Have created builders for insertion, removal, and single element replacement.
In adding Document.getMetadata which is nearly identical to Document.getData,
the two functions have been refactored to use a common static method
getDataSlice, with this.data/this.metadata as an argument.
Document.spliceMetadata has been added. It is essentially spliceData with
the data/metadata synchronisation issue.
Metadata cursor position is now tracked in the TransactionProcessor. Cursor
advancement has been moved to a function so the metadata cursor can be reset
every time the data cursor is moved.
There were unhit bugs in the TransactionProcessor run test section, where
the data being loaded into the test documents wasn't always being deep-copied,
and the assert was looking at getData instead of getFulldata (which wouldn't
be able to test metadata changes).
Change-Id: Ieb20ab3e7827bc7ff04148f147da6728eb1eb666
* New! Button and InputLabel widgets
* Using new buttons in the demo
* Moved styles around to generalize input styles
Change-Id: Ic42e133f8fe0fffcb61374c09dd5668db82a4799
TODO: Use these buttons other places! (like ve.init)
This means that <p data-foo="bar"> will now be converted to a paragraph
with attributes {"html/0/data-foo":"bar"} rather than {"html/foo":"bar"}
This paves the way for multi-element node (about group) handling in the
node API: nodes representing multiple DOM elements will have html/i/attr
to represent an attribute of the i'th DOM element.
Change-Id: Iea52bdccd721942ca708c8f9f47e934524809845
When encountering an inline node (i.e. content node that's not a text
node) within a branch node that's not a content branch node, the
converter should start a wrapper. But it doesn't do this, it only opens
wrappers for text nodes and annotations.
Fixed this in the converter, added a test for it, and fixed an existing
test that asserted the broken behavior.
Change-Id: I6e143e21e68b68f0d85b8772e24a2d3a5d465410
* Introduce context object as specified for
ve.dm.Node.static.toDataElement()
* Remove wrapping variable in favor of context.wrapping
* Remove wrappingIsOurs in favor of context.canCloseWrapper
* Introduce originallyExpectingContent and use it to repopulate
context.expectingContent after closing a wrapper
* Replace most uses of branchHasContent with context.expectingContent
** Except for two cases where we need originallyExpectingContent
These changes fix a case where a metaBlock was generated in an inline
position. Updated the tests to reflect this.
Change-Id: I6baf6053f8a3a0b7d91487f812b9235a7b2b3db1
Continues where Ibb682332a6084e357104183641a104e3ae1e253f left off, adding tests and removing inconsistencies between the behavior of the document constructor, which was adding empty text nodes to empty paragraphs, and correcting other tests which expected empty text nodes to be there as well.
Change-Id: I414d061cdd494b8023f14e944eda2910a4dab0d4
Objective: Simplify the registration and use of triggers
Changes:
* Renamed ve.Command to ve.Trigger
* Renamed command demo to trigger demo
* Removed language prefixing of triggers
* Generating trigger tooltips rather than hard-coding them in i18n
* Added documentation to clarify that only 'mac' and 'pc' are supported platforms, and how the default is chosen
* Simplified trigger registry's register command
* Updated trigger registrations
Change-Id: Ibab6ad5b5c86f24707f064967dc2119a81125392
Extension-specific types are RDFa types (or type regexes) that are
registered with the ModelRegistry separately. If an element has a type
that is extension-specific, then that element can only be matched by a
rule that asserts one of its extension-specific types.
For MediaWiki, we would call
ve.dm.modelRegistry.registerExtensionSpecificType(/^mw:/ ) .
So then an element like <span typeof="mw:foobar"> would either match a
rule specifically for mw:foobar, if one exists, or no rule at all; even
the rule for <span> would not match. The consequence of this is that
elements with unrecognized mw:-prefixed RDFa types are alienated.
Change-Id: Ia8ab1fe5dffb9f813689324372a168e8e4a3e0bc
ModelRegistry registers both annotations and nodes, and performs
matching on both at the same time. It also registers annotations with
the AnnotationFactory, and nodes with the NodeFactory.
Change-Id: I5e68e506a0e573cc0afe6304ccea058ffc20d1c8
Add static properties for matching, data<->DOM conversion, and name. Use
matchTagNames, toDataElement and toDOMElement. name isn't used yet.
Change-Id: I5e7df3303bbd65e6968e931b568c23d76003a9a4
* Made method descriptions imperative: "Do this" rather than "Does this"
* Changed use of "this object" to "the object" in method documentation
* Added missing documentation
* Fixed incorrect documentation
* Fixed incorrect debug method names (as in those VeDmClassName tags we add to functions so they make sense when dumped into in the console)
* Normalized use of package names throughout
* Normalized class descriptions
* Removed incorrect @abstract tags
* Added missing @method tags
* Lots of other minor cleanup
Change-Id: I4ea66a2dd107613e2ea3a5f56ff54d675d72957e
Objectives:
* Make the link inspector easier to use
* Try to resolve a few bugs (bug 43841, bug 43063, bug 42986)
* Stop using jquery.multiSuggest (which didn't really understand annotations)
* Better divide MediaWiki specifics from generic implementations
Changes:
VisualEditor.php, modules/ve/test/index.php, demos/ve/index.php
* Updated links to files
ve.Registry
* Fixed mistake where registry was initialized as an array - this didn't cause any errors because you can add arbitrary properties to an array and use it like any other object
ve.Factory
* Removed duplicate initialization of registry property
* Added entries property, which is an array that's appended to for tracking the order of registrations
ve.CommandRegistry
* Added mwLink command which opens the mwLink inspector
ve.ui.TextInputWidget
* Added basic widget class for text inputs
ve.ui.TextInputMenuWidget
* Added widget that provides a menu of options for a text input widget
ve.ui.MWLinkTargetInputWidget
* Added MediaWiki specific link target widget
ve.ui.MenuWidget
* Converted ve.ui.Menu into a widget
* Moved the body of onSelect to onMouseUp
ve.ui.LinkTargetInputWidget
* Added link target widget which adds link annotation functionality to a normal text input
ve.ui.InputWidget
* Added generic input widget which emits reliable and instant change events and synchronizes a value property with the DOM value
ve.ui.Widget
* Added base widget class
* Widgets can be used in any frame
ve.ui.Tool
* Fixed line length issues
ve.ui.InspectorFactory
* Made use of new entries property for factories to select the most recently added inspector if more than one match a given annotation
ve.ui.Inspector
* Added auto-focus on the first visible input element on open
* Moved afterClose event to after re-focus on document on close
* Added documentation
ve.ui.Frame
* Adjusted documentation
* Added binding of $$ to the frame context so it can be passed around
* Added documentation
ve.ui.Context
* Added ve.ui.Widget.css to iframes
* Updated code as per moving of ve.ui.Menu to ve.ui.MenuWidget
* Removed unused positionBelowOverlay method
* Added CSS settings to set overlay left and width properties according to context size
* Added documentation
ve.ui.DropdownTool
* Updated code as per moving of ve.ui.Menu to ve.ui.MenuWidget
ve.ui.FormatDropdownTool
* Added documentation
ve.ui.MWLinkButtonTool
* Added MediaWiki specific version of ve.ui.LinkButtonTool, which opens the mwLink inspector
ve.ui.Widget.css
* Added styles for all widgets
ve.ui.Tool.css, ve.init.sa.css, ve.init.mw.ViewPageTarget.css, ve.init.mw.ViewPageTarget-apex.css
* Updated code as per moving of ve.ui.Menu to ve.ui.MenuWidget
ve.ui.Menu.css
* Deleted (merged into ve.ui.Widget.css)
ve.ui.Menu.css
* Deleted suggest styles (no longer used)
pending.gif, pending.psd
* Added diagonal stripe animation to indicate a pending request to the API
ve.ui.MWLinkInspector
* Added MediaWiki specific inspector which uses MediaWiki specific annotations and widgets
ve.ui.LinkInspector
* Removed mw global hint (not needed anymore)
* Switched from comparing targets to annotations (since the target text is ambiguous in some situations)
* Switched to using input widget, which is configured using a static property
* Removed use of jquery.multiSuggest
* Moved MediaWiki specifics to their own class (ve.ui.MWLinkInspector)
ve.init.mw.ViewPageTarget
* Added MediaWiki specific toolbar and command options
Change-Id: I859b5871a9d2f17d970c002067c8ff24f3513e9f
Some of these stubs didn't inherit Node at all. Made them all inherit
LeafNode because their rules specify they can't have children.
Change-Id: If4afc8de350f67ee78a41307c426ec2aceeb884f
CE doesn't actually render meta nodes anymore now that we split them out
into the meta-linmod.
