Search Definition reference

This document lists the syntax and content of search definitions, document types and fields. This is a reference, read search definitions first for an overview. Find an example at the end.

There must be at least one search definition (.sd) file containing a search element in an application package.

Search definition syntax

Throughout this document, a string in square brackets represents some argument. The whole string, including the brackets, is replaced by a concrete string in a search definition.

Constructs in search definitions have a regular syntax. Each element starts by the element identifier, possibly followed by the name of this particular occurrence of the element, possibly followed by a space-separated list of interleaved attribute names and attribute values, possibly followed by the element body. Thus, one will find elements of these varieties:

[element-identifier] : [element-body]
[element-identifier] [element-name] : [element-body]
[element-identifier] [element-name] [attribute-name] [attribute-value]
[element-identifier] [element-name] [attribute-name] [attribute-value] {
    [element-body]
}
Note that one-line element values starts by a colon and ends by newline. Multiline values (for fields supporting them) are any block of text enclosed in curly brackets. Comments may be inserted anywhere and start with a hash (#).

Search definition elements

A search definition must contain no more than one search clause - elements:

search
    document
        struct
            field
                match
        field
            alias
            attribute
            bolding
            id
            index
            indexing
            indexing-rewrite
            match
            normalizing
            query-command
            rank
            rank-type
            sorting
            stemming
            struct-field
                indexing
                match
                query-command
                struct-fieldsummary
                summary-to
            summary
            summary-to
            weight
            weightedset
        compression
    index
    field
    fieldset
    rank-profile
        match-phase
            attribute
            order
            max-hits
            diversity
                attribute
                min-groups
        first-phase
            keep-rank-count
            rank-score-drop-limit
            expression
        ignore-default-rank-features
        num-threads-per-search
        rank
        rank-type
        rank-features
        constants
        rank-properties
        second-phase
            expression
            rerank-count
        summary-features
    constant
    stemming
    document-summary
        summary
    annotation
        field
    import field

The root element of search definitions. A search definition describes how some data should be stored, indexed, ranked and presented in results. A search definition must be defined in a file named [search-definition-name].sd.

search [name] {
    [body]
}
The body is mandatory and may contain:

NameDescriptionOccurrence
document A document defined in this search definition. One
field A field not contained in the document. Use fields outside documents when you want to derive new field values to be placed in the indexing structure from document fields. Zero to many
fieldset A field set to provide a way to group document fields together for searching. When you query a field set, you will get results from all the fields in the field set. Zero to many
rank-profile An explicitly defined set of ranking settings. Zero to many
constant A constant tensor located in a file used for ranking. Zero to many
stemming The default stemming setting. Default is shortest. Not applicable to streaming search Zero or one
document-summary An explicitly defined document summary. Zero to many

document

Contained in search. Describes a document type. This can also be the root of the search definition, if the document is not to be searched directly. A document type may inherit the fields of one or more other document types. If no document types are explicitly inherited, the document inherits the generic document type.

document [name] inherits [name-list] {
    [body]
}
The document name is optional, it defaults to the containing search element's name. If there is no containing search element, the document name is required.

The inherits attribute is optional and has as value a comma-separated list of names of other document types.

The body of a document type is optional and may contain:

NameDescriptionOccurrence
struct A struct type definition for this document.Zero to many
fieldA field of this document. Zero to many
compression Specifies compression options for documents of this document type in storage. Zero to one

struct

Contained in document. Defines a composite type. A struct consists of zero or more fields that the user can access together as one. The struct has to be defined before it is used as a type in a field specification.

struct [name] {
    [body]
}
The struct name should not contain any underscores.

Please note that struct types is not supported by indexed search and the behavior of operations over struct fields is undefined. Searching in struct elements is supported by streaming search.

The body of a struct is optional and may contain:

NameDescriptionOccurrence
fieldA field of this struct. Zero to many

field

Contained in search, document, struct or annotation. Defines a named value with a type and (optionally) how this field should be stored, indexed, searched, presented and how it should influence ranking.

field [name] type [type-name] {
    [body]
}
Do not use names that are used for other purposes in the indexing language or other places in the search definition file. Reserved names are:
  • attribute
  • body
  • case
  • context
  • documentid
  • else
  • header
  • hit
  • host
  • if
  • index
  • position
  • reference
  • relevancy
  • sddocname
  • summary
  • switch
  • tokenize
Other names not to use include any words that start with a number or includes special characters.

The type attribute is mandatory and has one of the following values:

NameType
annotationreference<annotationtype> Declares a reference to an annotation on a given string. Should only be used for fields declared inside annotation, or as a base type by the use of any of the compound types listed above, inside annotation.
array<element-type>An array of element-type. The element type can be any single value type.
weightedset<element-type>A weighted set: Like an array, but each element is also assigned an integer weight.
bytesigned 8-bit integer
double64-bit IEEE 754 floating point
float32-bit IEEE 754 floating point
intsigned 32-bit integer
longsigned 64-bit integer
positionDocument position in geographical coordinates, e.g. latitude and longitude.
predicateA boolean expression in predicate logic.
rawbinary data
stringany text
structnameDeclares a field with a specific struct type, given by the struct name. Indexing restrictions
map<key-type,value-type>A map using the given types as keys and values. Keys and values can be any type. Indexing restrictions
tensor(dimension-1,...,dimension-N)A tensor with a set of named dimensions and a set of values located in the space of those dimensions.
uriA Uniform Resource Identifier (a URL or any other unique string id)
reference<document-type> A reference to an instance of a document-type used in parent-child relationship.

