OASIS Logo


OData Version 4.02. Part 2: URL Conventions

Committee Specification Draft 01

28 February 2024

\(\hbox{}\)

This stage:

https://docs.oasis-open.org/odata/odata/v4.02/csd01/part2-url-conventions/odata-v4.02-csd01-part2-url-conventions.md (Authoritative)
https://docs.oasis-open.org/odata/odata/v4.02/csd01/part2-url-conventions/odata-v4.02-csd01-part2-url-conventions.html
https://docs.oasis-open.org/odata/odata/v4.02/csd01/part2-url-conventions/odata-v4.02-csd01-part2-url-conventions.pdf

Previous stage:

N/A

Latest stage:

https://docs.oasis-open.org/odata/odata/v4.02/odata-v4.02-part2-url-conventions.md (Authoritative)
https://docs.oasis-open.org/odata/odata/v4.02/odata-v4.02-part2-url-conventions.html
https://docs.oasis-open.org/odata/odata/v4.02/odata-v4.02-part2-url-conventions.pdf

Technical Committee:

OASIS Open Data Protocol (OData) TC

Chairs:

Ralf Handl (ralf.handl@sap.com), SAP SE
Michael Pizzo (mikep@microsoft.com), Microsoft

Editors:

Michael Pizzo (mikep@microsoft.com), Microsoft
Ralf Handl (ralf.handl@sap.com), SAP SE
Heiko Theißen (heiko.theissen@sap.com), SAP SE

Additional artifacts:

This prose specification is one component of a Work Product that also includes:

This specification replaces or supersedes:

This specification is related to:

Abstract:

The Open Data Protocol (OData) enables the creation of REST-based data services, which allow resources, identified using Uniform Resource Locators (URLs) and defined in an Entity Data Model (EDM), to be published and edited by Web clients using simple HTTP messages. This specification defines a set of recommended (but not required) rules for constructing URLs to identify the data and metadata exposed by an OData service as well as a set of reserved URL query string operators.

Status:

This document was last revised or approved by the OASIS Open Data Protocol (OData) TC on the above date. The level of approval is also listed above. Check the “Latest stage” location noted above for possible later revisions of this document. Any other numbered Versions and other technical work produced by the Technical Committee (TC) are listed at https://www.oasis-open.org/committees/tc_home.php?wg_abbrev=odata#technical.

TC members should send comments on this specification to the TC’s email list. Others should send comments to the TC’s public comment list, after subscribing to it by following the instructions at the “Send A Comment” button on the TC’s web page at https://www.oasis-open.org/committees/odata/.

This specification is provided under the RF on RAND Terms Mode of the OASIS IPR Policy, the mode chosen when the Technical Committee was established. For information on whether any patents have been disclosed that may be essential to implementing this specification, and any offers of patent licensing terms, please refer to the Intellectual Property Rights section of the TC’s web page (https://www.oasis-open.org/committees/odata/ipr.php).

Note that any machine-readable content (Computer Language Definitions) declared Normative for this Work Product is provided in separate plain text files. In the event of a discrepancy between any such plain text file and display content in the Work Product’s prose narrative document(s), the content in the separate plain text file prevails.

Key words:

The key words “MUST”, “MUST NOT”, “REQUIRED”, “SHALL”, “SHALL NOT”, “SHOULD”, “SHOULD NOT”, “RECOMMENDED”, “NOT RECOMMENDED”, “MAY”, and “OPTIONAL” in this document are to be interpreted as described in BCP 14 RFC2119 and RFC8174 when, and only when, they appear in all capitals, as shown here.

Citation format:

When referencing this specification the following citation format should be used:

[OData-v4.02-Part2]

OData Version 4.02. Part 2: URL Conventions. Edited by Michael Pizzo, Ralf Handl, and Heiko Theißen. 28 February 2024. OASIS Committee Specification Draft 01. https://docs.oasis-open.org/odata/odata/v4.02/csd01/part2-url-conventions/odata-v4.02-csd01-part2-url-conventions.html. Latest stage: https://docs.oasis-open.org/odata/odata/v4.02/odata-v4.02-part2-url-conventions.html.

Notices

Copyright © OASIS Open 2024. All Rights Reserved.

Distributed under the terms of the OASIS IPR Policy.

The name “OASIS” is a trademark of OASIS, the owner and developer of this specification, and should be used only to refer to the organization and its official outputs.

For complete copyright information please see the full Notices section in an Appendix below.


Table of Contents


1 Introduction

The Open Data Protocol (OData) enables the creation of REST-based data services, which allow resources, identified using Uniform Resource Locators (URLs) and defined in a data model, to be published and edited by Web clients using simple HTTP messages. This specification defines a set of recommended (but not required) rules for constructing URLs to identify the data and metadata exposed by an OData service as well as a set of reserved URL query string operators, which if accepted by an OData service, MUST be implemented as required by this document.

The OData-JSON document specifies the format of the resource representations that are exchanged using OData and the OData-Protocol document describes the actions that can be performed on the URLs (optionally constructed following the conventions defined in this document) embedded in those representations.

Services are encouraged to follow the URL construction conventions defined in this specification when possible as consistency promotes an ecosystem of reusable client components and libraries.

1.1 Changes from Earlier Versions

Section Feature / Change Issue
Section 5.1.1.7.1 New overload for function matchespattern with flags ODATA-1618

1.2 Glossary

1.2.1 Definitions of Terms

1.2.2 Acronyms and Abbreviations

1.2.3 Document Conventions

Keywords defined by this specification use this monospaced font.

Function signatures in this specification use the following paragraph style:

Type FunctionName(Type Param1, Type Param2)

Some sections of this specification are illustrated with non-normative examples.

Example 1: text describing an example uses this paragraph style

Non-normative examples use this paragraph style.

All examples in this document are non-normative and informative only.

All other text is normative unless otherwise labeled.


2 URL Components

A URL used by an OData service has at most three significant parts: the service root URL, the resource path, and query options. Additional URL constructs (such as a fragment) can be present in a URL used by an OData service; however, this specification applies no further meaning to such additional constructs.

Example 2: OData URL broken down into its component parts:

http://host:port/path/SampleService.svc/Categories(1)/Products?$top=2&$orderby=Name
\______________________________________/\____________________/ \__________________/
                  |                               |                       |
          service root URL                  resource path           query options

Mandated and suggested content of these three significant URL components used by an OData service are covered in sequence in the three following chapters.

2.1 URL Parsing

OData follows the URI syntax rules defined in RFC3986 and in addition assigns special meaning to several of the sub-delimiters defined by RFC3986, so special care has to be taken regarding parsing and percent-decoding.

RFC3986 defines three steps for URL processing that MUST be performed before percent-decoding:

After applying these steps defined by RFC3986 the following steps MUST be performed:

Note: neither RFC3986 nor this specification assign special meaning to “+” (octet 0x2B). Some implementations decode “+” (octet 0x2B) as space (octet 0x20), others take it literally.

Clients SHOULD percent-encode space (octet 0x20) as %20 and “+” (octet 0x2B) as %2B and avoid the ambiguous “+” (octet 0x2B) in URLs.

2.2 URL Syntax

The OData syntax rules for URLs are defined in this document and the OData-ABNF. Note that the ABNF is not expressive enough to define what a correct OData URL is in every imaginable use case. This specification document defines additional rules that a correct OData URL MUST fulfill. In case of doubt on what makes an OData URL correct the rules defined in this specification document take precedence. Note also that the rules in OData-ABNF assume that URLs and URL parts have been percent-encoding normalized as described in section 6.2.2.2 of RFC3986 before applying the grammar to them, i.e. all characters in the unreserved set (see rule unreserved in OData-ABNF) are plain literals and not percent-encoded. For characters outside of the unreserved set that are significant to OData the ABNF rules explicitly state whether the percent-encoded representation is treated identical to the plain literal representation. This is done to make the input strings in the ABNF test cases more readable.

For example, one of these rules is that single quotes within string literals are represented as two consecutive single quotes.

Example 3: valid OData URLs:

http://host/service/People('O''Neil')
http://host/service/People(%27O%27%27Neil%27)
http://host/service/People%28%27O%27%27Neil%27%29
http://host/service/Categories('Smartphone%2FTablet')

Example 4: invalid OData URLs:

http://host/service/People('O'Neil')
http://host/service/People('O%27Neil')
http://host/service/Categories('Smartphone/Tablet')

The first and second examples are invalid because a single quote in a string literal must be represented as two consecutive single quotes. The third example is invalid because forward slashes are interpreted as path segment separators and Categories('Smartphone is not a valid OData path segment, nor is Tablet').


3 Service Root URL

The service root URL identifies the root of an OData service. A GET request to this URL returns the format-specific service document, see OData-JSON.

The service root URL MUST terminate in a forward slash.

The service document enables simple hypermedia-driven clients to enumerate and explore the resources published by the OData service.


4 Resource Path

The rules for resource path construction as defined in this section are optional. OData services SHOULD follow the subsequently described URL path construction rules and are indeed encouraged to do so; as such consistency promotes a rich ecosystem of reusable client components and libraries.

Services that do not follow the resource path conventions for entity container children are strongly encouraged to document their resource paths by annotating entity container children with the term Core.ResourcePath defined in OData-VocCore. The annotation value is the URL of the annotated resource and may be relative to xml:base (if present), otherwise the request URL.

Resources exposed by an OData service are addressable by corresponding resource path URL components to enable interaction of the client with that resource aspect.

To illustrate the concept, some examples for resources might be: customers, a single customer, orders related to a single customer, and so forth. Examples of addressable aspects of these resources as exposed by the data model might be: collections of entities, a single entity, properties, links, operations, and so on.

An OData service MAY respond with 301 Moved Permanently or 307 Temporary Redirect from the canonical URL to the actual URL.

4.1 Addressing the Model for a Service

OData services expose their entity model according to OData-CSDLJSON or OData-CSDLXML at the metadata URL, formed by appending $metadata to the service root URL.

Example 5: Metadata document URL

http://host/service/$metadata

OData clients may request a particular format for metadata either through the Accept header or by using the $format system query option.

Example 6: Metadata document URL with specified format

http://host/service/$metadata?$format=json

4.2 Addressing the Batch Endpoint for a Service

OData services that support batch requests expose a batch URL formed by appending $batch to the service root URL.

Example 7: batch URL

http://host/service/$batch

4.3 Addressing Entities

The basic rules for addressing a collection (of entities), a single entity within a collection, a singleton, as well as a property of an entity are covered in the resourcePath syntax rule in OData-ABNF.

Below is a (non-normative) snippet from OData-ABNF:

resourcePath = entitySetName                  [collectionNavigation]
             / singletonEntity                [singleNavigation]
             / actionImportCall
             / entityColFunctionImportCall    [ collectionNavigation ]
             / entityFunctionImportCall       [ singleNavigation ]
             / complexColFunctionImportCall   [ collectionPath ]
             / complexFunctionImportCall      [ complexPath ]
             / primitiveColFunctionImportCall [ collectionPath ]
             / primitiveFunctionImportCall    [ primitivePath ]
             / functionImportCallNoParens     [ querySegment ]
             / crossjoin                      [ querySegment ]
             / %s"$all"                       [ "/" optionallyQualifiedEntityTypeName ]

Since OData has a uniform composable URL syntax and associated rules there are many ways to address a collection of entities, including, but not limited to:

Example 8:

http://host/service/Products

Example 9: function with parameters in resource path

http://host/service/ProductsByCategoryId(categoryId=2)

Example 10: function with parameters as query options

http://host/service/ProductsByColor(color=@color)?@color='red'

Likewise there are many ways to address a single entity.

Sometimes a single entity can be accessed directly, for example by:

Example 11:

http://host/service/BestProductEverCreated

Often however a single entity is accessed by composing more path segments to a resourcePath that identifies a collection of entities, for example by:

Example 12:

http://host/service/Categories(1)

Example 13:

http://host/service/Products/Model.MostExpensive()

These rules are recursive, so it is possible to address a single entity via another single entity, a collection via a single entity and even a collection via a collection; examples include, but are not limited to:

Example 14:

http://host/service/Products(1)/Supplier

Example 15:

http://host/service/Products(1)/Model.MostRecentOrder()

Example 16:

http://host/service/Categories(1)/Products

Example 17:

http://host/service/Categories(1)/Model.TopTenProducts()

Example 18:

http://host/service/Categories(1)/Products/Model.AllOrders()

Finally it is possible to compose path segments onto a resource path that identifies a primitive, complex instance, collection of primitives or collection of complex instances and bind an action or function that returns an entity or collections of entities.

