1
0
Fork 0
mirror of https://forgejo.ellis.link/continuwuation/continuwuity.git synced 2025-09-09 17:10:57 +00:00
continuwuity/src/service/rooms/state_compressor/mod.rs

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

488 lines
13 KiB
Rust
Raw Normal View History

2022-10-05 20:34:31 +02:00
use std::{
collections::{HashMap, HashSet},
fmt::Write,
2022-10-05 20:34:31 +02:00
mem::size_of,
sync::{Arc, Mutex},
2022-10-05 20:34:31 +02:00
};
use arrayvec::ArrayVec;
use conduit::{
at, checked, err, expected, utils,
utils::{bytes, math::usize_from_f64},
Result,
};
use database::Map;
2022-10-05 20:33:55 +02:00
use lru_cache::LruCache;
use ruma::{EventId, RoomId};
use crate::{
rooms,
rooms::short::{ShortId, ShortStateHash, ShortStateKey},
Dep,
};
pub struct Service {
pub stateinfo_cache: Mutex<StateInfoLruCache>,
db: Data,
services: Services,
}
struct Services {
short: Dep<rooms::short::Service>,
state: Dep<rooms::state::Service>,
}
struct Data {
shortstatehash_statediff: Arc<Map>,
}
#[derive(Clone)]
struct StateDiff {
parent: Option<ShortStateHash>,
added: Arc<CompressedState>,
removed: Arc<CompressedState>,
}
#[derive(Clone, Default)]
pub struct ShortStateInfo {
pub shortstatehash: ShortStateHash,
pub full_state: Arc<CompressedState>,
pub added: Arc<CompressedState>,
pub removed: Arc<CompressedState>,
}
#[derive(Clone, Default)]
pub struct HashSetCompressStateEvent {
pub shortstatehash: ShortStateHash,
pub added: Arc<CompressedState>,
pub removed: Arc<CompressedState>,
}
type StateInfoLruCache = LruCache<ShortStateHash, ShortStateInfoVec>;
type ShortStateInfoVec = Vec<ShortStateInfo>;
type ParentStatesVec = Vec<ShortStateInfo>;
pub(crate) type CompressedState = HashSet<CompressedStateEvent>;
pub(crate) type CompressedStateEvent = [u8; 2 * size_of::<ShortId>()];
impl crate::Service for Service {
fn build(args: crate::Args<'_>) -> Result<Arc<Self>> {
let config = &args.server.config;
let cache_capacity = f64::from(config.stateinfo_cache_capacity) * config.cache_capacity_modifier;
Ok(Arc::new(Self {
stateinfo_cache: LruCache::new(usize_from_f64(cache_capacity)?).into(),
db: Data {
shortstatehash_statediff: args.db["shortstatehash_statediff"].clone(),
},
services: Services {
short: args.depend::<rooms::short::Service>("rooms::short"),
state: args.depend::<rooms::state::Service>("rooms::state"),
},
}))
}
fn memory_usage(&self, out: &mut dyn Write) -> Result {
let (cache_len, ents) = {
let cache = self.stateinfo_cache.lock().expect("locked");
let ents = cache
.iter()
.map(at!(1))
.flat_map(|vec| vec.iter())
.fold(HashMap::new(), |mut ents, ssi| {
ents.insert(Arc::as_ptr(&ssi.added), compressed_state_size(&ssi.added));
ents.insert(Arc::as_ptr(&ssi.removed), compressed_state_size(&ssi.removed));
ents.insert(Arc::as_ptr(&ssi.full_state), compressed_state_size(&ssi.full_state));
ents
});
(cache.len(), ents)
};
let ents_len = ents.len();
let bytes = ents.values().copied().fold(0_usize, usize::saturating_add);
let bytes = bytes::pretty(bytes);
writeln!(out, "stateinfo_cache: {cache_len} {ents_len} ({bytes})")?;
Ok(())
}
fn clear_cache(&self) { self.stateinfo_cache.lock().expect("locked").clear(); }
fn name(&self) -> &str { crate::service::make_name(std::module_path!()) }
}
impl Service {
2021-08-12 23:04:00 +02:00
/// Returns a stack with info on shortstatehash, full state, added diff and
/// removed diff for the selected shortstatehash and each parent layer.
