mediawiki-extensions-Visual.../modules/ve/dm/ve.dm.Converter.js

1075 lines
36 KiB
JavaScript
Raw Normal View History

JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
/*!
* VisualEditor DataModel Converter class.
*
* @copyright 2011-2012 VisualEditor Team and others; see AUTHORS.txt
* @license The MIT License (MIT); see LICENSE.txt
*/
/**
* DataModel converter.
*
* Converts between HTML DOM and VisualEditor linear data.
*
* @class
* @constructor
* @param {Object} options Conversion options
*/
ve.dm.Converter = function VeDmConverter( nodeFactory, annotationFactory ) {
// Properties
this.nodeFactory = nodeFactory;
this.annotationFactory = annotationFactory;
this.elements = {
'toDomElement': {},
'toDataElement': {},
'dataElementTypes': {}
};
// Events
this.nodeFactory.addListenerMethod( this, 'register', 'onNodeRegister' );
};
/* Static Methods */
/**
* Get linear model data from a string optionally applying annotations
*
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* @static
* @param {string} text Plain text to convert
* @param {Array} [annotations] Array of annotation objects to apply
* @returns {Array} Linear model data, one element per character
*/
ve.dm.Converter.getDataContentFromText = function ( text, annotations ) {
var characters = text.split( '' ),
annotationSet = new ve.AnnotationSet(),
i;
if ( !annotations || annotations.length === 0 ) {
return characters;
}
// Build annotation set
for ( i = 0; i < annotations.length; i++ ) {
annotationSet.push( annotations[i] );
}
// Apply annotations to characters
for ( i = 0; i < characters.length; i++ ) {
// Make a shallow copy of the annotationSet object, otherwise adding an annotation to one
// character automatically adds it to all of others as well, annotations should be treated
// as immutable, so it's OK to share references, but annotation sets are not immutable, so
// it's not safe to share references - each annotated character needs its own set
characters[i] = [characters[i], annotationSet.clone()];
}
return characters;
};
/* Methods */
/**
* Handle register events from the node factory.
*
* FIXME
* If a node is special; such as document, alienInline, alienBlock and text; its {converters}
* property should be set to null, as to distinguish it from a new node type that someone has simply
* forgotten to implement converters for.
*
* @method
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* @param {string} type Node type
* @param {Function} constructor Node constructor
* @throws {Error} Missing conversion data in node implementation
*/
ve.dm.Converter.prototype.onNodeRegister = function ( dataElementType, constructor ) {
if ( !constructor.static.matchTagNames || !constructor.static.toDomElement ||
!constructor.static.toDataElement
) {
throw new Error( 'Missing static properties in node implementation of ' + dataElementType );
} else {
var i,
domElementTypes = constructor.static.matchTagNames,
toDomElement = constructor.static.toDomElement,
toDataElement = constructor.static.toDataElement;
// Registration
this.elements.toDomElement[dataElementType] = toDomElement;
for ( i = 0; i < domElementTypes.length; i++ ) {
this.elements.toDataElement[domElementTypes[i]] = toDataElement;
this.elements.dataElementTypes[domElementTypes[i]] = dataElementType;
}
}
};
/**
* Get the DOM element for a given linear model element.
*
* This invokes the toDomElement function registered for the element type.
*
* @method
* @param {Object} dataElement Linear model element
* @returns {HTMLElement|boolean} DOM element, or false if the element cannot be converted
*/
ve.dm.Converter.prototype.getDomElementFromDataElement = function ( dataElement ) {
var key, domElement, dataElementAttributes, wrapper,
dataElementType = dataElement.type;
if ( dataElementType === 'alienInline' || dataElementType === 'alienBlock' ) {
// Alien
// Create nodes from source
wrapper = document.createElement( 'div' );
wrapper.innerHTML = dataElement.attributes.html;
if ( wrapper.childNodes.length > 1 ) {
// Wrap the HTML in a single element, this makes
// it much easier to deal with. It'll be unwrapped
// at the end of getDomFromData().
