Enum SearchTokenization

Namespace

MongoDB.EntityFrameworkCore.Metadata.Search

Assembly

MongoDB.EntityFrameworkCore.dll

Tokenization strategies for MongoDB search string fields.

public enum SearchTokenization

Fields

EdgeGram = 0

Edge-gram tokenization from the left edge (prefix grams). Produces tokens that start at the beginning of the term and extend up to the configured maxGram. Typically used for autocomplete and prefix matching (e.g., "mon", "mong", "mongo").

RightEdgeGram = 1

Edge-gram tokenization from the right edge (suffix grams). Produces tokens that end at the end of the term and extend backward, enabling suffix matching (e.g., "bson", "son", "on"). Useful when you need to match endings such as file extensions or suffix-based queries.

NGram = 2

N-gram tokenization using a sliding window across the entire term (infix grams). Produces overlapping tokens of length between minGram and maxGram, enabling substring matching anywhere within a word. Useful for partial/infix search, with increased index size.

Remarks

Tokenization controls how text is split into smaller units (tokens) for indexing and matching. Use edge-gram tokenization for efficient prefix/suffix autocomplete, and n-gram tokenization for infix matching within words. Actual token lengths are controlled by tokenizer options such as minGram and maxGram in the analyzer configuration. See MongoDB search tokenizer docs for details: https://www.mongodb.com/docs/atlas/atlas-search/analyzers/tokenizers/edge-gram/, https://www.mongodb.com/docs/atlas/atlas-search/analyzers/tokenizers/n-gram/.