APIs: Query Parameters
←APIs - Load Balancing & Redundancy · Index↑ · Discovery→
The Query API supports a range of query string parameters which may be used as part of GET requests, or within Websocket subscriptions.
The following document describes the expected usage and behaviour of these query parameters alongside the RAML specification in order to aid implementers. A description of each individual query parameter is included within the RAML.
Pagination
Query APIs SHOULD support pagination of their API resources where the ‘paged’ trait is specified in the RAML documentation. A 501 HTTP status code MUST be returned where pagination is attempted against a Query API which does not implement it. Pagination is not used by Websocket subscriptions.
Query API clients MUST detect whether pagination is being used by examining the HTTP response headers for ‘X-Paging-Limit’ which MUST be returned in all cases where pagination is in use.
The following implementation notes should be observed:
-
The registry which backs the Query API is expected to maintain a ‘creation’ and ‘update’ timestamp alongside each registered resource. These values should not be returned to API clients in the response body, but will be made available via headers and used as pagination cursors.
-
In order to ensure that pagination does not result in resources being skipped, it is important that there are no duplicate creation or update timestamps stored against resources of the same type (Node, Source, Flow etc). It is suggested (although not mandated) that these timestamps are stored with nanosecond resolution using a TAI timebase, which will allow clients to navigate collections based on a common understanding of time.
-
The choice to page based on ‘creation’ or ‘update’ timestamp depends on the client’s intended use of the data. Paging based on ‘creation’ time from the start of a collection provides the best mechanism to identify all resources held in a registry. Paging based on ‘update’ time from the end of a collection provides the best mechanism to watch for changing resources in the case that Websocket subscriptions are not available.
-
Paging through a collection for a second time, using the same cursors as previously should never result in new data appearing in the response payloads when using ‘creation’ ordering, provided the requested paging times are less than or equal to the maximum paging time held in the registry. However, note that as the registry is dynamic and resources may be updated or deleted a given cursor may return fewer results than it did previously. When using ‘update’ ordering and a query filter, new data may appear when cursors are used for a second time.
-
When query parameters which perform filtering are used at the same time as paging, the filters must be applied by the implementation before applying paging parameters to the resulting data set.
-
A server may choose its own default value for paging limit (see Example 1).
-
Where both ‘since’ and ‘until’ parameters are specified, the ‘since’ value takes precedence where a resulting data set is constrained by the server’s value of ‘limit’ (see Example 5).
-
Servers should include ‘prev’ and ‘next’ Link headers and may include ‘first’ and ‘last’ Link headers in all responses.
Examples
The following examples aim to identify how pagination should behave in the presence of a set of registered data. In order to avoid displaying full resource representations, the only data listed here is the ‘update’ timestamp associated with each registered record. The same procedures can be applied where ‘creation’ timestamps and the ‘?paging.order=create’ parameter are used instead.
Where a paging limit is not specified in a request the server’s default is used.
Sample Data: Registered Node Update Timestamps (Comma-Separated)
[0:1, 0:2, 0:3, 0:4, 0:5, 0:6, 0:7, 0:8, 0:9, 0:10, 0:11, 0:12, 0:13, 0:14, 0:15, 0:16, 0:17, 0:18, 0:19, 0:20]
Each of the above corresponds to the update timestamp of a corresponding Node, in the format <seconds>:<nanoseconds> and displayed in ascending order. These will be used throughout the following examples.
Response payloads in the examples will show these values, but in a real implementation should be replaced by the corresponding JSON objects for the Nodes or other resources being queried.
Example 1: Initial /nodes Request
In this example there are no query parameters used in the request, but as the Query API supports pagination it returns a subset of the results with headers identifying how to page further into the collection.
Request
GET /x-nmos/v1.1/query/nodes
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:20&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:10&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:10
X-Paging-Until: 0:20
Payload Resources
[
0:20,
0:19,
0:18,
0:17,
0:16,
0:15,
0:14,
0:13,
0:12,
0:11
]
Notes
- The data set returned when no query parameters are specified must be from the most recently updated (or created) resources in the collection, returned in descending order.
- The X-Paging headers identify properties of the collection of data returned in the response. These parameters may be used to construct a URL which would return the same set of bounded data on consecutive requests (for example /x-nmos/query/v1.1/nodes?paging.since=0:10&paging.until=0:20).
- The Link header identifies the ‘next’ and ‘prev’ cursors which an application may use to make its next requests. An implementation may also provide ‘first’ and ‘last’ cursors to identify the beginning and end of a collection if this can be addressed via a consistent URL.
