github.com/ledgerwatch/erigon-lib@v1.0.0/compress/parallel_compress.go

/*
   Copyright 2021 Erigon contributors

   Licensed under the Apache License, Version 2.0 (the "License");
   you may not use this file except in compliance with the License.
   You may obtain a copy of the License at

       http://www.apache.org/licenses/LICENSE-2.0

   Unless required by applicable law or agreed to in writing, software
   distributed under the License is distributed on an "AS IS" BASIS,
   WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
   See the License for the specific language governing permissions and
   limitations under the License.
*/

package compress

import (
	"bufio"
	"container/heap"
	"context"
	"encoding/binary"
	"errors"
	"fmt"
	"io"
	"os"
	"sync"
	"sync/atomic"
	"time"

	"github.com/ledgerwatch/erigon-lib/common"
	"github.com/ledgerwatch/erigon-lib/common/assert"
	"github.com/ledgerwatch/erigon-lib/etl"
	"github.com/ledgerwatch/erigon-lib/patricia"
	"github.com/ledgerwatch/erigon-lib/sais"
	"github.com/ledgerwatch/log/v3"
	"golang.org/x/exp/slices"
)

// MinPatternScore is minimum score (per superstring) required to consider including pattern into the dictionary
const MinPatternScore = 1024

func optimiseCluster(trace bool, input []byte, mf2 *patricia.MatchFinder2, output []byte, uncovered []int, patterns []int, cellRing *Ring, posMap map[uint64]uint64) ([]byte, []int, []int) {
	matches := mf2.FindLongestMatches(input)

	if len(matches) == 0 {
		output = append(output, 0) // Encoding of 0 in VarUint is 1 zero byte
		output = append(output, input...)
		return output, patterns, uncovered
	}
	if trace {
		fmt.Printf("Cluster | input = %x\n", input)
		for _, match := range matches {
			fmt.Printf(" [%x %d-%d]", input[match.Start:match.End], match.Start, match.End)
		}
	}
	cellRing.Reset()
	patterns = append(patterns[:0], 0, 0) // Sentinel entry - no meaning
	lastF := matches[len(matches)-1]
	for j := lastF.Start; j < lastF.End; j++ {
		d := cellRing.PushBack()
		d.optimStart = j + 1
		d.coverStart = len(input)
		d.compression = 0
		d.patternIdx = 0
		d.score = 0
	}
	// Starting from the last match
	for i := len(matches); i > 0; i-- {
		f := matches[i-1]
		p := f.Val.(*Pattern)
		firstCell := cellRing.Get(0)
		maxCompression := firstCell.compression
		maxScore := firstCell.score
		maxCell := firstCell
		var maxInclude bool
		for e := 0; e < cellRing.Len(); e++ {
			cell := cellRing.Get(e)
			comp := cell.compression - 4
			if cell.coverStart >= f.End {
				comp += f.End - f.Start
			} else {
				comp += cell.coverStart - f.Start
			}
			score := cell.score + p.score
			if comp > maxCompression || (comp == maxCompression && score > maxScore) {
				maxCompression = comp
				maxScore = score
				maxInclude = true
				maxCell = cell
			} else if cell.optimStart > f.End {
				cellRing.Truncate(e)
				break
			}
		}
		d := cellRing.PushFront()
		d.optimStart = f.Start
		d.score = maxScore
		d.compression = maxCompression
		if maxInclude {
			if trace {
				fmt.Printf("[include] cell for %d: with patterns", f.Start)
				fmt.Printf(" [%x %d-%d]", input[f.Start:f.End], f.Start, f.End)
				patternIdx := maxCell.patternIdx
				for patternIdx != 0 {
					pattern := patterns[patternIdx]
					fmt.Printf(" [%x %d-%d]", input[matches[pattern].Start:matches[pattern].End], matches[pattern].Start, matches[pattern].End)
					patternIdx = patterns[patternIdx+1]
				}
				fmt.Printf("\n\n")
			}
			d.coverStart = f.Start
			d.patternIdx = len(patterns)
			patterns = append(patterns, i-1, maxCell.patternIdx)
		} else {
			if trace {
				fmt.Printf("cell for %d: with patterns", f.Start)
				patternIdx := maxCell.patternIdx
				for patternIdx != 0 {
					pattern := patterns[patternIdx]
					fmt.Printf(" [%x %d-%d]", input[matches[pattern].Start:matches[pattern].End], matches[pattern].Start, matches[pattern].End)
					patternIdx = patterns[patternIdx+1]
				}
				fmt.Printf("\n\n")
			}
			d.coverStart = maxCell.coverStart
			d.patternIdx = maxCell.patternIdx
		}
	}
	optimCell := cellRing.Get(0)
	if trace {
		fmt.Printf("optimal =")
	}
	// Count number of patterns
	var patternCount uint64
	patternIdx := optimCell.patternIdx
	for patternIdx != 0 {
		patternCount++
		patternIdx = patterns[patternIdx+1]
	}
	var numBuf [binary.MaxVarintLen64]byte
	p := binary.PutUvarint(numBuf[:], patternCount)
	output = append(output, numBuf[:p]...)
	patternIdx = optimCell.patternIdx
	lastStart := 0
	var lastUncovered int
	uncovered = uncovered[:0]
	for patternIdx != 0 {
		pattern := patterns[patternIdx]
		p := matches[pattern].Val.(*Pattern)
		if trace {
			fmt.Printf(" [%x %d-%d]", input[matches[pattern].Start:matches[pattern].End], matches[pattern].Start, matches[pattern].End)
		}
		if matches[pattern].Start > lastUncovered {
			uncovered = append(uncovered, lastUncovered, matches[pattern].Start)
		}
		lastUncovered = matches[pattern].End
		// Starting position
		posMap[uint64(matches[pattern].Start-lastStart+1)]++
		lastStart = matches[pattern].Start
		n := binary.PutUvarint(numBuf[:], uint64(matches[pattern].Start))
		output = append(output, numBuf[:n]...)
		// Code
		n = binary.PutUvarint(numBuf[:], p.code)
		output = append(output, numBuf[:n]...)
		atomic.AddUint64(&p.uses, 1)
		patternIdx = patterns[patternIdx+1]
	}
	if len(input) > lastUncovered {
		uncovered = append(uncovered, lastUncovered, len(input))
	}
	if trace {
		fmt.Printf("\n\n")
	}
	// Add uncoded input
	for i := 0; i < len(uncovered); i += 2 {
		output = append(output, input[uncovered[i]:uncovered[i+1]]...)
	}
	return output, patterns, uncovered
}

