Skip to main content

SqliteStore

Struct SqliteStore 

Source
pub struct SqliteStore { /* private fields */ }
Expand description

Every write (via with_txn or a direct execute) goes through writer, the single connection WAL reserves for the one process-wide writer. Reads go through readers: a pool of separate connections in file-backed mode, or the same connection as the writer for in_memory, where a second connection would just see an empty, unrelated database.

Implementations§

Source§

impl SqliteStore

Source

pub fn record_peer_view(&self, tag: &IdentityTag, view: &PeerView) -> Result<()>

Source

pub fn peer_views(&self, tag: &IdentityTag) -> Result<Vec<PeerView>>

Every view recorded under tag, ordered by peer id.

Source§

impl SqliteStore

Source

pub fn recovery_recipients(&self, tag: &IdentityTag) -> Result<Vec<Recipient>>

The recovery recipients this device wraps tag’s epochs to: the recipients of its envelopes that name no machine, less retired ones. Handshakes only store envelopes addressed to this machine, so a peer never adds one.

Source

pub fn recovery_envelopes(&self, recipient: &Recipient) -> Result<Vec<Envelope>>

Every envelope, under any identity, wrapped to recipient with no machine id.

Source

pub fn recovery_offered(&self) -> Result<bool>

Source

pub fn mark_recovery_offered(&self, now_micros: i64) -> Result<()>

Source§

impl SqliteStore

Source

pub fn put_retirement( &self, signed: &SignedRetirement, retirer_key: &VerifyingKey, ) -> Result<bool>

Keeps a retirement verified under its retirer’s key. Which retirers count is the caller’s rule. A fingerprint so far retired only by an epoch’s cause gains the signed record; otherwise the first stored under its identity stays and this returns false.

Source

pub fn note_named_retirements(&self, tag: &IdentityTag) -> Result<Vec<String>>

Records each well-formed fingerprint an epoch cause of tag names that no retirement here covers yet, unsigned: the epoch was taken, so the phrase must open nothing here even when the signed retirement cannot be verified. Returns the fingerprints recorded.

Source

pub fn retirements(&self, tag: &IdentityTag) -> Result<Vec<SignedRetirement>>

The signed retirements stored under tag, by fingerprint.

Source

pub fn retired_under( &self, tag: &IdentityTag, ) -> Result<Vec<(String, Option<SignedRetirement>)>>

Every fingerprint retired under tag, with its signed retirement when one was verified here; None for one only an epoch’s cause named.

Source

pub fn retired_fingerprints(&self) -> Result<BTreeSet<String>>

Every retired fingerprint, under any identity.

Source

pub fn rotating_retirements( &self, tag: &IdentityTag, ) -> Result<BTreeSet<String>>

The fingerprints retired under tag whose retirement calls for a new epoch and a new store key.

Source

pub fn mark_rotating_retirements( &self, tag: &IdentityTag, ) -> Result<Vec<String>>

Marks the retirements under tag, signed or named by a cause, whose phrase a device of tag not revoked lists in its row, or this device still wraps to, as calling for rotation; the mark is sticky. Made-up fingerprints listed by no one are kept but rotate nothing. Run before delete_retired_envelopes. Returns the newly marked fingerprints.

Source

pub fn epochs_wrapped_to_retired(&self, tag: &IdentityTag) -> Result<Vec<Epoch>>

The epochs of tag this store holds an envelope for to a retired recovery recipient.

Source

pub fn delete_retired_envelopes(&self) -> Result<usize>

Deletes every envelope, under any identity, wrapped to a retired recovery recipient. Returns how many went.

Source§

impl SqliteStore

Source

pub fn put_revocation( &self, signed: &SignedRevocation, revoker_key: &VerifyingKey, ) -> Result<bool>

Keeps a revocation verified under its revoker’s key on the target’s machine row, creating a bare row for a target this store has not seen. Which revokers count is the caller’s rule. Returns false when it was already stored. apply_revocations marks the devices it reaches.

Source

pub fn revocations(&self, tag: &IdentityTag) -> Result<Vec<SignedRevocation>>

Source

pub fn revoked_by_rule( &self, tag: &IdentityTag, extra: Option<&Revocation>, ) -> Result<BTreeSet<String>>

The devices of tag that its stored revocations reach, plus those extra would reach, from the first rows held now.

Source

pub fn revokers_of( &self, tag: &IdentityTag, machine_id: &str, ) -> Result<BTreeSet<String>>

The revokers of the stored revocations of tag whose rule reaches machine_id.

Source

pub fn apply_revocations(&self, tag: &IdentityTag) -> Result<Vec<String>>

Marks every device the stored revocations of tag reach as revoked, which is sticky, drops their log floor acks and pins, and returns those newly marked.

Source

pub fn frame_floor(&self, tag: &IdentityTag, peer: &str) -> Result<Option<u32>>

The lowest epoch number peer may still seal wire frames under.

Source

pub fn raise_frame_floor( &self, tag: &IdentityTag, peer: &str, number: u32, ) -> Result<bool>

Raises the floor for peer, never lowering it; returns true when it moved.

Source§

impl SqliteStore

Source

pub fn device_route(&self, machine_id: &str) -> Result<Option<DeviceRoute>>

Source

pub fn device_routes(&self) -> Result<Vec<DeviceRoute>>

Source

pub fn set_device_route(&self, route: DeviceRoute) -> Result<()>

Replaces what this store knows about reaching route.machine_id.

Source

pub fn merge_device_routes( &self, sender: &str, routes: &[DeviceRoute], ) -> Result<bool>

Takes the routes a peer sent in its handshake: its own addresses go first, ahead of those this store already knew for it, and one it relays for a third device only adds to what this store already knows. Routes to this device and to machines with no active row here are ignored. True when anything changed.

Source

pub fn legacy_routes(&self) -> Result<Vec<(String, LegacyRoute)>>

Source

pub fn set_legacy_route( &self, machine_id: &str, route: &LegacyRoute, ) -> Result<()>

Source

pub fn sync_peers_migrated(&self) -> Result<bool>

Set once every sync-peers line has been matched to a device, after which the file is no longer read.

Source

pub fn mark_sync_peers_migrated(&self) -> Result<()>

Source§

impl SqliteStore

Source

pub fn put_key_epoch( &self, tag: &IdentityTag, epoch: &Epoch, cause: Option<&str>, created_at: i64, ) -> Result<bool>

Records an epoch once, filling in its cause if a later call brings one; returns false when the epoch was already known.

Source

pub fn key_epochs(&self, tag: &IdentityTag) -> Result<Vec<StoredEpoch>>

Source

pub fn put_envelope(&self, envelope: &Envelope) -> Result<bool>

Keeps the first envelope stored for an epoch and recipient; age output is randomized, so a second wrap of the same key is a different blob carrying nothing new. Returns false when one was already stored.

Source

pub fn envelope( &self, tag: &IdentityTag, epoch: &Epoch, recipient: &str, ) -> Result<Option<Envelope>>

Source

pub fn envelopes_for_epoch( &self, tag: &IdentityTag, epoch: &Epoch, ) -> Result<Vec<Envelope>>

Source

pub fn envelopes_for_machine(&self, machine_id: &str) -> Result<Vec<Envelope>>

Source

pub fn insert_first_device_record( &self, signed: &SignedRecord, signer: &str, signer_key: &VerifyingKey, ) -> Result<bool>

Stores a device’s first record after checking it is signed by signer under signer_key. Which signer may introduce a device (the founder, a pairing device, a migration self-signature) is the caller’s rule. Only that verified signature is stored, and the same first record arriving under another signer gains its signature. Returns false when nothing new was stored.