I took a stab at consolidating metaBlock and metaInline (into simply
'meta'), but we can't do that with the current node API unless we put a
lot of meta-specific hacks in the converter. So I'm leaving this for the
node API rewrite.
Change-Id: Ie83413df718eabcaeb504316a2db0d24a1be2226
See CODING.md for how to run it.
Mistakes fixed:
* Warning: Unknown type function
-> Function
* Warning: Unknown type DOMElement
-> HTMLElement
* Warning: Unknown type DOM Node
-> HTMLElement
* Warning: Unknown type Integer
-> Mixed
* Warning: Unknown type Command
-> ve.Command
* Warning: Unknown type any
-> number
* Warning: Unknown type ve.Transaction
-> ve.dm.Transaction
* Warning: Unknown type ve.dm.AnnotationSet
-> ve.AnnotationSet
* Warning: Unknown type false
-> boolean
* Warning: Unknown type ve.dm.AlienNode
ve.dm doesn't have a generic AlienNode like ve.ce
-> Unknown type ve.dm.AlienInlineNode|ve.dm.AlienBlockNode
* Warning: Unknown type ve.ve.Surface
-> ve.ce.Surface
* ve.example.lookupNode:
-> Last @param should be @return
* ve.dm.Transaction.prototype.pushReplace:
-> @param {Array] should be @param {Array}
* Warning: ve.BranchNode.js:27: {@link ve.Node#hasChildren} links to non-existing member
-> (removed)
* Warning: ve.LeafNode.js:21: {@link ve.Node#hasChildren} links to non-existing member
-> (removed)
Differences fixed:
* Variadic arguments are like @param {Type...} [name]
instead of @param {Type} [name...]
* Convert all file headers from /** to /*! because JSDuck tries
to parse all /** blocks and fails to parse with all sorts of
errors for "Global property", "Unnamed property", and
"Duplicate property".
Find: \/\*\*([^@]+)(@copyright)
Replace: /*!$1$2
* Indented blocks are considered code examples.
A few methods had documentation with numbered lists that were
indented, which have now been updated to not be intended.
* The free-form text descriptions are parsed with Markdown,
which requires lists to be separated from paragraphs by an
empty line.
And we should use `backticks` instead of {braces} for inline
code in text paragraphs.
* Doc blocks for classes and their constructor have to be
in the correct order (@constructor, @param, @return must be
before @class, @abstract, @extends etc.)
* `@extends Class` must not have Class {wrapped}
* @throws must start with a {Type}
* @example means something else. It is used for an inline demo
iframe, not code block. For that simply indent with spaces.
* @member means something else.
Non-function properties are marked with @property, not @member.
* To create a link to a class or member, in most cases the name
is enough to create a link. E.g. Foo, Foo.bar, Foo.bar#quux,
where a hash stands for "instance member", so Foo.bar#quux,
links to Foo.bar.prototype.quux (the is not supported, as
"prototype" is considered an implementation detail, it only
indexes class name and method name).
If the magic linker doesn't work for some case, the
verbose syntax is {@link #target label}.
* @property can't have sub-properties (nested @param and @return
values are supported, only @static @property can't be nested).
We only have one case of this, which can be worked around by
moving those in a new virtual class. The code is unaltered
(only moved down so that it isn't with the scope of the main
@class block). ve.dm.TransactionProcessor.processors.
New:
* @mixins: Classes mixed into the current class.
* @event: Events that can be emitted by a class. These are also
inherited by subclasses. (+ @param, @return and @preventable).
So ve.Node#event-attach is inherited to ve.dm.BreakNode,
just like @method is.
* @singleton: Plain objects such as ve, ve.dm, ve.ce were missing
documentation causing a tree error. Documented those as a
JSDuck singleton, which they but just weren't documented yet.
NB: Members of @singleton don't need @static (if present,
triggers a compiler warning).
* @chainable: Shorthand for "@return this". We were using
"@return {classname}" which is ambiguous (returns the same
instance or another instance?), @chainable is specifically
for "@return this". Creates proper labels in the generated
HTML pages.
Removed:
* @mixin: (not to be confused with @mixins). Not supported by
JSDuck. Every class is standalone anyway. Where needed marked
them @class + @abstract instead.
Change-Id: I6a7c9e8ee8f995731bc205d666167874eb2ebe23
This happens when the <span> is the start of unwrapped content. The
converter logic to look at the tag name in wrapping mode doesn't kick in
because we're not yet in wrapping mode at that point.
The core issue was that previously, we relied on the document
structure/state to choose between alienBlock and alienInline, and only
used the tag name where the document structure was ambiguous (wrapping).
Changed this to be the other way around: we now rely primarily on the
tag name, and if that doesn't match what we expect based on the document
structure, we work around that if possible. Specifically:
* inline tag in our wrapper --> inline alien
* block tag in our wrapper --> close wrapper, block alien
* inline tag in wrapper that's not ours --> inline alien
* block tag in wrapper that's not ours --> *inline* alien
* inline tag in structural location --> open wrapper, inline alien
* block tag in structural location --> block alien
* inline tag in content location --> inline alien
* block tag in content location --> *inline* alien
only in the fourth and the last case do we need to use the "wrong" alien type to
preserve document validity, and it will always be inline where block was
expected, which should reduce UI issues.
The condensed version of the above, which is used in the code, is:
* If in a non-wrapper content location, use inline
* If in a wrapper that's not ours, use inline
* Otherwise, decide based on tag name
* Open or close wrapper if needed
ve.dm.Converter:
* Replace isInline logic in createAlien() with the above
* Factor out code to start wrapping (was duplicated) into startWrapping()
* Call startWrapping() if createAlien() returns an alienInline and we're
in a structural location
Tests:
* Add test cases with aliens at the start and end of unwrapped content
** The first one failed prior to these changes and now passes, the
second one was already passing
* Fix about group test case, was exhibiting the bug that this commit fixes
Change-Id: I657aa0ff5bc2b57cd48ef8a99c8ca930936c03b8
Although $.toJSON optimises heavily for modern browsers (it
becomes a direct reference to JSON.stringify), we still load the
extra plugin.
JSON is specified as part of ECMAScript 5, but most browsers
supported this one before they supported the rest of ES5.
http://caniuse.com/#search=JSON
Cut off for native JSON is IE7, Firefox 3.0 (3.6 supports it) and
Safari 3. Not any of our concern as VE will most likely never
support those (certainly not at this point in time, and less
likely as time goes on).
Change-Id: I4e8f26ac94763fa38d29e41264de0247f53a21e5
I noticed this bug on [[List of Presidents of the United States]]. When
there's HTML that looks like "<td>Foo\n<meta/></td>", the converter will
collect the newline in wrappedWhitespace, then attempt to splice it out
and store it in internal data. But instead, it ends up splicing out the
/metaBlock element, which causes strange unbalanced input, which causes
an empty table in the node tree.
Change-Id: I79ed2fa9a834cc42759c7d21250d8842f563d38f
ve.ce.Surface
* Switched to using getSlice instead of getData in copy and paste handlers
* Added try/catch which attempts to build a transaction with the unbalanced data first, but falls back on the balanced data otherwise
ve.dm.*Node
* Added default style attributes (now used by ve.dm.NodeFactory)
ve.dm.Document
* Fixed bugs in fixupInsertions where parentType was being set with an object rather than a string
* Made use of getDataElement
* Added adoption capability so that inserting a</h1><p>b into <p>c[cursor]d</p> results in <p>ca</p><p>bd</p> rather than throwing an exception
* Renamed getBalancedData to getSlice, now retuning a ve.dm.DocumentSlice object
ve.dm.DocumentSlice
* Introduced new container for balanced data and a range of the original context - useful for copy/paste
ve.dm.NodeFactory
* Added getDataElement method, which uses default attributes to create a boilerplate version of a data element
ve.dm.Document.test
* Updated getBalancedData test to be a getSlice test
demos/ve/index, VisualEditor, test/index
* Added references to ve.dm.DocumentSlice
Change-Id: Id9269a29e51ca213508de8f155d3feec5e5b0774
The converter was misbehaving when handling <p>s inside <span>s. This
can't be expressed in the linmod, but it would try to anyway. <span><p>
would result in too many paragraph closing elements, leading to an
exception in ve.dm.Document complaining about unbalanced input.
<span>\n<p> would result in an exception in the converter itself while
trying to perform whitespace preservation on the newline.