Example 2: Request With Custom Limit
This request is similar to Example 1, but the client has chosen to use a custom page size limit.
Request
GET /x-nmos/query/v1.1/nodes?paging.limit=5
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:20&paging.limit=5>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:15&paging.limit=5>; rel="prev"
X-Paging-Limit: 5
X-Paging-Since: 0:15
X-Paging-Until: 0:20
Payload Resources
[
0:20,
0:19,
0:18,
0:17,
0:16
]
Notes
- In this case the server has accepted the client’s paging size limit request. If the client had requested a page size which the server was unable to honour, the actual page size used would be returned in X-Paging-Limit.
Example 3: Request With Since Parameter
Request
GET /x-nmos/query/v1.1/nodes?paging.since=0:4
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:14&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:4&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:4
X-Paging-Until: 0:14
Payload Resources
[
0:14,
0:13,
0:12,
0:11,
0:10,
0:9,
0:8,
0:7,
0:6,
0:5
]
Example 4: Request With Until Parameter
Request
GET /x-nmos/query/v1.1/nodes?paging.until=0:16
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:16&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:6&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:6
X-Paging-Until: 0:16
Payload Resources
[
0:16,
0:15,
0:14,
0:13,
0:12,
0:11,
0:10,
0:9,
0:8,
0:7
]
Example 5: Request With Since & Until Parameters
Request
GET /x-nmos/query/v1.1/nodes?paging.since=0:4&paging.until=0:16
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:14&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:4&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:4
X-Paging-Until: 0:14
Payload Resources
[
0:14,
0:13,
0:12,
0:11,
0:10,
0:9,
0:8,
0:7,
0:6,
0:5
]
Notes
- Whilst both ‘since’ and ‘until’ are specified, as this server example has a default paging limit of ‘10’, the ‘since’ parameter must take precedence. As a result of this the value of X-Paging-Until is lower than requested and a further request must be made to retrieve any remaining data.
Edge Cases
When a client requests data which falls at the extreme ends of the stored data set it may be less clear what values should be returned in the ‘X-Paging-Limit’ and ‘X-Paging-Since’ headers. The following examples are intended to clarify the expectation in these cases.
Where a paging limit is not specified in a request the server’s default is used.
Example 1: Client request occurs at the beginning of the data set
Request
GET /x-nmos/query/v1.1/nodes?paging.until=0:20
In this case, assume that there are stored records for ‘0:21’ and ‘0:22’ but no earlier.
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:20&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:0&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:0
X-Paging-Until: 0:20
Payload Resources
[]
Example 2: Client request occurs at the end of the data set
Request
GET /x-nmos/query/v1.1/nodes?paging.since=0:20
In this case, assume that there are stored records for ‘0:19’ and ‘0:20’ but no later.
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.since=0:20&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?paging.until=0:20&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:20
X-Paging-Until: 0:20
Payload Resources
[]
Notes:
- In this situation the client is expected to re-perform the same request (as specified in the ‘next’ cursor) until data is returned. If the client were to increment the value of ‘since’ requested it would be in danger of moving ahead of the current time and missing records.
Example 3: Client request includes a query parameter resulting in a single result, but no paging parameters
Request
GET /x-nmos/query/v1.1/nodes?label=My%20Node
In this case, assume that the most recently created or updated Node held in the registry has a paging value of ‘0:20’ associated with it. Also assume that the Node with label ‘My Node’ has a paging time of ‘0:15’.
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?label=My%20Node&paging.since=0:20&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?label=My%20Node&paging.until=0:0&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:0
X-Paging-Until: 0:20
Payload Resources
[
0:15
]
Example 4: Client request includes a query parameter resulting in a no result, with no paging parameters
Request
GET /x-nmos/query/v1.1/nodes?label=My%20Invalid%20Node
In this case, assume that the most recently created or updated Node held in the registry has a paging value of ‘0:20’ associated with it.
Response
Headers
Link: <http://example.api.com/x-nmos/query/v1.1/nodes/?label=My%20Invalid%20Node&paging.since=0:20&paging.limit=10>; rel="next", <http://example.api.com/x-nmos/query/v1.1/nodes/?label=My%20Invalid%20Node&paging.until=0:0&paging.limit=10>; rel="prev"
X-Paging-Limit: 10
X-Paging-Since: 0:0
X-Paging-Until: 0:20
Payload Resources
[]
Downgrade Queries
Query APIs SHOULD support downgrade queries against their API resources where the ‘downgrade’ trait is specified in the RAML documentation. A 501 HTTP status code MUST be returned where a downgrade query is attempted against a Query API which does not implement it.