The body of a field is optional for search, document and struct, and disallowed for annotation. It may contain the following elements:

NameDescriptionOccurrence
alias Make an index or attribute available in searches under an additional name Zero to many
attribute Specify an attribute setting. Zero to many
bolding Specifies whether content of this field should be bolded. Zero to one
id Explicitly decide the numerical id of this field. Is normally not necessary, but can be used to save some disk space. Zero to one
index Specify a parameter of an index. Not applicable to streaming search Zero to many
indexing The indexing statements used to create index structure additions from this field. Zero to one
indexing-rewrite Determines the rewriting Vespa is allowed to do on the indexing statements of this field.Not applicable to streaming search Zero to one
match Set the matching type to use for this field. Zero to one
normalizing Specifies the kind of spelling normalizing to do on this field. Zero or one.
query-command Specifies a command which can be received by a plugin searcher in the Search Container. Zero to many
rank The high level ranking method to use for the field Zero or one
rank-type Selects the set of low-level rank settings to be used for this field when using default nativeRank. Zero to one
sorting The sort specification for this field. Zero or one.
stemming Specifies the kind of stemming to use for this field. Not applicable to streaming search Zero or one.
struct-field A subfield of a field of type struct. The struct must have been defined to contain this subfield in the struct definition. If you want the subfield to be handled differently from the rest of the struct, you may specify it within the body of the struct-field. Zero to many.
summary Sets a summary setting of this field, set to dynamic to make a dynamic summary. Zero to many
summary-to The list of document summary names this should be included in. Not applicable to streaming search, instead declare non-standard summaries in a document-summary tag outside of the document declaration Zero to one
weight The importance of a term boost field, a positive integer. Zero to one
weightedset Attributes of a weighted set type. Zero to one

If the field is part of a struct definition, i.e. contained in the struct element, only match may be specified.

If the field is of type struct, only indexing, match and query-command may be specified.

A field declared outside of a document tag (i.e. immediately within a search tag) is referred to as an extra-field. Such fields may not be set directly, not programmatically and not through a feed - doing so will cause the document to be rejected by the indexer. Extra-field may only be populated using indexing statements that input the value of proper fields (e.g. indexing: input my_document_field | normalize | summary | index).

struct-field

Contained in field or struct-field. Defines how this struct field (a subfield of a struct) should be stored, indexed, searched, presented and how it should influence ranking. The field in which this struct field is contained must be of type struct, which has defined this struct field. Note that struct fields are only available for streaming search.

field [name] {
    [body]
}
The body of a struct field is optional and may contain the following elements:
NameDescriptionOccurrence
indexing The indexing statements used to create index structure additions from this field. Zero to one
match Set the matching type to use for this field. Zero to one
query-command Specifies a command which can be received by a plugin searcher in the Search Container. Zero to many
struct-field A subfield of a field of type struct. The struct must have been defined to contain this subfield in the struct definition. If you want the subfield to be handled differently from the rest of the struct, you may specify it within the body of the struct-field. Zero to many.
summary Sets a summary setting of this field, set to dynamic to make a dynamic summary. Zero to many
summary-to The list of document summary names this should be included in. Zero to one
If this struct field is of type struct (i.e. a nested struct), only indexing, match and query-command may be specified.

fieldset

Contained in search. Note: this is not related to the Document fieldset.

Field sets provide a way to group fields together for searching. When you query a field set, you will get results from all the fields in the field set. Given the clause below:

fieldset myfieldset {
  fields: a,b,c
}
Using the query yql=select+*+from+sources+*+where+myfieldset+contains+"foo"%3B will return all the documents for which one or more of the fields a, b or c contain "foo". By naming the field set 'default', you can search those fields without specifying the field set in unstructured queries: query=foo.

The fields making up the field set should be as similar as possible in terms of indexing clause, matching etc. If they are not, you must test your application thoroughly. For example, it will work for a mix of attributes and indexes, but the matching for attribute fields will always be exact unless you are in streaming mode.

If you need specific match settings for the field set, such as exact, you must specify it using a match clause:

fieldset myfieldset {
  fields: a,b,c
  match {
    exact
  }
}
You may use query-commands in the field set to set search settings. Example:
fieldset myfieldset {
  fields: a,b,c
  query-command:"exact @@"
}

compression

Contained in document. If a compression level is set within this element, lz4 compression is enabled for whole documents.

compression {
    [body]
}
The body of a compression specification is optional and may contain:
NameDescriptionOccurrence
type LZ4 is the only valid compression method. Zero to one
level Enable compression. LZ4 is linear and 9 means HC(high compression) Zero to one
threshold A percentage (multiplied by 100) giving the maximum size that compressed data can have to keep the compressed value. If the resulting compressed data is higher than this, the document will be stored uncompressed. Default value is 95. Zero to one

rank-profile

Contained in search. A rank profile is a named set of rank settings which can be specified during queries (see the ranking parameter in the search API).

Rank profiles are used to specify an alternative ranking of the same data for different purposes, and to experiment with new rank settings. If no explicit rank profile is specified, one called "default" is implicitly created to hold the rank settings from each field. The "default" rank profile is always selected for queries which does not specify one. It is possible to add additional settings to the default rank profile by explicitly defining it.

rank-profile [name] inherits [rank-profile] {
    [body]
}
The inherits attribute is optional. If defined, it contains the name of one other rank profile in the same search definition. Values not defined in this rank profile will then be inherited as expected. It is possible to inherit the default rank profile, even if it is not explicitly listed.

In addition to the default rank profile, a profile named unranked is implicitly created. This rank-profile makes sure that the rank phases in the search backend are skipped and should be used for queries that only require matching and do not use ranking. If you are sorting on something different than rank score this is also the profile to use. Note that this profile should not be used if the query contains Wand search operators. Also note that using this profile will give better performance as the rank phases are skipped.

The body of a rank-profile may contain:

NameDescriptionOccurrence
match-phase Ranking configuration to be used for hit limitation during matching. Zero or one
first-phase The ranking config to be used for first-phase ranking. Zero or one
rank-features The rank features to be dumped when using the query-argument rankfeatures. Zero or more
second-phase The ranking config to be used for second-phase ranking. Zero or one
summary-features The rank features to be dumped for all queries. Zero or more
ignore-default-rank-features Do not dump the default set of rank features, only those explicitly specified with the rank-features command. Zero or one
Overrides the global persearch threads to a lower value. Zero or one
constants List of constant key/value pairs available in ranking expressions. Zero or one
rank-properties List of any rank property key-values to be used by rank features. Zero or one
macro [name] A way to simply ranking expression by defining named macros that can be referenced during ranking phase(s) and as part of the summary-features. Zero or more
rank The high level ranking method to use for a field in this profile. Zero or more
rank-type The rank type of a field in this profile. Zero or more
Refer to the rank profiles defined in the example below.

match-phase

Contained in rank-profile. The config specifying ranking to be used during matching. This is used to limit the result set in order to cut latency. It is particularly useful if the first-phase ranking is expensive. It can be used for sorting on numeric values to limit the evaluated result set.