4.3.1 Canonical URL

For OData services conformant with the addressing conventions in this section, the canonical form of an absolute URL identifying a non-contained entity is formed by adding a single path segment to the service root URL. The path segment is made up of the name of the entity set associated with the entity followed by the key predicate identifying the entity within the collection. No type-cast segment is added to the canonical URL, even if the entity is an instance of a type derived from the declared entity type of its entity set.

The canonical key predicate for single-part keys consists only of the key property value without the key property name. For multi-part keys the key properties appear in the same order they appear in the key definition in the service metadata.

Example 19: Non-canonical URL

http://host/service/Categories(ID=1)/Products(ID=1)

Example 20: Canonical URL for previous example:

http://host/service/Products(1)

4.3.2 Canonical URL for Contained Entities

For contained entities (i.e. related via a containment navigation property, see OData-CSDLJSON or OData-CSDLXML) the canonical URL is the canonical URL of the containing entity followed by:

If the containment navigation property is defined on a complex type used within an ordered collection, the canonical URL of the containing entity is the canonical URL for the collection of complex types followed by a segment containing the zero-based ordinal of the complex typed instance within the collection.

If a navigation property leading to a related entity type has a partner navigation property that specifies a referential constraint, then those key properties of the related entity that take part in the referential constraint MAY be omitted from URLs.

Example 21: full key predicate of related entity

https://host/service/Orders(1)/Items(OrderID=1,ItemNo=2)

Example 22: shortened key predicate of related entity

https://host/service/Orders(1)/Items(2)

The two above examples are equivalent if the navigation property Items from Order to OrderItem has a partner navigation property from OrderItem to Order with a referential constraint tying the value of the OrderID key property of the OrderItem to the value of the ID key property of the Order.

The shorter form that does not specify the constrained key parts redundantly is preferred. If the value of the constrained key is redundantly specified, then it MUST match the principal key value.

4.3.4 Resolving an Entity-Id

To resolve an entity-id into a representation of the identified entity, the client issues a GET request to the $entity resource located at the URL $entity relative to the service root URL. The entity-id MUST be specified using the system query option $id. The entity-id may be expressed as an absolute IRI or relative to the request root URL.

Example 23: request the entity representation for an entity-id

http://host/service/$entity?$id=Products(0)

The semantics of $entity are covered in the OData-Protocol document.

4.3.5 Alternate Keys

In addition to the canonical (primary) key an entity set or entity type can specify one or more alternate keys with the Core.AlternateKeys term (see OData-VocCore). Entities can be addressed via an alternate key using the same parentheses-style convention as for the canonical key, with one difference: single-part alternate keys MUST specify the key property name to unambiguously determine the alternate key.

Example 24: the same employee identified via the alternate key SSN, the canonical (primary) key ID using the non-canonical long form with specified key property name, and the canonical short form without key property name

http://host/service/Employees(SSN='123-45-6789')
http://host/service/Employees(ID='A1245')
http://host/service/Employees('A1245')

4.3.6 Key-as-Segment Convention

Services MAY support an alternate convention for addressing entities by appending a segment containing the unprefixed and unquoted key value to the URL of the collection containing the entity. Forward-slashes in key value segments MUST be percent-encoded; single quotes within key value segments are treated as part of the key value and do not need to be doubled or percent encoded.

Example 25: valid OData URLs:

http://host/service/Employees/A1245
http://host/service/People/O'Neil
http://host/service/People/O%27Neil
http://host/service/Categories/Smartphone%2FTablet
http://host/service/ThingyWithDurationKey/P12DT23H59M59.999999999999S
http://host/service/ThingyWithEnumerationKey/Yellow

For multi-part keys, the entity MUST be addressed by multiple segments applied, one for each key value, in the order they appear in the metadata description of the entity key.

Example 26: multi-part key predicate, parentheses-style and key-as-segment

https://host/service/OrderItems(OrderID=1,ItemNo=2)
https://host/service/OrderItems/1/2

If a navigation property leading to a related entity type has a partner navigation property that specifies a referential constraint, then those key properties of the related entity that take part in the referential constraint MUST be omitted from URLs using key-as-segment convention.

Example 27: key predicate of related entity — no key segments for key properties of related entity with a referential constraint to preceding key segments

https://host/service/Orders/1/Items/2

The above example assumes that the navigation property Items from Order to OrderItem has a partner navigation property from OrderItem to Order with a referential constraint tying the value of the OrderID key property of the OrderItem to the value of the ID key property of the Order.

Because representing key values as segments could be ambiguous with other URL construction conventions, services that support key-as segment MUST implement the following precedence rules:

If a segment following an entity collection:

  1. matches a defined OData segment (starting with “$”), treat it as such
  2. matches a qualified bound function, bound action, or type name, treat it as such
  3. matches an unqualified bound function, bound action, or type name defined in a default namespace (see OData-Protocol) treat it as such
  4. treat as a key value

For maximum interoperability, services that support the key-as-segment convention SHOULD also support the canonical parentheses-style convention for addressing an entity within a collection, otherwise they MUST specify the URL for each returned entity in a response, as specified by the particular format.

Note: the key-as-segment convention can only be used with the canonical (primary) key and cannot be used with alternate keys as the key property names are not present in the keys and an alternative key cannot be determined.

4.4 Addressing References between Entities

OData services are based on a data model that supports relationships as first class constructs. For example, an OData service could expose a collection of Products entities each of which are related to a Category entity.

References between entities are addressable in OData just like entities themselves are (as described above) by appending a navigation property name followed by /$ref to the entity URL.

Example 28: URL addressing the references between Categories(1) and Products

http://host/service/Categories(1)/Products/$ref

Resource paths addressing a single related entity reference can be used in DELETE requests to unrelate two entities. In OData 4.01, this includes resource paths that identify an individual entity reference within a related collection by key. In OData 4.0, resource paths addressing a collection of references MUST be followed by the system query option $id in order to identify a single entity reference within the collection to be removed. The entity-id specified by $id may be expressed absolute or relative to the request URL. For details see OData-Protocol.

Example 29: three ways of unrelating Categories(1) and Products(0); the second option is supported only in OData 4.01

DELETE http://host/service/Categories(1)/Products/$ref?$id=../../Products(0)
DELETE http://host/service/Categories(1)/Products(0)/$ref
DELETE http://host/service/Products(0)/Category/$ref

4.5 Addressing Operations

The semantic rules for addressing and invoking actions and functions are defined in the OData-Protocol document.

Services MAY additionally support the use of the unqualified name of an action or function in a URL by defining one or more default namespaces through the Core.DefaultNamespace term defined in OData-VocCore. For more information on default namespaces, see Default Namespaces in OData-Protocol.

4.5.1 Addressing Actions

The grammar for addressing and invoking actions is defined by the following syntax grammar rules in OData-ABNF:

4.5.2 Addressing Functions

The grammar for addressing and invoking functions is defined by syntax rules in OData-ABNF, in particular:

Note: there is no literal representation for Edm.Stream values in URLs, so it is not possible to pass Edm.Stream values to parameters of function imports or to non-binding parameters of bound functions used in the resource path. Function expressions within query options can use path expressions of type Edm.Stream as values of non-binding function parameters.

4.6 Addressing a Property

To address an entity property clients append a path segment containing the property name to the URL of the entity. If the property has a complex type value, properties of that value can be addressed by further property name composition.

4.7 Addressing a Raw Value

To address the raw value of a primitive property or operation result, clients append the path segment /$value to the property or operation URL.

Properties and operation results of type Edm.Stream already return the raw value of the media stream and do not support appending the /$value segment.

4.8 Addressing the Count of a Collection

To address the raw value of the number of items in a collection, clients append /$count to the resource path of the URL identifying the entity set or collection.

The /$count path suffix identifies the integer count of records in the collection and SHOULD NOT be combined with the system query options $top, $skip, $orderby, $expand, and $format. The count MUST NOT be affected by $top, $skip, $orderby, or $expand.

The count is calculated after applying any /$filter path segments, or $filter or $search system query options to the collection.

Example 30: the number of related entities

http://host/service/Categories(1)/Products/$count

Example 31: the number of entities in an entity set

http://host/service/Products/$count

Example 32: entity count in a $filter expression. Note that the spaces around gt are for readability of the example only; in real URLs they must be percent-encoded as %20.

http://host/service/Categories?$filter=Products/$count gt 0

Example 33: count of a filtered collection in a $filter expression; returns all Categories containing more than two products whose price is greater than 5.00.

http://host/service/Categories?$filter=Products/$count($filter=Price gt 5.00) gt 2

Example 34: entity count in an $orderby expression

http://host/service/Categories?$orderby=Products/$count

4.9 Addressing a Member within an Entity Collection

Collections of entities are modeled as entity sets, collection-valued navigation properties, or operation results.

For entity sets, results of operations associated with an entity set through an EntitySet or EntitySetPath declaration, or collection-valued navigation properties with a NavigationPropertyBinding or ContainsTarget=true specification, members of the collection can be addressed by convention by appending the parenthesized key to the URL specifying the collection of entities, or by using the key-as-segment convention if supported by the service.

For collection-valued navigation properties with navigation property bindings that end in a type-cast segment, a type-cast segment MUST be appended to the collection URL before appending the key segment.

Note: entity sets or collection-valued navigation properties annotated with the term Capabilities.IndexableByKey defined in OData-VocCap and a value of false do not support addressing their members by key.

4.10 Addressing a Member of an Ordered Collection

Collections can be annotated as ordered using the Core.Ordered term (see OData-VocCore). Individual items within an ordered collection of primitive or complex types can be addressed by appending a segment containing the zero-based ordinal to the URL of the collection. A negative ordinal number indexes from the end of the collection, with -1 representing the last item in the collection. Services MUST NOT specify a negative index when returning the address of a member of an ordered collection.

Entity types are stably addressable using their canonical URL and SHOULD NOT be accessed or accessible using an index.

Example 35: the first address in a list of addresses for MainSupplier

http://host/service/MainSupplier/Addresses/0

4.11 Addressing Derived Types

Any resource path or path expression identifying a collection of entities or complex type instances can be appended with a type-cast segment, that is a path segment containing the qualified name of a type derived from the declared item type of the collection. The result will be restricted to instances of the derived type and may be empty.

Any resource path or path expression identifying a single entity or complex type instance can be appended with a type-cast segment containing the qualified name of a type derived from the declared type of the identified resource. If used in a resource path and the identified resource is not an instance of the derived type, the request will result in a 404 Not Found response. If used in a path expression, the type cast will evaluate to null.

Services MAY additionally support the use of the unqualified name of a derived type in a URL by defining one or more default namespaces through the Core.DefaultNamespace term defined in OData-VocCore. For more information on default namespaces, see Default Namespaces in OData-Protocol.

Services MAY also support treating an instance as a type outside of the type hierarchy using the same syntax and semantics as when addressing a derived type. In this case, the set and values of properties of the addressed type may be different than the properties of the source type. The set of such possible target types outside of the type hierarchy SHOULD be called out using the Core.MayImplement annotation term, defined in OData-VocCore.

Example 36: entity set restricted to VipCustomer instances

http://host/service/Customers/Model.VipCustomer

Example 37: entity restricted to a VipCustomer instance, resulting in 404 Not Found if the customer with key 1 is not a VipCustomer

http://host/service/Customers/Model.VipCustomer(1)
http://host/service/Customers(1)/Model.VipCustomer

Example 38: cast the complex property Address to its derived type DetailedAddress, then get a property of the derived type

http://host/service/Customers(1)/Address/Model.DetailedAddress/Location

Example 39: filter expression with type cast; will evaluate to null for all non-VipCustomer instances and thus return only instances of VipCustomer

http://host/service/Customers?$filter=Model.VipCustomer/PercentageOfVipPromotionProductsOrdered gt 80

Example 40: expand the single related Customer only if it is an instance of Model.VipCustomer. For to-many relationships only Model.VipCustomer instances would be inlined,

http://host/service/Orders?$expand=Customer/Model.VipCustomer

4.12 Addressing a Subset of a Collection

Any resource path or path expression identifying a collection can be appended with a filter path segment consisting of /$filter followed by parentheses containing a parameter alias or a filter expression following the filter syntax rule in OData-ABNF. If the parentheses contain a parameter alias, a filter expression MUST be assigned to the parameter alias in the query part of the request URL. If the filter path segment appears in the resource path, the filter expression in parentheses MUST NOT use forward slashes, it must be specified with a parameter alias instead.