domElement = document.createElement( 'div' );
domElement.setAttribute( 'data-ve-multi-child-alien-wrapper', 'true' );
while ( wrapper.firstChild ) {
domElement.appendChild( wrapper.firstChild );
}
} else {
domElement = wrapper.firstChild;
}
return domElement;
}
if ( !( dataElementType in this.elements.toDomElement ) ) {
// Unsupported element
return false;
}
domElement = this.elements.toDomElement[dataElementType]( dataElement );
dataElementAttributes = dataElement.attributes;
if ( dataElementAttributes ) {
for ( key in dataElementAttributes ) {
// Only include 'html/*' attributes and strip the 'html/' from the beginning of the name
if ( key.indexOf( 'html/' ) === 0 ) {
domElement.setAttribute( key.substr( 5 ), dataElementAttributes[key] );
}
}
}
// Change markers
if (
dataElement.internal && dataElement.internal.changed &&
!ve.isEmptyObject( dataElement.internal.changed ) &&
ve.init.platform.useChangeMarkers()
) {
domElement.setAttribute( 'data-ve-changed',
JSON.stringify( dataElement.internal.changed )
);
}
return domElement;
};
/**
* Get the linear model data element for a given DOM element.
*
* This invokes the toDataElement function registered for the element type
*
* @method
* @param {HTMLElement} domElement DOM element
* @param {Array} annotations Annotations to apply if the node is a content node
* @returns {Object|boolean} Linear model element, or false if the node cannot be converted
*/
ve.dm.Converter.prototype.getDataElementFromDomElement = function ( domElement, annotations ) {
var dataElement, domElementAttributes, dataElementAttributes, domElementAttribute, i,
annotationSet, domElementType = domElement.nodeName.toLowerCase();
annotations = annotations || [];
if (
// Unsupported elements
!( domElementType in this.elements.toDataElement )
// TODO check for generated elements
) {
return false;
}
dataElement = this.elements.toDataElement[domElementType]( domElement );
domElementAttributes = domElement.attributes;
if ( domElementAttributes.length ) {
dataElementAttributes = dataElement.attributes = dataElement.attributes || {};
// Include all attributes and prepend 'html/' to each attribute name
for ( i = 0; i < domElementAttributes.length; i++ ) {
domElementAttribute = domElementAttributes[i];
dataElementAttributes['html/' + domElementAttribute.name] = domElementAttribute.value;
}
}
if ( ve.dm.nodeFactory.isNodeContent( dataElement.type ) && annotations.length > 0 ) {
// Build annotation set
annotationSet = new ve.AnnotationSet();
for ( i = 0; i < annotations.length; i++ ) {
annotationSet.push( annotations[i] );
}
dataElement.annotations = annotationSet;
}
return dataElement;
};
/**
* Check if an HTML DOM node represents an annotation, and if so, build an annotation object for it.
*
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* Annotation Object:
* { 'type': 'type', data: { 'key': 'value', ... } }
*
* @param {HTMLElement} domElement HTML DOM node
* @returns {Object|boolean} Annotation object, or false if the node is not an annotation
*/
ve.dm.Converter.prototype.getDataAnnotationFromDomElement = function ( domElement ) {
return ve.dm.annotationFactory.createFromElement( domElement ) || false;
};
/**
* Build an HTML DOM node for a linear model annotation.
*
* @method
* @param {Object} dataAnnotation Annotation object
* @returns {HTMLElement} HTML DOM node
*/
ve.dm.Converter.prototype.getDomElementFromDataAnnotation = function ( dataAnnotation ) {
var htmlData = dataAnnotation.toHTML(),
domElement = document.createElement( htmlData.tag );
ve.setDomAttributes( domElement, htmlData.attributes );
return domElement;
};
/**
* Convert an HTML DOM tree to a linear model.
*
* Do not use the annotations, dataElement and path parameters, they're used for internal
* recursion only.
*
* @method
* @param {HTMLElement} domElement Wrapper div containing the HTML to convert
* @param {Array} [annotations] Array of annotations (objects) to apply to the generated data
* @param {Object} [dataElement] Data element to wrap the returned data in
* @param {Array} [path] Array of linear model element types
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* @param {boolean} [alreadyWrapped] Whether the caller has already started wrapping bare content in a paragraph
* @returns {Array} Linear model data
*/
ve.dm.Converter.prototype.getDataFromDom = function ( domElement, annotations, dataElement, path, alreadyWrapped ) {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
function createAlien( domElement, branchHasContent, isWrapper ) {
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
// We generate alienBlock elements for block tags and alienInline elements for
// inline tags; unless we're in a content location, in which case we have no choice
// but to generate an alienInline element.
