AnalyzerType uses ConstantsTrait
ArangoSearch analyzer type discriminator, used as the `type` field of every payload sent to `POST /_api/analyzer` and returned by `GET /_api/analyzer/{name}`.
The exposed set covers the V1 must-have analyzers plus ngram
(substring / autocomplete indexing) and pipeline (an ordered chain
of sub-analyzers — typically norm → ngram for case- and
accent-insensitive autocomplete). The remaining types shipped by
arangojs (aql, classification, nearest_neighbors, geo_json,
geo_point, geo_s2, segmentation, collation, minhash,
delimiter, multi_delimiter, stopwords) are deferred to a later
follow-up.
Tags
Table of Contents
Constants
- IDENTITY : string = 'identity'
- Pass-through analyzer — emits its input verbatim, with no transformation. Useful as the default analyzer on every link and on every field that does not need language-aware normalisation.
- NGRAM : string = 'ngram'
- N-gram analyzer — emits the substrings (n-grams) of its input between a `min` and `max` length. The building block of substring / "as-you-type" autocomplete search, typically paired with a `text` analyzer on the same field (see {@see \oihana\arango\clients\analyzer\NgramAnalyzer}).
- NORM : string = 'norm'
- Locale-aware normaliser. Lower-cases / upper-cases the input and optionally strips diacritics. Does not tokenise — use {@see self::TEXT} for that.
- PIPELINE : string = 'pipeline'
- Pipeline analyzer — runs an ordered chain of sub-analyzers, each fed the output of the previous one. The typed way to compose analyzers the server otherwise only exposes individually — most notably `norm` → `ngram`, which normalises case and accents **before** the n-gram split so case-/accent-insensitive autocomplete actually matches (a standalone `ngram` normalises neither). See {@see \oihana\arango\clients\analyzer\PipelineAnalyzer}.
- STEM : string = 'stem'
- Locale-aware stemmer. Reduces inflected forms of a word to a common root (e.g. `running` → `run`). Single-token input only.
- TEXT : string = 'text'
- Full-text analyzer — tokenises on word boundaries, optionally lower-cases, removes stopwords, applies stemming and accent folding, and optionally emits edge n-grams for prefix search.
Constants
IDENTITY
Pass-through analyzer — emits its input verbatim, with no transformation. Useful as the default analyzer on every link and on every field that does not need language-aware normalisation.
public
string
IDENTITY
= 'identity'
NGRAM
N-gram analyzer — emits the substrings (n-grams) of its input between a `min` and `max` length. The building block of substring / "as-you-type" autocomplete search, typically paired with a `text` analyzer on the same field (see {@see \oihana\arango\clients\analyzer\NgramAnalyzer}).
public
string
NGRAM
= 'ngram'
NORM
Locale-aware normaliser. Lower-cases / upper-cases the input and optionally strips diacritics. Does not tokenise — use {@see self::TEXT} for that.
public
string
NORM
= 'norm'
PIPELINE
Pipeline analyzer — runs an ordered chain of sub-analyzers, each fed the output of the previous one. The typed way to compose analyzers the server otherwise only exposes individually — most notably `norm` → `ngram`, which normalises case and accents **before** the n-gram split so case-/accent-insensitive autocomplete actually matches (a standalone `ngram` normalises neither). See {@see \oihana\arango\clients\analyzer\PipelineAnalyzer}.
public
string
PIPELINE
= 'pipeline'
STEM
Locale-aware stemmer. Reduces inflected forms of a word to a common root (e.g. `running` → `run`). Single-token input only.
public
string
STEM
= 'stem'
TEXT
Full-text analyzer — tokenises on word boundaries, optionally lower-cases, removes stopwords, applies stemming and accent folding, and optionally emits edge n-grams for prefix search.
public
string
TEXT
= 'text'