In order to streamline upgrades from one API version to another a Query API may sit in front of a registry which holds registered data matching multiple API versions’ schemas. By default the Query API must only return data matching the API version specified in the request URL, however downgrade queries permit old-versioned responses to be provided to clients which are confident that they can handle any missing attributes between the specified API versions.
Downgrades MUST only be performed between minor API versions as major versions may remove or re-purpose attributes. Clients which support multiple major API versions should retrieve this data via multiple HTTP requests or websocket subscriptions.
Examples
Example 1: No Downgrade Parameter
Request
GET /x-nmos/query/v1.1/nodes
Response
- Returns all Nodes held in the registry which match the v1.1 schema.
- Returns Nodes with versions greater than or equal to v1.2, and less than v2.0. These Nodes must have all non-v1.1 keys stripped by the Query API before they are returned.
Example 2: Downgrade From v1.1 to v1.0
Request
GET /x-nmos/query/v1.1/nodes?query.downgrade=v1.0
Response
- Returns all Nodes held in the registry which match the v1.1 schema.
- Returns Nodes with versions greater than or equal to v1.2, and less than v2.0. These Nodes must have all non-v1.1 keys stripped by the Query API before they are returned.
- Returns all Nodes which are registered in v1.0 format.
Example 3: Downgrade From v1.3 to v1.1
Request
GET /x-nmos/query/v1.3/flows?query.downgrade=v1.1
Response
- Returns all Flows held in the registry which match the v1.3 schema.
- Returns Flows with versions greater than or equal to v1.4, and less than v2.0. These Flows must have all non-v1.3 keys stripped by the Query API before they are returned.
- Returns all Flows which are registered in v1.2 or v1.1 format.
Invalid Examples
Invalid Example 1: Downgrade Between Major API Versions
Request
GET /x-nmos/query/v3.0/flows?query.downgrade=v1.0
Response
- Returns an HTTP 400 error code as downgrade queries may not be performed between major API versions.
Basic Queries
Query APIs SHOULD support basic queries against their API resources. A 501 HTTP status code MUST be returned where a basic query is attempted against a Query API which does not implement it.
Basic queries make use of standard HTTP GET query parameters in the form ‘?key1=value1&key2=value2’. Keys may match any attribute which the API schemas indicate could be returned by a given resource.
Any attribute which may be returned by a particular API resource SHOULD be available to use as a query parameter, however the RAML documentation only explicitly identifies core attributes under ‘queryParameters’.
If a query parameter is requested which does not match an attribute found in any resource, an empty result set must be returned.
- Querying Within Objects
- Querying using attributes of objects is permitted via the use of a ‘.’ separator. For example, the /receivers resource may be queried with ‘?subscription.sender_id=2683ad14-642f-459d-a169-ef91c76cec6b’.
- Querying Within Arrays
- Querying using attributes of objects held within arrays is permitted via the use of a ‘.’ separator. For example, the /nodes resource may be queried with ‘?services.type=urn:x-manufacturer:service:tally’. More advanced queries within arrays may require use of the Resource Query Language ‘in()’ operator.
Examples
Example 1: Basic Query Using One Parameter
Request
GET /x-nmos/query/v1.0/senders?transport=urn:x-nmos:transport:rtp
Response
- Returns all Sender objects which have an attribute ‘transport’ which exactly matches the string ‘urn:x-nmos:transport:rtp’.
Example 2: Basic Query Using Two Parameters
Request
GET /x-nmos/query/v1.0/sources?format=urn:x-nmos:format:video&device_id=9126cc2f-4c26-4c9b-a6cd-93c4381c9be5
Response
- Returns all Source objects which have an attribute of ‘format’ which exactly matches ‘urn:x-nmos:format:video’ AND a ‘device_id’ attribute which exactly matches ‘9126cc2f-4c26-4c9b-a6cd-93c4381c9be5’.
Example 3: Querying Within Objects
Request
GET /x-nmos/query/v1.0/flows?tags.studio=HQ1
Response
- Returns all Flows which have a ‘tags’ attribute with a key of ‘studio’. The value of ‘tags.studio’ (which is an array in this case, see Flow schema) must contain ‘HQ1’ as one of its entries.
Example 4: Querying Within Arrays
Request
GET /x-nmos/query/v1.0/nodes?services.type=urn:x-manufacturer:service:myservice
Response
- The schema defines a Node’s ‘services’ as an array of objects, where ‘type’ is a key in these inner objects. As a result, this query returns all Nodes where one of these service objects has a ‘type’ of ‘urn:x-manufacturer:service:myservice’.