Source

pub fn update_device_record(&self, signed: &SignedRecord) -> Result<bool>

Replaces a device’s current record with a later version it signed itself under its current key, keeping only that signature. Returns false when that version is already stored unchanged.

Source

pub fn replace_device_record_for_new_identity( &self, signed: &SignedRecord, signer: &str, signer_key: &VerifyingKey, ) -> Result<bool>

Stores a first record that moves a device into another identity, after checking it is signed by signer under signer_key. The row held for that machine id under its earlier identity is replaced, and its revoked flag cleared, since revocations apply per identity; the caller reapplies the new identity’s revocations. A machine with no row gets this one as its first. Returns false when the record is already the stored first record.

Source

pub fn device_record( &self, machine_id: &str, ) -> Result<Option<StoredDeviceRecord>>

Source

pub fn device_records(&self) -> Result<Vec<StoredDeviceRecord>>

Source§

impl SqliteStore

Source

pub fn put_keys_durably( &self, tag: &IdentityTag, epochs: &[Epoch], envelopes: &[Envelope], created_at: i64, ) -> Result<()>

Commits the epochs and envelopes with the WAL synced to disk, so the caller may delete the only other copy of a key once this returns. Rows already stored are kept as they are.

Source

pub fn own_identity_envelopes( &self, recipient: &str, ) -> Result<Option<(IdentityTag, Vec<Envelope>)>>

This device’s identity, named by its own device record, and the envelopes of that identity wrapped to recipient. None until the store holds the record.

Source

pub fn delete_envelope( &self, tag: &IdentityTag, epoch: &Epoch, recipient: &str, ) -> Result<bool>

Returns false when no envelope was stored for that epoch and recipient.

Source

pub fn known_peer_ids(&self) -> Result<BTreeSet<String>>

Every other machine this store has synced with or holds records from: sync cursor keys (with any push: prefix removed) and the origins in the sync log.

Source§

impl SqliteStore

Source

pub fn live_counts(&self) -> Result<LiveCounts>

Source

pub fn outbox_pending(&self) -> Result<u64>

Outbox rows the index drain has not applied yet. While any are pending the index counts lag the store by design, so a count mismatch then is not drift.

Source

pub fn integrity_report( &self, indexes: Option<&Indexes>, ) -> Result<IntegrityReport>

Source

pub fn repair( &self, indexes: Option<&mut Indexes>, embed: Option<EmbedFn<'_>>, dry_run: bool, ) -> Result<RepairStats>

Restores every invariant in IntegrityReport::drift, each step in its own transactions so a failure part-way leaves earlier steps committed. With indexes, the outbox is drained before the key garbage collection and before the consistency check, so one pass leaves nothing for the next. dry_run counts without writing.

Source

pub fn backfill_embeddings( &self, embed: EmbedFn<'_>, dry_run: bool, ) -> Result<(u64, u64)>

Embeds hot rows that have none, over the same context-plus-content body the ingest path embeds. updated_at is bumped so the peer’s last-writer-wins check accepts the row that now carries a vector. Runs only this step of repair, so a caller that must not merge or retire rows can fill vectors on its own schedule. Returns (rows embedded, rows whose embedding failed or had the wrong dimension); with dry_run it counts rows awaiting a vector and writes nothing.

Source

pub fn backfill_chunk_embeddings( &self, embed: EmbedBatchFn<'_>, max_memories: usize, budget: Duration, ) -> Result<ChunkBackfill>

Embeds the chunk sets of up to max_memories memories that need them and stores each set whole via store_chunks, which queues the index upsert. Texts go to embed in batches of at most 16 (a padded ONNX batch of 64 held about 2.4 GB of arena, past the Pi’s cap), whole memories per batch; once budget has elapsed no further batch starts, so a caller on a shared thread gets it back promptly. A memory any of whose vectors fails or has the wrong dimension counts as failed and is left for a later call. Records the remaining pending count for chunk_progress.

Source

pub fn sync_log_row_exists(&self, local_seq: u64) -> Result<bool>

Whether a sync_log row with this local_seq still exists; a test and diagnostic hook for the sync log garbage collector.

Source§

impl SqliteStore

Source

pub fn near_duplicate_groups( &self, indexes: &Indexes, threshold: f32, limit: usize, ) -> Result<NearDupListing>

Groups of live distillates sharing a project and type whose cosine is at least threshold, closest group first, at most limit, with the total number of groups.

Source

pub fn near_duplicate_group_count( &self, indexes: &Indexes, threshold: f32, ) -> Result<usize>

Source

pub fn check_merge( &self, member_ids: &[Uuid], ) -> Result<Result<(), MergeRefusal>>

The member checks merge_distillates makes, read-only, so a caller can refuse a doomed merge before paying for its embedding.

Source

pub fn merge_distillates( &self, member_ids: &[Uuid], content: &str, embedding: Vec<f32>, model: &str, ) -> Result<Result<MergeReport, MergeRefusal>>

One transaction: store content as a distillate of the members’ project, type and scope, supersede every member by it and move each raw’s distillation marker onto it. Refused without writing anything unless the members are 2 to 4 distinct live distillates sharing all three. Merges for the store’s owner: members by different authors give a distillate authored by this store.

Source

pub fn record_near_duplicate_review( &self, member_ids: &[Uuid], verdict: ReviewVerdict, ) -> Result<Result<u64, MergeRefusal>>

Links every pair of member_ids with the verdict’s relation and context (see ReviewVerdict::relation); returns how many links were new. A pair holds one edge per relation type, so a later separate or unmerged verdict replaces the earlier one.

Source

pub fn reopen_near_duplicate_reviews( &self, verdicts: &[ReviewVerdict], ) -> Result<u64>

Deletes the review edges of verdicts, replicated as relation deletes, so their pairs are grouped again on every peer. Returns how many edges were removed; a rerun removes nothing.

Source§

impl SqliteStore

Source

pub fn backfill_session_at( &self, evidence: &dyn SessionEvidence, dry_run: bool, reemit: bool, ) -> Result<SessionBackfillReport>

Fills a NULL session_at from the strongest evidence each memory has (see SessionTier) and logs a sync Update per row, without touching updated_at: peers merge the date apart from last-write-wins, so the Update never beats a newer edit made elsewhere. A date already set is never replaced, so a second run writes nothing. reemit also relogs every date set before this run, for a peer that upgraded late.

Source§

impl SqliteStore

Source

pub fn at_rest_keys(&self, purpose: AtRestPurpose) -> Result<Option<AtRestKeys>>

The keys files for purpose are sealed under; None for a store with no store key, whose files stay plaintext. The first call mints them. Once the store key has changed since the current key was minted (a rotation, or the store’s first encryption), a new current key is minted and the old one is kept as retired until the files are resealed, so a copy of the old store key opens none of the files written after that.

Source

pub fn settle_at_rest_keys( &self, purpose: AtRestPurpose, current: SealKeyId, ) -> Result<()>

Records that every file for purpose is sealed under current, and forgets the retired keys. Nothing changes if current is no longer the current key.

Source§

impl SqliteStore

Source

pub fn ingest_memory_if_absent(&self, memory: &Memory) -> Result<bool>

store_memory_if_absent for ingest, which also skips a memory whose text a delete removed from the same source, so a source file still on disk cannot bring it back under a new id.

Source

pub fn ingest_blocked(&self, memory: &Memory) -> Result<bool>

Whether a delete removed this memory’s text from its source.

Source§

impl SqliteStore

Source

pub fn bootstrap_state(&self) -> Result<Option<BootstrapState>>

Source

pub fn set_bootstrap_state(&self, state: Option<BootstrapState>) -> Result<()>

None clears the state.

