feat: new put implementation

This commit is contained in:
2026-04-04 00:13:50 +02:00
parent 322d71f0be
commit 05b633afca
7 changed files with 111 additions and 67 deletions

115
table.go
View File

@@ -45,30 +45,61 @@ func (t Table[K, V]) load() float64 {
return float64(t.Size()) / float64(t.TotalCapacity())
}
// resize clears all buckets, changes the sizes of them to a specific capacity,
// and fills them back up again. It is a helper function for [Table.grow] and
// [Table.shrink]; use them instead.
func (t *Table[K, V]) resize(capacity uint64) error {
entries := make([]entry[K, V], 0, t.Size())
for k, v := range t.Entries() {
entries = append(entries, entry[K, V]{k, v})
// insert attempts to put/update an entry in the table, without modifying the
// size of the table. Returns a displaced entry and 'homeless = true' if an
// entry could not be placed after exhausting evictions.
func (t *Table[K, V]) insert(entry Entry[K, V]) (displaced Entry[K, V], homeless bool) {
if t.bucketA.update(entry.key, entry.value) {
return
}
t.bucketA.resize(capacity)
t.bucketB.resize(capacity)
if t.bucketB.update(entry.key, entry.value) {
return
}
for _, entry := range entries {
if err := t.Put(entry.key, entry.value); err != nil {
return err
for range t.maxEvictions() {
if entry, homeless = t.bucketA.insert(entry); !homeless {
return
}
if entry, homeless = t.bucketB.insert(entry); !homeless {
return
}
}
return nil
return entry, true
}
// resized creates an empty copy of the table, with a new capacity for each
// bucket.
func (t Table[K, V]) resized(capacity uint64) Table[K, V] {
return Table[K, V]{
growthFactor: t.growthFactor,
minLoadFactor: t.minLoadFactor,
bucketA: t.bucketA.resized(capacity),
bucketB: t.bucketB.resized(capacity),
}
}
// resize creates a new [Table.resized] with 'capacity', inserts all items into
// the array, and replaces the current table. It is a helper function for
// [Table.grow] and [Table.shrink]; use them instead.
func (t *Table[K, V]) resize(capacity uint64) bool {
updated := t.resized(capacity)
for k, v := range t.Entries() {
if _, failed := updated.insert(Entry[K, V]{k, v}); failed {
return false
}
}
*t = updated
return true
}
// grow increases the table's capacity by the [Table.growthFactor]. If the
// capacity is 0, it increases it to 1.
func (t *Table[K, V]) grow() error {
func (t *Table[K, V]) grow() bool {
var newCapacity uint64
if t.TotalCapacity() == 0 {
@@ -82,7 +113,7 @@ func (t *Table[K, V]) grow() error {
// shrink reduces the table's capacity by the [Table.growthFactor]. It may
// reduce it down to 0.
func (t *Table[K, V]) shrink() error {
func (t *Table[K, V]) shrink() bool {
return t.resize(t.bucketA.capacity / t.growthFactor)
}
@@ -106,36 +137,38 @@ func (t Table[K, V]) Has(key K) (exists bool) {
return err == nil
}
// Put sets the value for a key. Returns error if its value cannot be set.
func (t *Table[K, V]) Put(key K, value V) (err error) {
if t.bucketA.update(key, value) {
return nil
}
// Put sets the value for a key. If it cannot be set, an error is returned,
// along with the last displaced entry.
//
// On failure, the returned entry and the current table contents together
// preserve all previously inserted entries and the attempted entry.
func (t *Table[K, V]) Put(key K, value V) (displaced Entry[K, V], err error) {
var (
entry = Entry[K, V]{key, value}
homeless bool
)
if t.bucketB.update(key, value) {
return nil
}
entry, eviction := entry[K, V]{key, value}, false
for range t.maxEvictions() {
if entry, eviction = t.bucketA.evict(entry); !eviction {
return nil
for range defaultGrowthLimit {
if entry, homeless = t.insert(entry); !homeless {
return
}
if entry, eviction = t.bucketB.evict(entry); !eviction {
return nil
// Both this and the growth limit are necessary: this catches bad hashes
// early when the table is sparse, while the latter catches cases where
// growing never helps.
if t.load() < t.minLoadFactor {
return entry, fmt.Errorf("bad hash: resize on load %d/%d", t.Size(), t.TotalCapacity())
}
// It is theoretically possible to have a table with a larger capacity
// that is valid. But this chance is astronomically small, so we ignore
// it in this implementation.
if grew := t.grow(); !grew {
return entry, fmt.Errorf("bad hash: could not redistribute entries into larger table")
}
}
if t.load() < t.minLoadFactor {
return fmt.Errorf("bad hash: resize on load %d/%d = %f", t.Size(), t.TotalCapacity(), t.load())
}
if err := t.grow(); err != nil {
return err
}
return t.Put(entry.key, entry.value)
return entry, fmt.Errorf("bad hash: could not place entry after %d resizes", defaultGrowthLimit)
}
// Drop removes a value for a key in the table. Returns an error if its value
@@ -145,7 +178,9 @@ func (t *Table[K, V]) Drop(key K) (err error) {
t.bucketB.drop(key)
if t.load() < t.minLoadFactor {
return t.shrink()
// The error is not handled here, because table-shrinking is an internal
// optimization.
t.shrink()
}
return nil