var isInline =
// Force inline in content locations (but not wrappers)
( !wrapping && ( branchHasContent || annotations.length > 0 ) ) ||
// Also force inline in wrappers that aren't ours
( wrapping && !wrappingIsOurs ) ||
// Look at the tag name otherwise
!ve.isBlockElement( domElement ),
type = isInline ? 'alienInline' : 'alienBlock',
html, alien, annotationSet, i;
if ( isWrapper ) {
html = $( domElement ).html();
} else {
html = $( '<div>' ).append( $( domElement ).clone() ).html();
}
alien = [
{
'type': type,
'attributes': {
'html': html
}
},
{ 'type': '/' + type }
];
if ( annotations.length > 0 ) {
annotationSet = new ve.AnnotationSet();
for ( i = 0; i < annotations.length; i++ ) {
annotationSet.push( annotations[i] );
}
alien[0].annotations = annotationSet;
}
return alien;
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
function addWhitespace( element, index, whitespace ) {
if ( !element.internal ) {
element.internal = {};
}
// whitespace = [ outerPre, innerPre, innerPost, outerPost ]
// <tag> text </tag> <nextTag>
// ^^^^^^^^^^ ^^^^^^^^^ ^^^^^^^^^^^ ^^^^^^^^^^
// outerPre innerPre innerPost outerPost
if ( !element.internal.whitespace ) {
element.internal.whitespace = [];
}
if ( !element.internal.whitespace[index] ) {
element.internal.whitespace[index] = '';
}
element.internal.whitespace[index] = whitespace;
}
function processNextWhitespace( element ) {
// This function uses and changes nextWhitespace in the outer function's scope,
// which means it's not really a function but more of a shortcut.
if ( nextWhitespace !== '' ) {
addWhitespace( element, 0, nextWhitespace );
nextWhitespace = '';
}
}
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
function startWrapping() {
// Mark this paragraph as having been generated by
// us, so we can strip it on the way out
wrappingParagraph = {
'type': 'paragraph',
'internal': { 'generated': 'wrapper' }
};
data.push( wrappingParagraph );
wrapping = true;
wrappingIsOurs = true;
processNextWhitespace( wrappingParagraph );
}
function stopWrapping() {
if ( wrappedWhitespace !== '' ) {
// Remove wrappedWhitespace from data
data.splice( wrappedWhitespaceIndex, wrappedWhitespace.length );
addWhitespace( wrappingParagraph, 3, wrappedWhitespace );
nextWhitespace = wrappedWhitespace;
}
data.push( { 'type': '/paragraph' } );
wrappingParagraph = undefined;
wrapping = false;
wrappingIsOurs = false;
}
/**
* Helper function to group adjacent child elements with the same about attribute together.
* If there are multiple adjacent child nodes with the same about attribute, they are
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* wrapped in a `<div>` with the data-ve-aboutgroup attribute set.
*
* This function does not wrap single-element about groups, and does not descend into the
* child elements.
*
JSDuck: Generated code documentation! 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
2013-01-04 08:54:17 +00:00
* @private
* @param element {HTMLElement} Element to process
*/
function doAboutGrouping( element ) {
var child = element.firstChild, textNodes = [],
prevChild, aboutGroup, aboutWrapper, childAbout, nextChild, i;
while ( child ) {
nextChild = child.nextSibling;
if ( !child.getAttribute ) {
// Text nodes don't have a getAttribute() method. Thanks HTML DOM,
// that's really helpful ^^
textNodes.push( child );
child = nextChild;
continue;
}
childAbout = child.getAttribute( 'about' );
if ( childAbout && !aboutGroup ) {
// Start of a new about group
aboutGroup = childAbout;
} else if ( childAbout && childAbout === aboutGroup ) {
// Continuation of the current about group
if ( !aboutWrapper ) {
// This is the second child in this group, so the
// previous child is the first child in this group.