Source

pub fn holds_replicated_rows(&self) -> Result<bool>

True when any memory, entity, conversation or relation is stored.

Source§

impl SqliteStore

Source

pub fn chunks_needing_embedding(&self, limit: usize) -> Result<Vec<ChunkWork>>

Live memories that split into two or more chunks and whose stored set is absent or was split from other content, oldest first. Splitting happens here in Rust, so every candidate row is split on each call; the length floor only skips rows too short to yield two chunks.

Source

pub fn scan_chunk_work(&self, limit: usize) -> Result<(Vec<ChunkWork>, usize)>

chunks_needing_embedding, plus the total number of memories needing chunks. Counting past limit keeps scanning the whole table but holds on to nothing beyond the first limit.

Source

pub fn chunk_progress(&self) -> Result<(usize, Option<usize>)>

(live memories holding a chunk set, memories still needing one as of the last chunk backfill). The second is None before any backfill has run, because counting it fresh means splitting every live row.

Source

pub fn store_chunks( &self, memory_id: Uuid, content_hash: &str, chunks: &[(usize, String, Vec<f32>)], ) -> Result<bool>

Replaces the memory’s whole chunk set and queues an index upsert for it, in one transaction. Returns false, writing nothing, when the memory is no longer live or its content no longer hashes to content_hash. Chunks are derived locally on every peer, so nothing here touches sync_log.

Source

pub fn chunks_for(&self, memory_id: Uuid) -> Result<Vec<MemoryChunk>>

Source

pub fn delete_chunks(&self, memory_id: Uuid) -> Result<()>

Source

pub fn chunk_vec_key(&self, memory_id: Uuid, ord: usize) -> Result<u64>

Stable u64 key in chunks.usearch for one chunk, allocated on first use, mirroring vec_key.

Source

pub fn chunk_vec_keys_for(&self, memory_id: Uuid) -> Result<Vec<(usize, u64)>>

(ord, key) pairs allocated for the memory, by ord.

Source

pub fn delete_chunk_vec_key(&self, memory_id: Uuid, ord: usize) -> Result<()>

Drops the key row once its vector is gone from chunks.usearch, so chunk_vec_keys only holds keys the index can still resolve.

Source

pub fn chunk_for_vec_key(&self, key: u64) -> Result<Option<(Uuid, usize)>>

Source

pub fn embedded_chunk_count(&self, dim: usize) -> Result<usize>

Chunk rows of live memories whose embedding has exactly dim components: what chunks.usearch should hold once drained.

Source

pub fn live_chunked_memories(&self) -> Result<Vec<(Uuid, String)>>

Live memories holding a chunk set, each with the hash of its current content, which the set must match to still be indexed.

Source

pub fn purge_chunks_outside_live_set(&self) -> Result<()>

Deletes chunk rows and chunk keys of every memory outside the live set. Only a full rebuild calls this, because it resets the chunk index first and so has no vectors to remove per key.

The scoped counterpart of Indexes::chunk_search: a brute-force cosine scan of the embedded chunks of live in-scope memories, returning each memory’s best (memory id, cosine, ord) for the top limit memories, best first. Rows arrive grouped by memory (the primary key’s order), so only the running best of one memory and the top limit are ever held, however wide the scope.

Source§

impl SqliteStore

Source

pub fn attach_cold_tier(&self, tier: Arc<ColdTier>) -> bool

Gives the store the cold tier its cold rows live in. Only the first call takes effect; returns whether this one did.

Source

pub fn cold_tier(&self) -> Option<&Arc<ColdTier>>

Source

pub fn cold_record(&self, id: Uuid) -> Result<Option<ColdRecord>>

The full text and embedding of memory id from its cold segment, or None when the row is not cold. A segment that fails to open is an error, never an absent record.

Source

pub fn get_memory_full(&self, id: Uuid) -> Result<Option<Memory>>

get_memory_with_embedding with a cold row’s content and embedding read back from its segment. Store readers otherwise see the stub.

Source

pub fn seal_cold_segment(&self, older_than: DateTime<Utc>) -> Result<ColdSeal>

Moves up to one segment’s worth of archived raws older than older_than into a new cold segment. Each row keeps a stub of its content and loses its embedding; updated_at, the sync log and the index outbox are left alone, because archiving is local. A row that changed after it was read, its embedding included, stays as it is.

Source

pub fn cold_manifest(&self, segment: &str) -> Result<Option<String>>

Source

pub fn check_cold_segments(&self) -> Result<Vec<UnreadableSegment>>

Opens every segment’s header so one that would fail a recall shows up in status first. Returns the segments that failed.

Source

pub fn run_cold_pass( &self, now: DateTime<Utc>, pause: Duration, ) -> Result<ColdPassReport>

One full cold pass: deletes files earlier passes retired, retires or rewrites segments holding records no row points at, seals until nothing is old enough (pausing between segments so other writers get the writer lock), then checks every segment.

Source

pub fn cold_stats(&self) -> Result<ColdStats>

Fails when the store holds cold rows but no cold tier is attached, since none of them can then be read.

Source§

impl SqliteStore

Source

pub fn vacuum(&self) -> Result<(u64, u64)>

VACUUM cannot run inside a transaction and needs the writer alone; returns (bytes before, bytes after).

Source

pub fn vacuum_clearing_wal(&self) -> Result<(u64, u64)>

vacuum, then a checkpoint that truncates the WAL, because the frames VACUUM appended, and older ones past them, would otherwise keep every page’s previous contents on disk. Errors when readers keep the checkpoint from finishing, though the VACUUM itself went through.

Source

pub fn vacuum_space(&self) -> Result<VacuumSpace>

VACUUM writes a full copy of the live pages to SQLite’s temp directory and then the same again into the WAL, so both need room.

Source

pub fn stored_redaction_compacted(&self) -> Result<bool>

Source

pub fn mark_stored_redaction_compacted(&self) -> Result<()>

Source§

impl SqliteStore

Source

pub fn correct_memory( &self, old_id: Uuid, replacement: Memory, evidence: &str, ) -> Result<Result<Corrected, CorrectRefusal>>

Stores replacement in the corrected memory’s project and scope, supersedes the old memory with it and links them by a supersedes relation carrying evidence, in one transaction. The sync log gets the replacement’s Create before the relation, so a peer never sees a dangling edge.

Source

pub fn corrected_at(&self, ids: &[Uuid]) -> Result<HashMap<Uuid, DateTime<Utc>>>

When each of ids that replaced another memory was written, keyed by id; ids that replaced nothing are absent.

Source

pub fn correction_evidence(&self, replacement: Uuid) -> Result<Option<String>>

The evidence recorded when replacement corrected another memory.

Source§

impl SqliteStore

Source

pub fn delete_scope( &self, scope: &DeleteScope, archived_sessions: &[String], ) -> Result<DeleteReport>

Removes the scope’s memories, the distillates any of them was distilled into, their chunks, relations and markers, and the scope’s conversations, all in one transaction. Every removed memory and conversation is tombstoned and the delete logged, so it wins on every peer, and the earlier log records of each removed row are reclaimed so the log keeps none of its content. archived_sessions are sessions the caller found in the transcript archive under this scope; their conversations are tombstoned even when the store holds nothing of them.

Source

pub fn delete_purge_requested(&self) -> Result<Option<String>>

A token while a delete applied from a peer still waits for the index compaction and cold purge the deleting daemon ran itself.

Source

pub fn delete_purge_done(&self, token: &str) -> Result<()>

Clears the request token named, unless a delete applied since.

Source

pub fn preview_delete( &self, scope: &DeleteScope, archived_sessions: &[String], ) -> Result<DeleteReport>

What delete_scope would remove, counted without writing anything.

