Page Menu
Home
WickedGov Phorge
Search
Configure Global Search
Log In
Files
F7408706
ve.ce.TestRunner.js
No One
Temporary
Actions
View File
Edit File
Delete File
View Transforms
Subscribe
Flag For Later
Award Token
Size
9 KB
Referenced Files
None
Subscribers
None
ve.ce.TestRunner.js
View Options
/*!
* VisualEditor IME-like testing
*
* @copyright See AUTHORS.txt
*/
/**
* Text offset from start or end of string
*
* The offsets are measured in text node code unit boundaries. That is, there is an
* implicit start and end position in each text node, so there are eight positions in
* "he<b>ll</b>o". This will not always agree exactly with a real browser's cursor
* positions, because of grapheme clustering and positioning around tags, but is a
* useful test shorthand.
*
* @class
* @constructor
* @param {string} direction (forward counts from string start, backward from end)
* @param {number} offset Offset in text node code units
*/
ve
.
ce
.
TestOffset
=
function
VeCeTestOffset
(
direction
,
offset
)
{
if
(
direction
!==
'forward'
&&
direction
!==
'backward'
)
{
throw
new
Error
(
'Invalid direction "'
+
direction
+
'"'
);
}
this
.
direction
=
direction
;
this
.
offset
=
offset
;
this
.
lastText
=
null
;
this
.
lastSel
=
null
;
};
/**
* @typedef {Object} OffsetInfo
* @memberof ve.ce.TestOffset
* @property {number} [consumed] The number of code units consumed (if n out of range)
* @property {Node} [node] The node containing the offset (if n in range)
* @property {number} [offset] The offset in code units / child elements (if n in range)
* @property {string} [slice] String representation of the offset position (if n in range)
*/
/**
* Calculate the offset from each end of a particular HTML string
*
* @param {Node} node The DOM node with respect to which the offset is resolved
* @return {ve.ce.TestOffset.OffsetInfo} Offset information
*/
ve
.
ce
.
TestOffset
.
prototype
.
resolve
=
function
(
node
)
{
const
reversed
=
(
this
.
direction
!==
'forward'
);
return
ve
.
ce
.
TestOffset
.
static
.
findTextOffset
(
node
,
this
.
offset
,
reversed
);
};
/* Static Methods */
ve
.
ce
.
TestOffset
.
static
=
{};
/**
* Find text by offset in the given node.
* Returns the same as #resolve.
*
* @private
* @static
* @param {Node} node
* @param {number} n Offset
* @param {boolean} reversed
* @return {ve.ce.TestOffset.OffsetInfo} Offset information
*/
ve
.
ce
.
TestOffset
.
static
.
findTextOffset
=
function
(
node
,
n
,
reversed
)
{
if
(
node
.
nodeType
===
node
.
TEXT_NODE
)
{
// Test length >= n because one more boundaries than code units
if
(
node
.
textContent
.
length
>=
n
)
{
const
offset
=
reversed
?
node
.
textContent
.
length
-
n
:
n
;
const
slice
=
node
.
textContent
.
slice
(
0
,
offset
)
+
'|'
+
node
.
textContent
.
slice
(
offset
);
return
{
node
:
node
,
offset
:
offset
,
slice
:
slice
};
}
else
{
return
{
consumed
:
node
.
textContent
.
length
+
1
};
}
}
if
(
node
.
nodeType
!==
node
.
ELEMENT_NODE
)
{
return
{
consumed
:
0
};
}
// node is an ELEMENT_NODE below here
// TODO consecutive text nodes will cause an extra phantom boundary.
// In realistic usage, this can't always be avoided, because normalize() will
// close an IME.
const
childNodes
=
Array
.
prototype
.
slice
.
call
(
node
.
childNodes
);
if
(
childNodes
.
length
===
0
)
{
if
(
n
===
0
)
{
return
{
node
:
node
,
offset
:
0
,
slice
:
'|'
};
}
return
{
consumed
:
0
};
}
if
(
reversed
)
{
childNodes
.
reverse
();
}
let
consumed
=
0
;
for
(
let
i
=
0
,
len
=
childNodes
.
length
;
i
<
len
;
i
++
)
{
const
childNode
=
node
.
childNodes
[
i
];
const
found
=
ve
.
ce
.