Match-phase is a feature for performance optimization - how to rank documents using a quality attribute and using estimates to cut evaluation - read more in the sizing guide.
match-phase {
    attribute: [numeric single value attribute]
    order: [ascending | descending]
    max-hits: [integer]
    diversity
}
NameDescription
attribute Which attribute to use as the quality signal. The attribute referenced must be a single valued numeric attribute with fast-search enabled. No default.
order Whether the attribute should be used in descending order (prefer documents with a high score) or ascending order (prefer documents with a low value in the attribute). Usually it is not necessary to specify this, as the default value descending is by far the most common.
max-hits Requested hits per search node. Usually a number like 10000 works well here. The default is 1400. ToDo Check this
diversity Guarantee a minimum result set diversity.

diversity

Contained in match-phase. Diversity is used to specify diversity in different phases - supported in match-phase. It is used to guarantee a minimum result set diversity.

Specify the name of an attribute that will be used to provide diversity. Result sets are guaranteed to get at least min-groups unique values from the diversity attribute from this phase. A document is considered as a candidate if:

  • The query has not yet reached the max-hits number produced from this phase.
  • The query has not yet reached the max number of candidates in one group. This is computed by the max-hits of the phase divided by min-groups
diversity {
    attribute: [numeric attribute]
    min-groups: [integer]
}
NameDescription
attribute Which attribute to use when deciding diversity. The attribute referenced must be a single valued numeric or string attribute.
min-groups Specifies the minimum number of groups returned from the phase. Using this with match-phase often means one can reduce max-hits

first-phase

Contained in rank-profile. The config specifying the first phase of ranking. This is the initial ranking performed on all hits, and you should therefore avoid doing heavy rank-calculations here. By default, this will use the ranking feature nativeRank.

first-phase {
    [body]
}
The body of a firstphase-ranking statement consists of:
NameDescription
expression Specify the ranking expression to be used for first phase of ranking - see ranking expressions.
keep-rank-count How many documents to keep the first phase top rank values for. Default value is 10000.
rank-score-drop-limit Drop all hits with a first phase rank score less than or equal to this floating point number. Default value is -Double.MAX_VALUE.

expression

Contained in first-phase or second-phase. Specify a ranking expression. The expression can either be written directly or loaded from a file. When writing it directly the syntax is:

expression: [ranking expression]
or
expression {
    [ranking expression]
    [ranking expression]
    [ranking expression]
}
The second format is primarily a convenience feature when using long expressions, enabling them to be split over multiple lines.

Expressions can also be loaded from a separate file. This is useful when dealing with the very long expressions generated by e.g. MLR. The syntax is:

expression: file:[path-to-expressionfile]
The path is relative to the location of the search definition file (note: directories are not allowed in the path). The file itself must end with .expression. This suffix is optional in the sd-file. Therefore expression: file:mlrranking.expression and expression: file:mlrranking are identical. Both refer to a file called mlrranking.expression in the searchdefinition directory.

rank-features

Contained in rank-profile. List of extra rank features to be dumped when using the query-argument rankfeatures.

rank-features: [feature] [feature]
or
rank-features {
    [feature]
    [feature]
}
Any number of ranking features can be listed on each line, separated by space.

constants

Contained in rank-profile. List of constants available in ranking expressions, resolved and optimized at configuration time.

constants {
    key: value
}
NameDescription
key Name of the constants.
value A number or any string. Must be quoted if it contains spacing.

rank-properties

Contained in rank-profile. List of generic properties, in the form of key/value pairs to be used by ranking features.

rank-properties {
    key: value
}
NameDescription
key Name of the property.
value A number or any string. Must be quoted if it contains spacing.

macro (inline)? [name]

Contained in rank-profile. It is possible to define named expression macros that can be referenced as a part of the ranking expression. A macro accepts any number of arguments.

macro [name]([arg1], [arg2], [arg3]) {
  expression: …
}
or
macro [name] ([arg1], [arg2], [arg3]) {
  expression {
      [ranking expression]
      [ranking expression]
      …
}
Note that the parenthesis is required after the name. A rank-profile example is shown below:
rank-profile default inherits default {
  macro myfeature() {
    expression: fieldMatch(title) + freshness(timestamp)
  }
  macro otherfeature(foo) {
    expression{ nativeRank(foo, body) }
  }

  first-phase {
    expression: myfeature * 10
  }
  second-phase {
    expression: otherfeature(title) * myfeature
  }
  summary-features: myfeature
}
You can not include macros that accept arguments in summary features.

Adding the inline modifier will inline this macro in the calling expression if it also has no arguments. This is faster for very small and cheap macros (and more expensive for others).

second-phase

Contained in rank-profile. The config specifying the second phase of ranking. This is the optional reranking performed on the best hits from the first phase, and where you should put any advanced ranking calculations (e.g. MLR). By default, no second-phase ranking is performed. In streaming search we perform the second phase ranking on all hits. You can therefore put all the rank calculation in the first phase rank expression and just skip second phase.

second-phase {
    [body]
}
The body of a secondphase-ranking statement consists of:
NameDescription
expression Specify the ranking expression to be used for first phase of ranking. (for a description, see the ranking expression documentation.
rerank-count Optional argument. Specifies the number of hits to be reranked. Default value is 100

summary-features

Contained in rank-profile. List of rank features to be dumped for every query. Using many items will have a performance impact, a larger list to be returned only when requested can be specified in ranking features.

summary-features: [feature] [feature]…
or
summary-features {
    [feature]
    [feature]
}
Any number of ranking features can be listed on each line, separated by space.

constant

Contained in search. This defines a named constant tensor located in a file with a given type that can be used in ranking expressions via the rank feature constant. A constant with a given name is defined as follows:

constant [name] {
    [body]
}
The body of a constant must contain:
NameDescriptionOccurrence
file Path to the location of the file containing the constant tensor. The path is relative to the root of the application package containing this sd-file. The format of the file is JSON and is the same as when specifying a tensor field in a document put or update. Refer to the Document JSON Format for reference. Compression is supported - if the filename ends with ".json.lz4", Vespa assumes the tensor is LZ4 compressed.

