Transforming data with Analyzers

While AQL string functions allow for basic text manipulation, true text processing including tokenization, language-specific word stemming, case conversion and removal of diacritical marks (accents) from characters only become possible with Analyzers.

Analyzers parse input values and transform them into sets of sub-values, for example by breaking up text into words. If they are used in Views then the documents’ attribute values of the linked collections are used as input and additional metadata is produced internally. The data can then be used for searching and sorting to provide the most appropriate match for the specified conditions, similar to queries to web search engines.

Analyzers can be used on their own to tokenize and normalize strings in AQL queries with the TOKENS() function. The following example shows the creation of a custom Analyzer and how it transforms an example input:

var analyzers = require("@arangodb/analyzers")
var a = analyzers.save("custom", "text", {
  locale: "en",
  stopwords: ["a", "example"]
}, []);
db._query(`RETURN TOKENS("UPPER & lower, a Stemming Example.", "custom")`).toArray();

[ 
  [ 
    "upper", 
    "lower", 
    "stem" 
  ] 
]

How Analyzers process values depends on their type and configuration. The configuration is comprised of type-specific properties and list of features. The features control the additional metadata to be generated to augment View indexes, to be able to rank results for instance.

Analyzers can be managed via an HTTP API and through a JavaScript module.

Value Handling

While most of the Analyzer functionality is geared towards text processing, there is no restriction to strings as input data type when using them through Views or inverted indexes – your documents may have attributes of any data type after all.

Strings are processed according to the Analyzer, whereas other primitive data types (null, true, false, numbers) are generally left unchanged. Exceptions are Analyzers that specifically work with other data types, like geo-spatial or query-based Analyzers.

The elements of arrays are processed individually, regardless of the level of nesting, if you use Analyzers stand-alone. That is, strings are processed by the configured Analyzer(s) and other primitive values are returned as-is. This also applies if you use Analyzers in arangosearch Views, or in search-alias Views with inverted indexes that have the searchField option enabled. The array elements are unpacked, processed, and indexed individually.

If you use inverted indexes with the searchField option disabled, optionally through search-alias Views, array elements are not unpacked by default. Most Analyzers do not accept arrays as input in this context. You can unpack one array level and let the configured Analyzer process the individual elements by using [*] as a suffix for a field in the index definition. Primitive values other than strings are indexed as-is.

Analyzers cannot process objects as a whole. However, you can work with individual object attributes. You can use inverted indexes and Views to index specific object attributes or sub-attributes, or index all sub-attributes with the includeAllFields option enabled. Each non-object value is handled as described above. Sub-objects in arrays can be indexed, too (with limitations). However, only primitive values are added to the index. Arrays and objects cannot be searched for as a whole.

Also see:

• SEARCH operation on how to query indexed values such as numbers and nested values
• arangosearch Views and Inverted indexes for details about how compound data types (arrays, objects) get indexed

Analyzer Names

Each Analyzer has a name for identification with the following naming constraints:

• The name must only consist of the letters a to z (both in lower and upper case), the numbers 0 to 9, underscore (_) and dash (-) symbols. This also means that any non-ASCII names are not allowed.
• It must always start with a letter.
• The maximum allowed length of a name is 254 bytes.
• Analyzer names are case-sensitive.

Custom Analyzers are stored per database, in a system collection _analyzers. The names get prefixed with the database name and two colons, e.g. myDB::customAnalyzer.This does not apply to the globally available built-in Analyzers, which are not stored in an _analyzers collection.

Custom Analyzers stored in the _system database can be referenced in queries against other databases by specifying the prefixed name, e.g. _system::customGlobalAnalyzer. Analyzers stored in databases other than _system cannot be accessed from within another database however.

Analyzer Types

The following Analyzer types are available:

• identity: treats value as atom (no transformation)
• delimiter: splits into tokens at a user-defined character sequence
• multi_delimiter: splits into tokens at user-defined character sequences
• stem: applies stemming to the value as a whole
• norm: applies normalization to the value as a whole
• ngram: creates n-grams from the value with user-defined lengths
• text: tokenizes text strings into words, optionally with stemming, normalization, stop-word filtering and edge n-gram generation
• segmentation: tokenizes text in a language-agnostic manner, optionally with normalization
• wildcard: can apply another Analyzer and creates n-grams to enable fast partial matching for large strings
• aql: runs an AQL query to prepare tokens for index
• pipeline: chains multiple Analyzers
• stopwords: removes the specified tokens from the input
• collation: respects the alphabetic order of a language in range queries
• minhash: applies another Analyzer and then a locality-sensitive hash function, to find candidates for set comparisons based on the Jaccard index (Enterprise Edition only)
• classification: classifies the input text using a word embedding model (Enterprise Edition only)
• nearest_neighbors: finds the nearest neighbors of the input text using a word embedding model (Enterprise Edition only)
• geojson: breaks up a GeoJSON object into a set of indexable tokens
• geo_s2: like geojson but offers more efficient formats for indexing geo-spatial data (Enterprise Edition only)
• geopoint: breaks up JSON data describing a coordinate pair into a set of indexable tokens

The following table compares the Analyzers for text processing:

The available normalizations are case conversion and accents/diacritics removal. The segmentation Analyzer only supports case conversion.

The text Analyzer supports edge n-grams but not full n-grams.

Tokenization

The text and segmentation Analyzer types tokenize text into words (or a comparable concept of a word). See Word Boundaries  in the Unicode Standard Annex #29 about Unicode text segmentation for details.

These tokenizing Analyzers extract tokens, which removes characters like punctuation and whitespace. An exception is the segmentation Analyzer if you select "graphic" or "all" for the break option. They preserve @ and . characters of email addresses, for instance. There are also exceptions with both Analyzer types for sequences like numbers, for which decimal and thousands separators (. and ,) are preserved.

Normalization

The norm, text, and segmentation Analyzer types allow you to convert the input text to all lowercase or all uppercase for normalization purposes, namely case insensitive search. Case folding is not supported. Also see Case Mapping  in the ICU documentation.

The norm and text Analyzer types also allow you to convert characters with diacritical marks to the base characters. This normalization enables accent-insensitive search.

Analyzer Features

The features of an Analyzer determine what searching capabilities are available and are only applicable in the context of Views and inverted indexes.

The valid values for the features are dependant on both the capabilities of the underlying Analyzer type and the query filtering and sorting functions that the result can be used with. For example, the text type produces frequency + norm + position, and the PHRASE() AQL function requires frequency + position to be available.

The following features are supported:

• frequency: track how often a term occurs. Required for PHRASE(), NGRAM_MATCH(), BM25(), TFIDF(), and OFFSET_INFO().
• norm: calculate and store the field normalization factor that is used to score fairer if the same term is repeated, reducing its importance. Required for BM25() (except BM15) and TFIDF() (if called with normalization enabled). It is recommended to enable this feature for custom Analyzers.
• position: enumerate the tokens for position-dependent queries. Required for PHRASE(), NGRAM_MATCH(), and OFFSET_INFO(). If present, then the frequency feature is also required.
• offset: enable search highlighting capabilities (Enterprise Edition only). Required for OFFSET_INFO(). If present, then the position and frequency features are also required.