Invalid Examples
Invalid Example 1: Duplicate Query Parameters
Request
GET /x-nmos/query/v1.0/flows?tags.location=Salford&tags.location=London
Response
- This query specifies two matching query parameters with different values. As different programming frameworks may not make both of these values available, this is an invalid query for which the response is not defined and may vary by implementation. If this sort of query is required, it is suggested that Resource Query Language is used (see below).
Advanced (RQL) Queries (Optional)
Query APIs MAY support Resource Query Language (RQL https://github.com/persvr/rql/) queries against their API resources where the ‘rql’ trait is specified in the RAML documentation. A 501 HTTP status code MUST be returned where a RAML query is attempted using RQL functions or operators which are not supported by a Query API.
RQL should be formatted in the normalised form as opposed to using FIQL syntax, and passed via the query string using ‘?query.rql=…’.
Querying within objects and arrays should be performed as in the Basic Queries case by using the ‘.’ separator.
When an RQL query is specified, the Basic Queries format MAY be ignored in order to simplify implementation.
Note that as RQL permits the definition of very complex queries, the server may return a 400 error code to indicate that it is refusing to action the client’s request due to the query’s complexity.
Constraints
Not all RQL operators will be suitable for use with a Query API. It is suggested that the following operators are supported, however these are not mandated and implementations may choose to support their own subset.
1) Functions: in(), out(), select()
2) Logical Operators: and(), or(), not()
3) Relational Operators: eq(), ne(), gt(), ge(), lt(), le()
NB: It is suggested that sort() and limit() are not implemented as paging and associated limits is specified directly by the Query API.
Examples
Example 1: Simple Query
Request
GET /x-nmos/query/v1.1/senders?query.rql=eq(transport,urn%3Ax-nmos%3Atransport%3Artp)
Response
- Returns all Sender objects which have an attribute ‘transport’ which exactly matches the string ‘urn:x-nmos:transport:rtp’. This response matches that of ‘Example 1’ in basic querying.
Example 2: Advanced Query
Request
GET /x-nmos/query/v1.1/sources?query.rql=and(eq(format,urn%3Ax-nmos%3Aformat%3Avideo),in(tags.location,(Salford,London)))
Response
- Returns all Sources which have a ‘format’ attribute equal to ‘urn:x-nmos:format:video’ AND a ‘tags’ attribute with a key of ‘location’, which has values of ‘Salford’ or ‘London’.
Ancestry Queries (Optional)
Query APIs MAY support Source and Flow ancestry queries against their API resources where the ‘ancestry’ trait is specified in the RAML documentation. A 501 HTTP status code MUST be returned where an ancestry query is attempted against a Query API which does not implement it.
Sources and Flows list their ‘parents’ in an array. A Query API implementing ancestry tracking may be queried using ‘?query.ancestry_…’ parameters in order to identify parents or children of a given Source or Flow.
Examples
Example 1: Children Of A Source
Request
GET /x-nmos/query/v1.1/sources?query.ancestry_id=c1398579-15bc-468e-91ec-df5bbefe1cd3&query.ancestry_type=children
Response
Headers
X-Ancestry-Generations: 4
Payload
- Returns Sources which list an ID of ‘c1398579-15bc-468e-91ec-df5bbefe1cd3’ in their ‘parents’ attribute array. These are the first generation ancestors.
- For each first generation ancestor found, the ID of these Sources is used to look up the next generation in the ‘parents’ arrays of further Sources.
- This process continues up to the generation limit returned in the X-Ancestry-Generations header. The server may refuse to honour a requested ‘‘?query.ancestry_generations’ value using a 400 error code if it is deemed to be too resource intensive.
Notes
- Responses should not include the Source with ID specified in the ‘ancestry_id’ parameter.
Example 2: Parents Of A Flow
Request
GET /x-nmos/query/v1.1/flows?query.ancestry_id=ad14888a-3a98-444c-8aa8-4d87b77cbaa1&query.ancestry_type=parents&query.ancestry_generations=2
Response
Headers
X-Ancestry-Generations: 2
Payload
- Returns Flows which are listed in the ‘parents’ array of the Flow identified by the ID ‘ad14888a-3a98-444c-8aa8-4d87b77cbaa1’. These are the first generation parents.
- For each first generation parent found, their ‘parents’ array is used to identify the next generation of parents to be returned.
- This process continues up to the generation limit returned in the X-Ancestry-Generations header (in this case limited to 2 by a query parameter). The server may refuse to honour a requested ‘‘?query.ancestry_generations’ value using a 400 error code if it is deemed to be too resource intensive.
Notes
- Responses should not include the Flow with ID specified in the ‘ancestry_id’ parameter.