func reduceDictWorker(trace bool, inputCh chan *CompressionWord, outCh chan *CompressionWord, completion *sync.WaitGroup, trie *patricia.PatriciaTree, inputSize, outputSize *atomic.Uint64, posMap map[uint64]uint64) {
	defer completion.Done()
	var output = make([]byte, 0, 256)
	var uncovered = make([]int, 256)
	var patterns = make([]int, 0, 256)
	cellRing := NewRing()
	mf2 := patricia.NewMatchFinder2(trie)
	var numBuf [binary.MaxVarintLen64]byte
	for compW := range inputCh {
		wordLen := uint64(len(compW.word))
		n := binary.PutUvarint(numBuf[:], wordLen)
		output = append(output[:0], numBuf[:n]...) // Prepend with the encoding of length
		output, patterns, uncovered = optimiseCluster(trace, compW.word, mf2, output, uncovered, patterns, cellRing, posMap)
		compW.word = append(compW.word[:0], output...)
		outCh <- compW
		inputSize.Add(1 + wordLen)
		outputSize.Add(uint64(len(output)))
		posMap[wordLen+1]++
		posMap[0]++
	}
}

// CompressionWord hold a word to be compressed (if flag is set), and the result of compression
// To allow multiple words to be processed concurrently, order field is used to collect all
// the words after processing without disrupting their order
type CompressionWord struct {
	word  []byte
	order uint64
}

type CompressionQueue []*CompressionWord

func (cq CompressionQueue) Len() int {
	return len(cq)
}

func (cq CompressionQueue) Less(i, j int) bool {
	return cq[i].order < cq[j].order
}

func (cq *CompressionQueue) Swap(i, j int) {
	(*cq)[i], (*cq)[j] = (*cq)[j], (*cq)[i]
}

func (cq *CompressionQueue) Push(x interface{}) {
	*cq = append(*cq, x.(*CompressionWord))
}

func (cq *CompressionQueue) Pop() interface{} {
	old := *cq
	n := len(old)
	x := old[n-1]
	old[n-1] = nil
	*cq = old[0 : n-1]
	return x
}

