NewRegexpQuery cannot return the correct result about bleve HOT 4 CLOSED

Is there any other method can query accurately?

I hope that I can use a func like ‘regexp’ because it can search accurately.

For example, here is a sentence ‘I hope that I can fly in the future’, I save it as a predicate ‘sentence’,if I give dgraph ‘fut’,it can match {‘sentence’ : ‘I hope that I can fly in the future’} because ‘future’ contains a continuous ‘fut’.

from bleve.

@aokaywe thanks for reaching out!

From what I see, using the default index mapping results in your documents being indexed with the standard analyser which breaks down the referred document into tokens separated by space. Hence, the regexp matching 2 words returns no matches.
(To underscore my point, if your regex was ".*ext3*", that would return the expected document)

You should use a custom analyser with a "single" tokenizer since that doesn't split input text and indexes it as is.

I will work on a code sample which I can share to further illustrate my point.

P.S.
You could consider batching your index operations into a single Batch() operation.

from bleve.

@aokaywe
Turns out you can use the keyword analyser too, as an alternative to building a custom analyser with the "single" tokenizer.

Here's the code to do that:

package main

import (
	"fmt"

	"github.com/blevesearch/bleve/v2"
	"github.com/blevesearch/bleve/v2/mapping"
)

func buildIndexMapping() (mapping.IndexMapping, error) {
	indexMapping := bleve.NewIndexMapping()
	indexMapping.DefaultAnalyzer = "keyword"

	return indexMapping, nil
}

func main() {
	// open a new index
	mapping, err := buildIndexMapping()
	if err != nil {
		fmt.Println(err)
		return
	}
	index, err := bleve.New("example1.bleve", mapping)
	if err != nil {
		fmt.Println(err)
		return
	}

	dataList := []struct {
		Name string
	}{
		{Name: "find text1 in file"},
		{Name: "find text1 in file2"},
		{Name: "find text2 in document"},
		{Name: "find text3 in content"},
	}

	b := index.NewBatch()
	for i, data := range dataList {
		docID := fmt.Sprintf("doc%d", i)
		err = b.Index(docID, data)
		if err != nil {
			fmt.Printf("Error indexing document %s: %v\n", docID, err)
		}
	}
	err = index.Batch(b)
	if err != nil {
		fmt.Printf("Error indexing batch: %v \n", err)
		return
	}

	// search for some text
	query := bleve.NewRegexpQuery(".*ext3 in.*")

	search := bleve.NewSearchRequest(query)
	searchResults, err := index.Search(search)
	if err != nil {
		fmt.Println(err)
		return
	}

	fmt.Println(searchResults)

}

Pls let me know if you require any further info. Feel free to close the issue once your query is resolved.

thanks!

from bleve.

@metonymic-smokey Thank you, this works.

from bleve.