This change makes the converter detect these scenarios and alienate the
offending node. So <span><p>Foo</p></span> converts to a wrapper
paragraph containing an alienInline whose HTML is "<p>Foo</p>" and which
is annotated with a TextStyleSpanAnnotation.
ve.dm.Converter.getDomFromData():
* Change the criteria for alienBlock vs alienInline
** Only infer from the node type if we're in wrapping mode AND we're at
the same level where the wrapping started (wrappingIsOurs). If the
latter isn't the case, we can't split the wrapper in the block case
because we're at the wrong level.
** Use alienInline not only if the branch is a content branch, but also
if there are active annotations. This catches e.g. <li><b><p>
(and generally <span><p> on the top level).
* Before converting a child element, check that the child isn't "bad".
Bad children are non-content children in content branches, and
non-content children encountered within a wrapper that we can't split.
Only good children are converted, and bad children are alienated (cue
Santa/Sinterklaas jokes).
* Add childIsContent and rename branchIsContent to branchHasContent
Change-Id: If420ae80ab0777424a9a5517335ef9d0170e87ae
* Fix 404 error for ve.ui.Icons-{raster,vector}.css
Follows-up 9563f08 / I840f72426f9a.
makeStaticLoader.php:
* Clean up old code.
* Error out early for missing module.
* Put i18n stuff in the right place
(some modules access ve.msg from the global scope to
assign status variables, for standalone on demos this was
failing due to wrong load order)
Change-Id: Idbff4c5136d567da747d9ae373cd2f6c3ee7fb1c
Rewrite VisualEditorMessagesModule:
* Replace copy-paste dump of user-css module with stuff for
VisualEditor (class commend and module::$origin).
* Remove duplication between getMessages and getScript.
* Actually implement getModifiedTime so that the comment in
getMessages() about cache invalidation is actually true
Fixes bug 42670: ext.visualEditor.specialMessages cache broken
ve.init:
* Implement addParsedMessages and getParsedMessage so that we
don't mix up plain messages with raw html messages (minoredit
was previously overloaded in mw.msg storage with a parsed html
message and retrieved though ve.msg, which is documented as
retuning plain text, not raw html). This is now separated into
a different method.
* Improved documentation of the other msg methods to emphasise
their differences
* Removed redundant code in attachSaveDialog() that was
(partially) already done in setupSaveDialog() and moved the
remaining bits into it as well. Checked all callers of these and
they are both only called from ViewPageTarget.prototype.onLoad
* Also implement them in the standalone platform implementation,
with the html escaper based on mw.html.escape
* Update init.platform.getMessage to use undefined instead of
discouraged 'if-in' statement.
* Add test suite.
demos/test:
* Re-run makeStaticLoader.php on test to add ve.init.Platform.test
* Re-run makeStaticLoader.php on demos and update i18n caller
to use ve.init.platform.addParsedMessages (also moved out of the
auto-generated block for easier updating)
Change-Id: I7f26b47e9467e850c08b9c217c4f1098590de109
HTML DOM has annoying behavior for <pre>s where .innerHTML eats the
first newline in a <pre>. Work around this by explicitly adding a
newline in the data->DOM converter if the <pre> already contained a
newline.
There is a separate bug in Parsoid that causes the newline to be lost
anyway, filed as bug 42666
Change-Id: Ia26cd4a4c61afbe439b0562deb7f24ee8b8147d7
When the content rendering stuff was moved to ve.ce.ContentBranchNode the onUpdate methods being used to update the DOM wrapper in ve.ce.HeadingNode was overlooked, so heading were not rendered on update anymore.
Change-Id: I994b8c43123c3cd02b9a550d5d7eac7d5052418e
ve.ce.Surface
* Added ve.ce.Surface.adjustCursor, which replaces repetitive and buggy code that was handling left and right arrow key presses
* New method only affects the selection target, so it won't collapse the selection on you - this was what caused bug 42401
* Made hasSlugAtOffset() actually return a boolean
ve.dm.Document
* Fixed turn-around issue in ve.dm.Document.getRelativeOffset - if the offset is already valid and we can't move in the direction we want, we should just leave it be, not turn around
* Since this method was being used by ve.ce.Surface to correct the cursor position on arrow key presses, it was causing the strange cursor jumping when you pressed an arrow key while at the edge of a document
ve.dm.SurfaceFragment
* Fixed typo where getAnnotationRangeFromSelection was preserving selection direction, but checking the wrong properties
ve.dm.Document.test
* Added tests that verify turn-around issue is fixed
Change-Id: Iba55cfc3d531e7d1333b78c94912ff22179aace8
Moved annotation rendering from ce.Textnode into the new
ce.ContentBranchNode class. This allows us to render annotations that
span across multiple nodes.
* Add ce.ContentBranchNode, inheriting ce.BranchNode
* Make ce.{Paragraph,Heading,Preformatted}Node inherit ce.ContentBranchNode
* Made ce.ContentBranchNode render its child nodes with anntations,
using .getAnnotatedHtml() on the child nodes
* Put a default implementation for .getAnnotatedHtml() in ce.LeafNode
* Override this in ce.TextNode to do escaping and whitespace handling
* Removed rendering code from ce.TextNode (this.$ is now unused there)
* Removed ce.TextNode.onUpdate() and ce.BranchNode.clean(), now unneeded
* Have ce.BranchNode propagate update events from children, so
ce.ContentBranchNode can rerender when its children change
* Update tests, add test case for escaping of &<>'"
Change-Id: I4600e984b287c6ff9267f4281d2f09bab9e1ad95
Was broken both on the way in and on the way out.
* Move alien restoration (data->DOM) out of the main getDomFromData()
function and into getDomElementFromDataElement(). This means the
comment about District 9 is gone (sniff), but moving this here ensures
all code paths hit it (previously, it was assumed annotated nodes
could never be aliens).
* In the DOM->data converter, add annotation application to
getDataElementFromDomElement() (for content nodes) and createAlien()
(for aliens). Previously, these nodes would not get annotations.
** ve.AnnotationSet doesn't have a constructor that takes an array, we
should fix that.
Change-Id: I65f8e9a322111ca3af275bf9997b0b1e7ee93769
The transaction builder would step around inline content elements when
building annotation transactions. This is now fixed.
I also tweaked the processor to tolerate attempts to annotate inline
closings. This allows the builder to generate simpler transactions,
because it doesn't have to step around the closing.
Change-Id: I1e0d7f95b38bad1b35b3e125a53350d2d126a7de
This is cleaner than passing around the attributes separately, and it
allows us to access the annotations in dm.LeafNode as well.
Change-Id: Ie5b90988114835831cbe5cdccf63c7cd45719e31
* Added whitelist argument to setDomAttributes which allows filtering of attributes being set
* Added prefix argument to ve.dm.Node.getAttributes to allow extracting a subset of attributes by name prefix
* Added a whitelist to ve.ce.Node which was extracted from MediaWiki's Sanitizer class
* Replaced attribute copying code with a call to setDomAttributes using the whitelist argument, passing in attributes from a call to ve.dm.Node.getAttributes using the prefix argument
Also…
* Removed comment in constructor of ve.ce.Node, documentation for properties is usually in the getters/setters, and already was in this case
* Renamed ve.setDOMAttributes to ve.setDomAttributes
* Renamed ve.getDOMAttributes to ve.getDomAttributes
* Renamed ve.getDOMText to ve.getDomText
* Renamed ve.getDOMHash to ve.getDomHash
* Updated all callers of renamed methods
Change-Id: Id556172d5d18ea431044b9d402400e1f0e67a293
When editing a new page, or loading an empty page into the editor, the
converter generates a paragraph so the document isn't completely empty.
This paragraph is then unwrapped on the way out, potentially destroying
change markers and generally producing strange HTML output.
Mark this paragraph with generated=empty rather than generated=wrapper,
and only unwrap it on the way out if it's still empty. This means we
cleanly round-trip empty documents (and empty list items and the like),
but if the user enters text, we create a paragraph like we're supposed
to.
Change-Id: Id0241221a67b769445676b833b5741320d99ea5f
When alienating in wrapping mode, we need to look at the type of tag to
decide whether to create a wrapped alienInline, or to interrupt the
paragraph for an alienBlock.
This was being done just fine for the general alienation case
(unrecognized tag), but not for the special cases (mw:unrecognized,
about groups).
* Centralize the logic for ending a wrapper in stopWrapping()
* Move the wrapping-contingent block/inline detection logic into
createAlien()
* Simplify the terrible if statement to decide whether a future decision
requires us to stop wrapping. Instead, detect the cases in each code
path separately and call stopWrapping() as appropriate
* Add tests
Change-Id: I4054584ae05e7d5daa71edead3e6a6588cf5d3bb
<span typeof="mw:Entity"> tags are now correctly represented in the
model, and rendered in CE. There are still issues with cursor movement
etc. in CE.
