Pacote
    Preparing search index...

    Type Alias Options<Document, SummaryField, IndexField>

    Configuration options for creating a search index.

    type Options<
        Document extends Record<string, unknown>,
        SummaryField extends keyof Document = keyof Document,
        IndexField extends keyof Document = keyof Document,
    > = {
        errorRate?: number;
        fields: IndexField[] | Record<IndexField, number>;
        minSize?: number;
        ngrams?: number;
        preprocess?: PreprocessFunction<Document, IndexField>;
        seed?: number;
        stemmer?: StemmerFunction;
        stopwords?: StopwordsFunction;
        summary: SummaryField[];
        termFrequencyBuckets?: number[];
        tokenizer?: TokenizerFunction;
    }

    Type Parameters

    • Document extends Record<string, unknown>

      The type of document being indexed.

    • SummaryField extends keyof Document = keyof Document

      The document keys that can be returned in search results.

    • IndexField extends keyof Document = keyof Document

      The document keys to be indexed for searching.

    Index

    Properties

    errorRate? fields minSize? ngrams? preprocess? seed? stemmer? stopwords? summary termFrequencyBuckets? tokenizer?

    Properties

    errorRate?: number

    Determines the desired error rate. A lower number yields more reliable results but makes the index larger. The value defaults to 0.0001 (or 0.01%).

    fields: IndexField[] | Record<IndexField, number>

    The fields to index, provided as an array or as a record of field keys and weight values.

    minSize?: number

    Minimum term cardinality used to calculate the Bloom filter size. This can be used to reduce false positives when dealing with small documents with sparse term frequency distribution. The default value is 0.

    ngrams?: number

    Indexes n-grams beyond the single text tokens. A value of 2 indexes digrams, a value of 3 indexes digrams and trigrams, and so forth. This allows seaching the index for simple phrases (a phrase search is entered "between quotes"). Indexing n-grams will increase the size of the generated indices roughly by a factor of n. Default value is 1 (no n-grams are indexed).

    preprocess?: PreprocessFunction<Document, IndexField>

    Preprocessing function, executed before all others. The function serialises each field as a string and optionally process it before indexing. For example, you might use this function to strip HTML from a field value. By default, this class simply converts the field value into a string.

    seed?: number

    Hash seed to use in Bloom Filters, defaults to 0x00c0ffee.

    stemmer?: StemmerFunction

    Allows plugging in a custom stemming function. By default, this class does not change text tokens.

    stopwords?: StopwordsFunction

    Filters tokens so that words that are too short or too common may be excluded from the index. By default, no stopwords are excluded.

    summary: SummaryField[]

    Determines which fields in the document can be stored in the index and returned as a search result.

    termFrequencyBuckets?: number[]

    Optimises storage by grouping indexed terms into buckets according to term frequency in a document. Defaults to [1, 2, 3, 4, 8, 16, 32, 64].

    tokenizer?: TokenizerFunction

    Allows a custom tokenizer function. By default content is transformed to lowercase, split at every whitespace or hyphen, and stripped of any non-word (A-Z, 0-9, and _) characters.

    Settings

    Member Visibility

    On This Page

    Properties