The collection will be restricted to instances matching the filter expression assigned to the parameter alias and may be empty.

The /$filter path segment MAY occur multiple times and it MAY be combined with the $filter system query option.

Example 41: red products that cost less than 10 — combining path segment and system query option

GET Products/$filter(@foo)?@foo=Price lt 10&$filter=Color eq 'red'

Example 42: red products that cost less than 10 — combine two path segments

GET Products/$filter(@p)/$filter(@c)?@p=Price lt 10&@c=Color eq 'red'

Example 43: categories with less than ten products older than three

GET Categories?$filter=Products/$filter(Age gt 3)/$count lt 10

Note that the /$filter path segment filters the “input” to the operation, and the $filter system query option filters the result of the operation, so the two can be used interchangeably only for GET operations. For a PATCH operation, the /$filter path segment is evaluated on the items before the modification and determines which items are to be modified, whereas the $filter system query option is evaluated after the modification and determines which of the modified items are part of the response.

The /$filter path segment MAY be followed by a path segment identifying a bound action or bound function applicable to the collection.

Example 44: invoke the Special.Cluster action on all products older than 3

POST /service/Products/$filter(@foo)/Special.Cluster?@foo=Age gt 3

4.13 Addressing Each Member of a Collection

To apply a PATCH or DELETE request to each member of a collection, clients append the path segment /$each to a resource path identifying a collection.

To apply a bound action or function to each member of a collection, clients append the path segment /$each followed by a path segment identifying the bound action or function. The response is a collection of instances of the result type of the bound operation. If the bound operation returns a collection, the response is a collection of collections. System query options applied to the response can use $it to reference an item in the outermost collection, followed by /any, /all, or /$count.

The resource path of the collection preceding /$each MAY contain type-cast segments or filter path segments to subset the collection.

4.14 Addressing the Media Stream of a Media Entity

To address the media stream represented by a media entity, clients append /$value to the resource path of the media entity URL. Services may redirect from this canonical URL to the source URL of the media stream.

Example 45: request the media stream for the picture with the key value Sunset4321299432:

http://host/service/Pictures('Sunset4321299432')/$value

4.15 Addressing the Cross Join of Entity Sets

In addition to querying related entities through navigation properties defined in the entity model of a service, the cross join operator allows querying across unrelated entity sets.

The cross join is addressed by appending the path segment $crossjoin to the service root URL, followed by the parenthesized comma-separated list of joined entity sets. It returns the Cartesian product of all the specified entity sets, represented as a collection of instances of a virtual complex type. Each instance consists of one non-nullable, single-valued navigation property per joined entity set. Each such navigation property is named identical to the corresponding entity set, with a target type equal to the declared entity type of the corresponding entity set.

The $filter and $orderby query options can be specified using properties of the entities in the selected entity sets, prepended with the entity set as the navigation property name.

The $expand query option can be specified using the names of the selected entity sets as navigation property names. If a selected entity set is not expanded, it MUST be represented using the read URL of the related entity as a navigation link in the complex type instance.

The $count, $skip, and $top query options can also be used with no special semantics.

Example 46: if Sales had a structural property ProductID instead of a navigation property Product, a “cross join” between Sales and Products could be addressed

http://host/service/$crossjoin(Products,Sales)?$filter=Products/ID eq Sales/ProductID

and would result in

{
  "@odata.context": "http://host/service/$metadata#Collection(Edm.ComplexType)",
  "value": [
    {
      "Products@odata.navigationLink": "Products(0)",
      "Sales@odata.navigationLink": "Sales(42)",
    },
    {
      "Products@odata.navigationLink": "Products(0)",
      "Sales@odata.navigationLink": "Sales(57)",
    },
    
    {
      "Products@odata.navigationLink": "Products(99)",
      "Sales@odata.navigationLink": "Sales(21)",
    }
  ]
}

4.16 Addressing All Entities in a Service

The symbolic resource $all, located at the service root, identifies the collection of all entities in a service, i.e. the union of all entity sets plus all singletons.

This symbolic resource is of type Collection(Edm.EntityType) and allows the $search system query option plus all other query options applicable to collections of entities.

The $all resource can be appended with a path segment containing the qualified name of an entity type in order to restrict the collections to entities of that type. Query options such as $select, $filter, $expand, and $orderby can be applied to this restricted set according to the specified type.

Example 47: all entities in a service that somehow match red

http://host/service/$all?$search=red

Example 48: all Customer entities in a service whose name contains red

http://host/service/$all/Model.Customer?$filter=contains(Name,'red')

4.17 Passing Query Options in the Request Body

The query options part of an OData URL can be quite long, potentially exceeding the maximum length of URLs supported by components involved in transmitting or processing the request. One way to avoid this is wrapping the request in a batch request, which has the penalty of needing to construct a well-formed batch request body.

An easier alternative for GET requests is to append /$query to the resource path of the URL, use the POST verb instead of GET, and pass the query options part of the URL in the request body.

Requests to paths ending in /$query MUST use the POST verb. Query options specified in the request body and query options specified in the request URL are processed together.

The request body MUST use Content-Type: text/plain. It contains the query portion of the URL and MUST use the same percent-encoding as in URLs (especially: no spaces, tabs, or line breaks allowed) and MUST follow the syntax rules described in chapter Query Options.

Example 49: passing a filter condition in the request body

POST http://host/service/People/$query
Content-Type: text/plain

$filter=[FirstName,LastName]%20in%20[["John","Doe"],["Jane","Smith"]]

5 Query Options

The query options part of an OData URL specifies three types of information: system query options, custom query options, and parameter aliases. All OData services MUST follow the query string parsing and construction rules defined in this section and its subsections.

5.1 System Query Options

System query options are query string parameters that control the amount and order of the data returned for the resource identified by the URL. The names of all system query options are optionally prefixed with a dollar ($) character. 4.01 Services MUST support case-insensitive system query option names specified with or without the $ prefix. Clients that want to work with 4.0 services MUST use lower case names and specify the $ prefix.

For GET, PATCH, and PUT requests the following rules apply:

For POST requests to an action URL the return type of the action determines the applicable system query options that a service MAY support, following the same rules as GET requests.

POST requests to an entity set follow the same rules as GET requests that return a single entity.

System query options SHOULD NOT be applied to a DELETE request.

An OData service may support some or all of the system query options defined. If a data service does not support a system query option, it MUST reject any request that contains the unsupported option.

The same system query option, irrespective of casing or whether or not it is prefixed with a $, MUST NOT be specified more than once for any resource.

The semantics of all system query options are defined in the OData-Protocol document.

The grammar and syntax rules for system query options are defined in OData-ABNF.

Dynamic properties can be used in the same way as declared properties. If they are not defined on an instance, they evaluate to null.

5.1.1 Common Expression Syntax

The following operators, functions, and literals can be used in $filter, $orderby, and $compute expressions.

The OData-ABNF commonExpr syntax rule defines the formal grammar of common expressions.

5.1.1.1 Logical Operators

OData defines a set of logical operators that evaluate to true or false (i.e. a boolCommonExpr as defined in OData-ABNF). Logical operators are typically used to filter a collection of resources.

The syntax rules for the logical operators are defined in OData-ABNF. 4.01 Services MUST support case-insensitive operator names. Clients that want to work with 4.0 services MUST use lower case operator names.

The six comparison operators can be used with all primitive values except Edm.Binary, Edm.Stream, and the Edm.Geo types. Edm.Binary, Edm.Stream, and the Edm.Geo types can only be compared to the null value using the eq and ne operators.

When applied to operands of numeric types, numeric promotion rules are applied.

The eq, ne, and in operators can be used with collection-valued operands, and the eq and ne operators can be used with operands of a structured type.

5.1.1.1.1 Equals

The eq operator returns true if the left operand is equal to the right operand, otherwise it returns false.

When applied to operands of entity types, the eq operator returns true if both operands represent the same entity, or both operands represent null.

When applied to operands of complex types, the eq operator returns true if both operands have the same structure and same values, or both operands represent null.

When applied to ordered collections, the eq operator returns true if both operands have the same cardinality and each member of the left operand is equal to the corresponding member of the right operand.

For services that support comparing unordered collections, the eq operator returns true if both operands are equal after applying the same ordering on both collections.

Each of the special values null, -INF, and INF is equal to itself, and only to itself.

The special value NaN is not equal to anything, even to itself.

5.1.1.1.2 Not Equals

The ne operator returns true if the left operand is not equal to the right operand, otherwise it returns false.

When applied to operands of entity types, the ne operator returns true if the two operands do not represent the same entity.

When applied to operands of complex types, the ne operator returns true if the operands do not have the same structure and same values.

When applied to ordered collections, the ne operator returns true if both operands do not have the same cardinality or any member of the left operand is not equal to the corresponding member of the right operand.

For services that support comparing unordered collections, the ne operator returns true if both operands do not have the same cardinality or do not contain the same members, in any order.

Each of the special values null, -INF, and INF is not equal to any value but itself.

The special value NaN is not equal to anything, even to itself.

The null value is not equal to any value but itself.

5.1.1.1.3 Greater Than

The gt operator returns true if the left operand is greater than the right operand, otherwise it returns false.

The special value INF is greater than any number, and any number is greater than -INF.

The Boolean value true is greater than false.

Services SHOULD order language-dependent strings according to the Content-Language of the response, and SHOULD annotate string properties with language-dependent order with the term Core.IsLanguageDependent, see OData-VocCore.

If any operand is null, the operator returns false.

5.1.1.1.4 Greater Than or Equal

The ge operator returns true if the left operand is greater than or equal to the right operand, otherwise it returns false.

See rules for gt and eq for details.

5.1.1.1.5 Less Than

The lt operator returns true if the left operand is less than the right operand, otherwise it returns false.

The special value -INF is less than any number, and any number is less than INF.

The Boolean value false is less than true.

Services SHOULD order language-dependent strings according to the Content-Language of the response, and SHOULD annotate string properties with language-dependent order with the term Core.IsLanguageDependent, see OData-VocCore.

If any operand is null, the operator returns false.

5.1.1.1.6 Less Than or Equal

The le operator returns true if the left operand is less than or equal to the right operand, otherwise it returns false.

See rules for lt and eq for details.

5.1.1.1.7 And

The and operator returns true if both the left and right operands evaluate to true, otherwise it returns false.

The null value is treated as unknown, so if one operand evaluates to null and the other operand to false, the and operator returns false. All other combinations with null return null.

5.1.1.1.8 Or

The or operator returns false if both the left and right operands both evaluate to false, otherwise it returns true.

The null value is treated as unknown, so if one operand evaluates to null and the other operand to true, the or operator returns true. All other combinations with null return null.

5.1.1.1.9 Not

The not operator returns true if the operand returns false, otherwise it returns false.

The null value is treated as unknown, so not null returns null.

5.1.1.1.10 Has

The has operator returns true if the right operand is an enumeration value whose flag(s) are set on the left operand.

The null value is treated as unknown, so if one operand evaluates to null, the has operator returns null.

5.1.1.1.11 In

The in operator returns true if the left operand is a member of the right operand. The right operand MUST be either a comma-separated list of primitive values, enclosed in parentheses, or a single expression that resolves to a collection.

5.1.1.1.12 Logical Operator Examples

The following examples illustrate the use and semantics of each of the logical operators.