Source

pub fn conversation_deleted(&self, id: Uuid) -> Result<bool>

Source

pub fn session_deleted(&self, path: &str) -> Result<bool>

Whether the Claude Code session a transcript at path belongs to was deleted, in which case nothing of it may be archived or ingested.

Source

pub fn purge_cold_segments(&self) -> Result<ColdPurge>

Retires or rewrites every cold segment holding records no row points at any more, then deletes the files left without a manifest, so a deleted cold row’s text leaves the disk now rather than at the next cold pass. A store without a cold tier has nothing to purge.

Source§

impl SqliteStore

Source

pub fn delete_status(&self, key: &str) -> Result<Option<DeleteStatus>>

The delete key names, by its id or, newest first, by its target.

Source§

impl SqliteStore

Source

pub fn is_encrypted(&self) -> bool

Source

pub fn rewrite_under_key(&self, to: RewriteTo<'_>) -> Result<Rewrite>

Rewrites the whole file under to through sqlcipher_export into a copy beside it, checks the copy, and renames it over the store. Writes wait on the writer for the whole rewrite; reads continue until the swap. A failure before the rename leaves the store as it was; one after it is a NotReopened.

Source

pub fn rewrite_space(&self) -> Result<VacuumSpace>

A rewrite needs room beside the store for a copy of its live pages, with a quarter more for the copy’s own page overhead.

Source

pub fn write_plaintext_copy(&self, dir: &Path) -> Result<PathBuf>

Writes a plaintext copy of the store into dir for a tool that reads the file directly, under a fresh name created with O_EXCL and mode 0600. The copy is a deliberate local decryption; the caller removes it with remove_plaintext_copies.

Source§

impl SqliteStore

Source

pub fn record_encrypted(&self, at_micros: i64, pause: Duration) -> Result<()>

Source

pub fn encryption_awaited_since(&self, now_micros: i64) -> Result<i64>

When something first waited on this store’s encryption migration: now_micros on the first call, the stored time on every later one.

Source

pub fn encrypted_record(&self) -> Result<Option<EncryptedRecord>>

When the one-time migration rewrote this store under its key, and how long writes waited for it. None for a store created encrypted.

Source

pub fn revocations_after_store_key(&self) -> Result<Vec<String>>

Revocations of other devices, under this device’s identity, that arrived after the current store key was made. Each one means a device that left may hold K_id, which opens every older store.key.wrapped, so the store key must change.

Source

pub fn retirements_after_store_key(&self) -> Result<Vec<String>>

Fingerprints of recovery phrases retired, in a way that calls for rotation, after the current store key was made. A retired phrase yields K_id, which opens every older store.key.wrapped, whether or not this device trusted it.

Source§

impl SqliteStore

Source

pub fn advance_floor_uptime(&self, now_micros: i64) -> Result<()>

Source

pub fn record_peer_ack( &self, peer_id: &str, acked_seq: u64, now_micros: i64, ) -> Result<()>

Acks only rise; an ack that reaches a peer’s pin turns the pin into that ack.

Source

pub fn pin_peer(&self, peer_id: &str, seq: u64) -> Result<()>

A renewal restarts the pin’s day of uptime.

Source

pub fn record_peer_in(&self, peer_id: &str, seq: u64) -> Result<()>

The highest seq of peer_id’s log that its pushes into this store have applied; it only rises.

Source

pub fn peer_in(&self, peer_id: &str) -> Result<Option<u64>>

Source

pub fn push_report(&self, peer_id: &str) -> Result<Option<u64>>

The position on this store’s log that peer_id last reported in a push handshake.

Source

pub fn record_push_report( &self, peer_id: &str, reported: Option<u64>, ) -> Result<()>

Replaces the last report, a value that may fall: a lower one means the peer was restored.

Source

pub fn peer_acks(&self) -> Result<Vec<PeerAck>>

Source

pub fn log_floor(&self) -> Result<u64>

Source

pub fn log_head(&self) -> Result<u64>

The highest seq this store has minted.

Source

pub fn floor(&self) -> Result<u64>

F: the lowest ack or pin over live neighbours, or the head when there are none.

Source

pub fn floor_health(&self, now_micros: i64) -> Result<FloorHealth>

Source

pub fn floor_pass_due(&self, now_micros: i64) -> Result<bool>

Source

pub fn run_floor_pass( &self, truncation: Truncation, now_micros: i64, ) -> Result<FloorPass>

Source

pub fn seed_floor_neighbours(&self, peers: &[String]) -> Result<usize>

Seeds peers as neighbours that have not acked yet, so a device that just bootstrapped keeps the log tail it carries until they ack it.

Source

pub fn reset_floor_for_new_identity( &self, identity_tag: &IdentityTag, ) -> Result<()>

Drops every neighbour, ack and pin of a device outside identity_tag, so a store that founded a new identity doesn’t hold its floor for the old one’s devices.

Source§

impl SqliteStore

Source

pub fn ingested_sessions_since( &self, source: &str, since: DateTime<Utc>, ) -> Result<Vec<IngestedSession>>

Distinct (session key, file) pairs of keyed source files ingested at or after since, ordered by session key then file.

Source§

impl SqliteStore

Source

pub fn retire_note_covered_raws(&self) -> Result<usize>

Marks every undistilled Claude Code raw that a live agent note of its session covers, meaning the raw is timed no later than the session’s newest live note, then archives the retirable raws. A raw is timed by session_at, else created_at, and so is a note. Returns how many raws were marked.

Source

pub fn llm_distilled_since(&self, cutoff: DateTime<Utc>) -> Result<bool>

Whether an LLM distilled any raw at or after cutoff, on this machine or on a peer, since distillation markers replicate with their original time. Markers written by note cover are not LLM work, nor are the legacy copies repair gives their exact-text twins, which keep the agent-note model.

Source§

impl SqliteStore

Source

pub fn notes_for_session( &self, session_id: &str, kind: Option<NoteKind>, ) -> Result<Vec<(Memory, AgentNote)>>

Live agent notes filed under session_id, oldest first, each with its fields read back from the content. A row whose content no longer parses is skipped.

Source

pub fn latest_notes_for_project( &self, project: &str, kind: Option<NoteKind>, limit: usize, ) -> Result<Vec<(Memory, AgentNote)>>

At most limit live agent notes filed under project, newest first, with the kind filter applied before the limit. A row whose content no longer parses is skipped.

Source

pub fn notes_created_between( &self, since: DateTime<Utc>, until: DateTime<Utc>, project_path: Option<&str>, ) -> Result<Vec<(Memory, AgentNote)>>

Live agent notes created in [since, until), newest first, in one project when project_path names it. Rows that no longer parse are skipped.

Source

pub fn noted_session_ids(&self) -> Result<HashSet<String>>

Every session id any live agent note is filed under.

Source§

impl SqliteStore

Source

pub fn archived_raws_since(&self, since: DateTime<Utc>) -> Result<ArchivedRaws>

Archived raw memories split by whether their session time (their creation time when the session is unknown) is at or after since. These are the transcript chunks consolidation took out of recall, which the raw fallback index keeps searchable; retired and superseded rows are left out.

Source

pub fn is_archived_raw(&self, id: Uuid) -> Result<bool>

Whether id is still an archived raw the fallback may answer from.

Source

pub fn raw_memory_state(&self, id: Uuid) -> Result<RawMemoryState>

Whether the raw memory id still stands behind the transcript passage it was cut from. Forgetting, deleting or correcting it withdraws the passage; a supersede onto a byte-identical row, which is how maintenance collapses duplicates, does not, unless the row it collapsed onto is withdrawn in turn.