TestOffset
.
static
.
findTextOffset
(
childNode
,
n
-
consumed
,
reversed
);
if
(
found
.
node
)
{
return
found
;
}
consumed
+=
found
.
consumed
;
// Extra boundary after element, if not followed by a text node
if
(
childNode
.
nodeType
===
node
.
ELEMENT_NODE
)
{
if
(
i
+
1
===
len
||
childNodes
[
i
+
1
].
nodeType
!==
node
.
TEXT_NODE
)
{
consumed
+=
1
;
if
(
consumed
===
n
)
{
// TODO: create a reasonable 'slice' string
return
{
node
:
node
,
offset
:
i
+
1
,
slice
:
'XXX'
};
}
}
}
}
return
{
consumed
:
consumed
};
};
/**
* IME-like CE test class
*
* @class
* @constructor
* @param {ve.ce.Surface} surface The UI Surface
*/
ve
.
ce
.
TestRunner
=
function
VeCeTestRunner
(
surface
)
{
this
.
view
=
surface
;
this
.
model
=
surface
.
getModel
();
this
.
doc
=
surface
.
getElementDocument
();
this
.
nativeSelection
=
surface
.
nativeSelection
;
// TODO: The code assumes that the document consists of exactly one paragraph
this
.
lastText
=
this
.
getParagraph
().
textContent
;
// Turn off SurfaceObserver setTimeouts
surface
.
surfaceObserver
.
pollInterval
=
null
;
// Take control of eventSequencer 'setTimeouts'
ve
.
test
.
utils
.
hijackEventSequencerTimeouts
(
this
.
view
.
eventSequencer
);
};
/* Methods */
/**
* Get the paragraph node in which testing occurs
*
* TODO: The code assumes that the document consists of exactly one paragraph
*
* @return {Node} The paragraph node
*/
ve
.
ce
.
TestRunner
.
prototype
.
getParagraph
=
function
()
{
const
p
=
this
.
view
.
$element
.
find
(
'.ve-ce-attachedRootNode > p'
)[
0
];
if
(
p
===
undefined
)
{
if
(
this
.
view
.
$element
.
find
(
'.ve-ce-attachedRootNode'
)[
0
]
===
undefined
)
{
throw
new
Error
(
'no CE div'
);
}
throw
new
Error
(
'CE div but no p'
);
}
return
p
;
};
/**
* Run any pending postponed calls
*
* Exceptions thrown will leave postponed calls in an inconsistent state
*/
ve
.
ce
.
TestRunner
.
prototype
.
endLoop
=
function
()
{
this
.
view
.
eventSequencer
.
endLoop
();
};
/**
* Send an event to the ce.Surface eventSequencer
*
* @param {string} eventName DOM event name (such as 'keydown')
* @param {Object} ev Fake event object with any necessary properties
*/
ve
.
ce
.
TestRunner
.
prototype
.
sendEvent
=
function
(
eventName
,
ev
)
{
// Ensure ev has an originalEvent property.
ev
.
originalEvent
=
ev
.
originalEvent
||
{};
this
.
view
.
eventSequencer
.
onEvent
(
eventName
,
ev
);
};
/**
* Change the text
*
* TODO: it should be possible to add markup
*
* @param {string} text The new text
*/
ve
.
ce
.
TestRunner
.
prototype
.
changeText
=
function
(
text
)
{
// TODO: This method doesn't handle arbitrary text changes in a paragraph
// with non-text nodes. It just works for the main cases that are important
// in the existing IME tests.
// Remove all descendant text nodes
// This may clobber the selection, so the test had better call changeSel next.
const
paragraph
=
this
.
getParagraph
();
$
(
paragraph
).
find
(
'*'
).
addBack
().
contents
().
each
(
function
()
{
if
(
this
.
nodeType
===
Node
.
TEXT_NODE
)
{
this
.
parentNode
.
removeChild
(
this
);
}
}
);
const
range
=
document
.
createRange
();
if
(
text
===
''
)
{
range
.
setStart
(
paragraph
,
0
);
}
else
{
// Insert the text at the start of the paragraph, and put the cursor after
// the insertion, to ensure consistency across test environments.