Analyzer Properties

The valid attributes/values for the properties are dependant on what type is used. For example, the delimiter type needs to know the desired delimiting character(s), whereas the text type takes a locale, stop-words and more.

identity

An Analyzer applying the identity transformation, i.e. returning the input unmodified.

It does not support any properties and will ignore them.

Examples

Applying the identity Analyzers does not perform any transformations, hence the input is returned unaltered:

db._query(`RETURN TOKENS("UPPER lower dïäcríticš", "identity")`).toArray();

[ 
  [ 
    "UPPER lower dïäcríticš" 
  ] 
]

delimiter

An Analyzer capable of breaking up delimited text into tokens.

It follows RFC 4180  but without starting new records on newlines and letting you freely choose the delimiter. You can wrap tokens in the input string in double quote marks to quote the delimiter. For example, a delimiter Analyzer that uses , as delimiter and an input string of foo,"bar,baz" results in the tokens foo and bar,baz instead of foo, bar, and baz.

The properties allowed for this Analyzer are an object with the following attributes:

• delimiter (string): the delimiting character or character sequence. The whole string is considered as one delimiter.

Examples

Split comma-separated text into tokens but do not split quoted fields

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("delimiter_csv", "delimiter", {
  delimiter: ","
}, []);
db._query(`RETURN TOKENS('foo,bar,baz,"bar,baz"', "delimiter_csv")`).toArray();

[ 
  [ 
    "foo", 
    "bar", 
    "baz", 
    "bar,baz" 
  ] 
]

Split input strings into tokens at every character sequence of hyphen-minus, right angled bracket, and a space

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("delimiter_arrow", "delimiter", {
  delimiter: "-> "
}, []);
db._query(`RETURN TOKENS("some-> hand-picked-> words", "delimiter_arrow")`).toArray();

[ 
  [ 
    "some", 
    "hand-picked", 
    "words" 
  ] 
]

multi_delimiter

An Analyzer capable of breaking up text into tokens using multiple delimiters.

Unlike with the delimiter Analyzer, the multi_delimiter Analyzer does not support quoting fields.

The properties allowed for this Analyzer are an object with the following attributes:

• delimiters (array): a list of strings of which each is considered as one delimiter that can be one or multiple characters long. The delimiters must not overlap, which means that a delimiter cannot be a prefix of another delimiter.

Examples

Split at delimiting characters , and ;, as well as the character sequence || but not a single | character

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("delimiter_multiple", "multi_delimiter", {
  delimiters: [",", ";", "||"]
}, []);
db._query(`RETURN TOKENS("differently,delimited;words||one|token", "delimiter_multiple")`).toArray();

[ 
  [ 
    "differently", 
    "delimited", 
    "words", 
    "one|token" 
  ] 
]

Double quote marks have no effect on the splitting

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("delimiter_noquote", "multi_delimiter", {
  delimiters: [","]
}, []);
db._query(`RETURN TOKENS('foo,bar,baz,"bar,baz"', "delimiter_noquote")`).toArray();

[ 
  [ 
    "foo", 
    "bar", 
    "baz", 
    "\"bar", 
    "baz\"" 
  ] 
]

stem

An Analyzer capable of stemming the text, treated as a single token, for supported languages.

The properties allowed for this Analyzer are an object with the following attributes:

• locale (string): a locale in the format language, e.g. "de" or "en". The locale is forwarded to the Snowball stemmer without checks. An invalid locale does not prevent the creation of the Analyzer. Also see Supported Languages.

Examples

Apply stemming to the input string as a whole:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("stem_en", "stem", {
  locale: "en"
}, []);
db._query(`RETURN TOKENS("databases", "stem_en")`).toArray();

[ 
  [ 
    "databas" 
  ] 
]

norm

An Analyzer capable of normalizing the text, treated as a single token, i.e. case conversion and accent removal.

The properties allowed for this Analyzer are an object with the following attributes:

• locale (string): a locale in the format language[_COUNTRY][_VARIANT] (square brackets denote optional parts), e.g. "de", "en_US", or "es__TRADITIONAL". See the ICU Documentation  for details. The locale is forwarded to ICU without checks. An invalid locale does not prevent the creation of the Analyzer. Also see Supported Languages.
• accent (boolean, optional):
  • true to preserve accented characters (default)
  • false to convert accented characters to their base characters
• case (string, optional):
  • "lower" to convert to all lower-case characters
  • "upper" to convert to all upper-case characters
  • "none" to not change character case (default)

Examples

Convert input string to all upper-case characters:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("norm_upper", "norm", {
  locale: "en",
  case: "upper"
}, []);
db._query(`RETURN TOKENS("UPPER lower dïäcríticš", "norm_upper")`).toArray();

[ 
  [ 
    "UPPER LOWER DÏÄCRÍTICŠ" 
  ] 
]

Convert accented characters to their base characters:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("norm_accent", "norm", {
  locale: "en",
  accent: false
}, []);
db._query(`RETURN TOKENS("UPPER lower dïäcríticš", "norm_accent")`).toArray();

[ 
  [ 
    "UPPER lower diacritics" 
  ] 
]

Convert input string to all lower-case characters and remove diacritics:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("norm_accent_lower", "norm", {
  locale: "en",
  accent: false,
  case: "lower"
}, []);
db._query(`RETURN TOKENS("UPPER lower dïäcríticš", "norm_accent_lower")`).toArray();

[ 
  [ 
    "upper lower diacritics" 
  ] 
]

ngram

An Analyzer capable of producing n-grams from a specified input in a range of min..max (inclusive). Can optionally preserve the original input.

This Analyzer type can be used to implement substring matching. Note that it slices the input based on bytes and not characters by default (streamType). The “binary” mode supports single-byte characters only; multi-byte UTF-8 characters raise an Invalid UTF-8 sequence query error.

The properties allowed for this Analyzer are an object with the following attributes:

• min (number): unsigned integer for the minimum n-gram length
• max (number): unsigned integer for the maximum n-gram length
• preserveOriginal (boolean):
  • true to include the original value as well
  • false to produce the n-grams based on min and max only
• startMarker (string, optional): this value will be prepended to n-grams which include the beginning of the input. Can be used for matching prefixes. Choose a character or sequence as marker which does not occur in the input.
• endMarker (string, optional): this value will be appended to n-grams which include the end of the input. Can be used for matching suffixes. Choose a character or sequence as marker which does not occur in the input.
• streamType (string, optional): type of the input stream
  • "binary": one byte is considered as one character (default)
  • "utf8": one Unicode codepoint is treated as one character

Examples

With min = 4 and max = 5, the Analyzer will produce the following n-grams for the input string "foobar":

• "foob"
• "fooba"
• "foobar" (if preserveOriginal is enabled)
• "ooba"
• "oobar"
• "obar"

An input string "foo" will not produce any n-gram unless preserveOriginal is enabled, because it is shorter than the min length of 4.

