explicitly specified persistence
/// defaults to `'tick`.
///
/// `fold_keyed` can also be provided with two type arguments, the key type `K` and aggregated
/// output value type `V2`. This is required when using `'static` persistence if the compiler
/// cannot infer the types.
///
/// ```dfir
/// source_iter([("toy", 1), ("toy", 2), ("shoe", 11), ("shoe", 35), ("haberdashery", 7)])
/// -> fold_keyed(|| 0, |old: &mut u32, val: u32| *old += val)
/// -> assert_eq([("toy", 3), ("shoe", 46), ("haberdashery", 7)]);
/// ```
///
/// Example using `'tick` persistence:
/// ```rustbook
/// let (input_send, input_recv) = dfir_rs::util::unbounded_channel::<(&str, &str)>();
/// let mut flow = dfir_rs::dfir_syntax! {
/// source_stream(input_recv)
/// -> fold_keyed::<'tick, &str, String>(String::new, |old: &mut _, val| {
/// *old += val;
/// *old += ", ";
/// })
/// -> for_each(|(k, v)| println!("({:?}, {:?})", k, v));
/// };
///
/// input_send.send(("hello", "oakland")).unwrap();
/// input_send.send(("hello", "berkeley")).unwrap();
/// input_send.send(("hello", "san francisco")).unwrap();
/// flow.run_available();
/// // ("hello", "oakland, berkeley, san francisco, ")
///
/// input_send.send(("hello", "palo alto")).unwrap();
/// flow.run_available();
/// // ("hello", "palo alto, ")
/// ```
pub const FOLD_KEYED: OperatorConstraints = OperatorConstraints {
name: "fold_keyed",
categories: &[OperatorCategory::KeyedFold],
hard_range_inn: RANGE_1,
soft_range_inn: RANGE_1,
hard_range_out: RANGE_1,
soft_range_out: RANGE_1,
num_args: 2,
persistence_args: &(0..=1),
type_args: &(0..=2),
is_external_input: false,
// If this is set to true, the state will need to be cleared using `#context.set_state_tick_hook`
// to prevent reading uncleared data if this subgraph doesn't run.
// https://github.com/hydro-project/hydro/issues/1298
has_singleton_output: false,
flo_type: None,
ports_inn: None,
ports_out: None,
input_delaytype_fn: |_| Some(DelayType::Stratum),
write_fn: |wc @ &WriteContextArgs {
hydroflow,
context,
op_span,
ident,
inputs,
is_pull,
root,
op_inst:
OperatorInstance {
generics:
OpInstGenerics {
persistence_args,
type_args,
..
},
..
},
arguments,
..
},
_| {
assert!(is_pull);
let persistence = match persistence_args[..] {
[] => Persistence::Tick,
[a] => a,
_ => unreachable!(),
};
let generic_type_args = [
type_args
.first()
.map(ToTokens::to_token_stream)
.unwrap_or(quote_spanned!(op_span=> _)),
type_args
.get(1)
.map(ToTokens::to_token_stream)
.unwrap_or(quote_spanned!(op_span=> _)),
];
let input = &inputs[0];
let initfn = &arguments[0];
let aggfn = &arguments[1];
let groupbydata_ident = wc.make_ident("groupbydata");
let hashtable_ident = wc.make_ident("hashtable");
let (write_prologue, write_iterator, write_iterator_after) = match persistence {
Persistence::Tick => {
(
quote_spanned! {op_span=>
let #groupbydata_ident = #hydroflow.add_state(::std::cell::RefCell::new(#root::rustc_hash::FxHashMap::<#( #generic_type_args ),*>::default()));
},
quote_spanned! {op_span=>
let mut #hashtable_ident = #context.state_ref(#groupbydata_ident).borrow_mut();
{
#[inline(always)]
fn check_input<Iter, A, B>(iter: Iter) -> impl ::std::iter::Iterator<Item = (A, B)>
where
Iter: std::iter::Iterator<Item = (A, B)>,
A: ::std::clone::Clone,
B: ::std::clone::Clone
{
iter
}
/// A: accumulator type
/// T: iterator item type
/// O: output type
#[inline(always)]
fn call_comb_type<A, T, O>(a: &mut A, t: T, f: impl Fn(&mut A, T) -> O) -> O {
(f)(a, t)
}
for kv in check_input(#input) {
// TODO(mingwei): remove `unknown_lints` when `clippy::unwrap_or_default` is stabilized.
