mirror of
https://gitee.com/milvus-io/milvus.git
synced 2025-12-07 01:28:27 +08:00
26ec841feb
Ref #42053 This is the first PR for optimizing `LIKE` with ngram inverted index. Now, only VARCHAR data type is supported and only InnerMatch LIKE (%xxx%) query is supported. How to use it: ``` milvus_client = MilvusClient("http://localhost:19530") schema = milvus_client.create_schema() ... schema.add_field("content_ngram", DataType.VARCHAR, max_length=10000) ... index_params = milvus_client.prepare_index_params() index_params.add_index(field_name="content_ngram", index_type="NGRAM", index_name="ngram_index", min_gram=2, max_gram=3) milvus_client.create_collection(COLLECTION_NAME, ...) ``` min_gram and max_gram controls how we tokenize the documents. For example, for min_gram=2 and max_gram=4, we will tokenize each document with 2-gram, 3-gram and 4-gram. --------- Signed-off-by: SpadeA <tangchenjie1210@gmail.com> Signed-off-by: SpadeA-Tang <tangchenjie1210@gmail.com>
61 lines
1.8 KiB
Go
61 lines
1.8 KiB
Go
package indexparamcheck
|
|
|
|
import (
|
|
"fmt"
|
|
"strconv"
|
|
|
|
"github.com/milvus-io/milvus-proto/go-api/v2/schemapb"
|
|
"github.com/milvus-io/milvus/pkg/v2/util/merr"
|
|
"github.com/milvus-io/milvus/pkg/v2/util/typeutil"
|
|
)
|
|
|
|
const (
|
|
MinGramKey = "min_gram"
|
|
MaxGramKey = "max_gram"
|
|
)
|
|
|
|
type NgramIndexChecker struct {
|
|
scalarIndexChecker
|
|
}
|
|
|
|
func newNgramIndexChecker() *NgramIndexChecker {
|
|
return &NgramIndexChecker{}
|
|
}
|
|
|
|
func (c *NgramIndexChecker) CheckTrain(dataType schemapb.DataType, params map[string]string) error {
|
|
if dataType != schemapb.DataType_VarChar {
|
|
// todo(SpadeA): we may support it for json in the future
|
|
return merr.WrapErrParameterInvalidMsg("Ngram index can only be created on VARCHAR field")
|
|
}
|
|
|
|
minGramStr, minGramExist := params[MinGramKey]
|
|
maxGramStr, maxGramExist := params[MaxGramKey]
|
|
if !minGramExist || !maxGramExist {
|
|
return merr.WrapErrParameterInvalidMsg("Ngram index must specify both min_gram and max_gram")
|
|
}
|
|
|
|
minGram, err := strconv.Atoi(minGramStr)
|
|
if err != nil {
|
|
return merr.WrapErrParameterInvalidMsg("min_gram for Ngram index must be an integer, got: %s", minGramStr)
|
|
}
|
|
|
|
maxGram, err := strconv.Atoi(maxGramStr)
|
|
if err != nil {
|
|
return merr.WrapErrParameterInvalidMsg("max_gram for Ngram index must be an integer, got: %s", maxGramStr)
|
|
}
|
|
|
|
if minGram <= 0 || maxGram <= 0 || minGram > maxGram {
|
|
return merr.WrapErrParameterInvalidMsg("invalid min_gram or max_gram value for Ngram index, min_gram: %d, max_gram: %d", minGram, maxGram)
|
|
}
|
|
|
|
return c.scalarIndexChecker.CheckTrain(dataType, params)
|
|
}
|
|
|
|
func (c *NgramIndexChecker) CheckValidDataType(indexType IndexType, field *schemapb.FieldSchema) error {
|
|
dType := field.GetDataType()
|
|
if !typeutil.IsStringType(dType) {
|
|
return fmt.Errorf("ngram index can only be created on VARCHAR field")
|
|
}
|
|
return nil
|
|
}
|