Above example but with startMarker = "^" and endMarker = "$" would produce the following:

• "^foob"
• "^fooba"
• "^foobar" (if preserveOriginal is enabled)
• "foobar$" (if preserveOriginal is enabled)
• "ooba"
• "oobar$"
• "obar$"

Create and use a trigram Analyzer with preserveOriginal disabled:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("trigram", "ngram", {
  min: 3,
  max: 3,
  preserveOriginal: false,
  streamType: "utf8"
}, []);
db._query(`RETURN TOKENS("foobar", "trigram")`).toArray();

[ 
  [ 
    "foo", 
    "oob", 
    "oba", 
    "bar" 
  ] 
]

Create and use a bigram Analyzer with preserveOriginal enabled and with start and stop markers:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("bigram_markers", "ngram", {
  min: 2,
  max: 2,
  preserveOriginal: true,
  startMarker: "^",
  endMarker: "$",
  streamType: "utf8"
}, []);
db._query(`RETURN TOKENS("foobar", "bigram_markers")`).toArray();

[ 
  [ 
    "^fo", 
    "^foobar", 
    "foobar$", 
    "oo", 
    "ob", 
    "ba", 
    "ar$" 
  ] 
]

text

An Analyzer capable of breaking up strings into individual words while also optionally filtering out stop-words, extracting word stems, applying case conversion and accent removal.

The properties allowed for this Analyzer are an object with the following attributes:

• locale (string): a locale in the format language[_COUNTRY][_VARIANT] (square brackets denote optional parts), e.g. "de", "en_US", or "es__TRADITIONAL". See the ICU Documentation  for details. The locale is forwarded to ICU without checks. An invalid locale does not prevent the creation of the Analyzer. Also see Supported Languages.

• accent (boolean, optional):

  • true to preserve accented characters
  • false to convert accented characters to their base characters (default)

• case (string, optional):

  • "lower" to convert to all lower-case characters (default)
  • "upper" to convert to all upper-case characters
  • "none" to not change character case

• stemming (boolean, optional):

  • true to apply stemming on returned words (default)
  • false to leave the tokenized words as-is

• edgeNgram (object, optional): if present, then edge n-grams are generated for each token (word). That is, the start of the n-gram is anchored to the beginning of the token, whereas the ngram Analyzer would produce all possible substrings from a single input token (within the defined length restrictions). Edge n-grams can be used to cover word-based auto-completion queries with an index, for which you should set the following other options: accent: false, case: "lower" and most importantly stemming: false.

  • min (number, optional): minimal n-gram length
  • max (number, optional): maximal n-gram length
  • preserveOriginal (boolean, optional): whether to include the original token even if its length is less than min or greater than max

• stopwords (array, optional): an array of strings with words to omit from result. Default: load words from stopwordsPath. To disable stop-word filtering provide an empty array []. If both stopwords and stopwordsPath are provided then both word sources are combined.

• stopwordsPath (string, optional): path with a language sub-directory (e.g. en for a locale en_US) containing files with words to omit. Each word has to be on a separate line. Everything after the first whitespace character on a line will be ignored and can be used for comments. The files can be named arbitrarily and have any file extension (or none).

  Default: if no path is provided then the value of the environment variable IRESEARCH_TEXT_STOPWORD_PATH is used to determine the path, or if it is undefined then the current working directory is assumed. If the stopwords attribute is provided then no stop-words are loaded from files, unless an explicit stopwordsPat is also provided.

  Note that if the stopwordsPath cannot be accessed, is missing language sub-directories or has no files for a language required by an Analyzer, then the creation of a new Analyzer is refused. If such an issue is discovered for an existing Analyzer during startup then the server will abort with a fatal error.

The Analyzer uses a fixed order of operations:

1. Tokenization
2. Accent removal (if accent is set to false)
3. Case conversion (unless case is set to none)
4. Stop word removal (if any are defined)
5. Word stemming (if stemming is set to true)

If you require a different order, consider using a pipeline Analyzer.

Stop words are removed after case/accent operations but before stemming. The reason is that stemming could map multiple words to the same one, and you would not be able to filter out specific words only.

The case/accent operations are not applied to the stop words for performance reasons. You need to pre-process them accordingly, for example, using the TOKENS() function with a text Analyzer that has the same locale, case, and accent settings as the planned text Analyzer, but with stemming set to false and stopwords set to [].

Examples

The built-in text_en Analyzer has stemming enabled (note the word endings):

db._query(`RETURN TOKENS("Crazy fast NoSQL-database!", "text_en")`).toArray();

[ 
  [ 
    "crazi", 
    "fast", 
    "nosql", 
    "databas" 
  ] 
]

You may create a custom Analyzer with the same configuration but with stemming disabled like this:

var analyzers = require("@arangodb/analyzers")
var a = analyzers.save("text_en_nostem", "text", {
  locale: "en",
  case: "lower",
  accent: false,
  stemming: false,
  stopwords: []
}, [])
db._query(`RETURN TOKENS("Crazy fast NoSQL-database!", "text_en_nostem")`).toArray();

[ 
  [ 
    "crazy", 
    "fast", 
    "nosql", 
    "database" 
  ] 
]

Custom text Analyzer with the edge n-grams capability and normalization enabled, stemming disabled and "the" defined as stop-word to exclude it:

var a = analyzers.save("text_edge_ngrams", "text", {
  edgeNgram: { min: 3, max: 8, preserveOriginal: true },
  locale: "en",
  case: "lower",
  accent: false,
  stemming: false,
  stopwords: [ "the" ]
}, []);
db._query(`RETURN TOKENS(
  "The quick brown fox jumps over the dogWithAVeryLongName",
  "text_edge_ngrams"
)`).toArray();

[ 
  [ 
    "qui", 
    "quic", 
    "quick", 
    "bro", 
    "brow", 
    "brown", 
    "fox", 
    "jum", 
    "jump", 
    "jumps", 
    "ove", 
    "over", 
    "dog", 
    "dogw", 
    "dogwi", 
    "dogwit", 
    "dogwith", 
    "dogwitha", 
    "dogwithaverylongname" 
  ] 
]

collation

Introduced in: v3.9.0

An Analyzer capable of converting the input into a set of language-specific tokens. This makes comparisons follow the rules of the respective language, most notable in range queries against Views.

For example, the Swedish alphabet has 29 letters: a to z plus å, ä, and ö, in that order. Using a Swedish locale (like sv), the sorting order is å after z, whereas using an English locale (like en), it is å after a. This impacts queries with SEARCH expressions like doc.text < "c", excluding å when using a Swedish locale but including it when using an English locale.

The properties allowed for this Analyzer are an object with the following attributes:

• locale (string): a locale in the format language[_COUNTRY][_VARIANT][@keywords] (square brackets denote optional parts), e.g. "de", "en_US", "es__TRADITIONAL", or fr@collation=phonebook. See the ICU Documentation  for details. The locale is forwarded to ICU without checks. An invalid locale does not prevent the creation of the Analyzer. Also see Supported Languages.

Examples

In Swedish, the letter å (note the small circle above the a) comes after z. Other languages treat it like a regular a, putting it before b. Below example creates two collation Analyzers, one with an English locale (en) and one with a Swedish locale (sv). It then demonstrates the difference in alphabetical order using a simple range query that returns letters before c:

var analyzers = require("@arangodb/analyzers");
var en = analyzers.save("collation_en", "collation", { locale: "en" }, []);
var sv = analyzers.save("collation_sv", "collation", { locale: "sv" }, []);
var test = db._create("test");
var docs = db.test.save([
  { text: "a" },
  { text: "å" },
  { text: "b" },
  { text: "z" },
]);
var view = db._createView("view", "arangosearch",
  { links: { test: { analyzers: [ "collation_en", "collation_sv" ], includeAllFields: true }}});
db._query("FOR doc IN view SEARCH ANALYZER(doc.text < TOKENS('c', 'collation_en')[0], 'collation_en') RETURN doc.text").toArray();
db._query("FOR doc IN view SEARCH ANALYZER(doc.text < TOKENS('c', 'collation_sv')[0], 'collation_sv') RETURN doc.text").toArray();

[ 
  "a", 
  "å", 
  "b" 
]

[ 
  "a", 
  "b" 
]

aql

Introduced in: v3.8.0

An Analyzer capable of running a restricted AQL query to perform data manipulation / filtering.