Example 50: all products with a Name equal to Milk

http://host/service/Products?$filter=Name eq 'Milk'

Example 51: all products with a Name not equal to Milk

http://host/service/Products?$filter=Name ne 'Milk'

Example 52: all products with a Name greater than Milk:

http://host/service/Products?$filter=Name gt 'Milk'

Example 53: all products with a Name greater than or equal to Milk:

http://host/service/Products?$filter=Name ge 'Milk'

Example 54: all products with a Name less than Milk:

http://host/service/Products?$filter=Name lt 'Milk'

Example 55: all products with a Name less than or equal to Milk:

http://host/service/Products?$filter=Name le 'Milk'

Example 56: all products with a Name equal to Milk that also have a Price less than 2.55:

http://host/service/Products?$filter=Name eq 'Milk' and Price lt 2.55

Example 57: all products that either have a Name equal to Milk or have a Price less than 2.55:

http://host/service/Products?$filter=Name eq 'Milk' or Price lt 2.55

Example 58: all products that do not have a Name that ends with ilk:

http://host/service/Products?$filter=not endswith(Name,'ilk')

Example 59: all products whose style value includes Yellow:

http://host/service/Products?$filter=style has Sales.Pattern'Yellow'

Example 60: all products whose Name is Milk or Cheese:

http://host/service/Products?$filter=Name in ('Milk', 'Cheese')

5.1.1.2 Arithmetic Operators

OData defines a set of arithmetic operators that require operands that evaluate to numeric types. Arithmetic operators are typically used to filter a collection of resources. However, services MAY allow using arithmetic operators with the $orderby system query option.

If an operand of an arithmetic operator is null, the result is null.

The syntax rules for the arithmetic operators are defined in OData-ABNF. 4.01 Services MUST support case-insensitive operator names. Clients that want to work with 4.0 services MUST use lower case operator names.

5.1.1.2.1 Addition

The add operator adds the left and right numeric operands.

For operands of type Edm.Decimal the scale of the result is \({\rm scaleof}(A {\ \tt add\ } B) = \max({\rm scaleof}(A), {\rm scaleof}(B))\), or variable if any operand has variable scale.

The add operator is also valid for the following time-related operands:

The rules for time-related operands are defined in XML-Schema-2, section E.3.3. Specifically, for adding a duration to a date:

Thus, today plus a positive duration smaller than one day is today; today minus a positive duration smaller than one day is yesterday.

5.1.1.2.2 Subtraction

The sub operator subtracts the right numeric operand from the left numeric operand.

For operands of type Edm.Decimal the scale of the result is \({\rm scaleof}(A {\ \tt sub\ } B) = \max({\rm scaleof}(A), {\rm scaleof}(B))\), or variable if any operand has variable scale.

The sub operator is also valid for the following time-related operands:

The rules for time-related operands are defined in XML-Schema-2, section E.3.3. Specifically for subtracting a duration from a date see the preceding section.

5.1.1.2.3 Negation

The negation operator, represented by a minus (-) sign, changes the sign of its numeric or Duration operand.

5.1.1.2.4 Multiplication

The mul operator multiplies the left and right numeric operands. The mul operator is also valid for multiplying a Duration value with a numeric value.

For operands of type Edm.Decimal the scale of the result is \({\rm scaleof}(A {\ \tt mul\ } B) = {\rm scaleof}(A) + {\rm scaleof}(B)\), floating if any operand has floating scale, or else variable if any operand has variable scale.

5.1.1.2.5 Division

The div and divby operators divide the left numeric operand by the right numeric operand. They are also valid for dividing a Duration value by a numeric value.

If the left operand is of type Edm.Decimal with floating scale, Edm.Double, or Edm.Single, then positive div zero returns INF, negative div zero returns -INF, and zero div zero returns NaN. For all other types the request fails.

For operands of type Edm.Decimal the result is computed with maximal decimal scale. If any operand has floating scale, the result has floating scale, else if any operand has variable scale, the result has variable scale. Otherwise the resulting scale is service-specific, and clients can use cast to force the result to a specific scale.

The div and divby operators differ in their handling of integers. If both operands to div are of an integer type, the result is an integer representing the whole number of times the right operator fits into the left operator. The divby operator, on the other hand, promotes both operands to decimal before computing the result, may yield a fractional result, and does not fail for divby zero, returning -INF, INF, or NaN depending on the sign of the left operand.

5.1.1.2.6 Modulo

The mod operator returns the remainder when the left numeric operand is divided by the right numeric operand. The sign of the result is the same as the sign of the left operand. If the right operand is zero, the request fails.

For operands of type Edm.Decimal the scale of the result is \({\rm scaleof}(A {\ \tt mod\ } B) = \max({\rm scaleof}(A), {\rm scaleof}(B))\), or variable if any operand has variable scale.

5.1.1.2.7 Arithmetic Operator Examples

The following examples illustrate the use and semantics of each of the Arithmetic operators.

Example 61: all products with a Price of 2.55:

http://host/service/Products?$filter=Price add 2.45 eq 5.00

Example 62: all products with a Price of 2.55:

http://host/service/Products?$filter=Price sub 0.55 eq 2.00

Example 63: all products with a Price of 2.55:

http://host/service/Products?$filter=Price mul 2.0 eq 5.10

Example 64: all products with a Price of 2.55:

http://host/service/Products?$filter=Price div 2.55 eq 1

Example 65: all products with an integer Rating value of 4 or 5:

http://host/service/Products?$filter=Rating div 2 eq 2

Example 66: all products with an integer Rating value of 5:

http://host/service/Products?$filter=Rating divby 2 eq 2.5

Example 67: all products with a Rating exactly divisible by 5:

http://host/service/Products?$filter=Rating mod 5 eq 0

5.1.1.3 Grouping

The Grouping operator (open and close parenthesis “( )”) controls the evaluation order of an expression. The Grouping operator returns the expression grouped inside the parenthesis.

Example 68: all products because 9 mod 3 is 0

http://host/service/Products?$filter=(4 add 5) mod (4 sub 1) eq 0

5.1.1.4 Canonical Functions

In addition to operators, a set of functions is also defined for use with the $compute, $filter or $orderby system query options, or in parameter alias values. The following sections describe the available functions.

Note: ISNULL or COALESCE operators are not defined. Instead, OData defines a null literal that can be used in comparisons.

If a parameter of a canonical function is null, the function returns null.

The syntax rules for all functions are defined in OData-ABNF. 4.01 Services MUST support case-insensitive canonical function names. Clients that want to work with 4.0 services MUST use lower case canonical function names.

5.1.1.5 String and Collection Functions

5.1.1.5.1 concat

The concat function has three overloads, with the following signatures:

Edm.String concat(Edm.String,Edm.String)
Collection concat(Collection,Collection)
OrderedCollection concat(OrderedCollection,OrderedCollection)

The concat function with string parameter values returns a string that appends the second parameter string value to the first.

The concat function with collection parameter values returns a collection that appends all items of the second collection to the first. If both collections are ordered, the result is also ordered.

The concatMethodCallExpr syntax rule defines how the concat function is invoked.

Example 69: all customers from Berlin, Germany

http://host/service/Customers?$filter=concat(concat(City,', '),Country) eq 'Berlin, Germany'
5.1.1.5.2 contains

The contains function has two overloads, with the following signatures:

Edm.Boolean contains(Edm.String,Edm.String)
Edm.Boolean contains(OrderedCollection,OrderedCollection)

The contains function with string parameter values returns true if the second string is a substring of the first string, otherwise it returns false. String comparison is case-sensitive, case-insensitive comparison can be achieved in combination with tolower or toupper.

The contains function with ordered collection parameter values returns true if the first collection can be transformed into the second collection by removing zero or more items from the beginning or the end of the first collection.

The containsMethodCallExpr syntax rule defines how the contains function is invoked.

Example 70: all customers with a CompanyName that contains Alfreds

http://host/service/Customers?$filter=contains(CompanyName,'Alfreds')
5.1.1.5.3 endswith

The endswith function has two overloads, with the following signatures:

Edm.Boolean endswith(Edm.String,Edm.String)
Edm.Boolean endswith(OrderedCollection,OrderedCollection)

The endswith function with string parameter values returns true if the first string ends with the second string, otherwise it returns false. String comparison is case-sensitive, case-insensitive comparison can be achieved in combination with tolower or toupper.

The endswith function with ordered collection parameter values returns true if the first collection can be transformed into the second collection by removing zero or more items from the beginning of the first collection.

The endsWithMethodCallExpr syntax rule defines how the endswith function is invoked.

Example 71: all customers with a CompanyName that ends with Futterkiste

http://host/service/Customers?$filter=endswith(CompanyName,'Futterkiste')
5.1.1.5.4 indexof

The indexof function has two overloads, with the following signatures:

Edm.Int32 indexof(Edm.String,Edm.String)
Edm.Int32 indexof(OrderedCollection,OrderedCollection)

The indexof function with string parameter values returns the zero-based character position of the first occurrence of the second string in the first string, or -1 if the first string does not contain the second string. String comparison is case-sensitive, case-insensitive comparison can be achieved in combination with tolower or toupper.

The indexof function with ordered collection parameter values returns the zero-based index of the first occurrence of the second collection in the first collection, or -1 if the first collection does not contain the second collection.

The indexOfMethodCallExpr syntax rule defines how the indexof function is invoked.

Example 72: all customers with a CompanyName containing lfreds starting at the second character

http://host/service/Customers?$filter=indexof(CompanyName,'lfreds') eq 1
5.1.1.5.5 length

The length function has two overloads, with the following signatures:

Edm.Int32 length(Edm.String)
Edm.Int32 length(Collection)

The length function with a string parameter value returns the number of characters in the string.

The length function with a collection parameter value returns the number of itens in the collection.

The lengthMethodCallExpr syntax rule defines how the length function is invoked.

Example 73: all customers with a CompanyName that is 19 characters long

http://host/service/Customers?$filter=length(CompanyName) eq 19
5.1.1.5.6 startswith

The startswith function has two overloads, with the following signatures:

Edm.Boolean startswith(Edm.String,Edm.String)
Edm.Boolean startswith(Collection,Collection)

The startswith function with string parameter values returns true if the first string starts with the second string, otherwise it returns false. String comparison is case-sensitive, case-insensitive comparison can be achieved in combination with tolower or toupper.

The startswith function with ordered collection parameter values returns true if the first collection can be transformed into the second collection by removing zero or more items from the end of the first collection.

The startsWithMethodCallExpr syntax rule defines how the startswith function is invoked.

Example 74: all customers with a CompanyName that starts with Alfr

http://host/service/Customers?$filter=startswith(CompanyName,'Alfr')
5.1.1.5.7 substring

The substring function has four overloads, with the following signatures:

Edm.String substring(Edm.String,Edm.Int32)
Edm.String substring(Edm.String,Edm.Int32,Edm.Int32)
OrderedCollection substring(OrderedCollection,Edm.Int32)
OrderedCollection substring(OrderedCollection,Edm.Int32,Edm.Int32)

The two-parameter substring function with string parameter values returns a substring of the first parameter string value, starting at the Nth character and finishing at the last character (where N is the second parameter integer value). The three-parameter substring function with string parameter values returns a substring of the first parameter string value identified by selecting up to M characters starting at the Nth character (where N is the second parameter integer value and M is the third parameter integer value).

The two-parameter substring function with ordered collection parameter values returns an ordered collection consisting of all items of the first collection starting at the Nth item and finishing at the last item. The three-parameter substring function with ordered collection parameter values returns an ordered collection consisting of up to M items of the first collection starting at the Nth item (where N is the second parameter integer value and M is the third parameter integer value).

The start index N is zero-based.

If the start index N is larger than the length of the string/collection, an empty string/collection is returned.

If the length M is larger than the length of the remaining string/collection starting at the Nth character/item, as many characters/items as are available are returned.

A negative length M is a bad request.

A negative start index N, if supported, returns a string/collection starting N characters/items before the end of the string/collection.

The substringMethodCallExpr syntax rule defines how the substring function is invoked.