Because the prioritization mechanism for annotations vs nodes is broken
in the current "node API", I had to hack two special cases for mw:Entity
into the converter. I also had to change the converter to ignore the
children of inline nodes (this was a legitimate bug, but had never come
up before).
Change-Id: Ib9f70437c58b4ca06aa09f7272bf51d9c41b18f2
* Make converter generate meta nodes with 'style': 'comment'
* Handle style==='comment' in MetaBlockNode toDOM converter
* Add some comments to the meta test case
** Update other tests accordingly
* Change getDomElementSummary() to actually assert presence of comment
nodes (specifically, all non-text child nodes)
Change-Id: Ieef9418f4c47df3541477d9420aa2ab8df6e3df1
When the link inspector is used to create a new annotation, the text is annotated with the default link target derived from the selected text. Then if the inspector is used to change that value, yet another transaction is processed when the inspector is closed.
To avoid having to press undo 2x, this change makes the inspector undo it's first change before applying the changed annotation.
This change also introduces insert, remove and select content actions.
Change-Id: I3e29189158fb01336d6b053bc2a8bda2a91a0a46
AnnotationAction and SurfaceFragment now use insertAnnotations.
ve.dm.Surface.test
* Removed test for annotate method (not needed anymore)
ve.dm.SurfaceFragment
* Now using getInsertionAnnotations method
* Added support for modifying insertion annotations when annotating a zero-length selection
ve.dm.Surface
* Moved in insertion annotations state from document model
* Added insertion annotation interface (enable, disable, areEnabled, get, set, add, and remove)
* Simplified handling of annotations on change
* Removed annotate method (not used anymore)
ve.dm.Document
* Removed insertion annotations (moved it to surface model)
ve.ce.Surface
* Cleaned up handleInsertion and changed it to use the insertion annotations interface on the surface model
ve.AnnotationAction
* Moved insertion annotation handling out of here since it's now included in the surface fragment
Change-Id: I047d656acf7fa1c63f726ca2b0801e1476f84f96
TransactionProcessor was using parentOuterRange without checking whether
it was present, so it was exploding for indexInNode results.
Now checking for parentOuterRange presence, and falling back to
nodeOuterRange when missing.
This fix causes inconsistencies with zero-length text nodes. We should
fix these eventually, but for now I've just made the unit tests
tolerant of zero-length-text-node deviations.
Change-Id: Id9eadd57a0d5fcbaf009c0781da0a03928aebb31
Editing the text of a list item results in a change marker on the
paragraph within that list item. However, that paragraph usually isn't
present in the HTML, so the converter unwraps it when converting back to
HTML, and the change markers are lost. Instead, transfer the change
markers to the <li>.
Change-Id: Id675075d19c08d69bc8e990174841dc393b749fc
About groups are HTML structures like the following:
<div about="#mwt1">....</div>
<span about="#mwt1">...</span>
<div about="#mwt1">...</div>
When about groups are alienated, they are now merged into one alien
node, rather than producing a separate alien node for each sibling.
This is very basic about group handling, because it only works for
groups of directly adjacent siblings (text nodes are permitted in
between, but nothing else) assumes all about groups are aliens (which
is currently true).
* Before processing an element in the DOM->data converter, perform about
grouping on its children. This temporarily wraps about groups in
<div data-ve-aboutgroup="value of about attribute">
* Extended createAlien() to handle single nodes as well as wrappers
holding multiple nodes.
* In the data->DOM converter, temporarily wrap multi-node aliens in
<div data-ve-multi-child-alien-wrapper="true"> . This makes the rest
of the algorithm easier.
Change-Id: I2df5f62bc222b570fc11a89fe43d353f8363ead8
* Now ve.Factory inherits from the more general ve.Registry
* New class ve.CommandRegistry
* Refactored setupToolbar and command setup code into setupComands
Change-Id: Ic548e5de95b77889727362d3e66d7be83c12a603
The port of mousetrap wasn't really what we needed. This is much simpler, matches the rest of our code, and does exactly what we need.
Change-Id: I67f413e097fc2d4078336edb14dd9440e771f196
Sequences that were scheduled directly after each other, such as "a b c" and "1 2 3" would end up overlapping sometimes, producing "a b 1 c 2 3" which failed to trigger the correct commands.
Change-Id: I27bb60e856e9d692a21e1587dc227f8aeb5fcf4e
This causes the converter not to strip inner whitespace in them, and
causes CE to suppress the whitespace mangling logic that is normally
applied (↵ for newlines, ➞ for tabs, alternating s for spaces).
Change-Id: I738a750c91a4ca4836c485e282865bb7525bf30a
* Add map of change markers per offset to Transaction
* Map is populated by TransactionProcessor
* Markers are reversed on rollback
* Removals aren't marked, Parsoid can detect these using DSR
discontinuities
Change-Id: I2290886ab411c6ad6162044ed85c091313613e51
* ve.dm.Converter still generates metaInline/metaBlock elements as
before, it's not affected by this change
* ve.dm.Document constructor splits its input into "real" data and
metadata
** Metadata is stored in this.metadata (the meta-linmod) as a sparse
array of arrays, with an element for each offset in this.data
** this.data itself does not contain the metadata
** This means the node tree also doesn't contain the metadata
** Which means CE doesn't know about it at all
* All splice operations on the linear model are sent through
ve.dm.Document.spliceData(), which performs the splice and syncs the
meta-linmod
** Metadata in the removed range is reaped and added to the metadata for
the offset immediately following the removal
* ve.dm.Document.getFullData() splices the linmod and meta-linmod back
into each other; this "full data" is then fed back to ve.dm.Converter
Change-Id: Ief6dfd5b59cc13a8457993ed85c725413029c4fb
* Actually return the spliced data like the docs claim we do
* Remove false claim that offset can be negative
* Add that data=[] && remove=0 is invalid; native splice() doesn't allow
this, and there is a case where we call native splice() directly
* Add tests
Change-Id: I90e77c1b22ea1c36cb61e89ea47831885a0b1cb9
Previously copyObject and copyArray would silently drop null values,
which is bad, especially considering we have example data for meta nodes
that has { 'key': null } somewhere.
Also added a test case that failed prior to this change.
Change-Id: I4f233cce041fbf38f701c494f1f78ac3d8535d88
Tests were completely broken because the link inspector threw a JS error
when trying to register itself with the nonexistent inspector factory.
Change-Id: I8a47222f0a5a37348262ed939b37fbc47d14e222
Attempting to descend into a string or number would cause a JS error,
because we would attempt to create prop[arguments[i]] as an empty object
(which is ignored), then try to descend into it (which blows up because
it's undefined, even though we've just set it). Guard against this by
explicitly checking for non-object-ness.
Change-Id: Ie65550baaae0ab88476c9a1ff40cc136090740a0
* Adjust the range in the annotation synchronizer, otherwise we emit
events for the wrong node
* Expanded test suite to the point where it was able to catch the bug
caused by not adjusting annotated ranges
* Removed selection.length === 0 check, no longer needed because
selectNodes() now throws an exception in this case
* Added a FIXME comment about duplicate update events that occur when
length adjustments are combined with something else
* Add a few more comments
Change-Id: I84f0368b1d7b601ed0766806607152dc97f34603
* Store the applied state in the Transaction
* Store the Transaction in the TransactionProcessor (previously, only
its operations were stored)
* Have commit() and rollback() throw exceptions when passed transactions
with the wrong applied state
* Add tests for this behavior
Change-Id: I27b7a96fdf4d3555d78f64c05a03702ea560c802
The data array is now taken by reference, and the caller must perform
any copying required.
Changed tests to make a deep copy of shared data sets (mostly
ve.dm.example) before passing them to ve.dm.Document().
Change-Id: Iedc64f9fd9cd689640de9a19379cf5f3db94a2bb
There's no use case for keeping a deep copy of the 'internal' property
in the node tree, and it was breaking some of my new tests concerning
change markers. We could keep internal data in the node tree if we
wanted to, but to be correct we'd have to synchronize every time we
changed it, which is a pain.
Change-Id: I024de1ff8b6b6154da82c103c4bb21db8ff2ec14
Based on https://github.com/ccampbell/mousetrap.
Cleaned up to fit our coding standards, pass JSHint, and assume
jQuery's fixes where possible (e.g. no need for an addEvent
utility, no need for filling e.target, e.which, etc. cross-browser
which jQuery.Event already does).
Initially all were local functions in the constructor, but to
allow some customisations in subclasses moved various methods
to the prototype instead and marked with @private.