Source

pub fn restore_archived_embedding( &self, id: Uuid, embedding: &[f32], ) -> Result<bool>

Restores the vector of an archived row that lost it, which happens when a redacted row arrives through sync. It is derived from the content on this device, so it is written locally: no outbox, no sync log, and updated_at is left alone so it cannot win a merge.

Source§

impl SqliteStore

Source

pub fn rewrite_memory_content( &self, memory_id: Uuid, rewrite: &dyn Fn(&str) -> String, embed: Option<EmbedBatchFn<'_>>, ) -> Result<RewriteOutcome>

Rewrites a memory’s content, context and retirement reason through rewrite, whatever its tier. Vectors are computed before the write transaction, so the writer lock is never held across model inference; the transaction then stores the text, the memory vector, a fresh chunk set (live rows only), an index upsert and a sync Update carrying the new row. Without embed, or when embedding fails, the memory vector and chunks are dropped and the repair loop’s backfills restore them. A row outside the live set that had no vector gets none. A change to the reason alone keeps the vectors and chunks, which are not derived from it. A cold row is an error, since its row holds only a stub.

Source§

impl SqliteStore

Source

pub fn local_author(&self) -> Result<String>

The author a write records when its writer names none: this store’s identity, or its device until it joins one.

Source

pub fn get_memory_for( &self, id: Uuid, principal: &Principal, ) -> Result<Option<Memory>>

None when no such memory exists or principal may not read it, so a caller cannot tell another user’s private memory from a missing one.

Source

pub fn check_change(&self, id: Uuid, principal: &Principal) -> Result<()>

Fails with AccessDenied when the memory exists and principal may not change it; an absent memory passes, leaving the change itself to report it missing.

Source

pub fn set_memory_scope( &self, id: Uuid, scope: Scope, principal: &Principal, ) -> Result<bool>

Sets who may read the memory, logged as an Update so every peer converges on it. False when no such memory exists.

Source

pub fn export_memories_for( &self, principal: &Principal, sink: &mut dyn FnMut(Memory) -> Result<()>, ) -> Result<()>

Streams every memory principal may read, without embeddings; a cold memory carries its full text read back from its segment.

Source§

impl SqliteStore

Source

pub fn claude_code_sessions_between( &self, since: DateTime<Utc>, until: DateTime<Utc>, project_path: Option<&str>, ) -> Result<Vec<ActiveSession>>

Main-agent Claude Code sessions with a transcript line timed in [since, until), read from the raw chunks’ session_at, so archived and consolidated chunks still count. Subagent transcripts are left out because no note is ever asked of them.

Source

pub fn memories_from_session( &self, ids: &[Uuid], session: &str, ) -> Result<HashSet<Uuid>>

Which of ids came from Claude Code session session by session_of: a raw of its transcript or of a subagent’s beneath it, or a note filed under its id.

Source§

impl SqliteStore

Source

pub fn note_failed_record( &self, source: &str, record: &SyncRecord, error: &Error, now_micros: i64, ) -> Result<()>

Keeps a record from source’s log that failed to apply, or counts one more attempt on one already kept. A content-deterministic failure settles after SETTLE_AFTER_ATTEMPTS attempts spread over a day; a transient one never does.

Source

pub fn retry_failed_records( &self, source: &str, now_micros: i64, ) -> Result<RetryCounts>

Applies again every unsettled record kept from source; one that now applies or settles stops holding source’s ack back.

Source

pub fn applied_through(&self, source: &str, reached: u64) -> Result<u64>

The highest seq of source’s log applied here without a gap: below the first record from it that still fails and has not settled.

Source

pub fn failed_records(&self) -> Result<Vec<FailedRecord>>

Source

pub fn failed_record_counts(&self) -> Result<FailedCounts>

Source

pub fn replay_settled_records( &self, build: &str, ) -> Result<Option<ReplayCounts>>

Applies every settled record again once per build, so a bug fixed in a later binary or schema recovers records whose source log may no longer hold them. A record that now lands is logged as a relay under replay:<id>:<n>, numbered densely, so it reaches this store’s pullers. Each page of records is its own transaction and the build is marked done after the last page, so a retry after a crash replays what is left under a new id.

Source

pub fn prune_settled_records(&self, now_micros: i64) -> Result<PruneCounts>

Drops settled records that a newer build could not recover once they are 90 days old, and the oldest past SETTLED_CAP.

Source§

impl SqliteStore

Source

pub fn export_snapshot( &self, options: SnapshotOptions, sink: &mut dyn FnMut(SnapshotRow) -> Result<()>, ) -> Result<SnapshotState>

Streams every row a peer needs to converge with this store, from one read transaction, into sink. A caller that pins the reader does so at log_head before calling, which is at or below the seq read here, so no row above the pin is truncated meanwhile.

Source

pub fn snapshot_merge_progress(&self) -> Result<Option<MergeProgress>>

Source

pub fn begin_snapshot_merge( &self, snapshot_id: &str, exporter: &str, seq: u64, ) -> Result<SnapshotMerge<'_>>

Starts a merge of snapshot snapshot_id, or resumes the one already in progress for it. A merge of another snapshot is abandoned: rows it applied stay, since merging never removes anything.

Source

pub fn abandon_snapshot_merge(&self) -> Result<()>

Source§

impl SqliteStore

Source

pub fn redaction_phase(&self) -> Result<RedactionPhase>

Source

pub fn run_stored_redaction( &self, embed: Option<EmbedBatchFn<'_>>, ) -> Result<RedactionRun>

The one-time pass that redacts content stored before redaction existed: every memory row whatever its tier (through rewrite_memory_content, so peers receive the redacted row), chunk sets the rewrite did not replace, entity names and aliases, relation context, and the payloads already in sync_log. Progress is kept in index_meta, not a schema migration, so peers on the same binary still agree on the schema while their passes run. A row that cannot be parsed has its raw text redacted in place and is recorded (see RedactionReport::skipped) before the cursor moves past it; any other error, including one from rewriting a row that parsed, fails the run. Returns immediately when the pass has already finished on this store.

Source

pub fn stored_redaction_report(&self) -> Result<RedactionReport>

Counts what the pass would redact, by location and kind, without writing anything. Text is read leniently (invalid UTF-8 replaced, a blob read as text), as the pass reads rows it skips.

Source§

impl SqliteStore

Source§

impl SqliteStore

Source

pub fn restore_explicit_memories(&self) -> Result<ExplicitRestoreStats>

Undoes what consolidation did to explicit memories before EXPLICIT_SOURCES was excluded from it: every archived explicit raw returns to the hot tier and the indexes, every marker on an explicit raw is dropped so no later pass retires it again, and a distillate whose raws were all explicit is superseded by the oldest of them, since nothing it summarised is missing from recall any more. A distillate that also stands in for transcript raws stays live. Every change replicates the way consolidation’s own writes do.

Source

pub fn refile_nested_transcript_projects(&self) -> Result<u64>

Re-files memories that an older adapter filed under the directory holding the transcript file (subagents, agent-transcripts) rather than the project directory under the projects root. Raws re-file from their own source_id; a distillate re-files to the project most of its raws now carry, because a consolidation batch is grouped by conversation and one project is the only value that fits a row with a single project_path, or to the project in its own source_id when no raw is left to vote. Every changed row is replicated as an Update and queued for the indexes, exactly as update_memory does.

Source

pub fn refile_stray_project_slugs(&self) -> Result<StraySlugReport>

Re-files explicit memories that an earlier remember filed under the bare directory name it was given (wardrobe) instead of the slug ingest stores for that directory (-mnt-data-wardrobe). A bare name resolves to the stored slugs ending in -<name>, with name encoded the way Claude Code encodes a path; the two spellings Claude Code and Cursor give one directory (with and without the leading -) count as one project. A name matching more than one project is reported, never guessed, and transcript rows under a bare slug are left alone because an adapter filed them there on purpose.