// Wrap the previous child
aboutWrapper = document.createElement( 'div' );
aboutWrapper.setAttribute( 'data-ve-aboutgroup', aboutGroup );
element.insertBefore( aboutWrapper, prevChild );
aboutWrapper.appendChild( prevChild );
}
// Append any outstanding text nodes to the wrapper
for ( i = 0; i < textNodes.length; i++ ) {
aboutWrapper.appendChild( textNodes[i] );
}
// Append this child to the wrapper
aboutWrapper.appendChild( child );
} else if ( aboutGroup ) {
// This child isn't in the current about group
aboutGroup = undefined;
aboutWrapper = undefined;
if ( childAbout ) {
// Start of a new about group
aboutGroup = childAbout;
}
}
prevChild = child;
child = nextChild;
textNodes = [];
}
}
// Fallback to defaults
annotations = annotations || [];
path = path || ['document'];
var i, j, childDomElement, annotation, childDataElement, text, childTypes, matches,
wrappingParagraph, prevElement, alien, rdfaType, isLink,
data = [],
branchType = path[path.length - 1],
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
branchHasContent = ve.dm.nodeFactory.canNodeContainContent( branchType ),
childIsContent,
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
nextWhitespace = '',
wrappedWhitespace = '',
wrappedWhitespaceIndex,
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
wrapping = alreadyWrapped,
wrappingIsOurs = false;
// Open element
if ( dataElement ) {
data.push( dataElement );
}
// Do about grouping
// FIXME this assumes every about group is an alien
doAboutGrouping( domElement );
// Add contents
for ( i = 0; i < domElement.childNodes.length; i++ ) {
childDomElement = domElement.childNodes[i];
switch ( childDomElement.nodeType ) {
case Node.ELEMENT_NODE:
// Alienate about groups
if ( childDomElement.hasAttribute( 'data-ve-aboutgroup' ) ) {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
alien = createAlien( childDomElement, branchHasContent, true );
if ( wrapping && alien[0].type === 'alienBlock' ) {
stopWrapping();
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
} else if (
!wrapping && !branchHasContent &&
alien[0].type === 'alienInline'
) {
startWrapping();
}
data = data.concat( alien );
processNextWhitespace( alien[0] );
prevElement = alien[0];
break;
}
// HACK handle <meta>/<link> separately because of the
// metaInline/metaBlock distinction
if (
childDomElement.nodeName.toLowerCase() === 'meta' ||
childDomElement.nodeName.toLowerCase() === 'link'
) {
isLink = childDomElement.nodeName.toLowerCase() === 'link';
childDataElement = {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
'type': branchHasContent ? 'metaInline' : 'metaBlock',
'attributes': {
'style': isLink ? 'link' : 'meta',
'key': childDomElement.getAttribute( isLink ? 'rel' : 'property' )
}
};
if ( childDomElement.hasAttribute( isLink ? 'href' : 'content' ) ) {
childDataElement.attributes.value = childDomElement.getAttribute( isLink ? 'href' : 'content' );
}
// Preserve HTML attributes
// FIXME the following is duplicated from getDataElementFromDomElement()
// Include all attributes and prepend 'html/' to each attribute name
for ( j = 0; j < childDomElement.attributes.length; j++ ) {
// ..but exclude attributes we've already processed,
// because they'll be overwritten otherwise *sigh*
// FIXME this sucks, we need a new node type API so bad
if (
childDomElement.attributes[j].name !== ( isLink ? 'rel' : 'property' ) &&
childDomElement.attributes[j].name !== ( isLink ? 'href' : 'content' )
) {
childDataElement.attributes['html/' + childDomElement.attributes[j].name] = childDomElement.attributes[j].value;
}
}
data.push( childDataElement );
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
data.push( { 'type': branchHasContent ? '/metaInline' : '/metaBlock' } );
processNextWhitespace( childDataElement );
prevElement = childDataElement;
break;
}
// Alienate anything with a mw: type that isn't registered
// HACK because we don't actually have an RDFa type registry yet,
// this hardcodes the set of recognized types
rdfaType = childDomElement.getAttribute( 'rel' ) ||
childDomElement.getAttribute( 'typeof' ) ||
childDomElement.getAttribute( 'property' );
if (
rdfaType &&
rdfaType.match( /^mw:/ ) &&
!rdfaType.match( /^mw:WikiLink/ ) &&
!rdfaType.match( /^mw:ExtLink/ ) &&
!rdfaType.match( /^mw:Entity/ )
) {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
alien = createAlien( childDomElement, branchHasContent );
if ( wrapping && alien[0].type === 'alienBlock' ) {
stopWrapping();
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
} else if (
!wrapping && !branchHasContent &&
alien[0].type === 'alienInline'
) {
startWrapping();
}
data = data.concat( alien );
processNextWhitespace( alien[0] );
prevElement = alien[0];
break;
}
// Detect and handle annotated content
// HACK except for mw:Entity. We need a node API rewrite, badly
annotation = this.getDataAnnotationFromDomElement( childDomElement );
if ( annotation && rdfaType !== 'mw:Entity' ) {
// Start auto-wrapping of bare content
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
if ( !wrapping && !branchHasContent ) {
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
startWrapping();
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
prevElement = wrappingParagraph;
}
// Append child element data
data = data.concat(
this.getDataFromDom(
childDomElement, annotations.concat( [ annotation ] ), undefined, path, wrapping
)
);
break;
}
// Look up child element type
childDataElement = this.getDataElementFromDomElement( childDomElement, annotations );
if ( childDataElement ) {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
childIsContent = ve.dm.nodeFactory.isNodeContent( childDataElement.type );
// Check that something isn't terribly wrong
if ( !(
// Non-content child in a content container
( branchHasContent && !childIsContent ) ||
// Non-content child trying to break wrapping at
// the wrong level
( wrapping && !wrappingIsOurs && !childIsContent )
) ) {
// End auto-wrapping of bare content from a previously processed node
// but only if childDataElement is a non-content element
if ( wrapping && wrappingIsOurs && !childIsContent ) {
stopWrapping();
}
if ( ve.dm.nodeFactory.canNodeHaveChildren( childDataElement.type ) ) {
// Append child element data
data = data.concat(
this.getDataFromDom(
childDomElement,
[],
childDataElement,
path.concat( childDataElement.type ),
wrapping
)
);
} else {
// Append empty node
data.push( childDataElement );
data.push( { 'type': '/' + childDataElement.type } );
}
processNextWhitespace( childDataElement );
prevElement = childDataElement;
break;
}
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
// If something is wrong, fall through, and the bad child
// will be alienated below.
