fzf/src/merger.go
Junegunn Choi 2fe1e28220 Improvements in performance and memory usage
I profiled fzf and it turned out that it was spending significant amount
of time repeatedly converting character arrays into Unicode codepoints.
This commit greatly improves search performance after the initial scan
by memoizing the converted results.

This commit also addresses the problem of unbounded memory usage of fzf.
fzf is a short-lived process that usually processes small input, so it
was implemented to cache the intermediate results very aggressively with
no notion of cache expiration/eviction. I still think a proper
implementation of caching scheme is definitely an overkill. Instead this
commit introduces limits to the maximum size (or minimum selectivity) of
the intermediate results that can be cached.
2015-04-17 22:23:52 +09:00

119 lines
2.4 KiB
Go

package fzf
import "fmt"
// Merger with no data
var EmptyMerger = NewMerger([][]*Item{}, false, false)
// Merger holds a set of locally sorted lists of items and provides the view of
// a single, globally-sorted list
type Merger struct {
lists [][]*Item
merged []*Item
chunks *[]*Chunk
cursors []int
sorted bool
tac bool
final bool
count int
}
// PassMerger returns a new Merger that simply returns the items in the
// original order
func PassMerger(chunks *[]*Chunk, tac bool) *Merger {
mg := Merger{
chunks: chunks,
tac: tac,
count: 0}
for _, chunk := range *mg.chunks {
mg.count += len(*chunk)
}
return &mg
}
// NewMerger returns a new Merger
func NewMerger(lists [][]*Item, sorted bool, tac bool) *Merger {
mg := Merger{
lists: lists,
merged: []*Item{},
chunks: nil,
cursors: make([]int, len(lists)),
sorted: sorted,
tac: tac,
final: false,
count: 0}
for _, list := range mg.lists {
mg.count += len(list)
}
return &mg
}
// Length returns the number of items
func (mg *Merger) Length() int {
return mg.count
}
// Get returns the pointer to the Item object indexed by the given integer
func (mg *Merger) Get(idx int) *Item {
if mg.chunks != nil {
if mg.tac {
idx = mg.count - idx - 1
}
chunk := (*mg.chunks)[idx/chunkSize]
return (*chunk)[idx%chunkSize]
}
if mg.sorted {
return mg.mergedGet(idx)
}
if mg.tac {
idx = mg.count - idx - 1
}
for _, list := range mg.lists {
numItems := len(list)
if idx < numItems {
return list[idx]
}
idx -= numItems
}
panic(fmt.Sprintf("Index out of bounds (unsorted, %d/%d)", idx, mg.count))
}
func (mg *Merger) Cacheable() bool {
return mg.count < mergerCacheMax
}
func (mg *Merger) mergedGet(idx int) *Item {
for i := len(mg.merged); i <= idx; i++ {
minRank := Rank{0, 0, 0}
minIdx := -1
for listIdx, list := range mg.lists {
cursor := mg.cursors[listIdx]
if cursor < 0 || cursor == len(list) {
mg.cursors[listIdx] = -1
continue
}
if cursor >= 0 {
rank := list[cursor].Rank(false)
if minIdx < 0 || compareRanks(rank, minRank, mg.tac) {
minRank = rank
minIdx = listIdx
}
}
mg.cursors[listIdx] = cursor
}
if minIdx >= 0 {
chosen := mg.lists[minIdx]
mg.merged = append(mg.merged, chosen[mg.cursors[minIdx]])
mg.cursors[minIdx]++
} else {
panic(fmt.Sprintf("Index out of bounds (sorted, %d/%d)", i, mg.count))
}
}
return mg.merged[idx]
}