One
type The type of the constant tensor, refer to tensor-type-spec for reference. One
Constant tensor example:
constant my_constant_tensor {
    file: constants/my_constant_tensor_file.json
    type: tensor(x{},y{})
}
This example has a constant tensor with two mapped dimensions, x and y. An example JSON file with such tensor constant:
{
    "cells": [
        { "address": { "x": "a", "y": "b"}, "value": 2.0 },
        { "address": { "x": "c", "y": "d"}, "value": 3.0 }
    ]
}
When an application with tensor constants is deployed, the files are distributed to the content nodes before the new configuration is being used by the search nodes. Incremental changes to constant tensors is not supported. When changed, replace the old file with a new one and re-deploy the application or create a new constant with a new name in a new file.

document-summary

Contained in search. An explicitly defined document summary. By default, a document summary named default is created. Using this element, other document summaries containing a different set of fields can be created.

document-summary [name] {
    [body]
}
The body of a document summary consists of:
NameDescriptionOccurrence
summary A summary field in this document summary. Zero to many
Use the summary query parameter to choose a document summary in searches. See also document summaries.

stemming

Contained in field, search or index. Sets how to stem a field or an index, or how to stem by default. Read more on stemming. Note: Not applicable to streaming search.

stemming: [stemming-type]
The stemming types are:
TypeDescription
none No stemming: Keep words as they are received.
best Use the 'best' stem of each word according to some heuristic scoring.
shortestUse the shortest stem of each word. This is the default setting.
multipleUse multiple stems. Retains all stems returned from the linguistics library.
Note: When combining multiple fields in a fieldset, all fields should use the same stemming type.

normalizing

Contained in field. Sets the normalizing to be done on this field. Normalizing will cause accents and similar decorations which are often misspelled to be normalized the same way both in documents and queries. Not supported in streaming search.

normalizing: [normalizing-type]
The normalizing type available is:
TypeDescription
noneNo normalizing
If this is not set, normalization will be done for this field.

alias

Contained in attribute, field or index. Makes an index or attribute available under an additional name:

alias [index/attr-name]: [alias]
If the index/attribute name is skipped, the containing field or index name is used. Alias names can be any name string, dots are allowed as well.

attribute

Contained in field. Specifies a property of an index structure attribute:

attribute [attribute-name]: [property]
or
attribute [attribute-name] {
    [property]
    [property]
    …
}
The attribute name can be skipped, in which case the field name is used. Refer to search definitions for actions required when adding or modifying attributes. Read Attributes for an introduction to attributes. The following properties are available:
PropertyDescription
fast-searchAt the cost of memory, speed up search when there are few hits and there are no other limiting factors.
fast-access In an indexed content cluster with searchable-copies < redundancy this property can be set to make sure that this attribute is always kept in memory for fast access in the context of applying partial updates and when used in a selection expression for garbage collection. If redundancy == searchable-copies (default) this property is a no-op.
hugeDeprecated. This setting no longer have any effect. All multi-value attributes now use a more adaptive approach in how data is stored in memory, and up to 1 billion documents per node is supported.
aliasAn alias for the attribute. Add an attribute name before the colon to specify an alias for another attribute than the one given by field name.
sortingThe sort specification for this attribute.
[tensor-type-spec]The tensor type specification for this tensor attribute.
An attribute is multi-valued if assigning it multiple values during indexing, either by using e.g. split and for_each or by letting multiple fields write their value to the attribute field.

Note that normalizing and tokenization is not enabled by default for attribute fields. Queries in attribute fields are hence not normalized. Use index on fields to enable. Both index and attribute can be set on a field.

sorting

Contained in attribute or field. Specifies how sorting should be done.

sorting : [property]
or
sorting {
    [property]
    [property]
    …
}
PropertyDescription
order Either ascending or descending. Default is ascending. Used unless overridden in sortspec in query.
function The Sort function to be used. Implemented functions are raw, lowercase, and uca. The default is uca, but please note that if no language or locale is specified in the query sortspec, the field, or generally for the query, lowercase will be used instead. Used unless overridden in sortspec in query.
strength Sort strength to be used. Implemented levels are primary, secondary, tertiary, quaternary and identical. The default is primary. Used unless overridden in sortspec in query. Only applicable if function is set to uca.
locale Locale to be used. The default is none, indicating that it is inferred from query. It should only be set here if the attribute is filled with data that is in 1 language only. Used unless overridden in sortspec in query. Only applicable if function is set to uca.

tensor-type-spec

Contained in attribute or constant. Specifies the tensor type for a tensor. A tensor type contains a list of dimensions on the format:

tensor(dimension-1,dimension-2,...,dimension-N)
A dimension is specified as follows:
  • dimension-name{} - a mapped dimension.
  • dimension-name[size] - a bound indexed dimension with the given size.
  • dimension-name[] - an unbound indexed dimension.
The tensor type for a tensor with two mapped dimensions x and y looks like:
tensor(x{},y{})
Example tensor with this type:
{{x:a,y:b}:10, {x:c,y:d}:20}
The tensor type for a tensor with two bound indexed dimensions x and y with sizes 3 and 2 respectively looks like this:
tensor(x[3],y[2])
Example tensor with this type (representing a matrix):
{{x:0,y:0}:1, {x:0,y:1}:2,
 {x:1,y:0}:3, {x:1,y:1}:5,
 {x:2,y:0}:7, {x:2,y:1}:11}
Note that the labels are indexes in the range [0,dimension-size>

A tensor with both sparse and indexed dimensions:

tensor(x[2],y{})
Example:
{{x:0,y:a}:10, {x:0,y:b}:20,
 {x:1,y:a}:5,  {x:1,y:b}:7}

bolding

Contained in field. Highlight matching query terms in the summary:

bolding: on
Not applicable to streaming search. Instead use summary: dynamic.

The default is no bolding, set bolding: on to enable it. Note that this command is overridden by summary: dynamic, if both are specified, bolding will be ignored. The difference between using bolding instead of summary: dynamic is the latter will provide a dynamic abstract in addition to highlighting search terms while the first only does highlighting.