The query must not access the storage engine. This means no FOR loops over collections or Views, no use of the DOCUMENT() function, no graph traversals. AQL functions are allowed as long as they do not involve Analyzers (TOKENS(), PHRASE(), NGRAM_MATCH(), ANALYZER() etc.) or data access, and if they can be run on DB-Servers in case of a cluster deployment. User-defined functions are not permitted.

The input data is provided to the query via a bind parameter @param. It is always a string. The AQL query is invoked for each token in case of multiple input tokens, such as an array of strings.

The output can be one or multiple tokens (top-level result elements). They get converted to the configured returnType, either booleans, numbers or strings (default).

The properties allowed for this Analyzer are an object with the following attributes:

• queryString (string): AQL query to be executed
• collapsePositions (boolean):
  • true: set the position to 0 for all members of the query result array
  • false (default): set the position corresponding to the index of the result array member
• keepNull (boolean):
  • true (default): treat null like an empty string
  • false: discard nulls from View index. Can be used for index filtering (i.e. make your query return null for unwanted data). Note that empty results are always discarded.
• batchSize (integer): number between 1 and 1000 (default = 1) that determines the batch size for reading data from the query. In general, a single token is expected to be returned. However, if the query is expected to return many results, then increasing batchSize trades memory for performance.
• memoryLimit (integer): memory limit for query execution in bytes. (default is 1048576 = 1Mb) Maximum is 33554432U (32Mb)
• returnType (string): data type of the returned tokens. If the indicated type does not match the actual type then an implicit type conversion is applied (see TO_STRING(), TO_NUMBER(), TO_BOOL())
  • "string" (default): convert emitted tokens to strings
  • "number": convert emitted tokens to numbers
  • "bool": convert emitted tokens to booleans

Examples

Soundex Analyzer for a phonetically similar term search:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("soundex", "aql", { queryString: "RETURN SOUNDEX(@param)" }, []);
db._query("RETURN TOKENS('ArangoDB', 'soundex')").toArray();

[ 
  [ 
    "A652" 
  ] 
]

Concatenating Analyzer for conditionally adding a custom prefix or suffix:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("concat", "aql", { queryString:
  "RETURN LOWER(LEFT(@param, 5)) == 'inter' ? CONCAT(@param, 'ism') : CONCAT('inter', @param)"
}, []);
db._query("RETURN TOKENS('state', 'concat')").toArray();
db._query("RETURN TOKENS('international', 'concat')").toArray();

[ 
  [ 
    "interstate" 
  ] 
]

[ 
  [ 
    "internationalism" 
  ] 
]

Filtering Analyzer that ignores unwanted data based on the prefix "ir", with keepNull: false and explicitly returning null:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("filter", "aql", { keepNull: false, queryString:
  "RETURN LOWER(LEFT(@param, 2)) == 'ir' ? null : @param"
}, []);
db._query("RETURN TOKENS('regular', 'filter')").toArray();
db._query("RETURN TOKENS('irregular', 'filter')").toArray();

[ 
  [ 
    "regular" 
  ] 
]

[ 
  [ ] 
]

Filtering Analyzer that discards unwanted data based on the prefix "ir", using a filter for an empty result, which is discarded from the View index even without keepNull: false:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("filter", "aql", { queryString:
  "FILTER LOWER(LEFT(@param, 2)) != 'ir' RETURN @param"
}, []);
var coll = db._create("coll");
var doc1 = db.coll.save({ value: "regular" });
var doc2 = db.coll.save({ value: "irregular" });
var view = db._createView("view", "arangosearch",
  { links: { coll: { fields: { value: { analyzers: ["filter"] }}}}})
db._query("FOR doc IN view SEARCH ANALYZER(doc.value IN ['regular', 'irregular'], 'filter') RETURN doc").toArray();

[ 
  { 
    "_key" : "85148", 
    "_id" : "coll/85148", 
    "_rev" : "_hg5lGPy--_", 
    "value" : "regular" 
  } 
]

Custom tokenization with collapsePositions on and off: The input string "A-B-C-D" is split into an array of strings ["A", "B", "C", "D"]. The position metadata (as used by the PHRASE() function) is set to 0 for all four strings if collapsePositions is enabled. Otherwise the position is set to the respective array index, 0 for "A", 1 for "B" and so on.

var analyzers = require("@arangodb/analyzers");
var a1 = analyzers.save("collapsed", "aql", { collapsePositions: true, queryString:
  "FOR d IN SPLIT(@param, '-') RETURN d"
}, ["frequency", "position"]);
var a2 = analyzers.save("uncollapsed", "aql", { collapsePositions: false, queryString:
  "FOR d IN SPLIT(@param, '-') RETURN d"
}, ["frequency", "position"]);
var coll = db._create("coll");
var doc = db.coll.save({ text: "A-B-C-D" });
var view = db._createView("view", "arangosearch",
  { links: { coll: { analyzers: [ "collapsed", "uncollapsed" ], includeAllFields: true }}});
db._query("FOR d IN view SEARCH PHRASE(d.text, {TERM: 'B'}, 1, {TERM: 'D'}, 'uncollapsed') RETURN d");
db._query("FOR d IN view SEARCH PHRASE(d.text, {TERM: 'B'}, -1, {TERM: 'D'}, 'uncollapsed') RETURN d");
db._query("FOR d IN view SEARCH PHRASE(d.text, {TERM: 'B'}, 1, {TERM: 'D'}, 'collapsed') RETURN d");
db._query("FOR d IN view SEARCH PHRASE(d.text, {TERM: 'B'}, -1, {TERM: 'D'}, 'collapsed') RETURN d");

[object ArangoQueryCursor, count: 1, cached: false, hasMore: false]

[ 
  { 
    "_key" : "85329", 
    "_id" : "coll/85329", 
    "_rev" : "_hg5lGYW--_", 
    "text" : "A-B-C-D" 
  } 
]

[object ArangoQueryCursor, count: 1, cached: false, hasMore: false]

[ 
  { 
    "_key" : "85329", 
    "_id" : "coll/85329", 
    "_rev" : "_hg5lGYW--_", 
    "text" : "A-B-C-D" 
  } 
]

[object ArangoQueryCursor, count: 0, cached: false, hasMore: false]

[object ArangoQueryCursor, count: 1, cached: false, hasMore: false]

[ 
  { 
    "_key" : "85329", 
    "_id" : "coll/85329", 
    "_rev" : "_hg5lGYW--_", 
    "text" : "A-B-C-D" 
  } 
]

The position data is not directly exposed, but we can see its effects through the PHRASE() function. There is one token between "B" and "D" to skip in case of uncollapsed positions. With positions collapsed, both are in the same position, thus there is negative one to skip to match the tokens.

pipeline

Introduced in: v3.8.0

An Analyzer capable of chaining effects of multiple Analyzers into one. The pipeline is a list of Analyzers, where the output of an Analyzer is passed to the next for further processing. The final token value is determined by last Analyzer in the pipeline.

