1
  2
  3
  4
  5
  6
  7
  8
  9
 10
 11
 12
 13
 14
 15
 16
 17
 18
 19
 20
 21
 22
 23
 24
 25
 26
 27
 28
 29
 30
 31
 32
 33
 34
 35
 36
 37
 38
 39
 40
 41
 42
 43
 44
 45
 46
 47
 48
 49
 50
 51
 52
 53
 54
 55
 56
 57
 58
 59
 60
 61
 62
 63
 64
 65
 66
 67
 68
 69
 70
 71
 72
 73
 74
 75
 76
 77
 78
 79
 80
 81
 82
 83
 84
 85
 86
 87
 88
 89
 90
 91
 92
 93
 94
 95
 96
 97
 98
 99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
// Copyright 2015-2020 Parity Technologies (UK) Ltd.
// This file is part of OpenEthereum.

// OpenEthereum is free software: you can redistribute it and/or modify
// it under the terms of the GNU General Public License as published by
// the Free Software Foundation, either version 3 of the License, or
// (at your option) any later version.

// OpenEthereum is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
// GNU General Public License for more details.

// You should have received a copy of the GNU General Public License
// along with OpenEthereum.  If not, see <http://www.gnu.org/licenses/>.

//! Disk-backed `HashDB` implementation.

use std::{io, sync::Arc};

use bytes::Bytes;
use ethcore_db::{DBTransaction, DBValue, KeyValueDB};
use ethereum_types::H256;
use hash_db::{AsHashDB, HashDB};
use keccak_hasher::KeccakHasher;
use std::collections::{BTreeMap, HashMap};

/// expose keys of a hashDB for debugging or tests (slow).
pub trait KeyedHashDB: HashDB<KeccakHasher, DBValue> {
    /// Primarily use for tests, highly inefficient.
    fn keys(&self) -> HashMap<H256, i32>;
}

/// Upcast to `KeyedHashDB`
pub trait AsKeyedHashDB: AsHashDB<KeccakHasher, DBValue> {
    /// Perform upcast to KeyedHashDB.
    fn as_keyed_hash_db(&self) -> &dyn KeyedHashDB;
}

/// A `HashDB` which can manage a short-term journal potentially containing many forks of mutually
/// exclusive actions.
pub trait JournalDB: KeyedHashDB {
    /// Return a copy of ourself, in a box.
    fn boxed_clone(&self) -> Box<dyn JournalDB>;

    /// Returns heap memory size used
    fn get_sizes(&self, sizes: &mut BTreeMap<String, usize>);

    /// Returns the size of journalled state in memory.
    /// This function has a considerable speed requirement --
    /// it must be fast enough to call several times per block imported.
    fn journal_size(&self) -> usize {
        0
    }

    /// Check if this database has any commits
    fn is_empty(&self) -> bool;

    /// Get the earliest era in the DB. None if there isn't yet any data in there.
    fn earliest_era(&self) -> Option<u64> {
        None
    }

    /// Get the latest era in the DB. None if there isn't yet any data in there.
    fn latest_era(&self) -> Option<u64>;

    /// Journal recent database operations as being associated with a given era and id.
    // TODO: give the overlay to this function so journaldbs don't manage the overlays themeselves.
    fn journal_under(&mut self, batch: &mut DBTransaction, now: u64, id: &H256) -> io::Result<u32>;

    /// Mark a given block as canonical, indicating that competing blocks' states may be pruned out.
    fn mark_canonical(&mut self, batch: &mut DBTransaction, era: u64, id: &H256)
        -> io::Result<u32>;

    /// Commit all queued insert and delete operations without affecting any journalling -- this requires that all insertions
    /// and deletions are indeed canonical and will likely lead to an invalid database if that assumption is violated.
    ///
    /// Any keys or values inserted or deleted must be completely independent of those affected
    /// by any previous `commit` operations. Essentially, this means that `inject` can be used
    /// either to restore a state to a fresh database, or to insert data which may only be journalled
    /// from this point onwards.
    fn inject(&mut self, batch: &mut DBTransaction) -> io::Result<u32>;

    /// Whether this database is pruned.
    fn is_pruned(&self) -> bool {
        true
    }

    /// Get backing database.
    fn backing(&self) -> &Arc<dyn KeyValueDB>;

    /// Clear internal strucutres. This should called after changes have been written
    /// to the backing strage
    fn flush(&self) {}

    /// Consolidate all the insertions and deletions in the given memory overlay.
    fn consolidate(&mut self, overlay: ::memory_db::MemoryDB<KeccakHasher, DBValue>);

    /// State data query
    fn state(&self, id: &H256) -> Option<Bytes>;

    /// Commit all changes in a single batch
    #[cfg(test)]
    fn commit_batch(&mut self, now: u64, id: &H256, end: Option<(u64, H256)>) -> io::Result<u32> {
        let mut batch = self.backing().transaction();
        let mut ops = self.journal_under(&mut batch, now, id)?;

        if let Some((end_era, canon_id)) = end {
            ops += self.mark_canonical(&mut batch, end_era, &canon_id)?;
        }

        let result = self.backing().write(batch).map(|_| ops).map_err(Into::into);
        self.flush();
        result
    }

    /// Inject all changes in a single batch.
    #[cfg(test)]
    fn inject_batch(&mut self) -> io::Result<u32> {
        let mut batch = self.backing().transaction();
        let res = self.inject(&mut batch)?;
        self.backing().write(batch).map(|_| res).map_err(Into::into)
    }
}