linkrs-storage 0.1.0

Storage engine with CSR and memory-mapped containers
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110

use super::SyncWrapper;
use crate::engine::graph_storage::AutoCommitBatchWindow;
use crate::StorageClient;
use linkrs_core::StorageResult;
use std::sync::Arc;

impl<S: StorageClient + crate::AutoCommitBatchOps> crate::AutoCommitBatchOps for SyncWrapper<S> {
    fn begin_auto_commit_batch(&self) -> StorageResult<Arc<AutoCommitBatchWindow>> {
        self.inner.begin_auto_commit_batch()
    }

    fn bind_auto_commit_statement(
        &self,
        window: &Arc<AutoCommitBatchWindow>,
    ) -> StorageResult<Self> {
        let inner = self.inner.bind_auto_commit_statement(window)?;
        Ok(SyncWrapper {
            inner,
            sync_manager: self.sync_manager.clone(),
            enabled: self.enabled,
            auto_commit_owner: self.auto_commit_owner,
        })
    }

    fn finalize_auto_commit_batch(&self, window: &AutoCommitBatchWindow) -> StorageResult<()> {
        self.inner.finalize_auto_commit_batch(window)
    }
}

impl<S: StorageClient + crate::AutoCommitGroupOps> crate::AutoCommitGroupOps for SyncWrapper<S> {
    fn begin_auto_commit_group(&self) -> StorageResult<Arc<AutoCommitBatchWindow>> {
        self.inner.begin_auto_commit_group()
    }

    fn finalize_auto_commit_group(&self, window: &AutoCommitBatchWindow) -> StorageResult<()> {
        self.inner.finalize_auto_commit_group(window)
    }
}

impl<S: StorageClient + linkrs_transaction::UndoTarget + 'static>
    linkrs_transaction::TransactionCommitSink for SyncWrapper<S>
{
    fn commit_transaction(
        &self,
        transaction_id: linkrs_core::types::TransactionId,
    ) -> Result<linkrs_core::types::CommitLsn, String> {
        self.commit_transaction_fact(transaction_id)
            .map_err(|error| error.to_string())
    }

    fn commit_transaction_with_descriptor(
        &self,
        descriptor: &linkrs_transaction::TransactionCommitDescriptor,
    ) -> Result<linkrs_core::types::CommitLsn, String> {
        self.commit_transaction_fact_with_durability(
            descriptor.transaction_id,
            descriptor.durability,
        )
        .map_err(|error| error.to_string())
    }

    fn finalize_commit(
        &self,
        descriptor: &linkrs_transaction::TransactionCommitDescriptor,
        commit_lsn: linkrs_core::types::CommitLsn,
    ) -> Result<(), String> {
        self.finalize_commit_fact(descriptor.transaction_id, commit_lsn)
            .map_err(|error| error.to_string())
    }

    fn abort_transaction(
        &self,
        transaction_id: linkrs_core::types::TransactionId,
    ) -> Result<(), String> {
        self.abort_transaction_fact(transaction_id)
            .map_err(|error| error.to_string())
    }

    fn abort_transaction_with_descriptor(
        &self,
        descriptor: &linkrs_transaction::TransactionAbortDescriptor,
    ) -> Result<(), String> {
        descriptor
            .context
            .execute_undo_logs(&self.inner)
            .map_err(|error| error.to_string())?;
        descriptor
            .context
            .clear_undo_logs()
            .map_err(|error| error.to_string())?;
        self.abort_transaction_fact(descriptor.transaction_id)
            .map_err(|error| error.to_string())
    }

    fn auto_checkpoint_if_needed(&self) -> Result<(), String> {
        use crate::StoragePersistenceOps;
        let _ = <Self as StoragePersistenceOps>::auto_checkpoint_if_needed(self)
            .map_err(|error| error.to_string())?;
        Ok(())
    }

    fn recover_unfinalized_commits(&self) -> Result<usize, String> {
        // Page-level redo belongs to storage WAL replay on open; fact-table
        // post-commit finalization (`finalize_commit_fact`) is idempotent, so
        // same-process re-drive happens through the manager's pending queue
        // (`recover_pending_finalization`). Nothing extra is durable here yet,
        // report zero rather than silently inheriting the default no-op.
        Ok(0)
    }
}