// reduceDict reduces the dictionary by trying the substitutions and counting frequency for each word
func reducedict(ctx context.Context, trace bool, logPrefix, segmentFilePath string, cf *os.File, datFile *DecompressedFile, workers int, dictBuilder *DictionaryBuilder, lvl log.Lvl, logger log.Logger) error {
	logEvery := time.NewTicker(60 * time.Second)
	defer logEvery.Stop()

	// DictionaryBuilder is for sorting words by their freuency (to assign codes)
	var pt patricia.PatriciaTree
	code2pattern := make([]*Pattern, 0, 256)
	dictBuilder.ForEach(func(score uint64, word []byte) {
		p := &Pattern{
			score:    score,
			uses:     0,
			code:     uint64(len(code2pattern)),
			codeBits: 0,
			word:     word,
		}
		pt.Insert(word, p)
		code2pattern = append(code2pattern, p)
	})
	dictBuilder.Close()
	if lvl < log.LvlTrace {
		logger.Log(lvl, fmt.Sprintf("[%s] dictionary file parsed", logPrefix), "entries", len(code2pattern))
	}
	ch := make(chan *CompressionWord, 10_000)
	inputSize, outputSize := &atomic.Uint64{}, &atomic.Uint64{}

	var collectors []*etl.Collector
	defer func() {
		for _, c := range collectors {
			c.Close()
		}
	}()
	out := make(chan *CompressionWord, 1024)
	var compressionQueue CompressionQueue
	heap.Init(&compressionQueue)
	queueLimit := 128 * 1024

	// For the case of workers == 1
	var output = make([]byte, 0, 256)
	var uncovered = make([]int, 256)
	var patterns = make([]int, 0, 256)
	cellRing := NewRing()
	mf2 := patricia.NewMatchFinder2(&pt)

	var posMaps []map[uint64]uint64
	uncompPosMap := make(map[uint64]uint64) // For the uncompressed words
	posMaps = append(posMaps, uncompPosMap)
	var wg sync.WaitGroup
	if workers > 1 {
		for i := 0; i < workers; i++ {
			posMap := make(map[uint64]uint64)
			posMaps = append(posMaps, posMap)
			wg.Add(1)
			go reduceDictWorker(trace, ch, out, &wg, &pt, inputSize, outputSize, posMap)
		}
	}
	t := time.Now()

	var err error
	intermediatePath := segmentFilePath + ".tmp"
	defer os.Remove(intermediatePath)
	var intermediateFile *os.File
	if intermediateFile, err = os.Create(intermediatePath); err != nil {
		return fmt.Errorf("create intermediate file: %w", err)
	}
	defer intermediateFile.Close()
	intermediateW := bufio.NewWriterSize(intermediateFile, 8*etl.BufIOSize)

	var inCount, outCount, emptyWordsCount uint64 // Counters words sent to compression and returned for compression
	var numBuf [binary.MaxVarintLen64]byte
	totalWords := datFile.count