Example 75: all customers with a CompanyName of lfreds Futterkiste once the first character has been removed

http://host/service/Customers?$filter=substring(CompanyName,1) eq 'lfreds Futterkiste'

Example 76: all customers with a CompanyName that has lf as the second and third characters, e.g, Alfreds Futterkiste

http://host/service/Customers?$filter=substring(CompanyName,1,2) eq 'lf'

5.1.1.6 Collection Functions

5.1.1.6.1 hassubset

The hassubset function has the following signature:

Edm.Boolean hassubset(Collection, Collection)

The hassubset function returns true if the first collection can be transformed into the second collection by reordering and/or removing zero or more items. The hasSubsetMethodCallExpr syntax rule defines how the hassubset function is invoked.

Example 77: hassubset expressions that return true

hassubset([4,1,3],[4,1,3])
hassubset([4,1,3],[1,3,4])
hassubset([4,1,3],[3,1])
hassubset([4,1,3],[4,3])
hassubset([4,1,3,1],[1,1])

Example 78: hassubset expression that returns false: 1 appears only once in the left operand

hassubset([1,2],[1,1,2])
5.1.1.6.2 hassubsequence

The hassubsequence function has the following signature:

Edm.Boolean hassubsequence(OrderedCollection,OrderedCollection)

The hassubsequence function returns true if the first collection can be transformed into the second collection by removing zero or more items. The hasSubsequenceMethodCallExpr syntax rule defines how the hassubsequence function is invoked.

Example 79: hassubsequence expressions that return true

hassubsequence([4,1,3],[4,1,3])
hassubsequence([4,1,3],[4,1])
hassubsequence([4,1,3],[4,3])
hassubsequence([4,1,3,1],[1,1])

Example 80: hassubsequence expressions that return false

hassubsequence([4,1,3],[1,3,4])
hassubsequence([4,1,3],[3,1])
hassubsequence([1,2],[1,1,2])

5.1.1.7 String Functions

5.1.1.7.1 matchespattern

The matchespattern function has the following signatures:

Edm.Boolean matchespattern(Edm.String,Edm.String)
Edm.Boolean matchespattern(Edm.String,Edm.String,Edm.String)

The second parameter MUST evaluate to a string containing an ECMAScript (JavaScript) regular expression. The matchespattern function returns true if the first parameter evaluates to a string matching that regular expression, using syntax and semantics of ECMAScript regular expressions, otherwise it returns false. If the optional third parameter is provided, it MUST evaluate to a string consisting of ECMAScript regular expression flags to modify the match.

Example 81: all customers with a CompanyName that match the (percent-encoded) regular expression ^A.*e$

http://host/service/Customers?$filter=matchespattern(CompanyName,'%5EA.*e$')

Example 82: all customers with a FormattedAddress that contains a line ending with berg or ends with berg

http://host/service/Customers?$filter=matchespattern(FormattedAddress,'berg$','m')
5.1.1.7.2 tolower

The tolower function has the following signature:

Edm.String tolower(Edm.String)

The tolower function returns the input parameter string value with all uppercase characters converted to lowercase according to Unicode rules. The toLowerMethodCallExpr syntax rule defines how the tolower function is invoked.

Example 83: all customers with a CompanyName that equals alfreds futterkiste once any uppercase characters have been converted to lowercase

http://host/service/Customers?$filter=tolower(CompanyName) eq 'alfreds futterkiste'
5.1.1.7.3 toupper

The toupper function has the following signature:

Edm.String toupper(Edm.String)

The toupper function returns the input parameter string value with all lowercase characters converted to uppercase according to Unicode rules. The toUpperMethodCallExpr syntax rule defines how the toupper function is invoked.

Example 84: all customers with a CompanyName that equals ALFREDS FUTTERKISTE once any lowercase characters have been converted to uppercase

http://host/service/Customers?$filter=toupper(CompanyName) eq 'ALFREDS FUTTERKISTE'
5.1.1.7.4 trim

The trim function has the following signature:

Edm.String trim(Edm.String)

The trim function returns the input parameter string value with all leading and trailing whitespace characters, according to Unicode rules, removed. The trimMethodCallExpr syntax rule defines how the trim function is invoked.

Example 85: all customers with a CompanyName without leading or trailing whitespace characters

http://host/service/Customers?$filter=trim(CompanyName) eq CompanyName

5.1.1.8 Date and Time Functions

5.1.1.8.1 date

The date function has the following signature:

Edm.Date date(Edm.DateTimeOffset)

The date function returns the date part of the DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value.

5.1.1.8.2 day

The day function has the following signatures:

Edm.Int32 day(Edm.Date)
Edm.Int32 day(Edm.DateTimeOffset)

The day function returns the day component Date or DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value. The dayMethodCallExpr syntax rule defines how the day function is invoked.

Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone (for example UTC) MUST fail evaluation of the day function for literal Edm.DateTimeOffset values that are not stated in the time zone of the normalized values.

Example 86: all employees born on the 8th day of a month

http://host/service/Employees?$filter=day(BirthDate) eq 8
5.1.1.8.3 fractionalseconds

The fractionalseconds function has the following signatures:

Edm.Decimal fractionalseconds[(Edm.DateTimeOffset)
Edm.Decimal fractionalseconds(Edm.TimeOfDay)

The fractionalseconds function returns the fractional seconds component of the DateTimeOffset or TimeOfDay parameter value as a non-negative decimal value less than 1. The fractionalsecondsMethodCallExpr syntax rule defines how the fractionalseconds function is invoked.

Example 87: all employees born less than 100 milliseconds after a full second of any minute of any hour on any day

http://host/service/Employees?$filter=[fractionalseconds(BirthDate) lt 0.1
5.1.1.8.4 hour

The hour function has the following signatures:

Edm.Int32 hour(Edm.DateTimeOffset)
Edm.Int32 hour(Edm.TimeOfDay)

The hour function returns the hour component (0 to 23) of the DateTimeOffset or TimeOfDay parameter value, evaluated in the time zone of the DateTimeOffset parameter value. The hourMethodCallExpr syntax rule defines how the hour function is invoked.

Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone (for example UTC) MUST fail evaluation of the hour function for literal Edm.DateTimeOffset values that are not stated in the time zone of the normalized values.

Example 88: all employees born in hour 4, between 04:00 (inclusive) and 05:00 (exclusive)

http://host/service/Employees?$filter=hour(BirthDate) eq 4
5.1.1.8.5 maxdatetime

The maxdatetime function has the following signature:

Edm.DateTimeOffset maxdatetime()

The maxdatetime function returns the latest possible point in time as a DateTimeOffset value.

5.1.1.8.6 mindatetime

The mindatetime function has the following signature:

Edm.DateTimeOffset mindatetime()

The mindatetime function returns the earliest possible point in time as a DateTimeOffset value.

5.1.1.8.7 minute

The minute function has the following signatures:

Edm.Int32 minute(Edm.DateTimeOffset)
Edm.Int32 minute(Edm.TimeOfDay)

The minute function returns the minute component (0 to 59) of the DateTimeOffset or TimeOfDay parameter value, evaluated in the time zone of the DateTimeOffset parameter value. The minuteMethodCallExpr syntax rule defines how the minute function is invoked.

Example 89: all employees born in minute 40 of any hour on any day

http://host/service/Employees?$filter=minute(BirthDate) eq 40
5.1.1.8.8 month

The month function has the following signatures:

Edm.Int32 month(Edm.Date)
Edm.Int32 month(Edm.DateTimeOffset)

The month function returns the month component of the Date or DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value. The monthMethodCallExpr syntax rule defines how the month function is invoked.

Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone (for example UTC) MUST fail evaluation of the month function for literal Edm.DateTimeOffset values that are not stated in the time zone of the normalized values.

Example 90: all employees born in May

http://host/service/Employees?$filter=month(BirthDate) eq 5
5.1.1.8.9 now

The now function has the following signature:

Edm.DateTimeOffset now()

The now function returns the current point in time (date and time with time zone) as a DateTimeOffset value.

Services are free to choose the time zone for the current point, for example UTC. Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone SHOULD return a value in the normalized time zone (for example UTC).

5.1.1.8.10 second

The second function has the following signatures:

Edm.Int32 second(Edm.DateTimeOffset)
Edm.Int32 second(Edm.TimeOfDay)

The second function returns the second component (0 to 59 for regular seconds, and 60 for leap seconds, without the fractional part) of the DateTimeOffset or TimeOfDay parameter value. The secondMethodCallExpr syntax rule defines how the second function is invoked.

Example 91: all employees born in second 40 of any minute of any hour on any day

http://host/service/Employees?$filter=second(BirthDate) eq 40
5.1.1.8.11 time

The time function has the following signature:

Edm.TimeOfDay time(Edm.DateTimeOffset)

The time function returns the time part of the DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value.

Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone (for example UTC) MUST fail evaluation of the time function for literal Edm.DateTimeOffset values that are not stated in the time zone of the normalized values.

5.1.1.8.12 totaloffsetminutes

The totaloffsetminutes function has the following signature:

Edm.Int32 totaloffsetminutes(Edm.DateTimeOffset)

The totaloffsetminutes function returns the signed number of minutes in the time zone offset part of the DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value.

5.1.1.8.13 totalseconds

The totalseconds function has the following signature:

Edm.Decimal totalseconds(Edm.Duration)

The totalseconds function returns the duration of the value in total seconds, including fractional seconds.

5.1.1.8.14 year

The year function has the following signatures:

Edm.Int32 year(Edm.Date)
Edm.Int32 year(Edm.DateTimeOffset)

The year function returns the year component of the Date or DateTimeOffset parameter value, evaluated in the time zone of the DateTimeOffset parameter value. The yearMethodCallExpr syntax rule defines how the year function is invoked.

Services that are unable to preserve the offset of Edm.DateTimeOffset values and instead normalize the values to some common time zone (for example UTC) MUST fail evaluation of the year function for literal Edm.DateTimeOffset values that are not stated in the time zone of the normalized values.

Example 92: all employees born in 1971

http://host/service/Employees?$filter=year(BirthDate) eq 1971

5.1.1.9 Arithmetic Functions

5.1.1.9.1 ceiling

The ceiling function has the following signatures

Edm.Double ceiling(Edm.Double)
Edm.Decimal ceiling(Edm.Decimal)

The ceiling function rounds the input numeric parameter up to the nearest numeric value with no decimal component. The ceilingMethodCallExpr syntax rule defines how the ceiling function is invoked.

Example 93: all orders with freight costs that round up to 32

http://host/service/Orders?$filter=ceiling(Freight) eq 32
5.1.1.9.2 floor

The floor function has the following signatures

Edm.Double floor(Edm.Double)
Edm.Decimal floor(Edm.Decimal)

The floor function rounds the input numeric parameter down to the nearest numeric value with no decimal component. The floorMethodCallExpr syntax rule defines how the floor function is invoked.

Example 94: all orders with freight costs that round down to 32

http://host/service/Orders?$filter=floor(Freight) eq 32
5.1.1.9.3 round

The round function has the following signatures

Edm.Double round(Edm.Double)
Edm.Decimal round(Edm.Decimal)

The round function rounds the input numeric parameter to the nearest numeric value with no decimal component. The mid-point between two integers is rounded away from zero, i.e. 0.5 is rounded to 1 and -0.5 is rounded to -1. The roundMethodCallExpr syntax rule defines how the round function is invoked.

Example 95: all orders with freight costs that round to 32

http://host/service/Orders?$filter=round(Freight) eq 32

5.1.1.10 Type Functions

5.1.1.10.1 cast

The cast function has the following signatures:

type cast(type)
type cast(expression,type)

The single parameter cast function returns the current instance cast to the type specified. The two-parameter cast function returns the object referred to by the expression cast to the type specified.