Really, only .register() must be called from the outside. The rest
assumes normalisation etc. or might break things if called
incorrectly.
Change-Id: Ic69a3c70759052094aefbeab623d885f8b118e14
This is the outer range of the parent of the node, if known. We'll need
this for change marking: when resizing a text node, for instance, we
need to mark its containing parent. This way we get the containing
parent's element's offset for free (selectNodes already tracks it in
currentFrame) rather than having to compute it with another traversal.
Change-Id: Ia335d8080ea9d414ab9f89b943e2ea0cd11d7df3
Some tests were using the wrong node in the expected data, but because
only the summaries were compared, this would succeeed as long as the
type and length were equal (and paragraphs of length 1 are quite common
in our test data). Fixed equalNodeSelection() to compare each node by
reference as well as comparing the summaries. If one of the equality
tests fails, the summaries will still be displayed as expected/actual
data (even though they might be equal), and the message will have
"(reference equality for selection[3].node)" appended to it.
This change broke the tests because a few test cases had bad data, fixed
those in this commit as well.
Change-Id: Iab420cf29d47f7368c8a9ce79f6309efae75685c
For <p>1<br/>2</p>, selectNodes([2,2]) correctly returned the end of the
first text node, but selectNodes([4,4]) returned index 2 in the
paragraph (i.e. between the break node and the second text node). The
correct behavior is to return the start of the second text node, i.e.
the mirror image of the behavior for [2,2].
Fixed this by applying the startBetween/endBetween logic only if the
relevant adjacent node is wrapped (or if it's missing). In the code,
this is expressed as !(adjacent node present && adjacent node wrapped).
Change-Id: Ie3b7fdf1de38ee253a798a7a73bc89734f4ca4fa
The HTML "1<br/>2" was being converted to a linmod that looked like
"<p>1</p><br></br><p>2</p>". This commit fixes the wrapping logic such
that the result is "<p>1<br></br>2</p>" instead. In general, inline
nodes (content nodes) should not interrupt the wrapping, but block nodes
should.
This creates a problem for alien nodes: normally, we determine whether an
alien node is a block alien or an inline alien based on context, but if
we're in wrapping mode we're unsure of the context. We can't tell the
difference between "1<tt>Foo</tt>2" (should be wrapped as one, because
tt is inline) and "1<figure></figure>2" (1 and 2 should be wrapped
separately, because figure is block) using context alone, so in these
cases (and ONLY in these cases) we look up whether the HTML tag in
question is an inline tag or a block tag and use that to decide.
Change-Id: I75e7f3da387dd401d9b93e09a21751951eccbb83
* CenterNode missing in ResourceLoader registry
* UI classes and rangy not in static test/index.html
* Transaction and TransactionProcessor listed twice
Added a maintenance script that generates the <script> and <link> tags for all
files in the same order everywhere.
Change-Id: I5d22d33769b4e356e8065d295505f6f9a8b0bea8
Also changed from using "type" to "name" to make it less specific and added a test to make sure it's working.
Change-Id: I150a7ab1a57b3df85b459dbc411c2eaefe08b5bb
This changes ve.dm.LinkAnnotation to be a generic annotation for <a>
tags, and adds ve.dm.MWInternalLinkAnnotation and
ve.dm.MWExternalLinkAnnotation as MW-specific subclasses. This nicely
splits out the MW-specific parts in LinkAnnotation, and ideally we'd
also move these files somewhere else to reflect their MW-specificity,
but I haven't gotten to that yet.
Similarly, ve.dm.TextStyleAnnotation is now a generic base class for
simple tag-only-no-metadata annotations, and it has 11 subclasses, one
for each tag we support. This is quite a bit more verbose than the
previous code, but I think it's cleaner and more flexible. I considered
writing a function that would generate a TextStyleAnnotation subclass,
then calling that 11 times, but that's not possible if we want to keep
named functions for the constructors.
Change-Id: Ifba10153eef40280e44025dd72d4e9d9f33b0632
Fleshes out ve.dm.Annotation to a class. Annotations in the linear model
will be instances of a subclass of ve.dm.Annotation. Annotations are
defined by subclassing ve.dm.Annotation and registering this subclass
with ve.dm.AnnotationFactory.
ve.dm.AnnotationFactory keeps track of which annotation classes are known,
and has code to match an HTML element to an annotation class, for use in
the converter.
Change-Id: I68802bdb8736ced1f9e04ee49c623944b448141c
Utility function to generate an opening HTML tag. Needed to integrate
the new annotation API with ve.ce.TextNode
Change-Id: I6804bbf6f79346fde1887fa82d29ec5cd0342d60
Previously, Undo used a transaction's lengthDifference to calculate the selection to display after the transaction was undone. Now, translateOffset with the reversed boolean set to true will properly translate the inverse of a transaction's selection change. Fixes bug #40538
Change-Id: I110bc0cbb5824547842efd391b9f2948b037b758
We were populating empty content nodes with zero-length text nodes to
make round-trip tests in the test suite work (otherwise blanking a
paragraph leaves behind a zero-length text node whereas creating an
empty paragraph does not), but the empty nodes are causing problems in
CE apparently.
* Do not create empty text nodes when constructing a node tree
* Be more careful with text-only replacements:
** Don't resize a text node to zero, remove it instead
** There may not be a text node to resize at all, build it in that case
** Switch nodeRange to nodeOuterRange, this was probably broken before
Tests:
* Change test case for zero-length text node to assert that there is
*no* zero-length text node :)
* Remove a test case concerning an empty text node from the
ve.ce.TextNode suite
Change-Id: Ie677457f2f0a7823a517ba3077b844ef52a20fcc
Previously tests for inheritClass (and other object management
utilities) were absent (as they were copied from upstream K-js).
I've copied the upstream test suite for this method here and
extended it with tests for this new feature.
Had to add es5:true to .jshintrc due to a bug in JSHint.
Repeated the setting in ve.inheritClass for future reference.
Source: https://github.com/Krinkle/K-js/blob/master/test/K.test.js
Change-Id: I63ac620d6ce7832ebfee454ddf7b7c90f6eb6121
This works just fine, as also previously tested/proven by
ve.cloneObject, which uses the same concept of creating an object
identical to what invoking the constructor with "new" would do,
but without invoking the constructor function (which has side-
effects).
Except in this case we do invoke the constructor function, but
we can't use new in ve.Factory because of the arbitrary number of
arguments.
Added a test to assert that 3+ arguments and that instanceof
work as expected.
Change-Id: If0add3da7475886e476900044acda2ba7d01fb11
Add some missing constructor names and rename the ones with a
lowercase 'v'.
I previously changed Object.create and others to using hasOwn,
but that turned out to be useless. The thought at the time was
to only use the native one if it really is a native one (and not
a polyfill from another script), however in then hasOwn is only
relevant on prototypes and when negated. For static members it
would be an own-property either way.
Follows-up:
* Id6783fcfc35a896db088ff424ff9faaabcaff716 (metanode)
* Iab763954fb8cf375900d7a9a92dec1c755d5407e (object-management)
Change-Id: Ia6ef597e5e5453277472dfc23f25d2878b68b7f6
Rather than each tool requesting annotations, and nodes pertaining to selection,
Emitted event supplies annotations and nodes to each tool's update method.
Using select vs. of traverseLeafNodes for code optimization.
Better documentation for updateTools()
Removed unneeded code.
Change-Id: I7c0baa1cc0f7fb731d6e28b175a76e931e9e2961
* Added documentation for ve.AnnotationSet
* Replaced uses of "// Inheritance" with "// parent Constructor"
* Added "// Mixin constructor" where needed
* Added missing section comments like "/* Static Methods */"
* Cleaned up excessive newlines (matching /\n\n\n/g)
* Put unnecessarily multi-line statements on a single line
Change-Id: I2c9b47ba296f7dd3c9cc2985581fbcefd6d76325
* Commands for Sublime:
Find*: "(\* @[a-z]+) ([^{].*) \{(.*)\}"
Replace: "$1 {$3} $2"
Save all && Close all
Find: " function("
Replace: " function ("
Save all && Close all
Find: "Intialization"
Replace: "Initialization"
Save all && Close all
* Consistent use of types (documented in CODING.rm):
- Merged {Integer} into {Number}.
- Merged {DOM Node} into {DOMElement}.
* Remove work-around /*jshint newcap: false */ from ve.js
Calling Object() as a function to to use the internal
toObject no longer throws a newcap warning in JSHint.
It only does that normal functions now .
(e.g. var a = Cap(); or var a = new uncap();)
* Add missing annotations (@static, @method, ..).