	if err = datFile.ForEach(func(v []byte, compression bool) error {
		select {
		case <-ctx.Done():
			return ctx.Err()
		default:
		}
		if workers > 1 {
			// take processed words in non-blocking way and push them to the queue
		outer:
			for {
				select {
				case compW := <-out:
					heap.Push(&compressionQueue, compW)
				default:
					break outer
				}
			}
			// take processed words in blocking way until either:
			// 1. compressionQueue is below the limit so that new words can be allocated
			// 2. there is word in order on top of the queue which can be written down and reused
			for compressionQueue.Len() >= queueLimit && compressionQueue[0].order < outCount {
				// Blocking wait to receive some outputs until the top of queue can be processed
				compW := <-out
				heap.Push(&compressionQueue, compW)
			}
			var compW *CompressionWord
			// Either take the word from the top, write it down and reuse for the next unprocessed word
			// Or allocate new word
			if compressionQueue.Len() > 0 && compressionQueue[0].order == outCount {
				compW = heap.Pop(&compressionQueue).(*CompressionWord)
				outCount++
				// Write to intermediate file
				if _, e := intermediateW.Write(compW.word); e != nil {
					return e
				}
				// Reuse compW for the next word
			} else {
				compW = &CompressionWord{}
			}
			compW.order = inCount
			if len(v) == 0 {
				// Empty word, cannot be compressed
				compW.word = append(compW.word[:0], 0)
				uncompPosMap[1]++
				uncompPosMap[0]++
				heap.Push(&compressionQueue, compW) // Push to the queue directly, bypassing compression
			} else if compression {
				compW.word = append(compW.word[:0], v...)
				ch <- compW // Send for compression
			} else {
				// Prepend word with encoding of length + zero byte, which indicates no patterns to be found in this word
				wordLen := uint64(len(v))
				n := binary.PutUvarint(numBuf[:], wordLen)
				uncompPosMap[wordLen+1]++
				uncompPosMap[0]++
				compW.word = append(append(append(compW.word[:0], numBuf[:n]...), 0), v...)
				heap.Push(&compressionQueue, compW) // Push to the queue directly, bypassing compression
			}
		} else {
			outCount++
			wordLen := uint64(len(v))
			n := binary.PutUvarint(numBuf[:], wordLen)
			if _, e := intermediateW.Write(numBuf[:n]); e != nil {
				return e
			}
			if wordLen > 0 {
				if compression {
					output, patterns, uncovered = optimiseCluster(trace, v, mf2, output[:0], uncovered, patterns, cellRing, uncompPosMap)
					if _, e := intermediateW.Write(output); e != nil {
						return e
					}
					outputSize.Add(uint64(len(output)))
				} else {
					if e := intermediateW.WriteByte(0); e != nil {
						return e
					}
					if _, e := intermediateW.Write(v); e != nil {
						return e
					}
					outputSize.Add(1 + uint64(len(v)))
				}
			}
			inputSize.Add(1 + wordLen)
			uncompPosMap[wordLen+1]++
			uncompPosMap[0]++
		}
		inCount++
		if len(v) == 0 {
			emptyWordsCount++
		}

		select {
		case <-logEvery.C:
			if lvl < log.LvlTrace {
				logger.Log(lvl, fmt.Sprintf("[%s] Replacement preprocessing", logPrefix), "processed", fmt.Sprintf("%.2f%%", 100*float64(outCount)/float64(totalWords)), "ch", len(ch), "workers", workers)
			}
		default:
		}
		return nil
	}); err != nil {
		return err
	}
	close(ch)
	// Drain the out queue if necessary
	if inCount > outCount {
		for compressionQueue.Len() > 0 && compressionQueue[0].order == outCount {
			compW := heap.Pop(&compressionQueue).(*CompressionWord)
			outCount++
			if outCount == inCount {
				close(out)
			}
			// Write to intermediate file
			if _, e := intermediateW.Write(compW.word); e != nil {
				return e
			}
		}
		for compW := range out {
			heap.Push(&compressionQueue, compW)
			for compressionQueue.Len() > 0 && compressionQueue[0].order == outCount {
				compW = heap.Pop(&compressionQueue).(*CompressionWord)
				outCount++
				if outCount == inCount {
					close(out)
				}
				// Write to intermediate file
				if _, e := intermediateW.Write(compW.word); e != nil {
					return e
				}
			}
		}
	}
	if err = intermediateW.Flush(); err != nil {
		return err
	}
	wg.Wait()
	if lvl < log.LvlTrace {
		log.Log(lvl, fmt.Sprintf("[%s] Replacement preprocessing", logPrefix), "took", time.Since(t))
	}
	if _, err = intermediateFile.Seek(0, 0); err != nil {
		return fmt.Errorf("return to the start of intermediate file: %w", err)
	}

	//var m runtime.MemStats
	//common.ReadMemStats(&m)
	//logger.Info(fmt.Sprintf("[%s] Dictionary build done", logPrefix), "input", common.ByteCount(inputSize.Load()), "output", common.ByteCount(outputSize.Load()), "alloc", common.ByteCount(m.Alloc), "sys", common.ByteCount(m.Sys))
	posMap := make(map[uint64]uint64)
	for _, m := range posMaps {
		for l, c := range m {
			posMap[l] += c
		}
	}
	//fmt.Printf("posMap = %v\n", posMap)
	var patternList PatternList
	distribution := make([]int, maxPatternLen+1)
	for _, p := range code2pattern {
		if p.uses > 0 {
			patternList = append(patternList, p)
			distribution[len(p.word)]++
		}
	}
	slices.SortFunc(patternList, patternListLess)
	logCtx := make([]interface{}, 0, 8)
	logCtx = append(logCtx, "patternList.Len", patternList.Len())