Source

pub fn refile_memory(&self, id: Uuid, project_path: &str) -> Result<bool>

Moves one memory to project_path, replicated and re-indexed as every refile is. False when it was already filed there or is absent.

Source§

impl SqliteStore

Source§

impl SqliteStore

Source

pub fn open(path: &Path, machine_id: String) -> Result<Self>

Source

pub fn open_with_pool_size( path: &Path, machine_id: String, reader_pool_size: usize, ) -> Result<Self>

Same as open, with an explicit reader pool size instead of the parallelism-derived default. reader_pool_size is clamped to at least 1 so a caller passing 0 still gets a working store.

Source

pub fn open_with_key( path: &Path, machine_id: String, key: &StoreKey, ) -> Result<Self>

Opens or creates a store encrypted under key; a new file is encrypted from its first page. Fails when the file is plaintext or under another key, and in a build without the encryption feature.

Source

pub fn open_with_pool_size_keyed( path: &Path, machine_id: String, reader_pool_size: usize, key: &StoreKey, ) -> Result<Self>

Source

pub fn in_memory(machine_id: String) -> Result<Self>

Source

pub fn with_wall_clock(self, clock: Arc<dyn WallClock>) -> Self

Replaces the default SystemClock that every seq is minted from.

Source

pub fn write_count(&self) -> u64

Rows changed through the writer since the store opened; unchanged across an interval means nothing was written in it.

Source

pub fn machine_id(&self) -> &str

Source

pub fn raise_clock_floors(&self, floors: &ClockFloors) -> Result<()>

Raises the clock’s floors from a peer’s report and never lowers one. Fails with ClockFloorRejected, raising none, when a value lies past HLC_CEILING_MICROS (for held_origin_seq, when its incarnation does).

Source

pub fn highest_origin_seq(&self, origin: &str) -> Result<Option<i64>>

The highest origin_seq from origin this store holds, in its live log or among the keys kept after the log rows went.

Source

pub fn clock_state(&self) -> Result<ClockState>

Source

pub fn tombstone_exists(&self, id: Uuid) -> Result<bool>

Source

pub fn memory_content_exists(&self, content_prefix: &str) -> Result<bool>

Source

pub fn max_sync_seq(&self) -> Result<u64>

Source

pub fn schema_version(&self) -> Result<i64>

The highest migration applied to the database file, which exceeds SCHEMA_VERSION when a newer binary migrated it.

Source

pub fn get_conversation(&self, id: Uuid) -> Result<Option<Conversation>>

Source§

impl SqliteStore

Source

pub fn store_memory_if_absent(&self, memory: &Memory) -> Result<bool>

Stores memory only when its id is neither a row (in any state) nor tombstoned, checked in the writer transaction so two writers of the same id cannot both insert; insert_memory would otherwise replace a retired row. True when it wrote.

Source§

impl SqliteStore

Source

pub fn index_watermark(&self) -> Result<i64>

Source

pub fn set_index_watermark(&self, seq: i64) -> Result<()>

Source

pub fn outbox_after(&self, seq: i64) -> Result<Vec<OutboxRow>>

Source

pub fn prune_outbox(&self, upto: i64) -> Result<()>

Source

pub fn max_outbox_seq(&self) -> Result<i64>

Source

pub fn vec_key(&self, memory_id: Uuid) -> Result<u64>

Stable u64 usearch key for a memory id, allocated on first use.

Source

pub fn existing_vec_key(&self, memory_id: Uuid) -> Result<Option<u64>>

Source

pub fn delete_vec_key(&self, memory_id: Uuid) -> Result<()>

Drops the key row once the vector is gone from usearch, so vec_keys only ever holds keys the index can still resolve.

Source

pub fn find_exact_duplicate( &self, content: &str, project_path: Option<&str>, ) -> Result<Option<Memory>>

The live memory content duplicates under remember’s rule: equal once surrounding whitespace is trimmed and both sides are lowercased. With project_path, only memories filed under that project count; among several matches the oldest wins. SQLite’s lower() folds ASCII only, so a stored memory that differs from content just in the case of a non-ASCII letter is not found, which the substring shortlist this replaces could not find either.

Source

pub fn find_exact_duplicate_outside_source( &self, content: &str, project_path: Option<&str>, source: &str, ) -> Result<Option<Memory>>

As find_exact_duplicate, but never returns a row of source, for a source whose rows belong to something else (vault notes follow their files) and so cannot stand in for a new memory.

bm25_search’s match with a source/type/project scope inside the query, so limit counts only rows the scope admits instead of cutting before a caller’s filter could run. The match is a substring test over content on hot, unsuperseded rows, with no relevance score to order by (tantivy, not SQLite, holds the text index), so matches come back newest first. Case folding is SQLite’s lower(), which is ASCII only.

Source

pub fn scoped_count( &self, source: Option<&str>, memory_type: Option<MemoryType>, project_path: Option<&str>, ) -> Result<usize>

How many live memories a source/type/project scope holds. Answered by COUNT off the source, memory_type, and project_path indexes, so it stays O(1) in memory however large the scope grows.

Source

pub fn retain_in_scope( &self, ids: &[Uuid], source: Option<&str>, memory_type: Option<MemoryType>, project_path: Option<&str>, ) -> Result<HashSet<Uuid>>

Which of ids fall inside a source/type/project scope.

Asks the database about the candidates in hand rather than reading the scope out to test membership locally, so the cost tracks the caller’s list and not the scope. Batched because SQLite caps how many bound parameters one statement may carry.

Source

pub fn memory_id_for_vec_key(&self, key: u64) -> Result<Option<Uuid>>

Source

pub fn memory_for_indexing( &self, memory_id: Uuid, ) -> Result<Option<IndexableMemory>>

Source

pub fn all_indexable(&self) -> Result<Vec<(Uuid, Vec<f32>, String)>>

Source

pub fn indexable_page( &self, after: i64, limit: usize, ) -> Result<Vec<(i64, Uuid, Vec<f32>, String)>>

Up to limit hot rows after rowid after, in rowid order, so a rebuild never holds more than one page and indexes rows in the order they were written, which is how ties between equal vectors fall.

Source

pub fn outbox_page(&self, seq: i64, limit: usize) -> Result<Vec<OutboxRow>>

Source

pub fn vector_indexable_count(&self, dim: usize) -> Result<usize>

Hot rows whose stored embedding has exactly dim components: the only rows the vector index can hold, so a row still waiting for an embedding does not read as index drift.

Source

pub fn indexable_count(&self) -> Result<usize>

Source

pub fn supersede_memory(&self, memory_id: Uuid, by: Uuid) -> Result<()>

Zep-style soft invalidation: the row stays, the indexes drop it via the outbox, and peers learn of it through a sync Update carrying the row.

Source

pub fn archive_memory(&self, memory_id: Uuid) -> Result<()>

Archive tier: the row keeps its content and embedding but leaves both retrieval indexes and every recall predicate. archived_segment stays NULL until the cold pass moves an old raw into a segment. Replicated as an Update so peers retire the same row instead of keeping it hot beside the distillate.

Source

pub fn retire_memory( &self, memory_id: Uuid, reason: &str, ) -> Result<RetireOutcome>

forget’s default: archives the row as archive_memory does and records why, so the memory leaves recall on every peer while its content stays recoverable. Nothing is tombstoned, and an id this store does not hold changes nothing.