}
// We don't know what this is, fall back to alien.
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
alien = createAlien( childDomElement, branchHasContent );
if ( wrapping && alien[0].type === 'alienBlock' ) {
stopWrapping();
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
} else if (
!wrapping && !branchHasContent &&
alien[0].type === 'alienInline'
) {
startWrapping();
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
data = data.concat( alien );
processNextWhitespace( alien[0] );
prevElement = alien[0];
break;
case Node.TEXT_NODE:
text = childDomElement.data;
if ( text === '' ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Empty text node?!?
break;
}
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
if ( !branchHasContent ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Strip and store outer whitespace
if ( text.match( /^\s+$/ ) ) {
// This text node is whitespace only
if ( wrapping ) {
// We're already wrapping, so output this whitespace
// and store it in wrappedWhitespace (see
// comment about wrappedWhitespace below)
wrappedWhitespace = text;
wrappedWhitespaceIndex = data.length;
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
data = data.concat(
ve.dm.Converter.getDataContentFromText( wrappedWhitespace, annotations )
);
} else {
// We're not in wrapping mode, store this whitespace
if ( !prevElement ) {
if ( dataElement ) {
// First child, store as inner
// whitespace in the parent
addWhitespace( dataElement, 1, text );
}
// Else, WTF?!? This is not supposed to
// happen, but it's not worth
// throwing an exception over.
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
} else {
addWhitespace( prevElement, 3, text );
}
nextWhitespace = text;
wrappedWhitespace = '';
}
// We're done, no actual text left to process
break;
} else {
// This text node contains actual text
// Separate the real text from the whitespace
// HACK: . doesn't match newlines in JS, so use
// [\s\S] to match any character
matches = text.match( /^(\s*)([\s\S]*?)(\s*)$/ );
if ( !wrapping ) {
// Wrap the text in a paragraph and output it
(bug 43056) Inline tags like <span> are block-alienated sometimes 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
2012-12-20 00:59:58 +00:00
startWrapping();
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Only store leading whitespace if we just
// started wrapping
if ( matches[1] !== '' ) {
if ( !prevElement ) {
if ( dataElement ) {
// First child, store as inner
// whitespace in the parent
addWhitespace( dataElement, 1, matches[1] );
}
// Else, WTF?!? This is not supposed to
// happen, but it's not worth
// throwing an exception over.
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
} else {
addWhitespace( prevElement, 3, matches[1] );
}
addWhitespace( wrappingParagraph, 0, matches[1] );
}
} else {
// We were already wrapping in a paragraph,
// so the leading whitespace must be output
data = data.concat(
ve.dm.Converter.getDataContentFromText( matches[1], annotations )
);
}
// Output the text sans whitespace
data = data.concat(
ve.dm.Converter.getDataContentFromText( matches[2], annotations )
);
// Don't store this in wrappingParagraph.internal.whitespace[3]
// and nextWhitespace just yet. Instead, store it
// in wrappedWhitespace. There might be more text
// nodes after this one, so we output wrappedWhitespace
// for now and undo that if it turns out this was
// the last text node. We can't output it later
// because we have to apply the correct annotations.
wrappedWhitespace = matches[3];
wrappedWhitespaceIndex = data.length;
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
data = data.concat(
ve.dm.Converter.getDataContentFromText( wrappedWhitespace, annotations )
);
prevElement = wrappingParagraph;
break;
}
2012-08-10 21:09:04 +00:00
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Strip leading and trailing inner whitespace
// (but only in non-annotation nodes)
// and store it so it can be restored later.