	i := 0
	// Build Huffman tree for codes
	var codeHeap PatternHeap
	heap.Init(&codeHeap)
	tieBreaker := uint64(0)
	for codeHeap.Len()+(patternList.Len()-i) > 1 {
		// New node
		h := &PatternHuff{
			tieBreaker: tieBreaker,
		}
		if codeHeap.Len() > 0 && (i >= patternList.Len() || codeHeap[0].uses < patternList[i].uses) {
			// Take h0 from the heap
			h.h0 = heap.Pop(&codeHeap).(*PatternHuff)
			h.h0.AddZero()
			h.uses += h.h0.uses
		} else {
			// Take p0 from the list
			h.p0 = patternList[i]
			h.p0.code = 0
			h.p0.codeBits = 1
			h.uses += h.p0.uses
			i++
		}
		if codeHeap.Len() > 0 && (i >= patternList.Len() || codeHeap[0].uses < patternList[i].uses) {
			// Take h1 from the heap
			h.h1 = heap.Pop(&codeHeap).(*PatternHuff)
			h.h1.AddOne()
			h.uses += h.h1.uses
		} else {
			// Take p1 from the list
			h.p1 = patternList[i]
			h.p1.code = 1
			h.p1.codeBits = 1
			h.uses += h.p1.uses
			i++
		}
		tieBreaker++
		heap.Push(&codeHeap, h)
	}
	if codeHeap.Len() > 0 {
		root := heap.Pop(&codeHeap).(*PatternHuff)
		root.SetDepth(0)
	}
	// Calculate total size of the dictionary
	var patternsSize uint64
	for _, p := range patternList {
		ns := binary.PutUvarint(numBuf[:], uint64(p.depth))    // Length of the word's depth
		n := binary.PutUvarint(numBuf[:], uint64(len(p.word))) // Length of the word's length
		patternsSize += uint64(ns + n + len(p.word))
	}

	logCtx = append(logCtx, "patternsSize", common.ByteCount(patternsSize))
	for i, n := range distribution {
		if n == 0 {
			continue
		}
		logCtx = append(logCtx, fmt.Sprintf("%d", i), fmt.Sprintf("%d", n))
	}
	if lvl < log.LvlTrace {
		logger.Log(lvl, fmt.Sprintf("[%s] Effective dictionary", logPrefix), logCtx...)
	}
	cw := bufio.NewWriterSize(cf, 2*etl.BufIOSize)
	// 1-st, output amount of words - just a useful metadata
	binary.BigEndian.PutUint64(numBuf[:], inCount) // Dictionary size
	if _, err = cw.Write(numBuf[:8]); err != nil {
		return err
	}
	binary.BigEndian.PutUint64(numBuf[:], emptyWordsCount)
	if _, err = cw.Write(numBuf[:8]); err != nil {
		return err
	}
	// 2-nd, output dictionary size
	binary.BigEndian.PutUint64(numBuf[:], patternsSize) // Dictionary size
	if _, err = cw.Write(numBuf[:8]); err != nil {
		return err
	}
	//fmt.Printf("patternsSize = %d\n", patternsSize)
	// Write all the pattens
	slices.SortFunc(patternList, patternListLess)
	for _, p := range patternList {
		ns := binary.PutUvarint(numBuf[:], uint64(p.depth))
		if _, err = cw.Write(numBuf[:ns]); err != nil {
			return err
		}
		n := binary.PutUvarint(numBuf[:], uint64(len(p.word)))
		if _, err = cw.Write(numBuf[:n]); err != nil {
			return err
		}
		if _, err = cw.Write(p.word); err != nil {
			return err
		}
		//fmt.Printf("[comp] depth=%d, code=[%b], codeLen=%d pattern=[%x]\n", p.depth, p.code, p.codeBits, p.word)
	}