pub async fn load_shortstatehash_info(&self, shortstatehash: ShortStateHash) -> Result<ShortStateInfoVec> {
if let Some(r) = self
.stateinfo_cache
.lock()
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
.expect("locked")
.get_mut(&shortstatehash)
{
2021-08-15 06:46:00 +02:00
return Ok(r.clone());
}
2022-10-05 20:34:31 +02:00
let StateDiff {
parent,
added,
removed,
} = self.get_statediff(shortstatehash).await?;
let response = if let Some(parent) = parent {
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
let mut response = Box::pin(self.load_shortstatehash_info(parent)).await?;
let mut state = (*response.last().expect("at least one response").full_state).clone();
state.extend(added.iter().copied());
let removed = (*removed).clone();
2021-08-12 23:04:00 +02:00
for r in &removed {
state.remove(r);
}
response.push(ShortStateInfo {
shortstatehash,
full_state: Arc::new(state),
added,
removed: Arc::new(removed),
});
response
2021-08-12 23:04:00 +02:00
} else {
vec![ShortStateInfo {
shortstatehash,
full_state: added.clone(),
added,
removed,
}]
};
self.stateinfo_cache
.lock()
.expect("locked")
.insert(shortstatehash, response.clone());
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
Ok(response)
2021-08-12 23:04:00 +02:00
}
pub async fn compress_state_event(&self, shortstatekey: ShortStateKey, event_id: &EventId) -> CompressedStateEvent {
const SIZE: usize = size_of::<CompressedStateEvent>();
let shorteventid = self
.services
.short
.get_or_create_shorteventid(event_id)
.await;
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
let mut v = ArrayVec::<u8, SIZE>::new();
v.extend(shortstatekey.to_be_bytes());
v.extend(shorteventid.to_be_bytes());
v.as_ref()
.try_into()
.expect("failed to create CompressedStateEvent")
}
2022-09-07 13:25:51 +02:00
/// Returns shortstatekey, event id
#[inline]
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
pub async fn parse_compressed_state_event(
&self, compressed_event: CompressedStateEvent,
) -> Result<(ShortStateKey, Arc<EventId>)> {
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
use utils::u64_from_u8;
let shortstatekey = u64_from_u8(&compressed_event[0..size_of::<ShortStateKey>()]);
let shorteventid = u64_from_u8(&compressed_event[size_of::<ShortStateKey>()..]);
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
let event_id = self
.services
.short
.get_eventid_from_short(shorteventid)
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
.await?;
Ok((shortstatekey, event_id))
2021-08-12 23:04:00 +02:00
}
/// Creates a new shortstatehash that often is just a diff to an already
2022-10-05 20:33:55 +02:00
/// existing shortstatehash and therefore very efficient.
///
/// There are multiple layers of diffs. The bottom layer 0 always contains
2021-08-12 23:04:00 +02:00
/// the full state. Layer 1 contains diffs to states of layer 0, layer 2
/// diffs to layer 1 and so on. If layer n > 0 grows too big, it will be
/// combined with layer n-1 to create a new diff on layer n-1 that's
/// based on layer n-2. If that layer is also too big, it will recursively
/// fix above layers too.
///
2021-08-12 23:04:00 +02:00
/// * `shortstatehash` - Shortstatehash of this state
/// * `statediffnew` - Added to base. Each vec is shortstatekey+shorteventid
/// * `statediffremoved` - Removed from base. Each vec is
/// shortstatekey+shorteventid
/// * `diff_to_sibling` - Approximately how much the diff grows each time
/// for this layer
/// * `parent_states` - A stack with info on shortstatehash, full state,
/// added diff and removed diff for each parent layer
#[tracing::instrument(skip_all, level = "debug")]
pub fn save_state_from_diff(
&self, shortstatehash: ShortStateHash, statediffnew: Arc<HashSet<CompressedStateEvent>>,
statediffremoved: Arc<HashSet<CompressedStateEvent>>, diff_to_sibling: usize,
mut parent_states: ParentStatesVec,
) -> Result {
let statediffnew_len = statediffnew.len();
let statediffremoved_len = statediffremoved.len();
let diffsum = checked!(statediffnew_len + statediffremoved_len)?;
if parent_states.len() > 3 {
2021-08-12 23:04:00 +02:00
// Number of layers
// To many layers, we have to go deeper
let parent = parent_states.pop().expect("parent must have a state");
let mut parent_new = (*parent.added).clone();
let mut parent_removed = (*parent.removed).clone();
2021-08-12 23:04:00 +02:00
for removed in statediffremoved.iter() {
if !parent_new.remove(removed) {
// It was not added in the parent and we removed it
parent_removed.insert(*removed);
}
2021-08-14 19:07:50 +02:00
// Else it was added in the parent and we removed it again. We
// can forget this change
}
for new in statediffnew.iter() {
if !parent_removed.remove(new) {
2021-08-14 19:07:50 +02:00
// It was not touched in the parent and we added it
parent_new.insert(*new);
}
2021-08-14 19:07:50 +02:00
// Else it was removed in the parent and we added it again. We
// can forget this change
2021-08-12 23:04:00 +02:00
}
2021-08-12 23:04:00 +02:00
self.save_state_from_diff(
shortstatehash,
Arc::new(parent_new),
Arc::new(parent_removed),
2021-08-12 23:04:00 +02:00
diffsum,
parent_states,
)?;
2021-08-12 23:04:00 +02:00
return Ok(());
}
2021-08-31 21:20:03 +02:00
if parent_states.is_empty() {
2021-08-12 23:04:00 +02:00
// There is no parent layer, create a new state
self.save_statediff(
2022-10-05 20:34:31 +02:00
shortstatehash,
&StateDiff {
2022-10-05 20:34:31 +02:00
parent: None,
added: statediffnew,
removed: statediffremoved,
},
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
);
2021-08-12 23:04:00 +02:00
return Ok(());
};
2021-08-12 23:04:00 +02:00
// Else we have two options.