* Remove unused variables
* Remove null-assignments to variables that should just be
undefined. There's a few variables explicitly set to null
whereas they are set a few lines under and not used otherwise
(e.g. 'tx' in ve.ce.Surface.prototype.onPaste)
Change-Id: I0721a08f8ecd93c25595aedaa1aadb0e08b83799
This node type represents <meta> or <link> (transparently, based on the
style attribute). I had to make two node types for this and hack the
toData conversion code directly into ve.dm.Converter, because we don't
have native support for node types that can be both inline and block.
(We should add this in the node API rewrite.)
The CE implementation renders a placeholder (with the same styles as an
alien node) right now. I'm not sure how nice that is, but it's better
than rendering raw <meta>/<link> tags.
This whole thing is a total pile of hacks to make VE deal with
<meta>/<link> tags until we have a proper node types API.
Change-Id: Id6783fcfc35a896db088ff424ff9faaabcaff716
This was broken in three different ways:
* On the way in, we were applying whitespace to an array of elements
rather than the actual element, so the whitespace wasn't stored.
* Whitespace processing on the way out was skipped for aliens because
they had their own code path. Refactored this so alien openings and
regular openings share much more code, including whitespace output.
* Somewhat unrelatedly, innerPost output was broken for paragraphs
containing inline elements, because the inline elements' processing
polluted lastOuterPost. Discovered this because my test with inline
aliens also happened to be the first test of whitespace preservation
in paragraphs with inline content elements. Fixed by explicitly
skipping content nodes when outputting whitespace.
Fixed these issues and added a test case.
Change-Id: I8edb61a008e60ace886b1a841b3417682ec39c32
* For the most common case:
- replace ve.extendClass with ve.inheritClass (chose slightly
different names to detect usage of the old/new one, and I
like 'inherit' better).
- move it up to below the constructor, see doc block for why.
* Cases where more than 2 arguments were passed to
ve.extendClass are handled differently depending on the case.
In case of a longer inheritance tree, the other arguments
could be omitted (like in "ve.ce.FooBar, ve.FooBar,
ve.Bar". ve.ce.FooBar only needs to inherit from ve.FooBar,
because ve.ce.FooBar inherits from ve.Bar).
In the case of where it previously had two mixins with
ve.extendClass(), either one becomes inheritClass and one
a mixin, both to mixinClass().
No visible changes should come from this commit as the
instances still all have the same visible properties in the
end. No more or less than before.
* Misc.:
- Be consistent in calling parent constructors in the
same order as the inheritance.
- Add missing @extends and @param documentation.
- Replace invalid {Integer} type hint with {Number}.
- Consistent doc comments order:
@class, @abstract, @constructor, @extends, @params.
- Fix indentation errors
A fairly common mistake was a superfluous space before the
identifier on the assignment line directly below the
documentation comment.
$ ack "^ [^*]" --js modules/ve
- Typo "Inhertiance" -> "Inheritance".
- Replacing the other confusing comment "Inheritance" (inside
the constructor) with "Parent constructor".
- Add missing @abstract for ve.ui.Tool.
- Corrected ve.FormatDropdownTool to ve.ui.FormatDropdownTool.js
- Add function names to all @constructor functions. Now that we
have inheritance it is important and useful to have these
functions not be anonymous.
Example of debug shot: http://cl.ly/image/1j3c160w3D45
Makes the difference between
< documentNode;
> ve_dm_DocumentNode
...
: ve_dm_BranchNode
...
: ve_dm_Node
...
: ve_dm_Node
...
: Object
...
without names (current situation):
< documentNode;
> Object
...
: Object
...
: Object
...
: Object
...
: Object
...
though before this commit, it really looks like this
(flattened since ve.extendClass really did a mixin):
< documentNode;
> Object
...
...
...
Pattern in Sublime (case-sensitive) to find nameless
constructor functions:
"^ve\..*\.([A-Z])([^\.]+) = function \("
Change-Id: Iab763954fb8cf375900d7a9a92dec1c755d5407e
Introduced the ve.AnnotationSet class to manage sets of annotations. This
is a generalization of ve.OrderedHashSet, a class that manages a set
using an array and an object keyed by hash.
Converted everything that stores, tracks or passes around annotations to
use ve.AnnotationSet. In particular, this means the linear model now
contains AnnotationSets instead of hash-keyed objects.
This allows us to maintain the order of annotations in the linear model,
and will help fix bugs with annotation ordering and splitting.
Change-Id: I50975b0a95f4cc33017a0b59fdede9ed1eff0124
[jshint]
ce/ve.ce.Surface.js: line 670, col 9, Too many var statements.
ce/ve.ce.Surface.js: line 695, col 6, Missing semicolon.
ce/ve.ce.Surface.js: line 726, col 22, Expected '===' and instead saw '=='.
ce/ve.ce.Surface.js: line 726, col 41, Expected '===' and instead saw '=='.
ce/ve.ce.Surface.js: line 733, col 13, Too many var statements.
ce/ve.ce.Surface.js: line 734, col 24, Expected '===' and instead saw '=='.
ce/ve.ce.Surface.js: line 1013, col 13, Too many var statements.
ce/ve.ce.Surface.js: line 1019, col 17, Too many var statements.
ce/ve.ce.Surface.js: line 1023, col 18, Too many ar statements.
ce/ve.ce.Surface.js: line 1027, col 13, Too many var statements.
dm/annotations/ve.dm.LinkAnnotation.js: line 70, col 52, Insecure '.'.
dm/ve.dm.Converter.js: line 383, col 29, Empty block.
dm/ve.dm.Converter.js: line 423, col 33, Empty block.
Commands:
* jshint .
* ack '(if|else|function|switch|for|while)\('
* Sublime Text 2:
Find(*): (if|else|function|switch|for|while)\(
Replace: $1 (
* ack ' ' -Q # double spaces, except in certain comments
Change-Id: I8e34bf2924bc8688fdf8acef08bbc4f6707e93be
This will cause ve.dm.SurfaceFragment.prototype.insertContent() to place
the selection after the insertion as well.
Change-Id: Ifa7e627daceb90408422eb58c110d475f34ba1e2
* Replaces c8b4a28936
* Use Object() casting to detect objects instead of .constructor
(or instanceof). Both .constructor and instanceof compare by reference
the type "Object" which means if the object comes from another window
(where there is a different "Object" and "Object.prototype") it will
drop out of the system and go freewack.
Theory: If a variable casted to an object returns true when strictly compared
to the original, the input must be an object.
Which is true. It doesn't change the inheritance, it doesn't make it inherit
from this window's Object if the object is from another window's object. All it
does is cast to an object if not an object already.
So e.g. "Object(5) !== 5" because 5 is a primitive value as opposed to an instance
of Number.
And contrary to "typeof", it doesn't return true for "null".
* .constructor also has the problem that it only works this way if the
input is a plain object. e.g. a simple construtor function that creates
an object also get in the wrong side of the if/else case since it is
an instance of Object, but not directly (rather indirectly via another
constructor).
* Added unit tests for basic getHash usage, as well as regression tests
against the above two mentioned problems (these tests fail before this commit).
* While at it, also improved other utilities a bit.
- Use hasOwnProperty instead of casting to boolean
when checking for presence of native support.
Thanks to Douglas Crockford for that tip.
- Fix documentation for ve.getHash: Parameter is not named "obj".
- Add Object-check to ve.getObjectKeys per ES5 Object.keys spec (to match native behavior)
- Add Object-check to ve.getObjectValues to match ve.getObjectKeys
- Improved performance of ve.getObjectKeys shim. Tried several potential optimizations
and compared with jsperf. Using a "static" reference to hasOwn improves performance
(by not having to look it up 4 scopes up and 3 property levels deep).
Also using [.length] instead of .push() shared off a few ms.
- Added unit tests for ve.getObjectValues
Change-Id: If24d09405321f201c67f7df75d332bb1171c8a36
This commit fully utilizes all four positions in the internal.whitespace
array. Outer whitespace is now preserved as well, and is duplicated
either in the adjacent sibling (one node's outerPost is the next
sibling's outerPre) or in the parent (a branch node's innerPre is its
first child's outerPre, and its innerPost is its last child's
outerPost). Before restoring saved whitespace, we check that these two
agree with each other, and if they disagree we assume the user has been
moving stuff around and don't restore any whitespace in that spot. The
whitespace at the very beginning and the very end of the document (i.e.
the first node's outerPre and the last node's outerPost) isn't
duplicated anywhere, nor is inner whitespace in content nodes.
The basic outline of the implementation is:
* When we encounter whitespace, strip it and store it in the previous
node's outerPost. Also store it in nextWhitespace so we can put it in
the next node's outerPre once we encounter that node.