The Analyzer is designed for cases like the following:

• Normalize text for a case insensitive search and apply n-gram tokenization
• Split input with delimiter Analyzer, followed by stemming with the stem Analyzer

The properties allowed for this Analyzer are an object with the following attributes:

• pipeline (array): an array of Analyzer definition-like objects with type and properties attributes

Examples

Normalize to all uppercase and compute bigrams:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("ngram_upper", "pipeline", { pipeline: [
  { type: "norm", properties: { locale: "en", case: "upper" } },
  { type: "ngram", properties: { min: 2, max: 2, preserveOriginal: false, streamType: "utf8" } }
] }, []);
db._query(`RETURN TOKENS("Quick brown foX", "ngram_upper")`).toArray();

[ 
  [ 
    "QU", 
    "UI", 
    "IC", 
    "CK", 
    "K ", 
    " B", 
    "BR", 
    "RO", 
    "OW", 
    "WN", 
    "N ", 
    " F", 
    "FO", 
    "OX" 
  ] 
]

Split at delimiting character ,, then stem the tokens:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("delimiter_stem", "pipeline", { pipeline: [
  { type: "delimiter", properties: { delimiter: "," } },
  { type: "stem", properties: { locale: "en" } }
] }, []);
db._query(`RETURN TOKENS("delimited,stemmable,words", "delimiter_stem")`).toArray();

[ 
  [ 
    "delimit", 
    "stemmabl", 
    "word" 
  ] 
]

stopwords

Introduced in: v3.8.1

An Analyzer capable of removing specified tokens from the input.

It uses binary comparison to determine if an input token should be discarded. It checks for exact matches. If the input contains only a substring that matches one of the defined stop words, then it is not discarded. Longer inputs such as prefixes of stop words are also not discarded.

The properties allowed for this Analyzer are an object with the following attributes:

• stopwords (array): array of strings that describe the tokens to be discarded. The interpretation of each string depends on the value of the hex parameter.
• hex (boolean): If false (default), then each string in stopwords is used verbatim. If true, then the strings need to be hex-encoded. This allows for removing tokens that contain non-printable characters. To encode UTF-8 strings to hex strings you can use e.g.
  • AQL:

    FOR token IN ["and","the"] RETURN TO_HEX(token)

  • arangosh / Node.js:

    ["and","the"].map(token => Buffer(token).toString("hex"))

  • Modern browser:

    ["and","the"].map(token => Array.from(new TextEncoder().encode(token), byte => byte.toString(16).padStart(2, "0")).join(""))

Examples

Create and use a stopword Analyzer that removes the tokens and and the. The stop word array with hex-encoded strings for this looks like ["616e64","746865"] (a = 0x61, n = 0x6e, d = 0x64 and so on). Note that a and theater are not removed, because there is no exact match with either of the stop words and and the:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("stop", "stopwords", {
  stopwords: ["616e64","746865"], hex: true
}, []);
db._query("RETURN FLATTEN(TOKENS(SPLIT('the fox and the dog and a theater', ' '), 'stop'))").toArray();

[ 
  [ 
    "fox", 
    "dog", 
    "a", 
    "theater" 
  ] 
]

Create and use an Analyzer pipeline that normalizes the input (convert to lower-case and base characters) and then discards the stopwords and and the:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("norm_stop", "pipeline", { "pipeline": [
  { type: "norm", properties: { locale: "en", accent: false, case: "lower" } },
  { type: "stopwords", properties: { stopwords: ["and","the"], hex: false } },
]}, []);
db._query("RETURN FLATTEN(TOKENS(SPLIT('The fox AND the dog äñḏ a ţhéäter', ' '), 'norm_stop'))").toArray();

[ 
  [ 
    "fox", 
    "dog", 
    "a", 
    "theater" 
  ] 
]

segmentation

Introduced in: v3.9.0

An Analyzer capable of breaking up the input text into tokens in a language-agnostic manner as per Unicode Standard Annex #29 , making it suitable for mixed language strings. It can optionally preserve all non-whitespace or all characters instead of keeping alphanumeric characters only, as well as apply case conversion.

The properties allowed for this Analyzer are an object with the following attributes:

• break (string, optional):
  • "all": return all tokens
  • "alpha": return tokens composed of alphanumeric characters only (default). Alphanumeric characters are Unicode codepoints from the Letter and Number categories, see Unicode Technical Note #36 .
  • "graphic": return tokens composed of non-whitespace characters only. Note that the list of whitespace characters does not include line breaks:
    • U+0009 Character Tabulation
    • U+0020 Space
    • U+0085 Next Line
    • U+00A0 No-break Space
    • U+1680 Ogham Space Mark
    • U+2000 En Quad
    • U+2028 Line Separator
    • U+202F Narrow No-break Space
    • U+205F Medium Mathematical Space
    • U+3000 Ideographic Space
• case (string, optional):
  • "lower" to convert to all lower-case characters (default)
  • "upper" to convert to all upper-case characters
  • "none" to not change character case

Examples

Create different segmentation Analyzers to show the behavior of the different break options:

var analyzers = require("@arangodb/analyzers");
var all = analyzers.save("segment_all", "segmentation", { break: "all" }, []);
var alpha = analyzers.save("segment_alpha", "segmentation", { break: "alpha" }, []);
var graphic = analyzers.save("segment_graphic", "segmentation", { break: "graphic" }, []);
db._query(`LET str = 'Test\twith An_EMAIL-address+123@example.org\n蝴蝶。\u2028бутерброд'
  RETURN {
    "all": TOKENS(str, 'segment_all'),
    "alpha": TOKENS(str, 'segment_alpha'),
    "graphic": TOKENS(str, 'segment_graphic'),
  }
`).toArray();

[ 
  { 
    "all" : [ 
      "test", 
      "\t", 
      "with", 
      " ", 
      "an_email", 
      "-", 
      "address", 
      "+", 
      "123", 
      "@", 
      "example.org", 
      "\n", 
      "蝴", 
      "蝶", 
      "。", 
      "
", 
      "бутерброд" 
    ], 
    "alpha" : [ 
      "test", 
      "with", 
      "an_email", 
      "address", 
      "123", 
      "example.org", 
      "蝴", 
      "蝶", 
      "бутерброд" 
    ], 
    "graphic" : [ 
      "test", 
      "with", 
      "an_email", 
      "-", 
      "address", 
      "+", 
      "123", 
      "@", 
      "example.org", 
      "\n", 
      "蝴", 
      "蝶", 
      "。", 
      "бутерброд" 
    ] 
  } 
]

wildcard

Introduced in: v3.12.0

An Analyzer that creates n-grams to enable fast partial matching for wildcard queries if you have large string values, especially if you want to search for suffixes or substrings in the middle of strings (infixes) as opposed to prefixes.

It can apply an Analyzer of your choice before creating the n-grams, for example, to normalize text for case-insensitive and accent-insensitive search.

See Wildcard Search with ArangoSearch for an example of how to use this Analyzer with Views, and Inverted indexes for an example using a standalone inverted index.

