918732f3d8
Previously, unicode.Tokenize() would allocate a Token one-by-one, on an as-needed basis. This change allocates a "backing array" of Tokens, so that it goes to the runtime object allocator much less often. It takes a heuristic guess as to the backing array size by using the average token (segment) length seen so far. Results from micro-benchmark (null-firestorm, bleve-blast) seem to give perhaps less than ~0.5 MB/second throughput improvement.
96 lines
2.4 KiB
Go
96 lines
2.4 KiB
Go
// Copyright (c) 2014 Couchbase, Inc.
|
|
// 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 unicode
|
|
|
|
import (
|
|
"github.com/blevesearch/segment"
|
|
|
|
"github.com/blevesearch/bleve/analysis"
|
|
"github.com/blevesearch/bleve/registry"
|
|
)
|
|
|
|
const Name = "unicode"
|
|
|
|
type UnicodeTokenizer struct {
|
|
}
|
|
|
|
func NewUnicodeTokenizer() *UnicodeTokenizer {
|
|
return &UnicodeTokenizer{}
|
|
}
|
|
|
|
func (rt *UnicodeTokenizer) Tokenize(input []byte) analysis.TokenStream {
|
|
ta := []analysis.Token(nil)
|
|
taNext := 0
|
|
|
|
rv := make(analysis.TokenStream, 0)
|
|
|
|
segmenter := segment.NewWordSegmenterDirect(input)
|
|
start := 0
|
|
pos := 1
|
|
for segmenter.Segment() {
|
|
segmentBytes := segmenter.Bytes()
|
|
end := start + len(segmentBytes)
|
|
if segmenter.Type() != segment.None {
|
|
if taNext >= len(ta) {
|
|
avgSegmentLen := end / (len(rv) + 1)
|
|
if avgSegmentLen < 1 {
|
|
avgSegmentLen = 1
|
|
}
|
|
|
|
remainingLen := len(input) - end
|
|
remainingSegments := remainingLen / avgSegmentLen
|
|
if remainingSegments > 1000 {
|
|
remainingSegments = 1000
|
|
}
|
|
if remainingSegments < 1 {
|
|
remainingSegments = 1
|
|
}
|
|
|
|
ta = make([]analysis.Token, remainingSegments)
|
|
taNext = 0
|
|
}
|
|
|
|
token := &ta[taNext]
|
|
taNext++
|
|
|
|
token.Term = segmentBytes
|
|
token.Start = start
|
|
token.End = end
|
|
token.Position = pos
|
|
token.Type = convertType(segmenter.Type())
|
|
|
|
rv = append(rv, token)
|
|
pos++
|
|
}
|
|
start = end
|
|
}
|
|
return rv
|
|
}
|
|
|
|
func UnicodeTokenizerConstructor(config map[string]interface{}, cache *registry.Cache) (analysis.Tokenizer, error) {
|
|
return NewUnicodeTokenizer(), nil
|
|
}
|
|
|
|
func init() {
|
|
registry.RegisterTokenizer(Name, UnicodeTokenizerConstructor)
|
|
}
|
|
|
|
func convertType(segmentWordType int) analysis.TokenType {
|
|
switch segmentWordType {
|
|
case segment.Ideo:
|
|
return analysis.Ideographic
|
|
case segment.Kana:
|
|
return analysis.Ideographic
|
|
case segment.Number:
|
|
return analysis.Numeric
|
|
}
|
|
return analysis.AlphaNumeric
|
|
}
|