	var positionList PositionList
	pos2code := make(map[uint64]*Position)
	for pos, uses := range posMap {
		p := &Position{pos: pos, uses: uses, code: pos, codeBits: 0}
		positionList = append(positionList, p)
		pos2code[pos] = p
	}
	slices.SortFunc(positionList, positionListLess)
	i = 0
	// Build Huffman tree for codes
	var posHeap PositionHeap
	heap.Init(&posHeap)
	tieBreaker = uint64(0)
	for posHeap.Len()+(positionList.Len()-i) > 1 {
		// New node
		h := &PositionHuff{
			tieBreaker: tieBreaker,
		}
		if posHeap.Len() > 0 && (i >= positionList.Len() || posHeap[0].uses < positionList[i].uses) {
			// Take h0 from the heap
			h.h0 = heap.Pop(&posHeap).(*PositionHuff)
			h.h0.AddZero()
			h.uses += h.h0.uses
		} else {
			// Take p0 from the list
			h.p0 = positionList[i]
			h.p0.code = 0
			h.p0.codeBits = 1
			h.uses += h.p0.uses
			i++
		}
		if posHeap.Len() > 0 && (i >= positionList.Len() || posHeap[0].uses < positionList[i].uses) {
			// Take h1 from the heap
			h.h1 = heap.Pop(&posHeap).(*PositionHuff)
			h.h1.AddOne()
			h.uses += h.h1.uses
		} else {
			// Take p1 from the list
			h.p1 = positionList[i]
			h.p1.code = 1
			h.p1.codeBits = 1
			h.uses += h.p1.uses
			i++
		}
		tieBreaker++
		heap.Push(&posHeap, h)
	}
	if posHeap.Len() > 0 {
		posRoot := heap.Pop(&posHeap).(*PositionHuff)
		posRoot.SetDepth(0)
	}
	// Calculate the size of pos dictionary
	var posSize uint64
	for _, p := range positionList {
		ns := binary.PutUvarint(numBuf[:], uint64(p.depth)) // Length of the position's depth
		n := binary.PutUvarint(numBuf[:], p.pos)
		posSize += uint64(ns + n)
	}
	// First, output dictionary size
	binary.BigEndian.PutUint64(numBuf[:], posSize) // Dictionary size
	if _, err = cw.Write(numBuf[:8]); err != nil {
		return err
	}
	//fmt.Printf("posSize = %d\n", posSize)
	// Write all the positions
	slices.SortFunc(positionList, positionListLess)
	for _, p := range positionList {
		ns := binary.PutUvarint(numBuf[:], uint64(p.depth))
		if _, err = cw.Write(numBuf[:ns]); err != nil {
			return err
		}
		n := binary.PutUvarint(numBuf[:], p.pos)
		if _, err = cw.Write(numBuf[:n]); err != nil {
			return err
		}
		//fmt.Printf("[comp] depth=%d, code=[%b], codeLen=%d pos=%d\n", p.depth, p.code, p.codeBits, p.pos)
	}
	if lvl < log.LvlTrace {
		logger.Log(lvl, fmt.Sprintf("[%s] Positional dictionary", logPrefix), "positionList.len", positionList.Len(), "posSize", common.ByteCount(posSize))
	}
	// Re-encode all the words with the use of optimised (via Huffman coding) dictionaries
	wc := 0
	var hc HuffmanCoder
	hc.w = cw
	r := bufio.NewReaderSize(intermediateFile, 2*etl.BufIOSize)
	var l uint64
	var e error
	for l, e = binary.ReadUvarint(r); e == nil; l, e = binary.ReadUvarint(r) {
		posCode := pos2code[l+1]
		if posCode != nil {
			if e = hc.encode(posCode.code, posCode.codeBits); e != nil {
				return e
			}
		}
		if l == 0 {
			if e = hc.flush(); e != nil {
				return e
			}
		} else {
			var pNum uint64 // Number of patterns
			if pNum, e = binary.ReadUvarint(r); e != nil {
				return e
			}
			// Now reading patterns one by one
			var lastPos uint64
			var lastUncovered int
			var uncoveredCount int
			for i := 0; i < int(pNum); i++ {
				var pos uint64 // Starting position for pattern
				if pos, e = binary.ReadUvarint(r); e != nil {
					return e
				}
				posCode = pos2code[pos-lastPos+1]
				lastPos = pos
				if posCode != nil {
					if e = hc.encode(posCode.code, posCode.codeBits); e != nil {
						return e
					}
				}
				var code uint64 // Code of the pattern
				if code, e = binary.ReadUvarint(r); e != nil {
					return e
				}
				patternCode := code2pattern[code]
				if int(pos) > lastUncovered {
					uncoveredCount += int(pos) - lastUncovered
				}
				lastUncovered = int(pos) + len(patternCode.word)
				if patternCode != nil {
					if e = hc.encode(patternCode.code, patternCode.codeBits); e != nil {
						return e
					}
				}
			}
			if int(l) > lastUncovered {
				uncoveredCount += int(l) - lastUncovered
			}
			// Terminating position and flush
			posCode = pos2code[0]
			if e = hc.encode(posCode.code, posCode.codeBits); e != nil {
				return e
			}
			if e = hc.flush(); e != nil {
				return e
			}
			// Copy uncovered characters
			if uncoveredCount > 0 {
				if _, e = io.CopyN(cw, r, int64(uncoveredCount)); e != nil {
					return e
				}
			}
		}
		wc++
		select {
		case <-logEvery.C:
			if lvl < log.LvlTrace {
				logger.Log(lvl, fmt.Sprintf("[%s] Compressed", logPrefix), "processed", fmt.Sprintf("%.2f%%", 100*float64(wc)/float64(totalWords)))
			}
		default:
		}
	}
	if e != nil && !errors.Is(e, io.EOF) {
		return e
	}
	if err = intermediateFile.Close(); err != nil {
		return err
	}
	if err = cw.Flush(); err != nil {
		return err
	}
	return nil
}