The properties allowed for this Analyzer are an object with the following attributes:

• ngramSize (number, required): unsigned integer for the n-gram length, needs to be at least 2. It can be greater than the substrings between wildcards that you want to search for, e.g. 4 with an expected search pattern of %up%if%ref% (substrings of length 2 and 3 between %), but this leads to a slower search (for ref% with post-validation using the ICU regular expression engine). A value of 3 is a good default, 2 is better for short strings
• analyzer (object, optional): an Analyzer definition-like objects with type and properties attributes

You cannot set the offset Analyzer features for this Analyzer. You can create the Analyzer without features, or use ["frequency", "position"] which uses more memory but doesn’t need to read stored column values for prefix, suffix, infix, and exact queries (something%, %something, %something%, something, but not some%thing or %some%thin_)

The set of n-grams that are created for each input token includes one start n-gram and ngramSize - 1 end n-grams using "\uFFFD" as start marker respectively end marker. For example, with an ngramSize of 3 and the input "graph", the generated n-grams are:

• "�gr"
• "gra"
• "rap"
• "aph"
• "ph�"
• "h�"

Examples

Create a wildcard Analyzer with an n-gram size of 4 and without enabling Analyzer features:

var analyzers = require("@arangodb/analyzers");
var analyzerWildcard = analyzers.save("wildcard_4", "wildcard", { ngramSize: 4 }, []);

db._query(`RETURN TOKENS("The quick brown Foxx", "wildcard_4")`).toArray();

[ 
  [ 
    "�The", 
    "The ", 
    "he q", 
    "e qu", 
    " qui", 
    "quic", 
    "uick", 
    "ick ", 
    "ck b", 
    "k br", 
    " bro", 
    "brow", 
    "rown", 
    "own ", 
    "wn F", 
    "n Fo", 
    " Fox", 
    "Foxx", 
    "oxx�", 
    "xx�", 
    "x�" 
  ] 
]

Use a norm Analyzer in a wildcard Analyzer to normalize the input to lowercase and convert accented characters to their base characters before creating trigrams from the string:

var analyzers = require("@arangodb/analyzers");
var analyzerWildcard = analyzers.save("wildcard_3", "wildcard", { ngramSize: 3, "analyzer": {
  type: "norm", properties: { locale: "en", accent: false, case: "lower" }
} }, ["frequency","position"]);

db._query(`RETURN TOKENS("The quick brown Foxx", "wildcard_3")`).toArray();

[ 
  [ 
    "�th", 
    "the", 
    "he ", 
    "e q", 
    " qu", 
    "qui", 
    "uic", 
    "ick", 
    "ck ", 
    "k b", 
    " br", 
    "bro", 
    "row", 
    "own", 
    "wn ", 
    "n f", 
    " fo", 
    "fox", 
    "oxx", 
    "xx�", 
    "x�" 
  ] 
]

minhash

ArangoDB Enterprise Edition ArangoGraph

Introduced in: v3.10.0

An Analyzer that computes so called MinHash signatures using a locality-sensitive hash function. It applies an Analyzer of your choice before the hashing, for example, to break up text into words.

The properties allowed for this Analyzer are an object with the following attributes:

• analyzer (object, required): an Analyzer definition-like objects with type and properties attributes
• numHashes (number, required): the size of the MinHash signature. Must be greater or equal to 1. The signature size defines the probabilistic error (err = rsqrt(numHashes)). For an error amount that does not exceed 5% (0.05), use a size of 1 / (0.05 * 0.05) = 400.

Examples

Create a minhash Analyzers:

var analyzers = require("@arangodb/analyzers");
var analyzerMinHash = analyzers.save("minhash5", "minhash", { analyzer: { type: "segmentation", properties: { break: "alpha", case: "lower" } }, numHashes: 5 }, []);
var analyzerSegment = analyzers.save("segment", "segmentation", { break: "alpha", case: "lower" }, []);
db._query(`
  LET str1 = "The quick brown fox jumps over the lazy dog."
  LET str2 = "The fox jumps over the crazy dog."
  RETURN {
    approx: JACCARD(TOKENS(str1, "minhash5"), TOKENS(str2, "minhash5")),
    actual: JACCARD(TOKENS(str1, "segment"), TOKENS(str2, "segment"))
  }`).toArray();

[ 
  { 
    "approx" : 0.42857142857142855, 
    "actual" : 0.5555555555555556 
  } 
]

classification

ArangoDB Enterprise Edition ArangoGraph

Introduced in: v3.10.0

An Analyzer capable of classifying tokens in the input text.

It applies a user-provided supervised fastText  word embedding model to classify the input text. It is able to classify individual tokens as well as entire inputs.

The properties allowed for this Analyzer are an object with the following attributes:

• model_location (string): the on-disk path to the trained fastText supervised model. Note: if you are running this in an ArangoDB cluster, this model must exist on every machine in the cluster.
• top_k (number, optional): the number of class labels that will be produced per input (default: 1).
• threshold (number, optional): the probability threshold for which a label will be assigned to an input. A fastText model produces a probability per class label, and this is what will be filtered (default: 0.99).

Examples

Create and use a classification Analyzer with a stored “cooking” classifier to classify items.

var analyzers = require("@arangodb/analyzers");
var classifier_single = analyzers.save("classifier_single", "classification", { "model_location": "/path_to_local_fasttext_model_directory/model_cooking.bin" }, []);
var classifier_top_two = analyzers.save("classifier_double", "classification", { "model_location": "/path_to_local_fasttext_model_directory/model_cooking.bin", "top_k": 2 }, []);
db._query(`LET str = "Which baking dish is best to bake a banana bread ?"
    RETURN {
      "all": TOKENS(str, "classifier_single"),
      "double": TOKENS(str, "classifier_double")
    }
  `);

[
  {
    "all" : [
      "__label__baking"
    ],
    "double" : [
      "__label__baking",
      "__label__bananas"
    ]
  }
]

nearest_neighbors

ArangoDB Enterprise Edition ArangoGraph

Introduced in: v3.10.0

An Analyzer capable of finding nearest neighbors of tokens in the input.

It applies a user-provided supervised fastText  word embedding model to retrieve nearest neighbor tokens in the text. It is able to find neighbors of individual tokens as well as entire input strings. For entire input strings, the Analyzer will return nearest neighbors for each token within the input string.

The properties allowed for this Analyzer are an object with the following attributes:

• model_location (string): the on-disk path to the trained fastText supervised model. Note: if you are running this in an ArangoDB cluster, this model must exist on every machine in the cluster.
• top_k (number, optional): the number of class labels that will be produced per input (default: 1).

Examples

Create and use a nearest_neighbors Analyzer with a stored “cooking” classifier to find similar terms.

var analyzers = require("@arangodb/analyzers");
var nn_single = analyzers.save("nn_single", "nearest_neighbors", { "model_location": "/path_to_local_fasttext_model_directory/model_cooking.bin" }, []);
var nn_top_two = analyzers.save("nn_double", "nearest_neighbors", { "model_location": "/path_to_local_fasttext_model_directory/model_cooking.bin", "top_k": 2 }, []);
db._query(`LET str = "salt, oil"
    RETURN {
      "all": TOKENS(str, "nn_single"),
      "double": TOKENS(str, "nn_double")
    }
  `);

[
  {
    "all" : [
      "pepper",
      "olive"
    ],
    "double" : [
      "pepper",
      "table",
      "olive",
      "avocado"
    ]
  }
]

geojson

Introduced in: v3.8.0

An Analyzer capable of breaking up a GeoJSON object or coordinate array in [longitude, latitude] order into a set of indexable tokens for further usage with ArangoSearch Geo functions.