* When we encounter whitespace in wrapped bare text, we don't know in
advance if it's gonna be succeeded by more non-whitespace (in which
case it needs to be output verbatim), or not (in which case it's
leading whitespace and needs to be stripped and stored). The fact that
annotations are nodes in HTML makes this trickier. So we write the
whitespace to the temporary linmod and store it in wrappedWhitespace,
then if it turns out to be trailing whitespace we take it back out of
the data array and record it the usual way.
* Because text nodes can contain any combination of leading whitespace
actual text and trailing whitespace, and because we may or may not
already have opened a wrapping paragraph, there are a lot of different
combinations to handle. We handle all of them but the resulting code
is pretty dense and verbose.
More low-level list of changes:
In getDataFromDom():
* Added helper function addWhitespace() for storing whitespace for an
element
* Added helper function processNextWhitespace() for processing any
whitespace passed on from the previous node via the nextWhitespace var
* Rename paragraph to wrappingParagraph. Make wrapping default to
alreadyWrapped so we can simplify wrapping||alreadyWrapped and
!wrapping&&!alreadyWrapped. Add wrappingIsOurs to track whether the
wrapping originated in this recursion level (needed for deciding when
to close the wrapper).
* Add prevElement to track the previous element so we can propagate
whitespace to it, and nextWhitespace so we can propagate whitespace to
the next element.
* Remove previous newline stripping hacks
* Integrate the logic for wrapping bare content with the outer
whitespace preservation code
* Remove wrapperElement, no longer needed because we have a dedicated
variable for the wrapping paragraph now and what was previously inner
whitespace preservation for wrapper paragraphs is now covered by the
outer whitespace preservation code.
In getDomFromData():
* Reinsert whitespace where appropriate
** outerPre is inserted when opening the element
** This covers outerPost as well except for the last child's outerPost,
which is handled as the parent's innerPost when closing the parent.
** innerPre and innerPost are inserted when closing the element. Care is
taken not to insert these if they're duplicates of something else.
* Propagate each node's outerPost to the next node (either the next
sibling or the parent) using parentDomElement.lastOuterPost. We can't
get this using .lastChild because we will have destroyed that child's
.veInternal by then, and we can't tell whether a node will be its
parent's last child when we process it (all other processing,
including first child handling is done when processing the node itself,
but this cannot be).
* Special handling is needed for the last node's outerPost, which ends
up in the container's .lastOuterPost property.
Tests:
* Allow .html to be null in data<->DOM converter tests. This indicates
that the test is a one-way data->DOM test, not a DOM->data->DOM
round-trip test. The data will be converted to HTML and checked
against .normalizedHtml
* Update existing tests as needed
* Add tests for outer whitespace preservation and storage
* Add test for squashing of whitespace in case of disagreement (this
requires .html=null)
Change-Id: I4db4fe372a421182e80a2535657af7784ff15f95
data-mw-gc is ancient and unused. We do need to detect and alienate
generated nodes, but that is now based on RDFa types. Removing the
data-mw-gc stuff for now because it doesn't work anyway, will replace it
with proper detection later.
Replaced instances of data-mw-gc in the test suite with unregistered
node types.
Change-Id: If3f5898d382a436fa57929013264c53af5e840ba
Also:
* Made a fragment with a null range become a null fragment
* Fixed incorrect order of arguments for binding a handler to transact event
* Added getters for surface, document and range
* Fixed several instances of passing a document instead of a surface into the constructor of a new surface fragment
* Fixed closest mode in expandRange - needed to check if parent existed before checking for it's type
* Fixed uses of ve.Transaction (doesn't exist) that were supposed to be uses of ve.dm.Transaction (does exist)
Change-Id: Ide13d9d2d1637399188c98c2e8b6e0826caeecc4
When a document is created, it should take it upon itself to make sure it has a new reference to the data using slice, not place this on the caller. Callers that do not use slice will often find strange and mysterious things going on and not know why. The real reason is that multiple documents sharing a reference to the same data array leads to seriously messed up behavior.
Change-Id: Ic4e25fcd9bf3f41a805003520a8f38e2768f5dbf
domToData wraps bare content in paragraph elements, which were then
converted to <p> tags by domToData. With this fix, HTML with "missing"
<p> tags actually round-trips through the editor correctly now, rather
than having <p> tags added wherever VE believes they should exist.
* Mark generated paragraph elements with .internal.generated = 'wrapper'
** This signifies the wrapper was generated but its contents were not,
so the right thing to do when converting back to HTML is to remove
the wrapper and keep the contents. We might want to use other values
of generated in the future.
* Unwrap nodes with generated=wrapper when converting to HTML
Tests:
* Add 'generated': 'wrapper' as appropriate. Only affects 1 test
* Remove 'normalizedHtml' for this test because it is no longer needed
** Need to keep 'normalizedHtml' for now because we normalize hrefs
* Eventually the main example should test bare content, but that
requires touching a lot of stuff. The main example could use some
beefing up anyway.
Change-Id: I277ad5fe3f64e07c1bbf49007d6bbaecc90b7466
This allows us to put other internal data in there in the future. Also
passing it through the Node constructor properly now.
* ve.dm.Node
** Rename fringeWhitespace property to internal
** Add internal parameter to constructor
** Remove setFringeWhitespace()
* Increase the number of parameters passed through by ve.Factory to 3
* Pass through .internal from linmod to nodeFactory in ve.dm.Document
* ve.dm.Converter
** Rename .fringeWhitespace to .internal.whitespace and make it an array
** Store a temporary reference to .internal in domElement.veInternal
* Add internal to all node constructors except TextNode
Tests:
* Update for fringeWhitespace->internal rename
* Add third parameter to ve.Factory tests
* Add .internal to getNodeTreeSummary
Change-Id: If20c0bb78fee3efa55f72e51e7fc261283358de7
To do this, we are using the replacer callback of JSON.stringify, which is supported by all of our target browsers including IE8. We are also leaning on Object.keys and Array.reduce, the latter of which required adding a new fallback implementation for some browsers which do not support it yet.
Change-Id: Ifa285ca3da4d94d962464f09414591532bbea79c
Because the Parsoid prefix format changed from /mw:Foo to /mw/Foo , the
href format for internal links has changed from "/Foo" to "Foo". So the
href is now simply the title, except that it may be preceded by one or
more "../" if the title of the page we're on contains a '/'.
So instead of stripping the leading slash from internal link hrefs and
putting it back on the way out, only strip any leading "../"s and dump
the titles directly into the hrefs on the way out.
Also update the link test case for this, and add a test case for the ../
stripping.
Change-Id: I3e0bdde20f22cda34eb45fc351de5e780419b6a2
Annotation types with more than one slash such as 'link/ExtLink/URL'
weren't being processed correctly because .split( '/', 2 ) throws away
everything after the second slash. Instead, don't pass a limit to
.split(); the code for reconstructing a slash-separated string from
multiple components was already in place.
Also add test cases for URL links and numbered links.
(Do you like the lines-of-code to lines-of-test ratio in this commit,
Trevor? ;) )
Change-Id: I7add87396447a01b1c23a4f9bfd63d2e8fd861ce
Refactor:
* ve.indexOf
Renamed from ve.inArray.
This was named after the jQuery method which in turn has a longer
story about why it is so unfortunately named. It doesn't return
a boolean, but an index. Hence the native method being called
indexOf as well.
* ve.bind
Renamed from ve.proxy.
I considered making it use Function.prototype.bind if available.
As it performs better than $.proxy (which doesn't use to the native
bind if available). However since bind needs to be bound itself in
order to use it detached, it turns out with the "call()" and
"bind()" it is slower than the $.proxy shim:
http://jsperf.com/function-bind-shim-perf
It would've been like this:
ve.bind = Function.prototype.bind ?
Function.prototype.call.bind( Function.prototype.bind ) :
$.proxy;
But instead sticking to ve.bind = $.proxy;
* ve.extendObject
Documented the parts of jQuery.extend that we use. This makes it
easier to replace in the future.
Documentation:
* Added function documentation blocks.
* Added annotations to functions that we will be able to remove
in the future in favour of the native methods.
With "@until + when/how".
In this case "ES5". Meaning, whenever we drop support for browsers
that don't support ES5. Although in the developer community ES5 is
still fairly fresh, browsers have been aware for it long enough
that thee moment we're able to drop it may be sooner than we think.
The only blocker so far is IE8. The rest of the browsers have had
it long enough that the traffic we need to support of non-IE
supports it.
Misc.:
* Removed 'node: true' from .jshintrc since Parsoid is no longer in
this repo and thus no more nodejs files.