The default XML element used to highlight the search terms is <hi> - to override, set container.qr-searchers configuration. Example using <strong>:

<container>
  <search>
    <config name="container.qr-searchers">
      <tag>
        <bold>
          <open>&lt;strong&gt;</open>
          <close>&lt;/strong&gt;</close>
        </bold>
        <separator>...</separator>
      </tag>
    </config>
  <search>
<container>

id

Contained in field. Sets the numerical id of this field. All fields have a document-internal id internally for transfer and storage. Id's are usually determined programmatically as a 31-bit number. Some storage and transfer space can be saved by instead explicitly setting id's to a 7-bit number.

id: [positive integer]
An id must satisfy these requirements:

  • Must be a positive integer
  • Must be less than 100 or larger than 127
  • Must be unique within the document and all documents this document inherits

index

Contained in field or search. Sets index parameters. Content in fields with index are normalized and tokenized by default. This element can be single- or multi-valued:

index [index-name]: [property]
or
index [index-name] {
    [property]
    [property]
    …
}
The index name can be skipped inside fields, causing the index name to be the field name. Parameters:
PropertyDescriptionOccurrence
alias Specify an alias to this index to be available in searches. Zero to many
stemming Set the stemming of this index. Indexes without a stemming setting get their stemming setting from the fields added to the index. Setting this explicitly is useful if fields with conflicting stemming settings are added to this index. Zero to one
arity Set the arity value for a predicate field. The data type for the containing field must be predicate. One (mandatory for predicate fields), else zero.
lower-bound Set the lower bound value for a predicate field. The data type for the containing field must be predicate. Zero to one.
upper-bound Set the upper bound value for predicate fields. The data type for the containing field must be predicate. Zero to one.
dense-posting-list-threshold Set the dense posting list threshold value for predicate fields. The data type for the containing field must be predicate. Zero to one.

indexing

Contained in field or struct-field. One or more Indexing Language instructions used to produce index, attribute and summary data from this field. Indexing instructions has pipeline semantics similar to unix shell commands. The value of the field enters the pipeline during indexing and the pipeline puts the value into the desired index structures, possibly doing transformations and pulling in other values along the way.

indexing: [index-statement]
or
indexing {
    [indexing-statement];
    [indexing-statement];
    …
}
If the field containing this is defined outside the document, it must start by an indexing statement which outputs a value (either "field [fieldname]" to fetch a field value, or a literal). Fields in documents will use the value of the enclosing field as input (field [fieldname]) if one isn't explicitly provided.

Specify the operations separated by the pipe (|) character. For advanced processing needs, use the indexing language, or write a document processor. Supported expressions for fields are:

attribute Attribute is used to make a field available for sorting and grouping.
index Creates a searchable index for the values of this field. All strings are lower-cased before stored in the index. By default the index name will be the same as the name of the search definition field. Use a fieldset to combine fields in the same set for searching.
set_language Sets document language - details.
summary Includes the value of this field in a summary field. Modify summary output by using summary: (e.g. to generate dynamic teasers).

indexing-rewrite

Contained in field. Vespa will normally rewrite indexing statements extensively to implement the technical tasks which are required to carry out the intentions of the indexing statement. The rewriting done can be controlled using this element.

indexing-rewrite: none
Include this to let an indexing statement pass through unaltered. Note that such statements must begin with an input <fieldname>, get_var or constant expression. You should understand which rewrites Vespa does, and be certain that your indexing statement can do without them to use this. This statement must be placed somewhere below the indexing statement in the field.

match

Contained in field, fieldset or struct-field. Sets the matching method to use for this field to something else than the default token matching. Note the restrictions found in the column named Valid with.

match: [property]
or
match {
    [property]
    [property]
    …
}
Whether the match type is text, word or exact, all term matching will be done after NFKC normalization and locale independent lowercasing (in that order).

PropertyDescriptionValid withRemarks
text This field is matched per token. Tokens are created by splitting on whitespace and non-letter characters, as well as by segmentation for CJK. All characters which are not letters or digits are ignored using text matching. Previously also known as "token" matching. Indexes, streaming Default for indexes. Can not be combined with exact matching.
word

A single word is formed from the field input (or each item in the field input if it is an array or a weighted set). This word is matched exactly: Strings containing any characters whatsoever will be indexed and matched as-is. In queries, the word to match is heuristically parsed taking into account some usual query syntax characters; one can also use double quotes to include space, star, or exclamation marks. This is the default matching mode for string attributes.

Example: If artist is a string attribute, this (advanced syntax) query:

foo AND (artist:"'N Sync" OR artist:"*NSYNC" OR artist:A*teens OR artist:"Wham!")
will match documents containing foo and at least one of 'N Sync or *NSYNC or A*teens or Wham! as the artist field.

Note that without the quotes, the space in 'N Sync would end that word and would result in a search for just 'N, similarly the ! would mean to increase the weight of a Wham term if not quoted.

Indexes, attributes Default for attributes. Can not be combined with text (token) matching.
exact

This field is matched exactly: Strings containing any characters whatsoever will be indexed and matched as-is. In queries, the exact match string ends at the exact match terminator, which is @@ per default. As a side effect, a field with match: exact is considered to be a filter field, just as if rank: filter was specified. This is because you will only get one word per field (or per item in the case of multi-valued types such as array<string>), so there isn't much ranking information that you could get anyway. You can turn off the implicit rank: filter by adding an explicit rank: normal.

Example: If tag is an exact match field, this (advanced syntax) query:

someword AND (tag:!*!@@ OR tag:(kanoo)@@)
will match documents containing someword and either !*! or (kanoo) as a tag.

Note that without the @@ terminating the second tag string, the second tag value would be (kanoo)).

Indexes, attributes, streaming Can not be combined with text (token) matching.
exact-terminator

When using exact match, a terminator for use in queries should be specified. The default is @@, but if the strings to match can contain two at-signs in a row, a different terminator must be used. Alternately the "word match" feature can be used, see above.

Example

    match {
        exact
        exact-terminator: "@@"
    }
on the tag field will cause the query tag:a b c!@@ to match documents containing the exact string a b c!

Indexes, attributes, streaming Only valid if exact matching is chosen.
prefix This field supports prefix* searches. For streaming: this field uses prefix* searching for all search terms. Attributes, streaming Prefix searching is always enabled for attributes and in streaming. In these cases use the query syntax for prefix terms to get prefix searching even though the match method is not prefix.
substring This field uses *substring* searching for all search terms as default. Streaming Substring searching is always enabled in streaming. Use the query syntax for substring terms to get substring searching even though the match method is not substring.
suffix This field uses *suffix searching for all search terms as default. Streaming Suffix searching is always enabled in streaming. Use the query syntax for suffix terms to get suffix searching even though the match method is not suffix.
max-length This limits the length of the field that will be used for matching. Indexes, streaming If this value is set it is the max number of characters of a field that will be considered during search. If not the default fieldmatchmaxlength will be used.
gram This field is matched using n-grams. For example, with the default gram size 2 the string "hi blue" is tokenized to "hi bl lu ue" both in the index and in queries to the index.