if (
annotations.length === 0 && i === 0 && dataElement &&
!ve.dm.nodeFactory.doesNodeHaveSignificantWhitespace( dataElement.type )
) {
2012-08-10 21:09:04 +00:00
// Strip leading whitespace from the first child
matches = text.match( /^\s+/ );
if ( matches && matches[0] !== '' ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
addWhitespace( dataElement, 1, matches[0] );
2012-08-10 21:09:04 +00:00
text = text.substring( matches[0].length );
}
}
if (
annotations.length === 0 &&
i === domElement.childNodes.length - 1 &&
dataElement &&
!ve.dm.nodeFactory.doesNodeHaveSignificantWhitespace( dataElement.type )
2012-08-10 21:09:04 +00:00
) {
// Strip trailing whitespace from the last child
matches = text.match( /\s+$/ );
if ( matches && matches[0] !== '' ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
addWhitespace( dataElement, 2, matches[0] );
2012-08-10 21:09:04 +00:00
text = text.substring( 0,
text.length - matches[0].length );
}
}
// Annotate the text and output it
data = data.concat(
ve.dm.Converter.getDataContentFromText( text, annotations )
);
break;
case Node.COMMENT_NODE:
childDataElement = {
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
'type': branchHasContent ? 'metaInline' : 'metaBlock',
'attributes': {
'style': 'comment',
'text': childDomElement.data
}
};
data.push( childDataElement );
(bug 42487) Don't crash the converter for "<span>\n<p>Foo</p></span>" 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
2012-12-05 22:35:10 +00:00
data.push( { 'type': branchHasContent ? '/metaInline' : '/metaBlock' } );
processNextWhitespace( childDataElement );
prevElement = childDataElement;
break;
}
}
// End auto-wrapping of bare content
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
if ( wrapping && wrappingIsOurs ) {
stopWrapping();
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Don't set wrapping = false here because it's checked below
wrapping = true;
}
// If we're closing a node that doesn't have any children, but could contain a paragraph,
// add a paragraph. This prevents things like empty list items
childTypes = ve.dm.nodeFactory.getChildNodeTypes( branchType );
if ( branchType !== 'paragraph' && dataElement && data[data.length - 1] === dataElement &&
!wrapping && !ve.dm.nodeFactory.canNodeContainContent( branchType ) &&
!ve.dm.nodeFactory.isNodeContent( branchType ) &&
Refactor ve.js utilities and improve documentation 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
2012-08-11 08:14:56 +00:00
( childTypes === null || ve.indexOf( 'paragraph', childTypes ) !== -1 )
) {
data.push( { 'type': 'paragraph', 'internal': { 'generated': 'empty' } } );
data.push( { 'type': '/paragraph' } );
}
// Close element
if ( dataElement ) {
data.push( { 'type': '/' + dataElement.type } );
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Add the whitespace after the last child to the parent as innerPost
if ( nextWhitespace !== '' ) {
addWhitespace( dataElement, 2, nextWhitespace );
nextWhitespace = '';
}
}
// Don't return an empty document
if ( branchType === 'document' && data.length === 0 ) {
return [
{ 'type': 'paragraph', 'internal': { 'generated': 'empty' } },
{ 'type': '/paragraph' }
];
}
return data;
};
/**
* Convert linear model data to an HTML DOM
*
* @method
* @param {Array} data Linear model data
* @returns {HTMLElement} Wrapper div containing the resulting HTML
*/
ve.dm.Converter.prototype.getDomFromData = function ( data ) {
var text, i, j, k, annotations, annotation, annotationElement, dataElement, arr,
childDomElement, pre, ours, theirs, parentDomElement, startClosingAt,
isContentNode, changed, parentChanged,
container = document.createElement( 'div' ),
domElement = container,
annotationStack = new ve.AnnotationSet();
for ( i = 0; i < data.length; i++ ) {
if ( typeof data[i] === 'string' ) {
// Text
text = '';
// Continue forward as far as the plain text goes
while ( typeof data[i] === 'string' ) {
text += data[i];
i++;
}
// i points to the first non-text thing, go back one so we don't skip this later
i--;
// Add text
domElement.appendChild( document.createTextNode( text ) );
} else if (
ve.isArray( data[i] ) ||
(
data[i].annotations !== undefined &&
ve.dm.nodeFactory.isNodeContent( data[i].type )
)
) {
// Annotated text or annotated nodes
text = '';
while (
ve.isArray( data[i] ) ||
(
data[i].annotations !== undefined &&
ve.dm.nodeFactory.isNodeContent( data[i].type )
)
) {
annotations = data[i].annotations || data[i][1];
// Close annotations as needed
// Go through annotationStack from bottom to top (low to high),
// and find the first annotation that's not in annotations.