// processSuperstring is the worker that processes one superstring and puts results
// into the collector, using lock to mutual exclusion. At the end (when the input channel is closed),
// it notifies the waitgroup before exiting, so that the caller known when all work is done
// No error channels for now
func processSuperstring(ctx context.Context, superstringCh chan []byte, dictCollector *etl.Collector, minPatternScore uint64, completion *sync.WaitGroup, logger log.Logger) {
	defer completion.Done()
	dictVal := make([]byte, 8)
	dictKey := make([]byte, maxPatternLen)
	var lcp, sa, inv []int32
	for superstring := range superstringCh {
		select {
		case <-ctx.Done():
			return
		default:
		}

		if cap(sa) < len(superstring) {
			sa = make([]int32, len(superstring))
		} else {
			sa = sa[:len(superstring)]
		}
		//log.Info("Superstring", "len", len(superstring))
		//start := time.Now()
		if err := sais.Sais(superstring, sa); err != nil {
			panic(err)
		}
		//log.Info("Suffix array built", "in", time.Since(start))
		// filter out suffixes that start with odd positions
		n := len(sa) / 2
		filtered := sa[:n]
		//filtered := make([]int32, n)
		var j int
		for i := 0; i < len(sa); i++ {
			if sa[i]&1 == 0 {
				filtered[j] = sa[i] >> 1
				j++
			}
		}
		// Now create an inverted array
		if cap(inv) < n {
			inv = make([]int32, n)
		} else {
			inv = inv[:n]
		}
		for i := 0; i < n; i++ {
			inv[filtered[i]] = int32(i)
		}
		//logger.Info("Inverted array done")
		var k int
		// Process all suffixes one by one starting from
		// first suffix in txt[]
		if cap(lcp) < n {
			lcp = make([]int32, n)
		} else {
			lcp = lcp[:n]
		}
		for i := 0; i < n; i++ {
			/* If the current suffix is at n-1, then we don’t
			   have next substring to consider. So lcp is not
			   defined for this substring, we put zero. */
			if inv[i] == int32(n-1) {
				k = 0
				continue
			}

			/* j contains index of the next substring to
			   be considered  to compare with the present
			   substring, i.e., next string in suffix array */
			j := int(filtered[inv[i]+1])

			// Directly start matching from k'th index as
			// at-least k-1 characters will match
			for i+k < n && j+k < n && superstring[(i+k)*2] != 0 && superstring[(j+k)*2] != 0 && superstring[(i+k)*2+1] == superstring[(j+k)*2+1] {
				k++
			}
			lcp[inv[i]] = int32(k) // lcp for the present suffix.

			// Deleting the starting character from the string.
			if k > 0 {
				k--
			}
		}
		//log.Info("Kasai algorithm finished")
		// Checking LCP array

		if assert.Enable {
			for i := 0; i < n-1; i++ {
				var prefixLen int
				p1 := int(filtered[i])
				p2 := int(filtered[i+1])
				for p1+prefixLen < n &&
					p2+prefixLen < n &&
					superstring[(p1+prefixLen)*2] != 0 &&
					superstring[(p2+prefixLen)*2] != 0 &&
					superstring[(p1+prefixLen)*2+1] == superstring[(p2+prefixLen)*2+1] {
					prefixLen++
				}
				if prefixLen != int(lcp[i]) {
					logger.Error("Mismatch", "prefixLen", prefixLen, "lcp[i]", lcp[i], "i", i)
					break
				}
				l := int(lcp[i]) // Length of potential dictionary word
				if l < 2 {
					continue
				}
			}
		}
		//logger.Info("LCP array checked")
		// Walk over LCP array and compute the scores of the strings
		var b = inv
		j = 0
		for i := 0; i < n-1; i++ {
			// Only when there is a drop in LCP value
			if lcp[i+1] >= lcp[i] {
				j = i
				continue
			}
			prevSkipped := false
			for l := int(lcp[i]); l > int(lcp[i+1]) && l >= minPatternLen; l-- {
				if l > maxPatternLen ||
					l > 20 && (l&(l-1)) != 0 { // is power of 2
					prevSkipped = true
					continue
				}