N-gram matching is useful mainly as an alternative to segmentation in CJK languages. Typically it results in increased recall and lower precision. However, as Vespa usually uses proximity in ranking the precision offset may not be of much importance. Grams consumes more resources than other matching methods because both indexes and queries will have more terms, and the terms contains repetition of the same letters. On the other hand, CPU intensive CJK segmentation is avoided.

It may also be used for substring matching in general.

Indexes
gram-size Sets the gram size when gram matching is used. The default size (if this is not present) is 2.

Example

    match {
        gram
        gram-size: 3
    }
Indexes This may be any positive number larger than 0.

rank

Contained in field or rank-profile. Set the kind of ranking calculations which will be done for the field. Even though the actual ranking expressions decide the ranking, this settings tells Vespa which preparatory calculations and which data structures are needed for the field.

rank [field-name]: [ranking settings]
or
rank {
    [ranking setting]
}
The field name should only be specified when used inside a rank-profile. The following ranking settings are supported in addition to the default:
Ranking settingDescription
filter Indicates that matching in this field should use fast bit vector data structures only. This saves a lot of CPU during matching, but only a few simple ranking features will be available for the field. This setting is appropriate for fields typically used for filtering or simple boosting purposes, like filtering or boosting on the language of the document.
normal The reverse of "filter", indicates that matching in this field should use normal data structures and give normal match information for ranking. Used to turn off implicit rank: filter when using match: exact. If both "filter" and "normal" are set somehow, the effect is as if only "normal" was specified.

query-command

Contained in fieldset, field or struct-field. Specifies a function to be performed on query terms to the indexes of this field when searching. The Search Container server has support for writing Vespa Searcher plugins which processes these commands.

query-command: [any string]
If you write a plugin searcher which needs some index-specific configuration parameter, that parameter can be set here.

rank-type

Contained in field or rank-profile. Selects the low-level rank settings to be used for this field when using nativeRank.

rank-type [field-name]: [rank-type-name]
The field name can be skipped inside fields. Defined rank types are:
TypeDescription
identity Used for fields which contains only what this document is, e.g. "Title". Complete identity hits will get a very high rank.
about Some text which is (only) about this document, e.g. "Description". About hits get high rank on partial matches and higher for matches early in the text and repetitive matches. This is the default rank type.
tags Used for simple tag fields of type tag. The tags rank type uses a logarithmic table to give more relative boost in the low range: As tags are added they should have significant impact on rank score, but as more and more tags are added, each new tag should contribute less.
empty Gives no relevancy effect on matches. Used for fields you just want to treat as filters.
For nativeRank you can specify a rank type per field. If the supported rank types do not meet your requirements you can explicit configure the native rank features using rank-properties. See the native rank reference for more information.

summary-to

Contained in field or struct-field. Specifies the name of the document summaries which should contain this field.

summary-to: [summary-name], [summary-name], …
If this is not specified, the field or struct-field will be included in the default document summary. See also document summaries.

summary

Contained in field or document-summary or struct-field. Declares a summary field.

summary: [property]
or
summary [name] type [type] {
    [body]
}
The summary name can be skipped if this is set inside a field. The name will then be the same as the name of the source field. In fields, the summary type can also be skipped, in which case the type will be determined by the field type. The summary data types available are the same as the document field data types. full summary is the default. Long field values (like document content fields) should be made dynamic. The body of a summary may contain:
NameDescriptionOccurrence
full Returns the full field value in the summary (the default). Zero to one
dynamic Make the value returned in results from this summary field be a dynamic abstract of the source summary field by extracting fragments of text around matching words. Matching words will also be highlighted, in similarity with the bolding feature. This highlighting is not affected by the query-argument bolding. The default XML element used to highlight query terms is <hi> - refer to bolding for how to configure. Zero to one
source Specifies the name of the field or fields from which the value of this summary field should be fetched. If multiple fields are specified, the value will be taken from the first field if that has a value, from the second if the first one is empty and so on.
source: [field-name], [field-name], …
When this is not specified, the source field is assumed to be the field with the same name as the summary field.
Zero to one
to Specifies the name of the document summaries this should be included in. An alternative form to summary-to in summaries.
to: [document-summary-name], [document-summary-name], …
This can only be specified in fields, not in explicit document summaries. When this is not specified, the field will go to the default document summary.
Zero to one
Read more about document summaries.

weight

Contained in field. The weight of a field - the default is 100. The field weight is used when calculating the rank scores.

weight: [positive integer]

weightedset

Contained in field of type weightedset. Properties of a weighted set.

weightedset: [property]
or
weightedset {
    [property]
    [property]
    …
}
PropertyDescriptionOccurrence
create-if-nonexistent If the weight of a key is adjusted in a document using a partial update increment or decrement command, but the key is currently not present, the command will be ignored by default. Set this to make keys to be created in this case instead. This is useful when the weight is used to represent the count of the key. Zero to one
remove-if-zero This is the companion of create-if-nonexistent for the converse case: By default keys may have zero as weight. With this turned on, keys whose weight is adjusted (or set) to zero, will be removed. Zero to one

annotation

Contained in search. Defines an annotation type, to be used by the Annotations API. A name of the annotation is mandatory, the body is optional.

annotation [name] {
    [body]
}

import field

Contained in search. Using a reference to a document type, import a field from that document type into this search definition to be used for matching, ranking, grouping and sorting. Refer to document references.

Only attribute fields of non-tensor type can be imported. The imported field inherit all but the following properties from the parent field:

  • attribute: fast-access

To use an imported field in summary, you need to create an explicit document summary containing that field. See document references for an example.