- This unraveled two lint errors: Usage of 'module' and 'console'.
(both were considered 'safe globals' due to nodejs, but not in
browser code).
* Replaced usage (before renaming):
- $.inArray -> ve.inArray
- Function.prototype.bind -> ve.proxy
- Array.isArray -> ve.isArray
- [].indexOf -> ve.inArray
- $.fn.bind/live/delegate/unbind/die/delegate -> $.fn.on/off
Change-Id: Idcf1fa6a685b6ed3d7c99ffe17bd57a7bc586a2c
This makes things like
== Foo ==
* Bar
render without the leading and trailing spaces, while still
round-tripping those spaces.
* Added a .fringeWhitespace property to the linear model and ve.dm.Node
** Object containing innerPre, innerPost, outerPre, outerPost
** Only inner* are used right now, outer* are planned for future use
** Like .attributes , it's suppressed if it's an empty object
* In getDataFromDom():
** Store the stripped whitespace in .fringeWhitespace
** Move emptiness check up: empty elements with .fringeWhitespace have
to be preserved
** Move paragraph wrapping up: .fringeWhitespace has to be applied to
the generated paragraph, not its parent
** Add wrapperElement to keep track of the element .fringeWhitespace has
to be added to; this is either dataElement or the generated paragraph
or nothing, but we can't modify dataElement because it's used later
* In getDomFromData():
** When processing an opening, store the fringeWhitespace data in the
generated DOM node
** When processing a closing, add the stored whitespace back in
* In the ve.dm.Document constructor, pass through .fringeWhitespace from
the linear model data to the generated nodes
Tests:
* Change one existing test case to account for this change
* Add three new test cases for this behavior
* Add normalizedHtml field so I can test behavior with bare content
Change-Id: I0411544652dd72b923c831c495d69ee4322a2c14
* Restricting "camelcase":
No changes, we were passing all of these already
* Explicitly unrestricting "forin" and "plusplus"
These are off by default in node-jshint, but some distro of jshint
and editors that use their own wrapper around jshint instead of
node-jshint (Eclipse?) may have different defaults. Therefor
setting them to false explicitly. This also serves as a reminder
for the future so we'll always know we don't pass that, in case
we would want to change that.
* Fix order ("quotemark" before "regexp")
* Restricting "unused"
We're not passing all of this, which is why I've set it to false
for now. But I did put it in .jshintrc as placeholder.
I've fixed most of them, there's some left where there is no clean
solution.
* While at it fix a few issues:
- Unused variables ($target, $window)
- Bad practices (using jQuery context for find instead of creation)
- Redundant /*global */ comments
- Parameters that are not used and don't have documentation either
- Lines longer than 100 chars @ 4 spaces/tab
* Note:
- ve.ce.Surface.prototype.onChange takes two arguments but never
uses the former. And even the second one can be null/undefined.
Aside from that, the .change() function emits
another event for the transaction already. Looks like this
should be refactored a bit, two more separated events probably
or one that is actually used better.
- Also cleaned up a lot of comments, some of which were missing,
others were incorrect
- Reworked the contentChange event so we are no longer using the
word new as an object key; expanded a complex object into multiple
arguments being passed through the event to make it easier to work
with and document
Change-Id: I8490815a508c6c379d5f9a743bb4aefd14576aa6
Stack traces, line numbers, etc. All the approaches I've seen are bad hacks. This is the best way to go.
Change-Id: Ib12e9d2ecfe610bcc89d046005e35cc13efa3d99
Throwing strings is bad because it doesn't include a lot of important
information that an error object does, such as a stack trace or where
the error was actually thrown from.
ve.Error inherits directly from Error. In the future we may create
more specific subclasses and/or do custom stuff.
Some interesting reading on the subject:
* http://www.devthought.com/2011/12/22/a-string-is-not-an-error/
Change-Id: Ib7c568a1dcb98abac44c6c146e84dde5315b2826
* There were only 3 files with single quotes, fixed them all.
* Added option to .jshintrc (be sure to use the latest version of
node-jshint since this is a fairly new addition to the library).
Change-Id: I8bf8895ce56bf86e3bed244279a9d32269e44763
This just cleans up a few places where single quotes should have been used instead of double quotes.
Change-Id: I6c53652e71ab96842ed5bb41fb1e0b8c923eb25d
Also:
* Removed a lot of dead code in Surface that was used in the now dead and gone sandbox.
* Changed from throwing an exception when calling getBalancedData on a range that produces no results from selectNodes to just returning []
Change-Id: Icf27094724eae5b90eec21308f9e26afe877e3ee
* Also update test/index.html with latest minimalistic format
* Update test suite hardcoded paths to match the definition in
VisualEditor.php for ResourceLoader
* Issues:
- 'jquery/jquery.json.js' should not be loaded directly,
using a dependency instead.
- Load scripts from the <head> in test/index.html so that
code that depends on document being ready is catches instead of
silently being ignored.
Change-Id: I5ad7390137f4d17c153a1bf69f19c4869c08e323
'''Kranitor commits''' are commits by Krinkle with his janitor hat on.
Must never contain functional changes mixed with miscellaneous changes.
.gitignore:
* Add .DS_Store to the ignore list so that browsing the directories
on Mac OS X, will not add these files to the list of untracked
files.
* Fix missing newline at end of file
.jshintrc
* raises -> throws
* +module (QUnit.module)
* remove 'Node' (as of node-jshint 1.7.2 this is now part of
'browser:true', as it should be)
Authors:
* Adding myself
MWExtension/VisualEditor.php
* Fix default value of wgVisualEditorParsoidURL to not
point to the experimental instance in WMF Labs.
Issues:
* ve.ce.TextNode:
- Fix TODO: Don't perform a useless clone of an already-jQuerified object.
- Use .html() to set html content instead of encapsulating between
two strings. This is slightly faster but more importantly safer,
and prevents situations where the resulting jQuery collection
actually contains 2 elements instead of 1, thus messing up
what .contents() is iterating over.
* ve.ce.Document.test.js
- Fix: ReferenceError: assert is not defined
* ve.dm.Document.test.js
- Fix: ReferenceError: assert is not defined
* ve.dm.Transaction.test.js
- Fix: ReferenceError: assert is not defined
* ve.dm.TransactionProcessor.test.js
- Fix: ReferenceError: assert is not defined
* ext.visualEditor.viewPageTarget
- Missing dependency on 'mediawiki.Title'
Code conventions / Misc cleanup
* Various JSHint warnings.
* Whitespace
* jQuery(): Use '<tag>' for element creation,
use '<valid><xml/></valid>' for parsing
* Use the default operator instead of ternary when the condition and
first value are the same.
x = foo ? foo : bar; -> x = foo || bar;
Because contrary to some programming language (PHP...), in JS the
default operator does not enforce a boolean result but returns the
original value, hence it being called the 'default' operator, as
opposed to the 'or' operator.
* No need to call addClass() twice, it takes a space-separated list
(jQuery splits by space and adds if needed)
* Use .on( event[, selector], fn ) instead of the deprecated
routers to it such as .bind(), .delegate() and .live().
All these three are now built-in and fully compatible with .on()
* Add 'XXX:' comments for suspicious code that I don't want to change
as part of a clean up commit.
* Remove unused variables (several var x = this; where x was not
used anywhere, possibly from boilerplate copy/paste)
* Follows-up Trevor's commit that converts test suites to the new
QUnit format. Also removed the globals since we no longer use those
any more.
Change-Id: I7e37c9bff812e371c7f65a6fd85d9e2af3e0a22f
Convert underscores in the href attribute to spaces in the linear model,
and back to underscores when going back to HTML. This ensures the link
targets displayed to and edited by the user look nice
Change-Id: I4855fce28ad8b724284c53881abc7b99b59b9079
This means we don't have to rely on data-rt.sHref. It also means that
we'll now be showing the canonical link target in the link inspector
rather than the link target as entered by the user, but that's fine.
Also change test to have href differ from sHref to show that we use
href.
Change-Id: Idabdbf2579663ef1efb47d6a73f39743c9f64f3b
This is ugly but makes things work again. I intend to clean this up once
we have a better attribute API
* Recognize mw:WikiLink, mw:SimpleWikiLink, mw:ExtLink,
mw:NumberedExtLink and mw:UrlLink
* Support is incomplete because we can't get to the annotation text with
the current API
* Preserve all unhandled attributes rather than special-casing data-mw
* Update remaining code using data-mw (sHref and stx extraction) to
account for the data-mw -> data-rt rename
* Update tests accordingly
Change-Id: Ia13d3008a6d4cdc8828f9acda5aa797566bc597f