// 1. We add the current diff on top of the parent layer.
// 2. We replace a layer above
let parent = parent_states.pop().expect("parent must have a state");
let parent_added_len = parent.added.len();
let parent_removed_len = parent.removed.len();
let parent_diff = checked!(parent_added_len + parent_removed_len)?;
if checked!(diffsum * diffsum)? >= checked!(2 * diff_to_sibling * parent_diff)? {
2021-08-12 23:04:00 +02:00
// Diff too big, we replace above layer(s)
let mut parent_new = (*parent.added).clone();
let mut parent_removed = (*parent.removed).clone();
for removed in statediffremoved.iter() {
if !parent_new.remove(removed) {
2021-08-14 19:07:50 +02:00
// It was not added in the parent and we removed it
parent_removed.insert(*removed);
2021-08-12 23:04:00 +02:00
}
2021-08-14 19:07:50 +02:00
// Else it was added in the parent and we removed it again. We
// can forget this change
}
for new in statediffnew.iter() {
if !parent_removed.remove(new) {
2021-08-14 19:07:50 +02:00
// It was not touched in the parent and we added it
parent_new.insert(*new);
2021-08-14 19:07:50 +02:00
}
// Else it was removed in the parent and we added it again. We
// can forget this change
2021-08-12 23:04:00 +02:00
}
2021-08-12 23:04:00 +02:00
self.save_state_from_diff(
shortstatehash,
Arc::new(parent_new),
Arc::new(parent_removed),
2021-08-12 23:04:00 +02:00
diffsum,
parent_states,
)?;
} else {
// Diff small enough, we add diff as layer on top of parent
self.save_statediff(
2022-10-05 20:34:31 +02:00
shortstatehash,
&StateDiff {
parent: Some(parent.shortstatehash),
2022-10-05 20:34:31 +02:00
added: statediffnew,
removed: statediffremoved,
},
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
);
2021-08-12 23:04:00 +02:00
}
2020-09-13 22:24:36 +02:00
Ok(())
}
2022-06-20 12:08:58 +02:00
/// Returns the new shortstatehash, and the state diff from the previous
/// room state
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
pub async fn save_state(
&self, room_id: &RoomId, new_state_ids_compressed: Arc<HashSet<CompressedStateEvent>>,
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
) -> Result<HashSetCompressStateEvent> {
let previous_shortstatehash = self
.services
.state
.get_room_shortstatehash(room_id)
.await
.ok();
let state_hash = utils::calculate_hash(
&new_state_ids_compressed
.iter()
.map(|bytes| &bytes[..])