Field types

string Use for a text field of any length. String fields may only contain text characters, as defined by isTextCharacter in com.yahoo.text.Text
Indexing By default, strings are tokenized before indexing. Tokenization removes any non-word characters, and splits the string into tokens on each word boundary. In addition, CJK tokens are split using a segmentation algorithm. The resulting tokens are what is becoming searchable in the index. To index strings as-is (that is, avoid tokenization), use indexing-rewrite: none. By default, strings are also normalized and stemmed.
Attribute Added as-is. match exact or prefix is supported types of searches in string attributes. Searches are however case-insensitive. A query for BritneY.spears will match a document containing BrItNeY.SpEars
Summary Added as-is.
int Use for single 32-bit integers.
Indexing Not supported. An attribute will automatically be used instead.
Attribute Becomes integer attributes, which supports range grouping and range searches.
Summary Added as a 32-bit integer.
long Use for single 64-bit integers.
Indexing Not supported. An attribute will automatically be used instead.
Attribute Becomes a 64-bit integer attribute, which supports range grouping and range searches.
Summary Added as a 64-bit integer.
byte Use for single 8-bit numbers.
Indexing Not supported. An attribute will automatically be used instead.
Attribute Added as a byte which supports range searches.
Summary Added as a byte.
float Use for floating point numbers (32-bit IEEE 754 float).
Indexing Not supported. An attribute will automatically be used instead.
Attribute Added as a 32-bit IEEE 754 float which supports range searches.
Summary Added as a 32-bit IEEE 754 float.
double Use for high precision floating point numbers (64-bit IEEE 754 double).
Indexing Not supported. An attribute will automatically be used instead.
Attribute Added as a 64-bit IEEE 754 double which supports range searches.
Summary Added as a 64-bit IEEE 754 double.
position Used to filter and/or rank documents by distance to a position in the query. See Geo search. The input format is a string with latitude;longitude, where the following are valid formats:
  1. S22.4532;W123.9887 - at output, this format is used
  2. N72°23'52;E26°04'22
  3. N72o20.92;E26o08.54
This format is also used in queries with position. A semicolon is used as separator between latitude and longitude - remember to URL encode the semicolon as %3B. Latitude is prefixed by N or S, and longitude by E or W. The angular measurement can either be expressed as degrees with a decimal fraction (this is the recommended way), or as degrees subdivided in minutes and seconds. It is also valid to express minutes with a decimal fraction, supporting regular GPS output format. Small letter o may be used as a replacement for degrees.
  • Position fields in search results render differently - returned is an item with an XML element with the latitude;longitude string plus X/Y coordinates:
    "mypos.position": "<position x=\"-121996000\" y=\"37401000\" latlong=\"N37.401000;W121.996000\" />"
    
    The X/Y coordinates are in millionths of degrees - see coordinate system in summary fields. This document also lists options for rendering X / Y.
  • When using the document api, position fields are rendered like:
    "mypos": {
        "x": -123988700,
        "y": -22453200
    },
    "mypos_zcurve": -6533494969659888,
    
    Use fieldsets to control which fields to output, adding --fieldset 'music:[document]' skips extra fields like _zcurve.
Indexing Not supported.
Attribute Added as an interleaved 64-bit integer (see Z-order curve).
Summary Added as-is.
predicate Use to match queries to a set of boolean constraints. See querying predicate fields.
Indexing Indexed in a variable size binary format that is optimized for application during query evaluation.
Attribute Not supported.
Summary Added as-is.
raw Use for binary data
Indexing Not supported.
Attribute Not supported.
Summary Added as raw data.
uri

Use for URLs.

Indexing

The URL is split into the different components which are indexed separately. Note that only URLs can be indexed this way, not other URIs. The different components are as defined by the HTTP standard: Scheme, hostname, port, path, query and fragment. Example:

http://mysite.mydomain.com:8080/path/shop?d=hab&id=1804905709&cat=100#frag1
schemehttp
hostnamemysite.mydomain.com (indexed as "mysite", "mydomain" and "com")
port8080 (note that port numbers 80 and 443 are not indexed, as they are the normal port numbers)
path/path/shop (indexed as "path" and "shop")
queryd=hab&id=1804905709&cat=100 (indexed as "d", "hab", "id", "1804905709", "cat" and "100")
fragmentfrag1
The syntax for searching these different components is:
[field-name].[component-name]:term
Example: In a uri field sourceurl, search for documents from slashdot:
query=sourceurl.hostname:slashdot
URL hostnames also support anchored searching, see search in URL fields.

It is not possible to index uri-typed fields into a common index, i.e. it has to be indexed separately from other fields. If you need to combine URLs with other fields you could store it in a string-field instead, but then you can not search in the different parts of the URL (scheme, hostname, port, path, query and fragment).

Aliasing also works different for URL fields - you are allowed to create aliases both to the index (as usual) and to the components of it. Use

alias [component]: [alias]
to create an alias to a component. For example, given this field:
field surl type uri {
    indexing: summary | index
    alias: url
    alias hostname: site
}
a search in "surl" and "url" will search in the entire url, while "surl.hostname" or "site" will search the hostname.

Attribute Added as-is as a string.
Summary Added as-is as a string.
array<element-type> Use to create a multi-value field of the element type. The element type can be any single value type, but not a nested array or weighted set.
Indexing Each element is indexed separately.
Attribute Added as a multi-value attribute.
Summary Added as a multi-value summary field.
weightedset<element-type>

Use to create a multi-value field of the element type, where each element is assigned a signed 32-bit integer weight. The element type can be any single value type, but not a nested array or weighted set. The weights may be assigned any semantics by the application. Two main use cases:

  1. The weight symbolizes the number of occurrences
  2. The weight specifies another value type, for instance the importance of the document
The weight of a matching value is by default used in nativeRank directly as the rank score of the field. It is also possible to create a rank type which uses a rank boost table, weightboost to calculate the rank value from the weight (the tags rank type does this by default).

The weights are returned in the summary for the attribute. The format of the field in the attribute summary is like:

<field name="some_field">
    <item weight="1">a</item>
    <item weight="10">b</item>
    <item weight="100">c</item>
</field>
It is possible to specify that a new key should be created if it does not exist before the update, and that it should be removed if the weight is set to zero. This is only usable together with the increment and decrement operations, see document updates.

