Class: Runner

reftest.Runner()

new Runner()

Source:

Methods

(async) run(testUrl, references, expected, timeout) → {Object}

Run a specific reftest.

The assumed semantics are those of web-platform-tests where references form a tree and each test must meet all the conditions to reach one leaf node of the tree in order for the overall test to pass.

Parameters:
Name Type Description
testUrl string

URL of the test itself.
references Array.<Array>

Array representing a tree of references to try.

Each item in the array represents a single reference node and
has the form <code>[referenceUrl, references, relation]</code>,
where <var>referenceUrl</var> is a string to the URL, relation
is either <code>==</code> or <code>!=</code> depending on the
type of reftest, and references is another array containing
items of the same form, representing further comparisons treated
as AND with the current item. Sibling entries are treated as OR.

For example with testUrl of T:

<pre><code>
  references = [[A, [[B, [], ==]], ==]]
  Must have T == A AND A == B to pass

  references = [[A, [], ==], [B, [], !=]
  Must have T == A OR T != B

  references = [[A, [[B, [], ==], [C, [], ==]], ==], [D, [], ]]
  Must have (T == A AND A == B) OR (T == A AND A == C) OR (T == D)
</code></pre>
expected string

Expected test outcome (e.g. PASS, FAIL).
timeout number

Test timeout in milliseconds.
Source:
Returns:

Result object with fields status, message and extra.

Type
Object

(async) setup(urlCount, screenshotMode)

Setup the required environment for running reftests.

This will open a non-browser window in which the tests will be loaded, and set up various caches for the reftest run.

Parameters:
Name Type Description
urlCount Object.<Number>

Object holding a map of URL: number of times the URL will be opened during the reftest run, where that's greater than 1.
screenshotMode string

String enum representing when screenshots should be taken
Source: