Static
createStreamLink configuration as JSON.
A StreamLink object.
The configuration format is JSON, which can contain the following attributes:
Attribute name | Type | Default | Optional | Description |
---|---|---|---|---|
liberator_urls | string or [string] or [[string]] or function() -> string | None | No | The Liberator URL groups to connect to (see below), or a function to return the next Liberator URL to try. |
credentials_provider | CredentialsProvider | None | Yes | The implementation of the CredentialsProvider interface. This custom Credentials Provider that will be used to send credentials to log into Liberator. |
username | string | None | Yes | Liberator login user-name to use if logins are not being managed by KeyMaster. |
password | string | None | Yes | Liberator login password to use if logins are not being managed by KeyMaster. |
keymaster_url | string | None (KeyMaster not used) | Yes | URL of KeyMaster for secure authentication of Liberator login requests. |
keymaster_poll_url | string | None (KeyMaster not used) | Yes | URL that StreamLink uses to check the connection to KeyMaster every
keymaster_poll_interval milliseconds. |
keymaster_poll_interval | number | 60000 milliseconds (1 minute) |
Yes | Interval in milliseconds at which StreamLink will poll KeyMaster
(on keymaster_poll_url ) to keep the KeyMaster session open. |
keymaster_poll_response_listener | KeymasterPollResponseListener | None | Yes | The implementation of the KeymasterPollResponseListener interface. This receives information about responses to keep-alive requests made to KeyMaster. |
keymaster_token_transport_method | string | POST | Yes | WARNING: We strongly recommend that you do not set this attribute to GET for security reasons. This attribute sets the transport method that KeyMaster will use to request the KeyMaster token, can be set to either GET or POST. |
keymaster_token_transport_method_poll | string | POST | Yes | WARNING: We strongly recommend that you do not set this attribute to GET for security reasons. This attribute sets the transport method that KeyMaster will use to poll the KeyMaster token, can be set to either GET or POST. |
keymaster_cross_domain_authentication | boolean | false | Yes | Enable cross-site (CORS) XMLHttpRequest invocations for keymaster requests. The server must be CORS enabled and include the Access-Control-Allow-Credentials: true http header. |
keymaster_http_header_provider | KeymasterHttpHeaderProvider | None | Yes | An implementation of the caplin.streamlink.KeymasterHttpHeaderProvider interface. Used to add http headers to KeyMaster requests. |
poll_interval | number | 1000 milliseconds (1 second) |
Yes | Time interval in milliseconds between polls of the Liberator. This only applies to HTTP polling type connections (see 'RTTP connection types' in the document StreamLink Overview). |
clock_offset_interval | number | 10000 milliseconds (10 seconds) |
Yes | StreamLink calculates a new clock offset value every clock_offset_interval milliseconds.
The clock offset is Liberator time minus StreamLink time; it is a measure of how much the Liberator
and client clocks differ and is used in the calculation of latency statistics, such as average latency
(see the latency_interval attribute). Also see StatisticsEvent.getClockOffset() . |
clock_offset_window | number | 120000 milliseconds (2 minutes) |
Yes | Time in milliseconds a calculated clock offset is valid for use in producing statistics. For the definition
of clock offset, see the clock_offset_interval attribute. You should not normally need to change
this attribute; if you think it needs to be changed, please contact Caplin Support for advice. |
latency_interval | number | 10000 milliseconds (10 seconds) |
Yes | Time in milliseconds between successive calculations of average latency.
Latency is the time taken for a StreamLink heartbeat message to be sent from the Liberator to the client.
The shorter the latency_interval , the more accurate are the average latency statistics.
However, too small a value can adversely impact performance. |
latency_window | number | 120000 milliseconds (2 minutes) |
Yes | Time in milliseconds for which a calculated latency value is valid for use in producing latency statistics. The shorter the time, the faster changes in average latency will show up in the statistics. However, too small a value can result in the average varying in a way that makes it hard to see trends. |
request_timeout | number | 10000 milliseconds (10 seconds) |
Yes | Time in milliseconds allowed for the response to each request before StreamLink reconnects to the Liberator. |
http_timeout | number | 3000 milliseconds (3 seconds) |
Yes | Time in milliseconds allowed for an HTTP Request to finish successfully before StreamLink times out. |
inactivity_timeout | number | -1 (disabled) | Yes | Set the duration (in milliseconds) that Liberator will use to disconnect this client should no messages have been received from StreamLink |
disable_polling | boolean | false | Yes | Set to true if you don't want to use polling connection when the rttp(s) protocol is specified. |
disable_sse_connection | boolean | false | Yes | Set to true if you don't want to use server sent events connection when the rttp(s) protocol is specified. |
reconnect_delay | number | 1000 milliseconds (1 second) |
Yes | Delay in milliseconds before retrying, after failing to connect. |
reconnect_attempts | number | 0 | Yes | Sets the number of times that Streamlink should retry a connection attempt for each connection type (eg. ws, https, poll etc.) Defaults to 0 i.e. only one connection attempt is made per connection type. |
reconnect_cycle_delays | number | 2000, 4000, 8000, 16000 | Yes | Array of backoff delays in milliseconds before retrying, after failing to connect with all available connection options. The entries are treated as a backoff where each entry is used once and the last entry is the delay that will keep getting used forever |
forever_frame_reconnect_count | number | 10000 | Yes | After receiving this number of messages, a forever frame connection will perform a fast reconnect to prevent memory leaks. |
application_id | string | None | Yes | Application identifier. This is used by the Caplin Platform to distinguish StreamLink requests from different client applications. It is recorded in StreamLink log files, and is used by Liberator in monitoring software licence compliance and by the Caplin Permissioning Service to restrict the applicability of permissioning rules to specific applications. |
http_pad_length | number | 4096 bytes | Yes | The amount of padding to use when establishing a streaming connection using the http(s):// connection type. Increasing this value can help should there be a buffering network component between the client and the Liberator. |
log_window_log_level | LogLevel | LogLevel.FINE | Yes | When the debug log window is shown, (see log_window_visible), only messages
at or above the log level defined by log_window_log_level appear in the window. |
log_window_visible | boolean | none | Yes | When true, the debug log window is shown when StreamLink first connects. When false, the debug log window cannot be shown. When left undefined the debug window can be shown if the page url contains a fragment of the form ?debug=[LogLevel name] .