.collect::<Vec<_>>(),
);
let (new_shortstatehash, already_existed) = self
.services
.short
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
.get_or_create_shortstatehash(&state_hash)
.await;
2022-06-20 12:08:58 +02:00
if Some(new_shortstatehash) == previous_shortstatehash {
return Ok(HashSetCompressStateEvent {
shortstatehash: new_shortstatehash,
..Default::default()
});
2022-06-20 12:08:58 +02:00
}
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
let states_parents = if let Some(p) = previous_shortstatehash {
self.load_shortstatehash_info(p).await.unwrap_or_default()
} else {
ShortStateInfoVec::new()
Database Refactor combine service/users data w/ mod unit split sliding sync related out of service/users instrument database entry points remove increment crap from database interface de-wrap all database get() calls de-wrap all database insert() calls de-wrap all database remove() calls refactor database interface for async streaming add query key serializer for database implement Debug for result handle add query deserializer for database add deserialization trait for option handle start a stream utils suite de-wrap/asyncify/type-query count_one_time_keys() de-wrap/asyncify users count add admin query users command suite de-wrap/asyncify users exists de-wrap/partially asyncify user filter related asyncify/de-wrap users device/keys related asyncify/de-wrap user auth/misc related asyncify/de-wrap users blurhash asyncify/de-wrap account_data get; merge Data into Service partial asyncify/de-wrap uiaa; merge Data into Service partially asyncify/de-wrap transaction_ids get; merge Data into Service partially asyncify/de-wrap key_backups; merge Data into Service asyncify/de-wrap pusher service getters; merge Data into Service asyncify/de-wrap rooms alias getters/some iterators asyncify/de-wrap rooms directory getters/iterator partially asyncify/de-wrap rooms lazy-loading partially asyncify/de-wrap rooms metadata asyncify/dewrap rooms outlier asyncify/dewrap rooms pdu_metadata dewrap/partially asyncify rooms read receipt de-wrap rooms search service de-wrap/partially asyncify rooms user service partial de-wrap rooms state_compressor de-wrap rooms state_cache de-wrap room state et al de-wrap rooms timeline service additional users device/keys related de-wrap/asyncify sender asyncify services refactor database to TryFuture/TryStream refactor services for TryFuture/TryStream asyncify api handlers additional asyncification for admin module abstract stream related; support reverse streams additional stream conversions asyncify state-res related Signed-off-by: Jason Volk <jason@zemos.net>
2024-08-08 17:18:30 +00:00
};
2022-06-20 12:08:58 +02:00
let (statediffnew, statediffremoved) = if let Some(parent_stateinfo) = states_parents.last() {
let statediffnew: HashSet<_> = new_state_ids_compressed
.difference(&parent_stateinfo.full_state)
.copied()
.collect();
let statediffremoved: HashSet<_> = parent_stateinfo
.full_state
.difference(&new_state_ids_compressed)
.copied()
.collect();
(Arc::new(statediffnew), Arc::new(statediffremoved))
2022-06-20 12:08:58 +02:00
} else {
(new_state_ids_compressed, Arc::new(HashSet::new()))
2022-06-20 12:08:58 +02:00
};
2022-06-20 12:08:58 +02:00
if !already_existed {
self.save_state_from_diff(
new_shortstatehash,
statediffnew.clone(),
statediffremoved.clone(),
2022-06-20 12:08:58 +02:00
2, // every state change is 2 event changes on average
states_parents,
)?;
};
Ok(HashSetCompressStateEvent {
shortstatehash: new_shortstatehash,
added: statediffnew,
removed: statediffremoved,
})
2022-06-20 12:08:58 +02:00
}
async fn get_statediff(&self, shortstatehash: ShortStateHash) -> Result<StateDiff> {
const BUFSIZE: usize = size_of::<ShortStateHash>();
const STRIDE: usize = size_of::<ShortStateHash>();
let value = self
.db
.shortstatehash_statediff
.aqry::<BUFSIZE, _>(&shortstatehash)
.await
.map_err(|e| err!(Database("Failed to find StateDiff from short {shortstatehash:?}: {e}")))?;
let parent = utils::u64_from_bytes(&value[0..size_of::<u64>()])
.ok()
.take_if(|parent| *parent != 0);
debug_assert!(value.len() % STRIDE == 0, "value not aligned to stride");
let num_values = value.len() / STRIDE;
let mut add_mode = true;
let mut added = HashSet::with_capacity(num_values);
let mut removed = HashSet::with_capacity(num_values);
let mut i = STRIDE;
while let Some(v) = value.get(i..expected!(i + 2 * STRIDE)) {
if add_mode && v.starts_with(&0_u64.to_be_bytes()) {
add_mode = false;
i = expected!(i + STRIDE);
continue;
}
if add_mode {
added.insert(v.try_into()?);
} else {
removed.insert(v.try_into()?);
}
i = expected!(i + 2 * STRIDE);
}
added.shrink_to_fit();
removed.shrink_to_fit();
Ok(StateDiff {
parent,
added: Arc::new(added),
removed: Arc::new(removed),
})
}
fn save_statediff(&self, shortstatehash: ShortStateHash, diff: &StateDiff) {
let mut value = Vec::<u8>::with_capacity(
2_usize
.saturating_add(diff.added.len())
.saturating_add(diff.removed.len()),
);
let parent = diff.parent.unwrap_or(0_u64);
value.extend_from_slice(&parent.to_be_bytes());
for new in diff.added.iter() {
value.extend_from_slice(&new[..]);
}
if !diff.removed.is_empty() {
value.extend_from_slice(&0_u64.to_be_bytes());
for removed in diff.removed.iter() {
value.extend_from_slice(&removed[..]);
}
}
self.db
.shortstatehash_statediff
.insert(&shortstatehash.to_be_bytes(), &value);
}
}
#[inline]
fn compressed_state_size(compressed_state: &CompressedState) -> usize {
compressed_state
.len()
.checked_mul(size_of::<CompressedStateEvent>())
.expect("CompressedState size overflow")
}