The cast function follows these assignment rules:

  1. The null value can be cast to any type.
  2. Primitive types are cast to Edm.String or a type definition based on it by using the literal representation used in payloads, and WKT (well-known text) format for Geo types, see rules fullCollectionLiteral, fullLineStringLiteral, fullMultiPointLiteral, fullMultiLineStringLiteral, fullMultiPolygonLiteral, fullPointLiteral, and fullPolygonLiteral in OData-ABNF. The cast fails if the target type specifies an insufficient MaxLength.
  3. Edm.String, or a type definition based on Edm.String, can be cast to a primitive type if the string contains a literal representation for the target type.
  4. Numeric primitive types are cast to each other with appropriate rounding. The cast fails if the integer part doesn’t fit into the target type.
  5. Edm.DateTimeOffset, Edm.Duration, and Edm.TimeOfDay values can be cast to the same type with a different precision with appropriate rounding.
  6. Non-Unicode string values can be cast to a Unicode string type definition. Unicode string values can be cast to a non-Unicode string type definition if the Unicode string only contains ASCII characters.
  7. Structured types are assignable to their type or a direct or indirect base type.
  8. Collections are cast item by item.
  9. Enumeration types are cast to integer types based on the numeric value of the enumeration member. The cast fails if the numeric value is not in the value range of the target type.
  10. Integer types are cast to enumeration types based on the numeric value of the enumeration members of the target type. For non-flag enumeration types the cast fails if there is no enumeration member with the same numeric value as the integer value. For flag enumeration types the cast fails if the integer value is not in the value range of the underlying integer type of the enumeration type.
  11. String values containing a representation of a date-time value according to XML-Schema-2, section 3.3.7 dateTime, can be cast to Edm.DateTimeOffset. If the string value does not contain a time-zone offset, it is treated as UTC.

The cast function is optional for primitive values (first five rules) and up-casts (seventh rule).

If the cast fails, the cast function returns null.

5.1.1.10.2 isof

The isof function has the following signatures

Edm.Boolean isof(type)
Edm.Boolean isof(expression,type)

The single parameter isof function returns true if the current instance is assignable to the type specified, according to the assignment rules for the cast function, otherwise it returns false.

The two parameter isof function returns true if the object referred to by the expression is assignable to the type specified, according to the same rules, otherwise it returns false.

The isofExpr syntax rule defines how the isof function is invoked.

Example 96: orders that are also BigOrders

http://host/service/Orders?$filter=isof(NorthwindModel.BigOrder)
http://host/service/Orders?$filter=isof($it,NorthwindModel.BigOrder)

Example 97: orders of a customer that is a VIPCustomer

http://host/service/Orders?$filter=isof(Customer,NorthwindModel.VIPCustomer)

5.1.1.11 Geo Functions

5.1.1.11.1 geo.distance

The geo.distance function has the following signatures:

Edm.Double geo.distance(Edm.GeographyPoint,Edm.GeographyPoint)
Edm.Double geo.distance(Edm.GeometryPoint,Edm.GeometryPoint)

The geo.distance function returns the shortest distance between the two points in the coordinate reference system signified by the two points’ SRIDs.

5.1.1.11.2 geo.intersects

The geo.intersects function has the following signatures:

Edm.Boolean geo.intersects(Edm.GeographyPoint,Edm.GeographyPolygon)
Edm.Boolean geo.intersects(Edm.GeometryPoint,Edm.GeometryPolygon)

The geo.intersects function returns true if the specified point lies within the interior or on the boundary of the specified polygon, otherwise it returns false.

5.1.1.11.3 geo.length

The geo.length function has the following signatures:

Edm.Double geo.length(Edm.GeographyLineString)
Edm.Double geo.length(Edm.GeometryLineString)

The geo.length function returns the total length of its line string parameter in the coordinate reference system signified by its SRID.

5.1.1.12 Conditional Functions

5.1.1.12.1 case

The case function has the following signature:

expression case(Edm.Boolean:expression, …, Edm.Boolean:expression)

Each parameter is a pair of expressions separated by a colon (:), where the first expression — the condition — MUST be a Boolean expression, and the second expression — the result — may evaluate to any type.

The case function evaluates the condition in each pair, starting with the leftmost pair, and stops as soon as a condition evaluates to true. It then returns the value of the result of this pair. It returns null if none of the conditions in any pair evaluates to true. Clients can specify a last pair whose condition is true to get a non-null “default/else/otherwise” result.

Boolean expressions containing DateTimeOffset or TimeOfDay literals without the optional seconds part will introduce ambiguity for parsers. Clients SHOULD use whitespace or parentheses to avoid ambiguity.

Clients SHOULD ensure that the results in all pairs are compatible. If all results are of the same type, the type of the case expression is of that type. If all results are of numeric type, then the type of the case expression is a numeric type capable of representing any of these expressions according to standard type promotion rules.

Services MAY support case expressions containing parameters of incompatible types, in which case the case expression is treated as Edm.Untyped and its value has the type of the parameter expression selected by the case statement.

Example 98: compute signum(X)

$compute=case(X gt 0:1,X lt 0:-1,true:0) as SignumX

5.1.1.13 Lambda Operators

OData defines two operators that evaluate a Boolean expression on a collection. Both must be prepended with a path expression that identifies a collection.

4.01 Services MUST support case-insensitive lambda operator names. Clients that want to work with 4.0 services MUST use lower case lambda operator names.

The argument of a lambda operator is a case-sensitive lambda variable name followed by a colon (:) and a Boolean expression that uses the lambda variable name to refer to properties of the instance or of members of the collection identified by the path expression.

If the name chosen for the lambda variable matches a property name of the current resource referenced by the resource path, the lambda variable takes precedence. Clients can prefix properties of the current resource referenced by the resource path with $it.

Other path expressions in the Boolean expression neither prefixed with the lambda variable nor $it are evaluated in the scope of the instance or of members of the collection at the origin of the path expression prepended to the lambda operator.

5.1.1.13.1 any

The any operator applies a Boolean expression to each member of a collection and returns true if and only if the expression is true for any member of the collection, otherwise it returns false. This implies that the any operator always returns false for an empty collection.

The any operator can be used without an argument expression. This short form returns false if and only if the collection is empty.

Example 99: all Orders that have any Items with a Quantity greater than 100

http://host/service/Orders?$filter=Items/any(d:d/Quantity gt 100)

Example 100: all customers having an order with a deviating shipping address. The Address in the argument expression is evaluated in the scope of the Customers collection.

http://host/service/Customers?$filter=Orders/any(o:o/ShippingAddress ne Address)

Example 101: all categories along with their products used in some order with a deviating unit price. The unprefixed UnitPrice in the argument expression is evaluated in the scope of the expanded Products.

http://host/service/Categories?$expand=Products($filter=OrderItems/any(oi:oi/UnitPrice ne UnitPrice))
5.1.1.13.2 all

The all operator applies a Boolean expression to each member of a collection and returns true if the expression is true for all members of the collection, otherwise it returns false. This implies that the all operator always returns true for an empty collection.

The all operator cannot be used without an argument expression.

Example 102: all Orders that have only Items with a Quantity greater than 100

http://host/service/Orders?$filter=Items/all(d:d/Quantity gt 100)

5.1.1.14 Literals

5.1.1.14.1 Primitive Literals

Primitive literals can appear in the resource path as key property values, and in the query part, for example, as operands in $filter expressions. They are represented according to the primitiveLiteral rule in OData-ABNF.

Example 103: expressions using primitive literals

NullValue eq null
TrueValue eq true
FalseValue eq false
Custom.Base64UrlDecode(binary'T0RhdGE') eq 'OData'
IntegerValue lt -128
DoubleValue ge 0.31415926535897931e1
SingleValue eq INF
DecimalValue eq 34.95
StringValue eq 'Say Hello,then go'
DateValue eq 2012-12-03
DateTimeOffsetValue eq 2012-12-03T07:16:23Z
DurationValue eq duration'P12DT23H59M59.999999999999S'
DurationValue eq 'P12DT23H59M59.999999999999S'
TimeOfDayValue eq 07:59:59.999
GuidValue eq 01234567-89ab-cdef-0123-456789abcdef
Int64Value eq 0
ColorEnumValue eq Sales.Pattern'Yellow'
ColorEnumValue eq 'Yellow'
geo.distance(Location,geography'SRID=0;Point(142.1 64.1)')

Duration literals in OData 4.0 required prefixing with “duration”. Enumeration literals in OData 4.0 required prefixing with the qualified type name of the enumeration.

In OData 4.01, services MUST support duration and enumeration literals with or without the type prefix. OData clients that want to operate across OData 4.0 and OData 4.01 services should always include the prefix for duration and enumeration types.

5.1.1.14.2 Complex and Collection Literals

Complex literals and collection literals in URLs are represented as JSON objects and arrays according to the arrayOrObject rule in OData-ABNF. Such literals MUST NOT appear in the path portion of the URL but can be passed to bound functions and function imports in path segments by using parameter aliases.

Object member values and array items can be expressions, including other objects and arrays, arithmetic expressions, property names, and of course primitive values.

Note that the special characters {, }, [, ], and " MUST be percent-encoded in URLs although some browsers will accept and pass them on unencoded.

Example 104: collection of string literals

http://host/service/ProductsByColors(colors=@c)?@c=["red","green"]

Example 105: check whether a pair of properties has one of several possible pair values

$filter=[FirstName,LastName] in [["John","Doe"],["Jane","Smith"]]
5.1.1.14.3 null

The null literal can be used to compare a value to null, or to pass a null value to a function.

5.1.1.14.4 $it

The $it literal can be used in expressions to refer to the current instance of the resource identified by the resource path. For a collection-valued resource the current instance is the instance currently being evaluated by the system query option. For a single-valued resource it is the resource itself.

It allows comparing properties of related entities to properties of the current instance in expressions within lambda operators or in $filter expressions nested within $expand or $select.

It also can be used in $filter and $orderby expressions on collections of primitive types to refer to the current primitive instance, but using the $this literal is preferred as $this can also be used in $filter and $orderby expressions nested within $select.

Note: property names and property paths in $filter expressions nested within $expand are evaluated in the context of the declared type of the expanded navigation property, so property names and property paths for the current instance of the collection identified by the resource path MUST be prefixed with $it/.

The $it literal can also be used as a path prefix to invoke a bound function overload on the current instance within an expression. Function names without a path prefix refer to an unbound function overload.

Example 106: email addresses ending with .com assuming EmailAddresses is a collection of strings

http://host/service/Customers(1)/EmailAddresses?$filter=endswith($it,'.com')

Example 107: customers along with their orders that shipped to the same city as the customer’s address. The nested filter expression is evaluated in the context of Orders; $it allows referring to values in the outer context of Customers. Note: the nested filter condition could equivalently be expressed as $it/Address/City eq $this/ShipTo/City.

http://host/service/Customers?$expand=Orders($filter=$it/Address/City eq ShipTo/City)

Example 108: products with at least 10 positive reviews. Model.PositiveReviews is a function bound to Model.Product returning a collection of reviews.

http://host/service/Products?$filter=$it/Model.PositiveReviews()/$count ge 10
5.1.1.14.5 $root

The $root literal can be used in expressions to refer to resources of the same service.

Example 109: all employees with the same last name as employee A1235

http://host/service/Employees?$filter=LastName eq $root/Employees('A1245')/LastName

Example 110: products ordered by a set of customers, where the set of customers is passed as a JSON array containing the resource paths from $root to each customer

http://host/service/ProductsOrderedBy(Customers=@c)?@c=[$root/Customers('ALFKI'),$root/Customers('BLAUS')]

Example 111: function call returning the average rating of a given employee by their peers (employees in department D1)

http://host/service/Employees('A1245')/self.AvgRating(RatedBy=@peers)?@peers=$root/Employees/$filter(Department eq 'D1')
5.1.1.14.6 $this

The $this literal can be used in $filter and $orderby expressions nested within $expand and $select for collection-valued properties and navigation properties. It refers to the current instance of the collection.

Example 112: select only email addresses ending with .com

http://host/service/Customers?$select=EmailAddresses($filter=endswith($this,'.com'))

5.1.1.15 Path Expressions

Properties and navigation properties of the structured type on which a common expression is evaluated can be used as operands or function parameters, as shown in the preceding examples.

Properties of complex properties can be used via the same syntax as in resource paths, i.e. by specifying the name of a complex property, followed by a forward slash (/) and the name of a property of the complex property, and so on,

Properties and navigation properties of entities related with a target cardinality 0..1 or 1 can be used by specifying the navigation property, followed by a forward slash (/) and the name of a property of the related entity, and so on.

If a complex property is null, or no entity is related (in case of target cardinality 0..1), its value, and the values of its components, are treated as null.

Example 113: similar behavior whether HeadquarterAddress is a nullable complex type or a nullable navigation property

Companies(1)/HeadquarterAddress/Street

To access properties of derived types, the property name MUST be prefixed with the qualified name of the derived type on which the property is defined, followed by a forward slash (/), see addressing derived types. If the current instance is not of the specified derived type, the path expression returns null.