				// Go back
				var isNew bool
				for j > 0 && int(lcp[j-1]) >= l {
					j--
					isNew = true
				}

				if !isNew && !prevSkipped {
					break
				}

				window := i - j + 2
				copy(b, filtered[j:i+2])
				slices.Sort(b[:window])
				repeats := 1
				lastK := 0
				for k := 1; k < window; k++ {
					if b[k] >= b[lastK]+int32(l) {
						repeats++
						lastK = k
					}
				}

				if (l < 8 || l > 64) && repeats < int(minPatternScore) {
					prevSkipped = true
					continue
				}

				score := uint64(repeats * (l))
				if score < minPatternScore {
					prevSkipped = true
					continue
				}

				dictKey = dictKey[:l]
				for s := 0; s < l; s++ {
					dictKey[s] = superstring[(int(filtered[i])+s)*2+1]
				}
				binary.BigEndian.PutUint64(dictVal, score)
				if err := dictCollector.Collect(dictKey, dictVal); err != nil {
					logger.Error("processSuperstring", "collect", err)
				}
				prevSkipped = false //nolint
				break
			}
		}
	}
}

func DictionaryBuilderFromCollectors(ctx context.Context, logPrefix, tmpDir string, collectors []*etl.Collector, lvl log.Lvl, logger log.Logger) (*DictionaryBuilder, error) {
	dictCollector := etl.NewCollector(logPrefix+"_collectDict", tmpDir, etl.NewSortableBuffer(etl.BufferOptimalSize), logger)
	defer dictCollector.Close()
	dictCollector.LogLvl(lvl)

	dictAggregator := &DictAggregator{collector: dictCollector, dist: map[int]int{}}
	for _, collector := range collectors {
		if err := collector.Load(nil, "", dictAggregator.aggLoadFunc, etl.TransformArgs{Quit: ctx.Done()}); err != nil {
			return nil, err
		}
		collector.Close()
	}
	if err := dictAggregator.finish(); err != nil {
		return nil, err
	}
	db := &DictionaryBuilder{limit: maxDictPatterns} // Only collect 1m words with highest scores
	if err := dictCollector.Load(nil, "", db.loadFunc, etl.TransformArgs{Quit: ctx.Done()}); err != nil {
		return nil, err
	}
	db.finish()

	db.Sort()
	return db, nil
}

func PersistDictrionary(fileName string, db *DictionaryBuilder) error {
	df, err := os.Create(fileName)
	if err != nil {
		return err
	}
	w := bufio.NewWriterSize(df, 2*etl.BufIOSize)
	db.ForEach(func(score uint64, word []byte) { fmt.Fprintf(w, "%d %x\n", score, word) })
	if err = w.Flush(); err != nil {
		return err
	}
	if err := df.Sync(); err != nil {
		return err
	}
	return df.Close()
}

func ReadSimpleFile(fileName string, walker func(v []byte) error) error {
	// Read keys from the file and generate superstring (with extra byte 0x1 prepended to each character, and with 0x0 0x0 pair inserted between keys and values)
	// We only consider values with length > 2, because smaller values are not compressible without going into bits
	f, err := os.Open(fileName)
	if err != nil {
		return err
	}
	defer f.Close()
	r := bufio.NewReaderSize(f, etl.BufIOSize)
	buf := make([]byte, 4096)
	for l, e := binary.ReadUvarint(r); ; l, e = binary.ReadUvarint(r) {
		if e != nil {
			if errors.Is(e, io.EOF) {
				break
			}
			return e
		}
		if len(buf) < int(l) {
			buf = make([]byte, l)
		}
		if _, e = io.ReadFull(r, buf[:l]); e != nil {
			return e
		}
		if err := walker(buf[:l]); err != nil {
			return err
		}
	}
	return nil
}