#[allow(unknown_lints, clippy::unwrap_or_default)]
let entry = #hashtable_ident.entry(kv.0).or_insert_with(#initfn);
#[allow(clippy::redundant_closure_call)] call_comb_type(entry, kv.1, #aggfn);
}
}
let #ident = #hashtable_ident.drain();
},
Default::default(),
)
}
Persistence::Static => {
(
quote_spanned! {op_span=>
let #groupbydata_ident = #hydroflow.add_state(::std::cell::RefCell::new(#root::rustc_hash::FxHashMap::<#( #generic_type_args ),*>::default()));
},
quote_spanned! {op_span=>
let mut #hashtable_ident = #context.state_ref(#groupbydata_ident).borrow_mut();
{
#[inline(always)]
fn check_input<Iter, A, B>(iter: Iter) -> impl ::std::iter::Iterator<Item = (A, B)>
where
Iter: std::iter::Iterator<Item = (A, B)>,
A: ::std::clone::Clone,
B: ::std::clone::Clone
{
iter
}
/// A: accumulator type
/// T: iterator item type
/// O: output type
#[inline(always)]
fn call_comb_type<A, T, O>(a: &mut A, t: T, f: impl Fn(&mut A, T) -> O) -> O {
(f)(a, t)
}
for kv in check_input(#input) {
// TODO(mingwei): remove `unknown_lints` when `clippy::unwrap_or_default` is stabilized.
#[allow(unknown_lints, clippy::unwrap_or_default)]
let entry = #hashtable_ident.entry(kv.0).or_insert_with(#initfn);
#[allow(clippy::redundant_closure_call)] call_comb_type(entry, kv.1, #aggfn);
}
}
// Play everything but only on the first run of this tick/stratum.
// (We know we won't have any more inputs, so it is fine to only play once.
// Because of the `DelayType::Stratum` or `DelayType::MonotoneAccum`).
let #ident = #context.is_first_run_this_tick()
.then_some(#hashtable_ident.iter())
.into_iter()
.flatten()
.map(
// TODO(mingwei): remove `unknown_lints` when `suspicious_double_ref_op` is stabilized.
#[allow(unknown_lints, suspicious_double_ref_op, clippy::clone_on_copy)]
|(k, v)| (
::std::clone::Clone::clone(k),
::std::clone::Clone::clone(v),
)
);
},
quote_spanned! {op_span=>
#context.schedule_subgraph(#context.current_subgraph(), false);
},
)
}
Persistence::Mutable => {
(
quote_spanned! {op_span=>
let #groupbydata_ident = #hydroflow.add_state(::std::cell::RefCell::new(#root::rustc_hash::FxHashMap::<#( #generic_type_args ),*>::default()));
},
quote_spanned! {op_span=>
let mut #hashtable_ident = #context.state_ref(#groupbydata_ident).borrow_mut();
{
#[inline(always)]
fn check_input<Iter: ::std::iter::Iterator<Item = #root::util::PersistenceKeyed::<K, V>>, K: ::std::clone::Clone, V: ::std::clone::Clone>(iter: Iter)
-> impl ::std::iter::Iterator<Item = #root::util::PersistenceKeyed::<K, V>> { iter }
#[inline(always)]
/// A: accumulator type
/// T: iterator item type
/// O: output type
fn call_comb_type<A, T, O>(a: &mut A, t: T, f: impl Fn(&mut A, T) -> O) -> O {
f(a, t)
}
for item in check_input(#input) {
match item {
Persist(k, v) => {
let entry = #hashtable_ident.entry(k).or_insert_with(#initfn);
#[allow(clippy::redundant_closure_call)] call_comb_type(entry, v, #aggfn);
},
Delete(k) => {
#hashtable_ident.remove(&k);
},
}
}
}
let #ident = #hashtable_ident
.iter()
.map(#[allow(suspicious_double_ref_op, clippy::clone_on_copy)] |(k, v)| (k.clone(), v.clone()));
},
quote_spanned! {op_span=>
#context.schedule_subgraph(#context.current_subgraph(), false);
},
)
}
};
Ok(OperatorWriteOutput {
write_prologue,
write_iterator,
write_iterator_after,
})
},
};