startClosingAt = undefined;
arr = annotationStack.get();
for ( j = 0; j < arr.length; j++ ) {
annotation = arr[j];
if ( !annotations.contains( annotation ) ) {
startClosingAt = j;
break;
}
}
if ( startClosingAt !== undefined ) {
// Close all annotations from top to bottom (high to low)
// until we reach startClosingAt
for ( j = annotationStack.getLength() - 1; j >= startClosingAt; j-- ) {
// Add text if needed
if ( text.length > 0 ) {
domElement.appendChild( document.createTextNode( text ) );
text = '';
}
// Traverse up
domElement = domElement.parentNode;
// Remove from annotationStack
annotationStack.removeAt( j );
}
}
// Open annotations as needed
arr = annotations.get();
for ( j = 0; j < arr.length; j++ ) {
annotation = arr[j];
if ( !annotationStack.contains( annotation ) ) {
// Add text if needed
if ( text.length > 0 ) {
domElement.appendChild( document.createTextNode( text ) );
text = '';
}
// Create new node and descend into it
annotationElement = this.getDomElementFromDataAnnotation( annotation );
domElement.appendChild( annotationElement );
domElement = annotationElement;
// Add to annotationStack
annotationStack.push( annotation );
}
}
if ( data[i].annotations === undefined ) {
// Annotated text
text += data[i][0];
} else {
// Annotated node
// Add text if needed
if ( text.length > 0 ) {
domElement.appendChild( document.createTextNode( text ) );
text = '';
}
// Insert the element
domElement.appendChild( this.getDomElementFromDataElement( data[i] ) );
// Increment i once more so we skip over the closing as well
i++;
}
i++;
}
// We're now at the first non-annotated thing, go back one so we don't skip this later
i--;
// Add any gathered text
if ( text.length > 0 ) {
domElement.appendChild( document.createTextNode( text ) );
text = '';
}
// Close any remaining annotation nodes
for ( j = annotationStack.getLength() - 1; j >= 0; j-- ) {
// Traverse up
domElement = domElement.parentNode;
}
// Clear annotationStack
annotationStack = new ve.AnnotationSet();
} else if ( data[i].type !== undefined ) {
dataElement = data[i];
// Element
if ( dataElement.type.charAt( 0 ) === '/' ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
parentDomElement = domElement.parentNode;
isContentNode = ve.dm.nodeFactory.isNodeContent( data[i].type.substr( 1 ) );
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Process whitespace
// whitespace = [ outerPre, innerPre, innerPost, outerPost ]
if (
!isContentNode &&
domElement.veInternal &&
domElement.veInternal.whitespace
) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Process inner whitespace. innerPre is for sure legitimate
// whitespace that should be inserted; if it was a duplicate
// of our child's outerPre, we would have cleared it.
pre = domElement.veInternal.whitespace[1];
2012-08-10 21:09:04 +00:00
if ( pre ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
if (
domElement.firstChild &&
domElement.firstChild.nodeType === 3
) {
2012-08-10 21:09:04 +00:00
// First child is a TextNode, prepend to it
domElement.firstChild.insertData( 0, pre );
} else {
// Prepend a TextNode
domElement.insertBefore(
document.createTextNode( pre ),
domElement.firstChild
);
}
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
ours = domElement.veInternal.whitespace[2];
if ( domElement.lastOuterPost === undefined ) {
// This node didn't have any structural children
// (i.e. it's a content-containing node), so there's
// nothing to check innerPost against
theirs = ours;
} else {
theirs = domElement.lastOuterPost;
}
if ( ours && ours === theirs ) {
if (
domElement.lastChild &&
domElement.lastChild.nodeType === 3
) {
2012-08-10 21:09:04 +00:00
// Last child is a TextNode, append to it
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
domElement.lastChild.appendData( ours );
2012-08-10 21:09:04 +00:00
} else {
// Append a TextNode
domElement.appendChild(
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
document.createTextNode( ours )
2012-08-10 21:09:04 +00:00
);
}
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Tell the parent about our outerPost
parentDomElement.lastOuterPost = domElement.veInternal.whitespace[3] || '';
} else if ( !isContentNode ) {
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Use empty string, because undefined means there were no
// structural children
parentDomElement.lastOuterPost = '';
2012-08-10 21:09:04 +00:00
}
// else don't touch lastOuterPost
// If closing a generated wrapper node, unwrap it
// It would be nicer if we could avoid generating in the first
// place, but then remembering where we have to skip ascending
// to the parent would be tricky.