The Analyzer can be used for two different coordinate representations:

• a GeoJSON feature like a Point or Polygon, using a JSON object like the following:

  {
    "type": "Point",
    "coordinates": [ -73.97, 40.78 ] // [ longitude, latitude ]
  }

• a coordinate array with two numbers as elements in the following format:

  [ -73.97, 40.78 ] // [ longitude, latitude ]
  

The properties allowed for this Analyzer are an object with the following attributes:

• type (string, optional):

  • "shape" (default): index all GeoJSON geometry types (Point, Polygon etc.)
  • "centroid": compute and only index the centroid of the input geometry
  • "point": only index GeoJSON objects of type Point, ignore all other geometry types

• options (object, optional): options for fine-tuning geo queries. These options should generally remain unchanged

  • maxCells (number, optional): maximum number of S2 cells (default: 20)
  • minLevel (number, optional): the least precise S2 level (default: 4)
  • maxLevel (number, optional): the most precise S2 level (default: 23)

• legacy (boolean, optional): This option controls how GeoJSON Polygons are interpreted (introduced in v3.10.5). Also see Legacy Polygons and GeoJSON interpretation.

  • If legacy is true, the smaller of the two regions defined by a linear ring is interpreted as the interior of the ring and a ring can at most enclose half the Earth’s surface.
  • If legacy is false, the area to the left of the boundary ring’s path is considered to be the interior and a ring can enclose the entire surface of the Earth.

  The default is false.

  If you use geojson Analyzers and upgrade from a version below 3.10 to a version of 3.10 or higher, the interpretation of GeoJSON Polygons changes.

  If you have polygons in your data that mean to refer to a relatively small region but have the boundary running clockwise around the intended interior, they are interpreted as intended prior to 3.10, but from 3.10 onward, they are interpreted as “the other side” of the boundary.

  Whether a clockwise boundary specifies the complement of the small region intentionally or not cannot be determined automatically. Please test the new behavior manually.

  If you require the old behavior, upgrade to at least 3.10.5, drop your geojson Analyzers, and create new ones with legacy set to true. If these Analyzers are used in arangosearch Views, then they need to be dropped as well before dropping the Analyzers, and recreated after creating the new Analyzers.

You should not set any of the Analyzer features as they cannot be utilized for Geo Analyzers.

Examples

Create a collection with GeoJSON Points stored in an attribute location, a geojson Analyzer with default properties, and a View using the Analyzer. Then query for locations that are within a 3 kilometer radius of a given point and return the matched documents, including the calculated distance in meters. The stored coordinate pairs and the GEO_POINT() arguments are expected in longitude, latitude order:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("geo_json", "geojson", {}, []);
var coll = db._create("geo");
var docs = db.geo.save([
  { location: { type: "Point", coordinates: [6.937, 50.932] } },
  { location: { type: "Point", coordinates: [6.956, 50.941] } },
  { location: { type: "Point", coordinates: [6.962, 50.932] } },
]);
var view = db._createView("geo_view", "arangosearch", {
  links: {
    geo: {
      fields: {
        location: {
          analyzers: ["geo_json"]
        }
      }
    }
  }
});
db._query(`LET point = GEO_POINT(6.93, 50.94)
  FOR doc IN geo_view
    SEARCH ANALYZER(GEO_DISTANCE(doc.location, point) < 2000, "geo_json")
    RETURN MERGE(doc, { distance: GEO_DISTANCE(doc.location, point) })`).toArray();

[ 
  { 
    "_id" : "geo/86036", 
    "_key" : "86036", 
    "_rev" : "_hg5lHQK---", 
    "location" : { 
      "type" : "Point", 
      "coordinates" : [ 
        6.937, 
        50.932 
      ] 
    }, 
    "distance" : 1015.8355739436823 
  }, 
  { 
    "_id" : "geo/86037", 
    "_key" : "86037", 
    "_rev" : "_hg5lHQK--_", 
    "location" : { 
      "type" : "Point", 
      "coordinates" : [ 
        6.956, 
        50.941 
      ] 
    }, 
    "distance" : 1825.1307183571266 
  } 
]

geo_s2

ArangoDB Enterprise Edition ArangoGraph

Introduced in: v3.10.5

An Analyzer capable of breaking up a GeoJSON object or coordinate array in [longitude, latitude] order into a set of indexable tokens for further usage with ArangoSearch Geo functions.

The Analyzer is similar to the geojson Analyzer, but it internally uses a format for storing the geo-spatial data that is more efficient. You can choose between different formats to make a tradeoff between the size on disk, the precision, and query performance.

The Analyzer can be used for two different coordinate representations:

• a GeoJSON feature like a Point or Polygon, using a JSON object like the following:

  {
    "type": "Point",
    "coordinates": [ -73.97, 40.78 ] // [ longitude, latitude ]
  }

• a coordinate array with two numbers as elements in the following format:

  [ -73.97, 40.78 ] // [ longitude, latitude ]
  

The properties allowed for this Analyzer are an object with the following attributes:

• format (string, optional): the internal binary representation to use for storing the geo-spatial data in an index
  • "latLngDouble" (default): store each latitude and longitude value as an 8-byte floating-point value (16 bytes per coordinate pair). This format preserves numeric values exactly and is more compact than the VelocyPack format used by the geojson Analyzer.
  • "latLngInt": store each latitude and longitude value as an 4-byte integer value (8 bytes per coordinate pair). This is the most compact format but the precision is limited to approximately 1 to 10 centimeters.
  • "s2Point": store each longitude-latitude pair in the native format of Google S2 which is used for geo-spatial calculations (24 bytes per coordinate pair). This is not a particular compact format but it reduces the number of computations necessary when you execute geo-spatial queries. This format preserves numeric values exactly.
• type (string, optional):
  • "shape" (default): index all GeoJSON geometry types (Point, Polygon etc.)
  • "centroid": compute and only index the centroid of the input geometry
  • "point": only index GeoJSON objects of type Point, ignore all other geometry types
• options (object, optional): options for fine-tuning geo queries. These options should generally remain unchanged
  • maxCells (number, optional): maximum number of S2 cells (default: 20)
  • minLevel (number, optional): the least precise S2 level (default: 4)
  • maxLevel (number, optional): the most precise S2 level (default: 23)

You should not set any of the Analyzer features as they cannot be utilized for Geo Analyzers.