Indexing Each token present in the field is indexed separately. Information indexed includes element number, element weight and a list of token occurrence positions for each element in which the token is present.
Attribute Added as a multi-value weighted attribute.
Summary Added as a multi-value summary field if this is an attribute.
tensor(dimension-1,...,dimension-N) Use to create a tensor field with the given tensor type spec that can be used for ranking. A tensor field is NOT searchable. See Tensor Evaluation Reference for definition of tensors and Document JSON Format for the JSON feed format for tensors.
Indexing Not supported.
Attribute Added as-is in an attribute to be used for ranking.
Summary Added as-is. The JSON result format (presentation.format=json) should be used when returning a summary class containing a tensor field as part of search.
struct Use to define a field with a struct datatype. To define a struct field, first create a struct definition. The struct must be defined inside the document definition. To declare a struct field in a document or struct, use the struct name to identify the field type:
field foo type address {}
Indexing Structs have the same indexing restrictions as maps.
Attribute Each field may or may not be an attribute according to its configuration.
Summary Each field in the struct has its own summary configuration
map<key-type,value-type> Use to create a map of values. Use any other Vespa type as both key and value. These configurations are identical:
field mymap type map<string, string> {
}

struct mystruct {
    field key type string { }
    field value type string { }
}
field mymap type array<mystruct> {
}
Indexing A map is handled as a struct with a key and value field with key-type and value-type as types, respectively - see below. Indexing restrictions:
  • Structs and maps can only be stored in the document summary store and hence can be used in search result summaries.
  • Structs and maps fields cannot be indexed.
  • Structs and maps can only be searched in streaming search mode.
  • Structs and maps cannot be attributes.
Attribute ToDo
Summary Both key and value have their own summary configuration.
annotationreference Use to define a field (inside annotation, or inside e.g. a struct used by a field in an annotation) with a reference to another annotation. To define a such a field, you must first create an annotation type. The struct must be defined inside the search definition. To declare an annotationreference field in an annotation, use the annotation name to identify the field type:
annotation foo {
    field baz type annotationreference<bar> { }
}

annotation bar { }
Indexing N/A.
Attribute N/A.
Summary N/A.
reference<document-type> A reference<document-type> field is a reference to an instance of a document-type - i.e. a foreign key. The reference is the document id of the document-type instance, hence a string. References are used to join documents in a parent-child relationship. A reference can only be made to global documents.
Indexing Invalid - deployment will fail.
Attribute As string - a reference must be an attribute.
Summary As string

Note that it is possible to make a document field of one type into one or more instances of another search field, by declaring a field outside the document, which uses other fields as input. For example, to create an integer attribute for a string containing a comma-separated list of integers in the document, do like this:

search example {
  document example {
    field yearlist type string { # Comma-separated years
      …
    }
    …
  }

  field year type array<int> { # Search field using the yearlist value
    indexing: input yearlist | split "," | attribute
  }
}

Config overrides affecting searchdefinition

maxtermoccurrences

Limits the number of occurrences of a term that will indexed. By default it is 100. This is a global setting that will be used for all fields in all document types.

<config name='vespa.configdefinition.ilscripts'>
    <maxtermoccurrences>100</maxtermoccurrences>
</config>

fieldmatchmaxlength

This is the max length of a field that is tokenized and indexed. By default it is 1000000. This is a global setting that will be used for all fields in all document types. For individual control see max-length

<config name='vespa.configdefinition.ilscripts'>
    <fieldmatchmaxlength>1000000</fieldmatchmaxlength>
</config>

Example

search example {

  # This documents inherits the fields of two others
  document example inherits mysuperdocument, another {

    field title type string {
      indexing: summary | index
      alias: analias.totitle
      alias default: analias_todefault
    }

    field description type string {
      indexing: summary | index
    }

    field author type string {
      indexing: summary | index
      # author names only, so no stemming
      stemming: none
    }

    field category type string {
      indexing: summary | attribute
      attribute: fast-search
      match: exact
    }

    field popularity type int {
      indexing: summary | attribute
    }

    field measurement type int {
      indexing: summary | attribute
    }

    # Index each category token separately, do not tokenize categories
    field categories type string {
      indexing {
        input categories | split "," | normalize | join " " | index
      }
      indexing-rewrite: none;
    }

    # Categories as an array - preferable
    field morecategories type array<string> {
      indexing: index
    }

    #restrict document access
    field permissions type array<byte> {
      indexing: input permissions . input geoperms . input adultness | summary | attribute
      attribute: fast-search
    }
    #placeholder - elements added to global permissions
    field geoperms type array<byte> {

    }
    #placeholder - elements added to global permissions
    field adultness type array<byte> {

    }
  }

  fieldset default {
    fields: title, description
  }

  # Additional default settings
  rank-profile default inherits default {
    first-phase {
        expression: nativeRank
    }
    second-phase {
      expression {
          0.5 * 0.5 * (0.1 * attribute(popularity) + fieldMatch(description))
          + 0.2 * attributeMatch(category)
          + 0.3 * fieldMatch(title)
      }
      rerank-count: 200
    }
  }

  # Some experimental ranking changes
  rank-profile experimental inherits default {
    second-phase {
      expression {
          0.5 * 0.5 * (attribute(measurement) * attribute(popularity) + fieldMatch(description))
          + 0.2 * attributeMatch(category)
          + 0.3 * fieldMatch(title)
      }
    }
  }

  # Ranking expression from separate file (filename mlrrank.expression, in the same directory as the sd-file)
  rank-profile mlrraning inherits default {
    second-phase {
      expression: file:mlrrank
    }
  }

  rank-profile other inherits experimental {
    second-phase {
      rerank-count: 100
    }
  }

  rank-profile justthebest {
    match-phase {
      attribute: popularity
      max-hits: 10000
      diversity {
        attribute: category
        min-groups: 20
      }
    }
    first-phase {
      expression: nativeRank + 0.1 * attribute(popularity)
    }
    second-phase {
      expression {
          0.5 * 0.5 * (0.1 * attribute(popularity) + fieldMatch(description))
          + 0.2 * attributeMatch(category)
          + 0.3 * fieldMatch(title)
      }
    }
  }

  # A derived field defined outside a document,
  # note how each indexing statement
  # fetches a value to index
  field exact type string {
    indexing {
      input title | summary | index;
      input category | summary | index;
    }
    stemming: none
    normalizing: none
    match {
      exact
      exact-terminator: "@@"
    }
  }

}