If the property or navigation property is not defined for the type of the resource and that type supports dynamic properties or navigation properties, then the property or navigation property is treated as null for all instances on which it has no value.

If the property or navigation property is not defined for the type of the resource and that type does not support dynamic properties or navigation properties, then the request may be considered malformed.

5.1.1.16 Annotation Values in Expressions

Services MAY support the use of annotation values as operands or function parameters, and they MAY advertise this by annotating the entity container with term Capabilities.AnnotationValuesInQuerySupported, see OData-VocCap.

Annotation values are referenced by the annotation name which consists of an at sign (@) followed by the qualified term name, optionally followed by a percent-encoded hash (%23) and an annotation qualifier. The annotation name can be prefixed with a path expression leading to the annotated resource or property.

If an annotation is not applied to the resource or property, then its value, and the values of its components, are treated as null.

Example 114: Return Products that have prices in Euro

http://host/service/Products?$filter=Price/@Measures.Currency eq 'EUR'

Example 115: Return Employees that have any error messages in the Core.Messages annotation

http://host/service/Employees?$filter=@Core.Messages/any(m:m/severity eq 'error')

Services MAY additionally support the use of the unqualified term name by defining one or more default namespaces through the Core.DefaultNamespace annotation term defined in OData-VocCore. For more information on default namespaces, see Default Namespaces in OData-Protocol. This short notation however uses the same name pattern as parameter aliases. If a query option is specified as a parameter alias, then any occurrence of the parameter alias name in an expression MUST evaluate to the parameter alias value and MUST NOT evaluate to the annotation value of an identical unqualified term name.

5.1.1.17 Operator Precedence

OData services MUST use the following operator precedence for supported operators when evaluating $filter and $orderby expressions. Operators are listed by category in order of precedence from highest to lowest. Operators in the same category have equal precedence:

Group Operator Description ABNF Expression
Grouping ( ) Precedence grouping parenExpr
boolParenExpr
Primary / Navigation firstMemberExpr
memberExpr
has Enumeration Flags hasExpr
in Is a member of inExpr
xxx( ) Method Call methodCallExpr
boolMethodCallExpr
functionExpr
Unary - Negation negateExpr
not Logical Negation notExpr
cast( ) Type Casting castExpr
Multiplicative mul Multiplication mulExpr
div Division divExpr
divby Decimal Division divbyExpr
mod Modulo modExpr
Additive add Addition addExpr
sub Subtraction subExpr
Relational gt Greater Than gtExpr
ge Greater than or Equal geExpr
lt Less Than ltExpr
le Less than or Equal leExpr
isof Type Testing isofExpr
Equality eq Equal eqExpr
ne Not Equal neExpr
Conditional AND and Logical And andExpr
Conditional OR or Logical Or orExpr

5.1.1.18 Numeric Promotion

Services SHOULD NOT require explicit cast operations between numeric types used in comparison expressions. Wherever possible, such comparisons should be performed using underlying types of sufficient size.

Services MAY support numeric promotion for arithmetic operations or when comparing two operands of comparable types by applying the following rules, in order:

Each of these promotions uses the same semantics as a castExpression to promote an operand to the target type.

OData does not define an implicit conversion between string and numeric types.

5.1.2 System Query Option $filter

The $filter system query option allows clients to filter a collection of resources that are addressed by a request URL. The expression specified with $filter is evaluated for each resource in the collection, and only items where the expression evaluates to true are included in the response. Resources for which the expression evaluates to false or to null, or which reference properties that are unavailable due to permissions, are omitted from the response.

The OData-ABNF filter syntax rule defines the formal grammar of the $filter query option.

5.1.3 System Query Option $expand

The $expand system query option specifies the related resources or media streams to be included in line with retrieved resources.

The OData-ABNF expand syntax rule defines the formal grammar of the $expand query option.

The value of $expand is a comma-separated list of expand items. Each expand item is evaluated relative to the retrieved resource being expanded. An expand item is either a path or one of the symbols * or $value.

A path consists of segments separated by a forward slash (/). Segments are either names of single- or collection-valued complex properties, instance annotations, or type-cast segments consisting of the qualified name of a structured type that is derived from the type identified by the preceding path segment to reach properties defined on the derived type.

A path can end with

If a structured type traversed by the path supports neither dynamic properties nor instance annotations, then a corresponding property segment MUST specify a declared property of that structured type. Otherwise, if a traversed type does support dynamic navigation properties or instance annotations and the corresponding property segment does not specify a declared property, then the expanded property appears only for those instances on which it has a value.

Example 116: expand a navigation property of an entity type

http://host/service/Products?$expand=Category

Example 117: expand a navigation property of a complex type

http://host/service/Customers?$expand=Addresses/Country

A path MUST NOT appear in more than one expand item.

Query options can be applied to an expanded navigation property by appending a semicolon-separated list of query options, enclosed in parentheses, to the navigation property name. Allowed system query options are $compute, $select, $expand, and $levels for all navigation properties, plus $filter, $orderby, $skip, $top, $count, and $search for collection-valued navigation properties.

Example 118: all categories and for each category all related products with a discontinued date equal to null

http://host/service/Categories?$expand=Products($filter=DiscontinuedDate eq null)

The $count segment can be appended to a navigation property name or type-cast segment following a navigation property name to return just the count of the related entities. The $filter and $search system query options can be used to limit the number of related entities included in the count.

Example 119: all categories and for each category the number of all related products

http://host/service/Categories?$expand=Products/$count

Example 120: all categories and for each category the number of all related blue products

http://host/service/Categories?$expand=Products/$count($search=blue)

To retrieve entity references instead of the related entities, append /$ref to the navigation property name or type-cast segment following a navigation property name. The system query options $filter, $search, $skip, $top, and $count can be used to limit the number of expanded entity references.

Example 121: all categories and for each category the references of all related products

http://host/service/Categories?$expand=Products/$ref

Example 122: all categories and for each category the references of all related products of the derived type Sales.PremierProduct

http://host/service/Categories?$expand=Products/Sales.PremierProduct/$ref

Example 123: all categories and for each category the references of all related premier products with a current promotion equal to null

http://host/service/Categories?$expand=Products/Sales.PremierProduct/$ref($filter=CurrentPromotion eq null)

Cyclic navigation properties (whose target type is identical or can be cast to its source type) can be recursively expanded using the special $levels option. The value of the $levels option is either a positive integer to specify the number of levels to expand, or the literal string max to specify the maximum expansion level supported by that service. A $levels option with a value of 1 specifies a single expand with no recursion.

Example 124: all employees with their manager, manager’s manager, and manager’s manager’s manager

http://host/service/Employees?$expand=ReportsTo($levels=3)

It is also possible to expand all declared and dynamic navigation properties using a star (*). To retrieve references to all related entities use */$ref, and to expand all related entities with a certain distance use the star operator with the $levels option. The star operator can be combined with explicitly named navigation properties, which take precedence over the star operator.

The star operator does not implicitly include stream properties.

Example 125: expand Supplier and include references for all other related entities

http://host/service/Categories?$expand=*/$ref,Supplier

Example 126: expand all related entities and their related entities

http://host/service/Categories?$expand=*($levels=2)

Specifying a stream property includes the media stream inline according to the specified format.

Example 127: include Employee’s Photo stream property along with other properties of the customer

http://host/service/Employees?$expand=Photo

Specifying $value for a media entity includes the media entity’s stream value inline according to the specified format.

Example 128: Include the Product’s media stream along with other properties of the product

http://host/service/Products?$expand=$value

5.1.4 System Query Option $select

The $select system query option allows clients to request a specific set of properties for each entity or complex type.

The $select query option is often used in conjunction with the $expand system query option, to define the extent of the resource graph to return ($expand) and then specify a subset of properties for each resource in the graph ($select). Expanded navigation properties MUST be returned, even if they are not specified in $select.

The OData-ABNF select syntax rule defines the formal grammar of the $select query option.

The value of $select is a comma-separated list of select items. Each select item is one of the following:

A path consists of segments separated by a forward slash (/). Segments are either names of single- or collection-valued complex properties, instance annotations, or type-cast segments consisting of the qualified name of a structured type that is derived from the type identified by the preceding path segment to reach properties defined on the derived type.

A path can end with

The $select system query option is interpreted relative to the entity type or complex type of the resources identified by the resource path section of the URL. Each select item in the $select clause indicates that the response MUST include the declared or dynamic properties, actions and functions identified by that select item. If a select item is a path expression traversing an entity or complex property that is null on an instance, then the null-valued entity or complex property is included and represented as null. The simplest form of a select item explicitly requests a property defined on the entity type of the resources identified by the resource path section of the URL.

Example 129: rating and release date of all products

http://host/service/Products?$select=Rating,ReleaseDate

It is also possible to request all declared and dynamic structural properties using a star (*).

Example 130: all structural properties of all products

http://host/service/Products?$select=*

If the select item is not defined for the type of the resource, and that type supports dynamic properties or instance annotations, then the property is treated as null for all instances on which it is not defined.

If the select item is not defined for the type of the resource, and that type does not support dynamic properties or instance annotations, then the request is considered malformed.

If the select item is an instance annotation of type entity or collection of entities, then the request is considered malformed. Entity-valued annotations can be included using $expand.

If the select item is a navigation property, then the corresponding navigation link is represented in the response. If the navigation property also appears in an $expand query option, then it is additionally represented as inline content. This inline content can itself be restricted with a nested $select query option, see section 5.1.2.

Example 131: name and description of all products, plus name of expanded category

http://host/service/Products?$select=Name,Description&$expand=Category($select=Name)

The select item MUST be prefixed with a qualified structured type name in order to select a property defined on a type derived from the type of the resource segment.

A select item that is a complex type or collection of complex type can be followed by a forward slash, an optional type-cast segment, and the name of a property of the complex type (and so on for nested complex types).

Example 132: the AccountRepresentative property of any supplier that is of the derived type Namespace.PreferredSupplier, together with the Street property of the complex property Address, and the Location property of the derived complex type Namespace.AddressWithLocation

http://host/service/Suppliers?$select=Namespace.PreferredSupplier/AccountRepresentative,Address/Street,Address/Namespace.AddressWithLocation/Location

Query options can be applied to a select item that is a path to a single complex value or a collection of primitive or complex values by appending a semicolon-separated list of query options, enclosed in parentheses, to the select item. The allowed system query options depend on the type of the resource identified by the select item, see section System Query Options, with the exception of $expand. The same property MUST NOT have select options specified in more than one place in a request and MUST NOT be specified in more than one expand.

Example 133: select up to five addresses whose City starts with an H, sorted, and with the Country expanded

http://host/service/Customers?$select=Addresses($filter=startswith(City,'H');$top=5;$orderby=Country/Name,City,Street)&$expand=Addresses/Country

Any structural property, non-expanded navigation property, or operation not requested as a select item (explicitly or via a star) SHOULD be omitted from the response.

Annotations requested in $select MUST be included in the response; $select overrules the include-annotations preference (see OData-Protocol) for the explicitly requested annotations. Additional annotations matching the preference can be included even if not requested via $select. The Preference-Applied response header only reflects the set of annotations included due to the include-annotations preference and not those only included due to $select.

If any select item (including a star) is specified, actions and functions SHOULD be omitted unless explicitly requested.

If an action or function is requested as a select item, either explicitly by using its qualified name, or implicitly by requesting all operations in a schema, then the service includes information about how to invoke that operation for each entity identified by the last path segment in the request URL for which the operation can be bound.

If an action or function is requested in a select item using its qualified name and that operation cannot be bound to the entities requested, the service MUST ignore the select item.

Example 134: the ID property, the ActionName action defined in Model and all actions and functions defined in the Model2 for each product if those actions and functions can be bound to that product

http://host/service/Products?$select=ID,Model.ActionName,Model2.*

When multiple select item exist in a $select clause, then the total set of properties, open properties, navigation properties, actions and functions to be returned is equal to the union of the set of those identified by each select item.

5.1.5 System Query Option $orderby

The $orderby system query option allows clients to request resources in a particular order.

The semantics of $orderby are covered in the OData-Protocol document.