|
log_buffer_size | number | -1 (disabled) | Yes | Sets the number of log lines to be cached in the log buffer. |
webworker_timer_url | string | None | Yes | Sets the url location of the webworker js code used to prevent throttling of background timers. |
fetch_credentials | string | None | Yes | Sets the credentials option on connections using an underlying Fetch request. Possible values are: "omit", "same-origin", "include" (see: https://developer.mozilla.org/en-US/docs/Web/API/Request/credentials). |
sse_with_credentials | boolean | None | Yes | If set to true will include the 'credentials: "include"' parameter on underlying server-sent events connection requests. |
clientside_refiner_enable | boolean | false | Yes | Enables clientside Refiner for all sorted/filtered containers, to specify which containers are refined client/server use the clientside_refiner_fn_use_for callback function. |
clientside_refiner_fn_use_for | function | always refine on the client | Yes | If clientside_refiner_enable is true this function will be called to determine if a particular container subscription should be refined on the client or server. The signature of the function is (subject :String, parameters: SubscriptionParameters) => boolean, returning true will refine on the client. |
clientside_refiner_custom_sorts | object | none | Yes | A Javascript object containing a mapping between custom sort names and their functions. For example {tenorSort: tenorSortFunction} where tenorSortFunction is a comparator function of the form (a,b) => number, number should be -1, 0, 1 for a<b, a==b, a>b |
clientside_refiner_custom_filters | object | none | Yes | A Javascript object containing a mapping between custom filter names and their functions. For example {evennumbers: evenNumberFunction}. The mapped functions can be of 2 forms, 1: a comparator like the custom sort functions in clientside_refiner_custom_sorts, 2: a simple function of the form (fieldProvider) => boolean, where fieldProvider is an object with a single method, getField(fieldName) for retrieving a field from the record being filtered. If the filter expression is of the form where:customFilterName:<fieldname>:<operator>:<value> (e.g where:tenorFilter:FIELD1<1m) then the filter is expected to be of type 1 (a comparator), if however the filter expression is of the form where:evennumbers then the filter is expected to be of type 2 (a simple filter). |
custom_attributes | object | None | Yes | A set of custom attributes provided in JSON object format (key-value pairs). These are used to annotate connections made using this StreamLink instance with any other useful identifying information. The attributes are recorded in the Liberator's RTTP request and traffic logs when the connection is established. |
json_handler | object | None | Yes | Sets a JsonHandler that will be used to parse and patch JS objects. |
Failover configuration using liberator_urls.
The liberator_urls
configuration attribute defines the Liberator URL groups to connect
to in failover situations. A URL group is selected at random and StreamLink tries to
connect to the URLs in the group sequentially. If StreamLink fails to connect to
any of the URLs in a group, the next group is tried, and so on.
For more information about how to configure a failover strategy for StreamLink, see the "Resilience, failover, and load balancing" section of the StreamLink Overview.
Each URL specified must include the protocol; one of: rttp, rttps, http, https, ws, wss, poll, polls.
Examples:
The following example shows the JSON configuration to connect to a single Liberator instance running on the
host liberator1
on port 8080. The credentials to login to the Liberator are
specified using the username
and password
attributes.
StreamLinkFactory.create({ "liberator_urls": "rttp://liberator1:8080", "username": "test@example.com", "password": "my password" });
The following example shows the JSON configuration for failover between two
Liberator pairs. The credentials to login to the Liberator are
specified using the username
and password
attributes.
StreamLinkFactory.create({ "liberator_urls": [ ["rttp://liberator1:8080", "rttp://liberator2:8080"], ["rttp://liberator3:8080", "rttp://liberator4:8080"] ], "username": "test@example.com", "password": "my password" });
liberator_urls can alternatively return a function that will return the next Liberator URL to try connecting to, this can be used to dynamically control the Liberator sequence:
var index = -1; var fixedLiberators = ["rttp://liberator1:8080", "rttp://liberator2:8080"]; StreamLinkFactory.create({ "liberator_urls": function(){ index = (++index) % fixedLiberators.length; return fixedLiberators[index]; }, "username": "test@example.com", "password": "my password" });
Generated using TypeDoc
Creates a StreamLink object that has the supplied StreamLink configuration settings.