Source

pub fn mark_archived(&self, memory_id: Uuid, segment: &str) -> Result<()>

Cold-tier flip: content becomes a 200-char stub, the embedding is dropped (the full row lives in the archived segment), and the outbox evicts the memory from both indexes.

Trait Implementations§

Source§

impl Store for SqliteStore

Source§

fn traverse( &self, start_id: Uuid, depth: u32, ) -> Result<Vec<(Memory, Vec<Relation>)>>

Live memories within depth hops of start_id, newest first, capped at TRAVERSE_MAX_RESULTS. Ingest writes no memory-to-memory edges (only mentions, derived_from and distilled_from), so a hop is any of: a memory-to-memory relation in either direction, a shared mentioned entity, or a shared conversation.

Source§

fn store_memory(&self, memory: &Memory) -> Result<()>

Source§

fn get_memory(&self, id: Uuid) -> Result<Option<Memory>>

Source§

fn get_memory_with_embedding(&self, id: Uuid) -> Result<Option<Memory>>

Source§

fn delete_memory(&self, id: Uuid) -> Result<()>

Source§

fn remove_memory(&self, id: Uuid) -> Result<bool>

delete_memory that also reports whether a row was there to remove, so a forget of an unknown id can be answered as such instead of silently succeeding.
Source§

fn store_entity(&self, entity: &Entity) -> Result<()>

Source§

fn get_entity(&self, id: Uuid) -> Result<Option<Entity>>

Source§

fn find_entity_by_name(&self, name: &str) -> Result<Option<Entity>>

Source§

fn store_conversation(&self, conversation: &Conversation) -> Result<()>

Source§

fn store_relation(&self, relation: &Relation) -> Result<()>

Source§

fn get_relations( &self, node_id: Uuid, relation_type: Option<RelationType>, ) -> Result<Vec<Relation>>

Cosine search restricted to memories matching source, memory_type, and/or project_path, scanned exhaustively over that subset. Read more
Source§

fn memories_by_source(&self, source: &str) -> Result<Vec<Memory>>

Source§

fn source_admits(&self, memory: &Memory, source: &str) -> Result<bool>

Whether memory answers a recall scoped to source. Consolidation leaves the raws archived and the distillate tagged ‘consolidation’, so a store that tracks distillation must let a distillate answer for the sources of its raws, or a source filter finds nothing once its raws are consolidated. The default is the plain column comparison for stores that never consolidate.
Source§

fn memories_by_type(&self, memory_type: MemoryType) -> Result<Vec<Memory>>

Source§

fn chunk_window(&self, memory_id: Uuid, ord: usize) -> Result<Option<String>>

The text around a memory’s chunk ord: that chunk joined with one neighbour on each side. None for a store that keeps no chunks.
Source§

fn update_memory(&self, memory: &Memory) -> Result<()>

Source§

fn record_access(&self, memory: &Memory) -> Result<()>

Source§

fn record_accesses(&self, ids: &[Uuid]) -> Result<()>

Count one access on each of ids and stamp them with the current time; confidence is left alone. A recall touches every result it returns, so an implementation should batch these into one write rather than take the writer lock per row. Ids the store does not hold are ignored.
Source§

fn memory_count(&self) -> Result<usize>

Source§

fn all_memory_ids(&self) -> Result<Vec<Uuid>>

Source§

fn all_relations(&self) -> Result<Vec<Relation>>

Source§

impl StoreAdmin for SqliteStore

Source§

fn session_deleted(&self, path: &str) -> Result<bool>

Whether the Claude Code session the transcript at path belongs to was deleted, so nothing of it may be captured again. The default knows of no deletes.
Source§

fn register_machine(&self, identity: &MachineIdentity) -> Result<()>

Source§

fn get_all_machines(&self) -> Result<HashMap<String, String>>

Source§

fn diagnostic(&self) -> Result<String>

Source§

fn find_or_create_entity(&self, name: &str, entity_type: &str) -> Result<Entity>

Source§

fn memory_content_exists(&self, content_prefix: &str) -> Result<bool>

Source§

fn store_memory_if_absent(&self, memory: &Memory) -> Result<bool>

Writes memory only when no memory holds its id in any state and the id is not tombstoned, checked in the insert’s own transaction so two concurrent writers cannot both pass. True when it wrote. The default cannot check and always writes.
Source§

fn ingest_memory_if_absent(&self, memory: &Memory) -> Result<bool>

store_memory_if_absent for ingest, which may also refuse text a delete removed from the same source. The default refuses nothing more.
Source§

fn ingest_blocked(&self, memory: &Memory) -> Result<bool>

Whether ingest must skip memory because a delete removed its text from its source. A read outside any write, so it only spares work ahead of ingest_memory_if_absent. The default knows of no deletes.
Source§

fn store_conversation_if_absent( &self, conversation: &Conversation, ) -> Result<bool>

store_memory_if_absent for conversations: an existing row keeps its fields and nothing is logged for sync.
Source§

fn memory_id_taken(&self, id: Uuid) -> Result<bool>

Whether a memory holds id in any state or the id is tombstoned. A read outside any write, so it can only spare work ahead of store_memory_if_absent, never replace it. The default knows nothing.
Source§

fn is_file_changed(&self, file_path: &str, file_hash: &str) -> Result<bool>

Source§

fn mark_ingested( &self, file_path: &str, file_hash: &str, memory_count: usize, source: &str, ) -> Result<()>

Source§

fn ingested_file_count(&self) -> Result<usize>

Source§

fn clear_ingest_log(&self) -> Result<usize>

Source§

fn ingest_log_entry(&self, file_path: &str) -> Result<Option<IngestLogEntry>>

Source§

fn ingested_session( &self, session_key: &str, content_hash: &str, ) -> Result<Option<IngestLogEntry>>

Any row, at any path, recording a file with this session key and content hash.
Source§

fn legacy_ingest_entries_with_stem( &self, stem: &str, ) -> Result<Vec<IngestLogEntry>>

Legacy rows still holding an old-format hash whose file stem is exactly stem.
Source§

fn upgrade_ingest_hash( &self, file_path: &str, content_hash: &str, session_key: Option<&str>, ) -> Result<()>

Replaces a row’s hash with its SHA-256 and session key in place, keeping its legacy flag.
Source§

fn record_ingested(&self, record: &IngestRecord<'_>) -> Result<()>

Source§

fn clear_reparseable_ingest_log(&self) -> Result<usize>

Clears the rows a reingest may re-parse, keeping legacy rows: their raws deleted under random ids would come back under deterministic ones.
Source§

fn delete_memories_by_source(&self, source: &str) -> Result<usize>

Source§

fn delete_memories_by_source_id(&self, source_id: &str) -> Result<usize>

Source§

fn rebuild_vector_index(&self) -> Result<()>

Source§

fn backfill_project_paths(&self) -> Result<u64>

Source§

fn backfill_sync_log(&self) -> Result<u64>

Source§

fn backfill_tombstones(&self) -> Result<u64>

Source§

fn prune_dangling_relations(&self) -> Result<u64>

Source§

fn compact_sync_log( &self, ops: &[SyncOp], dry_run: bool, ) -> Result<CompactStats>

Collapses re-logged duplicate sync_log rows down to one per (node_type, node_id, op), keeping the highest local_seq in each group.
Source§

fn memories_by_project_path(&self, project_path: &str) -> Result<Vec<Memory>>

Source§

fn distinct_project_paths(&self) -> Result<Vec<(String, u64)>>

Every distinct non-null project_path value with its live memory count, so a caller can discover the real slug recall’s project scope expects instead of guessing.
Source§

fn all_memories_with_embeddings(&self) -> Result<Vec<Memory>>

Source§

fn consolidation_stats(&self) -> Result<ConsolidationStats>

Source§

fn unconsolidated_memories(&self, limit: usize) -> Result<Vec<Memory>>

Hot, unsuperseded raws with no distillation_log row, ordered by (source_id, created_at) so a batch reads as conversations.
Source§

fn mark_consolidated( &self, raw_id: Uuid, distilled_id: Uuid, model: &str, ) -> Result<()>