The OData-ABNF orderby syntax rule defines the formal grammar of the $orderby query option.

5.1.6 System Query Options $top and $skip

The $top system query option requests the number of items in the queried collection to be included in the result. The $skip query option requests the number of items in the queried collection that are to be skipped and not included in the result. A client can request a particular page of items by combining $top and $skip.

The semantics of $top and $skip are covered in the OData-Protocol document. The OData-ABNF top and skip syntax rules define the formal grammar of the $top and $skip query options respectively.

5.1.7 System Query Option $count

The $count system query option allows clients to request a count of the matching resources included with the resources in the response. The $count query option has a Boolean value of true or false.

The semantics of $count is covered in the OData-Protocol document.

The $search system query option allows clients to request items within a collection matching a free-text search expression.

The $search query option can be applied to a URL representing a collection of entity, complex, or primitive typed instances, to return all matching items within the collection. Applying the $search query option to the $all resource requests all matching entities in the service.

If both $search and $filter are applied to the same request, the results include only those items that match both criteria.

The OData-ABNF search syntax rule defines the formal grammar of the $search query option.

Example 135: all products that are blue or green. It is up to the service to decide what makes a product blue or green.

http://host/service/Products?$search=blue OR green

5.1.8.1 Search Expressions

Search expressions are used within the $search system query option to request entities matching the specified expression. Leading and trailing spaces are not considered part of the search expression.

Terms can be any single word to be matched within the expression.

Terms enclosed in double-quotes comprise a phrase.

Each individual term or phrase comprises a Boolean expression that returns true if the term or phrase is matched, otherwise false. The semantics of what is considered a match is dependent upon the service.

Expressions enclosed in parenthesis comprise a group expression.

The search expression can contain any number of terms, phrases, or group expressions, along with the case-sensitive keywords NOT, AND, and OR, evaluated in that order.

Expressions prefaced with NOT evaluate to true if the expression is not matched, otherwise false.

Two expressions not enclosed in quotes and separated by a space are equivalent to the same two expressions separated by the AND keyword. Such expressions evaluate to true if both expressions evaluate to true, otherwise false.

Expressions separated by an OR evaluate to true if either of the expressions evaluate to true, otherwise false.

To support type-ahead use cases, incomplete search expressions can be sent as OData string literals enclosed in single-quotes, and single-quotes within the search expression doubled. Such an expression can also be used to search for double quotes: ?$search='"'.

The OData-ABNF searchExpr syntax rule defines the formal grammar of the search expression.

5.1.9 System Query Option $format

The $format system query option allows clients to request a response in a particular format and is useful for clients without access to request headers for standard content-type negotiation. Where present $format takes precedence over standard content-type negotiation.

The semantics of $format is covered in the OData-Protocol document.

The OData-ABNF format syntax rule defines the formal grammar of the $format query option.

5.1.10 System Query Option $compute

The $compute system query option allows clients to define computed properties that can be used in a $select or within a $filter or $orderby expression.

The $compute system query option is interpreted relative to the entity type or complex type of the resources identified by the resource path section of the URL.

The value of $compute is a comma-separated list of compute instructions, each consisting of a common expression followed by the keyword as, followed by the name for the computed dynamic property. This name MUST differ from the names of declared or dynamic properties of the identified resources. Services MAY support compute instructions that address dynamic properties added by other compute instructions within the same $compute system query option, provided that the service can determine an evaluation sequence.

The OData-ABNF compute syntax rule defines the formal grammar of the $compute query option.

Computed properties SHOULD be included as dynamic properties in the result and MUST be included if $select is specified with the computed property name, or star (*).

Example 136: compute total price for order items

http://host/service/Orders(10)/Items
  ?$select=Product/Description,Total
  &$filter=Total gt 100
  &$orderby=Total
  &$compute=Product/Price mul Quantity as Total

5.1.11 System Query Option $index

The $index system query option allows clients to do a positional insert into a collection annotated with the Core.PositionalInsert term (see OData-VocCore). The value of the $index system query option is the zero-based ordinal position where the item is to be inserted. The ordinal of items within the collection greater than or equal to the inserted position are increased by one. A negative ordinal indexes from the end of the collection, with -1 representing an insert at the end of the collection.

The OData-ABNF index syntax rule defines the formal grammar of the $index query option.

5.1.12 System Query Option $schemaversion

The $schemaversion system query option allows clients to specify the version of the schema against which the request is made. The semantics of $schemaversion is covered in the OData-Protocol document.

The OData-ABNF schemaversion syntax rule defines the formal grammar of the $schemaversion query option

5.2 Custom Query Options

Custom query options provide an extensible mechanism for service-specific information to be placed in a URL query string. A custom query option is any query option of the form shown by the rule customQueryOption in OData-ABNF.

Custom query options MUST NOT begin with a $ or @ character.

Example 137: service-specific custom query option debug-mode

http://host/service/Products?debug-mode=true

5.3 Parameter Aliases

Parameter aliases can be used in place of literal values in entity keys, function parameters, or within a $filter or $orderby expression.

Parameter aliases MUST start with an @ character, see rule parameterAlias in OData-ABNF.

The semantics of parameter aliases are covered in OData-Protocol. The OData-ABNF rule aliasAndValue defines the formal grammar for passing parameter alias values as query options.

Example 138:

http://host/service/Movies?$filter=contains(@word,Title)&@word='Black'

Example 139:

http://host/service/Movies?$filter=Title eq @title&@title='Wizard of Oz'

Example 140: JSON array of strings as parameter alias value — note that [, ], and " need to be percent-encoded in real URLs, the clear-text representation used here is just for readability

http://host/service/Products/Model.WithIngredients(Ingredients=@i)?@i=["Carrots","Ginger","Oranges"]

6 Conformance

The conformance requirements for OData clients and services are described in OData-Protocol.


Appendix A. References

This appendix contains the normative and informative references that are used in this document.

While any hyperlinks included in this appendix were valid at the time of publication, OASIS cannot guarantee their long-term validity.

A.1 Normative References

The following documents are referenced in such a way that some or all of their content constitutes requirements of this document.

[OData-ABNF]

ABNF components: OData ABNF Construction Rules Version 4.02 and OData ABNF Test Cases.
See link in “Related work” section on cover page.

[OData-CSDL]

OData Common Schema Definition Language (CSDL) JSON Representation Version 4.02.
See link in “Related work” section on cover page.

OData Common Schema Definition Language (CSDL) XML Representation Version 4.02.
See link in “Related work” section on cover page.

[OData-JSON]

OData JSON Format Version 4.02.
See link in “Related work” section on cover page.

[OData-Protocol]

OData Version 4.02. Part 1: Protocol.
See link in “Related work” section on cover page.

[OData-VocCap]

OData Vocabularies Version 4.0: Capabilities Vocabulary.
See link in “Related work” section on cover page.

[OData-VocCore]

OData Vocabularies Version 4.0: Core Vocabulary.
See link in “Related work” section on cover page.

[RFC2119]

Bradner, S., “Key words for use in RFCs to Indicate Requirement Levels”, BCP 14, RFC 2119, DOI 10.17487/RFC2119, March 1997. https://www.rfc-editor.org/info/rfc2119.

[RFC3986]

Berners-Lee, T., Fielding, R., and L. Masinter, “Uniform Resource Identifier (URI): Generic Syntax”, STD 66, RFC 3986, DOI 10.17487/RFC3986, January 2005. https://www.rfc-editor.org/info/rfc3986.

[RFC8174]

Leiba, B., “Ambiguity of Uppercase vs Lowercase in RFC 2119 Key Words”, BCP 14, RFC 8174, DOI 10.17487/RFC8174, May 2017. https://www.rfc-editor.org/info/rfc8174.

[XML-Schema-2]

W3C XML Schema Definition Language (XSD) 1.1 Part 2: Datatypes. D. Peterson, S. Gao, C. M. Sperberg-McQueen, H. S. Thompson, P. V. Biron, A. Malhotra, Editors, W3C Recommendation, 5 April 2012.
http://www.w3.org/TR/2012/REC-xmlschema11-2-20120405/. Latest version available at http://www.w3.org/TR/xmlschema11-2/.

A.2 Informative References

[ECMAScript]

ECMAScript 2023 Language Specification, 14th Edition, June 2023. Standard ECMA-262. https://www.ecma-international.org/publications-and-standards/standards/ecma-262/.


Appendix B. Safety, Security and Privacy Considerations


Appendix C. Acknowledgments

C.1 Participants

OData TC Members:

First Name Last Name Company
George Ericson Dell
Hubert Heijkers IBM
Ling Jin IBM
Stefan Hagen Individual
Michael Pizzo Microsoft
Christof Sprenger Microsoft
Ralf Handl SAP SE
Gerald Krause SAP SE
Heiko Theißen SAP SE

Appendix D. Revision History

Revision Date Editor Changes Made
Committee Specification Draft 01 2024-02-28 Michael Pizzo
Ralf Handl
Heiko Theißen
Import material from OData Version 4.01 Part 2: URL Conventions
Changes listed in section 1.1

Appendix E. Notices

Copyright © OASIS Open 2024. All Rights Reserved.

All capitalized terms in the following text have the meanings assigned to them in the OASIS Intellectual Property Rights Policy (the “OASIS IPR Policy”). The full Policy may be found at the OASIS website.

This document and translations of it may be copied and furnished to others, and derivative works that comment on or otherwise explain it or assist in its implementation may be prepared, copied, published, and distributed, in whole or in part, without restriction of any kind, provided that the above copyright notice and this section are included on all such copies and derivative works. However, this document itself may not be modified in any way, including by removing the copyright notice or references to OASIS, except as needed for the purpose of developing any document or deliverable produced by an OASIS Technical Committee (in which case the rules applicable to copyrights, as set forth in the OASIS IPR Policy, must be followed) or as required to translate it into languages other than English.

The limited permissions granted above are perpetual and will not be revoked by OASIS or its successors or assigns.

This document and the information contained herein is provided on an “AS IS” basis and OASIS DISCLAIMS ALL WARRANTIES, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO ANY WARRANTY THAT THE USE OF THE INFORMATION HEREIN WILL NOT INFRINGE ANY OWNERSHIP RIGHTS OR ANY IMPLIED WARRANTIES OF MERCHANTABILITY OR FITNESS FOR A PARTICULAR PURPOSE.

As stated in the OASIS IPR Policy, the following three paragraphs in brackets apply to OASIS Standards Final Deliverable documents (Committee Specification, Candidate OASIS Standard, OASIS Standard, or Approved Errata).

[OASIS requests that any OASIS Party or any other party that believes it has patent claims that would necessarily be infringed by implementations of this OASIS Standards Final Deliverable, to notify OASIS TC Administrator and provide an indication of its willingness to grant patent licenses to such patent claims in a manner consistent with the IPR Mode of the OASIS Technical Committee that produced this deliverable.]

[OASIS invites any party to contact the OASIS TC Administrator if it is aware of a claim of ownership of any patent claims that would necessarily be infringed by implementations of this OASIS Standards Final Deliverable by a patent holder that is not willing to provide a license to such patent claims in a manner consistent with the IPR Mode of the OASIS Technical Committee that produced this OASIS Standards Final Deliverable. OASIS may include such claims on its website, but disclaims any obligation to do so.]

[OASIS takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this OASIS Standards Final Deliverable or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any effort to identify any such rights. Information on OASIS’ procedures with respect to rights in any document or deliverable produced by an OASIS Technical Committee can be found on the OASIS website. Copies of claims of rights made available for publication and any assurances of licenses to be made available, or the result of an attempt made to obtain a general license or permission for the use of such proprietary rights by implementers or users of this OASIS Standards Final Deliverable, can be obtained from the OASIS TC Administrator. OASIS makes no representation that any information or list of intellectual property rights will at any time be complete, or that any claims in such list are, in fact, Essential Claims.]

The name “OASIS” is a trademark of OASIS, the owner and developer of this specification, and should be used only to refer to the organization and its official outputs. OASIS welcomes reference to, and implementation and use of, specifications, while reserving the right to enforce its marks against misleading uses. Please see https://www.oasis-open.org/policies-guidelines/trademark/ for above guidance.