// See T176453
const
textNode
=
document
.
createTextNode
(
text
);
paragraph
.
insertBefore
(
textNode
,
paragraph
.
firstChild
);
range
.
setStart
(
textNode
,
text
.
length
);
}
this
.
nativeSelection
.
removeAllRanges
();
this
.
nativeSelection
.
addRange
(
range
);
this
.
lastText
=
text
;
};
/**
* @typedef {Object} PlainRange
* @property {Node} startNode The node at the start of the selection
* @property {number} startOffset The start offset within the node
* @property {Node} endNode The node at the endof the selection
* @property {number} endOffset The endoffset within the node
*/
/**
* Select text by offset in concatenated text nodes
*
* @param {ve.ce.TestOffset|number} start The start offset
* @param {ve.ce.TestOffset|number} end The end offset
* @return {PlainRange} Selected range
*/
ve
.
ce
.
TestRunner
.
prototype
.
changeSel
=
function
(
start
,
end
)
{
if
(
typeof
start
===
'number'
)
{
start
=
new
ve
.
ce
.
TestOffset
(
'forward'
,
start
);
}
if
(
typeof
end
===
'number'
)
{
end
=
new
ve
.
ce
.
TestOffset
(
'forward'
,
end
);
}
const
foundStart
=
start
.
resolve
(
this
.
getParagraph
()
);
const
foundEnd
=
start
.
resolve
(
this
.
getParagraph
()
);
if
(
!
foundStart
.
node
)
{
throw
new
Error
(
'Bad start offset: '
+
start
.
offset
);
}
if
(
!
foundEnd
.
node
)
{
throw
new
Error
(
'Bad end offset: '
,
end
.
offset
);
}
const
nativeRange
=
this
.
doc
.
createRange
();
nativeRange
.
setStart
(
foundStart
.
node
,
foundStart
.
offset
);
nativeRange
.
setEnd
(
foundEnd
.
node
,
foundEnd
.
offset
);
this
.
nativeSelection
.
removeAllRanges
();
this
.
getParagraph
().
focus
();
this
.
nativeSelection
.
addRange
(
nativeRange
,
false
);
this
.
lastSel
=
[
start
,
end
];
return
{
startNode
:
foundStart
.
node
,
endNode
:
foundEnd
.
node
,
startOffset
:
foundStart
.
offset
,
endOffset
:
foundEnd
.
offset
};
};
/**
* Call assert.equal to check the IME test has updated the DM correctly
*
* @param {Object} assert The QUnit assertion object
* @param {string} testName The name of the test scenario
* @param {number} sequence The sequence number in the test scenario
*/
ve
.
ce
.
TestRunner
.
prototype
.
testEqual
=
function
(
assert
,
testName
,
sequence
)
{
const
comment
=
testName
+
' seq='
+
sequence
+
': "'
+
this
.
lastText
+
'"'
;
assert
.
strictEqual
(
this
.
model
.
getDocument
().
data
.
getText
(
false
),
this
.
lastText
,
comment
);
};
/**
* Call assert.notStrictEqual to check the IME test has not updated the DM correctly
*
* @param {Object} assert The QUnit assertion object
* @param {string} testName The name of the test scenario
* @param {number} sequence The sequence number in the test scenario
*/
ve
.
ce
.
TestRunner
.
prototype
.
testNotEqual
=
function
(
assert
,
testName
,
sequence
)
{
const
comment
=
testName
+
' seq='
+
sequence
+
': "'
+
this
.
lastText
+
'"'
;
assert
.
notStrictEqual
(
this
.
model
.
getDocument
().
data
.
getText
(
false
),
this
.
lastText
,
comment
);
};
ve
.
ce
.
TestRunner
.
prototype
.
ok
=
function
(
assert
,
testName
,
sequence
)
{
const
comment
=
testName
+
' seq='
+
sequence
+
': "'
+
this
.
lastText
+
'"'
;
assert
.
true
(
true
,
comment
);
};
ve
.
ce
.
TestRunner
.
prototype
.
failDied
=
function
(
assert
,
testName
,
sequence
,
ex
)
{
const
comment
=
testName
+
' seq='
+
sequence
+
': "'
+
this
.
lastText
+
'" ex='
+
ex
+
' stack=<'
+
ex
.
stack
+
'>'
;
assert
.
true
(
false
,
comment
);
};
File Metadata
Details
Attached
Mime Type
text/plain
Expires
Sat, Nov 21, 21:51 (1 d, 13 h)
Storage Engine
local-disk
Storage Format
Raw Data
Storage Handle
6f/f3/4c94211c5da648d6ab1d42490c7a
Default Alt Text
ve.ce.TestRunner.js (9 KB)
Attached To
Mode
rMWPROD MediaWiki Production
Attached
Detach File
Event Timeline
Log In to Comment