Examples

Create a collection with GeoJSON Points stored in an attribute location, a geo_s2 Analyzer with the latLngInt format, and a View using the Analyzer. Then query for locations that are within a 3 kilometer radius of a given point and return the matched documents, including the calculated distance in meters. The stored coordinate pairs and the GEO_POINT() arguments are expected in longitude, latitude order:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("geo_efficient", "geo_s2", { format: "latLngInt" }, []);
var coll = db._create("geo");
var docs = db.geo.save([
  { location: { type: "Point", coordinates: [6.937, 50.932] } },
  { location: { type: "Point", coordinates: [6.956, 50.941] } },
  { location: { type: "Point", coordinates: [6.962, 50.932] } },
]);
var view = db._createView("geo_view", "arangosearch", {
  links: {
    geo: {
      fields: {
        location: {
          analyzers: ["geo_efficient"]
        }
      }
    }
  }
});
db._query(`LET point = GEO_POINT(6.93, 50.94)
  FOR doc IN geo_view
    SEARCH ANALYZER(GEO_DISTANCE(doc.location, point) < 2000, "geo_efficient")
    RETURN MERGE(doc, { distance: GEO_DISTANCE(doc.location, point) })`).toArray();

[ 
  { 
    "_id" : "geo/86066", 
    "_key" : "86066", 
    "_rev" : "_hg5lHTO---", 
    "location" : { 
      "type" : "Point", 
      "coordinates" : [ 
        6.937, 
        50.932 
      ] 
    }, 
    "distance" : 1015.8355739436823 
  }, 
  { 
    "_id" : "geo/86067", 
    "_key" : "86067", 
    "_rev" : "_hg5lHTO--_", 
    "location" : { 
      "type" : "Point", 
      "coordinates" : [ 
        6.956, 
        50.941 
      ] 
    }, 
    "distance" : 1825.1307183571266 
  } 
]

The calculated distance between the reference point and the point stored in the second document is 1825.1307…. If you change the search condition to < 1825.1303, then the document is still returned despite the distance being higher than this value. This is due to the precision limitations of the latLngInt format. The returned distance is unaffected because it is calculated independent of the Analyzer. If you use either of the other two formats which preserve the exact coordinate values, then the document is filtered out as expected.

geopoint

Introduced in: v3.8.0

An Analyzer capable of breaking up a coordinate array in [latitude, longitude] order or a JSON object describing a coordinate pair using two separate attributes into a set of indexable tokens for further usage with ArangoSearch Geo functions.

The Analyzer can be used for two different coordinate representations:

• an array with two numbers as elements in the following format:

  [ 40.78, -73.97 ] // [ latitude, longitude ]
  

• two separate numeric attributes, one for latitude and one for longitude, as shown below:

  { "location": { "lat": 40.78, "lon": -73.97 } }

  The attributes cannot be at the top level of the document, but must be nested like in the example, so that the Analyzer can be defined for the field location with the Analyzer properties { "latitude": ["lat"], "longitude": ["lon"] }.

The properties allowed for this Analyzer are an object with the following attributes:

• latitude (array, optional): array of strings that describes the attribute path of the latitude value relative to the field for which the Analyzer is defined in the View
• longitude (array, optional): array of strings that describes the attribute path of the longitude value relative to the field for which the Analyzer is defined in the View
• options (object, optional): options for fine-tuning geo queries. These options should generally remain unchanged
  • maxCells (number, optional): maximum number of S2 cells (default: 20)
  • minLevel (number, optional): the least precise S2 level (default: 4)
  • maxLevel (number, optional): the most precise S2 level (default: 23)

You should not set any of the Analyzer features as they cannot be utilized for Geo Analyzers.

Examples

Create a collection with coordinate pairs stored in an attribute location, a geopoint Analyzer with default properties, and a View using the Analyzer. Then query for locations that are within a 3 kilometer radius of a given point. The stored coordinate pairs are in latitude, longitude order, but GEO_POINT() and GEO_DISTANCE() expect longitude, latitude order:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("geo_pair", "geopoint", {}, []);
var coll = db._create("geo");
var docs = db.geo.save([
  { location: [50.932, 6.937] },
  { location: [50.941, 6.956] },
  { location: [50.932, 6.962] },
]);
var view = db._createView("geo_view", "arangosearch", {
  links: {
    geo: {
      fields: {
        location: {
          analyzers: ["geo_pair"]
        }
      }
    }
  }
});
db._query(`LET point = GEO_POINT(6.93, 50.94)
  FOR doc IN geo_view
    SEARCH ANALYZER(GEO_DISTANCE(doc.location, point) < 2000, "geo_pair")
    RETURN MERGE(doc, { distance: GEO_DISTANCE([doc.location[1], doc.location[0]], point) })`).toArray();

[ 
  { 
    "_id" : "geo/86096", 
    "_key" : "86096", 
    "_rev" : "_hg5lHVa---", 
    "location" : [ 
      50.932, 
      6.937 
    ], 
    "distance" : 1015.8355739436823 
  }, 
  { 
    "_id" : "geo/86097", 
    "_key" : "86097", 
    "_rev" : "_hg5lHVa--_", 
    "location" : [ 
      50.941, 
      6.956 
    ], 
    "distance" : 1825.1307183571266 
  } 
]

Create a collection with coordinate pairs stored in an attribute location as separate nested attributes lat and lng, a geopoint Analyzer that specifies the attribute paths to the latitude and longitude attributes (relative to location attribute), and a View using the Analyzer. Then query for locations that are within a 3 kilometer radius of a given point:

var analyzers = require("@arangodb/analyzers");
var a = analyzers.save("geo_latlng", "geopoint", {
  latitude: ["lat"],
  longitude: ["lng"]
}, []);
var coll = db._create("geo");
var docs = db.geo.save([
  { location: { lat: 50.932, lng: 6.937 } },
  { location: { lat: 50.941, lng: 6.956 } },
  { location: { lat: 50.932, lng: 6.962 } },
]);
var view = db._createView("geo_view", "arangosearch", {
  links: {
    geo: {
      fields: {
        location: {
          analyzers: ["geo_latlng"]
        }
      }
    }
  }
});
db._query(`LET point = GEO_POINT(6.93, 50.94)
  FOR doc IN geo_view
    SEARCH ANALYZER(GEO_DISTANCE(doc.location, point) < 2000, "geo_latlng")
    RETURN MERGE(doc, { distance: GEO_DISTANCE([doc.location.lng, doc.location.lat], point) })`).toArray();

[ 
  { 
    "_id" : "geo/86126", 
    "_key" : "86126", 
    "_rev" : "_hg5lHXi--_", 
    "location" : { 
      "lat" : 50.932, 
      "lng" : 6.937 
    }, 
    "distance" : 1015.8355739436823 
  }, 
  { 
    "_id" : "geo/86127", 
    "_key" : "86127", 
    "_rev" : "_hg5lHXi--A", 
    "location" : { 
      "lat" : 50.941, 
      "lng" : 6.956 
    }, 
    "distance" : 1825.1307183571266 
  } 
]

Built-in Analyzers

There is a set of built-in Analyzers which are available by default for convenience and backward compatibility. They cannot be removed.

The identity Analyzer has no properties and the frequency and norm features. The Analyzers of type text all tokenize strings with stemming enabled, no stop words configured, accent removal and case conversion to lowercase turned on and the frequency, norm and position features

Note that locale, case, accent, stemming and stopwords are Analyzer properties. text_zh does not have actual stemming support for Chinese despite what the property value suggests.

Supported Languages

Tokenization and Normalization

Analyzers rely on ICU  for tokenization and normalization, which is language-dependent. The ICU data file icudtl.dat that ArangoDB ships with contains information for a lot of languages, which are technically all supported.

Setting an unsupported or invalid locale does not raise a warning or error. ICU will fall back to a locale without the requested variant, country, or script, or use its default locale if neither of the former is valid.

Stemming

Stemming support is provided by Snowball , which supports the following languages:

* Introduced in: v3.10.0