Source§

fn retire_consolidated_raws(&self) -> Result<usize>

Moves every distilled or skip-marked raw that is still hot to the archive tier. Nothing is deleted: the raw keeps its content and embedding and only leaves recall, so a distillate can always be traced back to what it was distilled from.
Source§

fn memories_created_between( &self, start: &DateTime<Utc>, end: &DateTime<Utc>, ) -> Result<Vec<Memory>>

Source§

fn sync_log_since(&self, after_seq: u64) -> Result<Vec<SyncEntry>>

Source§

fn sync_log_page( &self, after_seq: u64, limit: Option<usize>, ) -> Result<Vec<SyncEntry>>

Source§

fn sync_log_export_page( &self, after_seq: u64, limit: Option<usize>, ) -> Result<Vec<SyncEntry>>

A page to send a peer. A store that truncates its log refuses a page that starts below its floor.
Source§

fn max_sync_seq(&self) -> Result<u64>

Source§

fn get_sync_state(&self, peer_id: &str) -> Result<Option<SyncState>>

Source§

fn set_sync_state(&self, state: &SyncState) -> Result<()>

Source§

fn get_all_sync_states(&self) -> Result<Vec<SyncState>>

Source§

fn tombstone_exists(&self, id: Uuid) -> Result<bool>

Source§

fn apply_record(&self, record: &SyncRecord) -> Result<ApplyOutcome>

Applies one remote sync record with LWW/tombstone conflict semantics, in its own transaction.
Source§

fn apply_batch(&self, records: &[SyncRecord]) -> Result<Vec<ApplyOutcome>>

Applies a batch atomically: either every record commits or none do, so a caller can fall back to per-record apply_record on error without double-applying a prefix.

Auto Trait Implementations§

Blanket Implementations§

Source§

impl<T> Any for T
where T: 'static + ?Sized,

Source§

fn type_id(&self) -> TypeId

Gets the TypeId of self. Read more
Source§

impl<T> Borrow<T> for T
where T: ?Sized,

Source§

fn borrow(&self) -> &T

Immutably borrows from an owned value. Read more
Source§

impl<T> BorrowMut<T> for T
where T: ?Sized,

Source§

fn borrow_mut(&mut self) -> &mut T

Mutably borrows from an owned value. Read more
Source§

impl<T> Downcast for T
where T: Any,

Source§

fn into_any(self: Box<T>) -> Box<dyn Any>

Converts Box<dyn Trait> (where Trait: Downcast) to Box<dyn Any>, which can then be downcast into Box<dyn ConcreteType> where ConcreteType implements Trait.
Source§

fn into_any_rc(self: Rc<T>) -> Rc<dyn Any>

Converts Rc<Trait> (where Trait: Downcast) to Rc<Any>, which can then be further downcast into Rc<ConcreteType> where ConcreteType implements Trait.
Source§

fn as_any(&self) -> &(dyn Any + 'static)

Converts &Trait (where Trait: Downcast) to &Any. This is needed since Rust cannot generate &Any’s vtable from &Trait’s.
Source§

fn as_any_mut(&mut self) -> &mut (dyn Any + 'static)

Converts &mut Trait (where Trait: Downcast) to &Any. This is needed since Rust cannot generate &mut Any’s vtable from &mut Trait’s.
Source§

impl<T> DowncastSend for T
where T: Any + Send,

Source§

fn into_any_send(self: Box<T>) -> Box<dyn Any + Send>

Converts Box<Trait> (where Trait: DowncastSend) to Box<dyn Any + Send>, which can then be downcast into Box<ConcreteType> where ConcreteType implements Trait.
Source§

impl<T> DowncastSync for T
where T: Any + Send + Sync,

Source§

fn into_any_sync(self: Box<T>) -> Box<dyn Any + Send + Sync>

Converts Box<Trait> (where Trait: DowncastSync) to Box<dyn Any + Send + Sync>, which can then be downcast into Box<ConcreteType> where ConcreteType implements Trait.
Source§

fn into_any_arc(self: Arc<T>) -> Arc<dyn Any + Send + Sync> ⓘ

Converts Arc<Trait> (where Trait: DowncastSync) to Arc<Any>, which can then be downcast into Arc<ConcreteType> where ConcreteType implements Trait.
Source§

impl<T> From<T> for T

Source§

fn from(t: T) -> T

Returns the argument unchanged.

Source§

impl<T> Fruit for T
where T: Send + Downcast,

Source§

impl<T> Instrument for T

Source§

fn instrument(self, span: Span) -> Instrumented<Self> ⓘ

Instruments this type with the provided Span, returning an Instrumented wrapper. Read more
Source§

fn in_current_span(self) -> Instrumented<Self> ⓘ

Instruments this type with the current Span, returning an Instrumented wrapper. Read more
Source§

impl<T, U> Into<U> for T
where U: From<T>,

Source§

fn into(self) -> U

Calls U::from(self).

That is, this conversion is whatever the implementation of From<T> for U chooses to do.

Source§

impl<T> IntoEither for T

Source§

fn into_either(self, into_left: bool) -> Either<Self, Self> ⓘ

Converts self into a Left variant of Either<Self, Self> if into_left is true. Converts self into a Right variant of Either<Self, Self> otherwise. Read more
Source§

fn into_either_with<F>(self, into_left: F) -> Either<Self, Self> ⓘ
where F: FnOnce(&Self) -> bool,

Converts self into a Left variant of Either<Self, Self> if into_left(&self) returns true. Converts self into a Right variant of Either<Self, Self> otherwise. Read more
Source§

impl<T> Pointable for T

Source§

const ALIGN: usize

The alignment of pointer.
Source§

type Init = T

The type for initializers.
Source§

unsafe fn init(init: <T as Pointable>::Init) -> usize

Initializes a with the given initializer. Read more
Source§

unsafe fn deref<'a>(ptr: usize) -> &'a T

Dereferences the given pointer. Read more
Source§

unsafe fn deref_mut<'a>(ptr: usize) -> &'a mut T

Mutably dereferences the given pointer. Read more
Source§

unsafe fn drop(ptr: usize)

Drops the object pointed to by the given pointer. Read more
Source§

impl<T> Same for T

Source§

type Output = T

Should always be Self
Source§

impl<T, U> TryFrom<U> for T
where U: Into<T>,

Source§

type Error = !

The type returned in the event of a conversion error.
Source§

fn try_from(value: U) -> Result<T, !>

Performs the conversion.
Source§

impl<T, U> TryInto<U> for T
where U: TryFrom<T>,

Source§

type Error = <U as TryFrom<T>>::Error

The type returned in the event of a conversion error.
Source§

fn try_into(self) -> Result<U, <U as TryFrom<T>>::Error>

Performs the conversion.
Source§

impl<V, T> VZip<V> for T
where V: MultiLane<T>,

Source§

fn vzip(self) -> V

Source§

impl<T> WithSubscriber for T

Source§

fn with_subscriber<S>(self, subscriber: S) -> WithDispatch<Self> ⓘ
where S: Into<Dispatch>,

Attaches the provided Subscriber to this type, returning a WithDispatch wrapper. Read more
Source§

fn with_current_subscriber(self) -> WithDispatch<Self> ⓘ

Attaches the current default Subscriber to this type, returning a WithDispatch wrapper. Read more