// We unwrap all nodes with generated=wrapper, as well as nodes that
// have generated=empty and are empty.
if (
domElement.veInternal && (
domElement.veInternal.generated === 'wrapper' || (
domElement.veInternal.generated === 'empty' &&
domElement.childNodes.length === 0
)
)
) {
while ( domElement.firstChild ) {
parentDomElement.insertBefore(
domElement.firstChild,
domElement
);
}
// Transfer change markers
changed = domElement.getAttribute( 'data-ve-changed' );
if ( changed ) {
parentChanged = parentDomElement.getAttribute( 'data-ve-changed' );
if ( parentChanged ) {
changed = $.parseJSON( changed );
parentChanged = $.parseJSON( parentChanged );
for ( k in changed ) {
if ( k in parentChanged ) {
parentChanged[k] += changed[k];
} else {
parentChanged[k] = changed[k];
}
}
parentDomElement.setAttribute( 'data-ve-changed',
JSON.stringify( parentChanged ) );
} else {
parentDomElement.setAttribute( 'data-ve-changed',
changed );
}
}
parentDomElement.removeChild( domElement );
}
delete domElement.veInternal;
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
delete domElement.lastOuterPost;
// Ascend to parent node
domElement = parentDomElement;
} else {
// Create node from data
childDomElement = this.getDomElementFromDataElement( dataElement );
// Add reference to internal data
if ( dataElement.internal ) {
childDomElement.veInternal = dataElement.internal;
2012-08-10 21:09:04 +00:00
}
// Add element
domElement.appendChild( childDomElement );
// Descend into child node
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
parentDomElement = domElement;
domElement = childDomElement;
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Process outer whitespace
// Every piece of outer whitespace is duplicated somewhere:
// each node's outerPost is duplicated as the next node's
// outerPre, the first node's outerPre is the parent's
// innerPre, and the last node's outerPost is the parent's
// innerPost. For each piece of whitespace, we verify that
// the duplicate matches. If it doesn't, we take that to
// mean the user has messed with it and don't output any
// whitespace.
if ( domElement.veInternal && domElement.veInternal.whitespace ) {
// Process this node's outerPre
ours = domElement.veInternal.whitespace[0];
theirs = undefined;
if ( domElement.previousSibling ) {
// Get previous sibling's outerPost
theirs = parentDomElement.lastOuterPost;
} else if ( parentDomElement === container ) {
// outerPre of the very first node in the document, this one
// has no duplicate
theirs = ours;
} else {
// First child, get parent's innerPre
if (
parentDomElement.veInternal &&
parentDomElement.veInternal.whitespace
) {
theirs = parentDomElement.veInternal.whitespace[1];
// Clear after use so it's not used twice
parentDomElement.veInternal.whitespace[1] = undefined;
}
// else theirs=undefined
}
if ( ours && ours === theirs ) {
// Matches the duplicate, insert a TextNode
parentDomElement.insertBefore(
document.createTextNode( ours ),
domElement
);
}
}
}
}
}
Preserve whitespace between elements 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
2012-08-21 00:37:42 +00:00
// Process the outerPost whitespace of the very last node
if ( container.lastOuterPost !== undefined ) {
if ( container.lastChild && container.lastChild.nodeType === 3 ) {
// Last child is a TextNode, append to it
container.lastChild.appendData( container.lastOuterPost );
} else {
// Append a TextNode
container.appendChild( document.createTextNode( container.lastOuterPost ) );
}
delete container.lastOuterPost;
}
// Unwrap multi-child alien wrappers
$( container ).find( '[data-ve-multi-child-alien-wrapper]' ).each( function() {
$( this ).replaceWith( $( this ).contents() );
} );
// Workaround for bug 42469: if a <pre> starts with a newline, that means .innerHTML will
// screw up and stringify it with one fewer newline. Work around this by adding a newline.
// If we don't see a leading newline, we still don't know if the original HTML was
// <pre>Foo</pre> or <pre>\nFoo</pre> , but that's a syntactic difference, not a semantic
// one, and handling that is Parsoid's job.
$( container ).find( 'pre' ).each( function() {
var matches;
if ( this.firstChild.nodeType === Node.TEXT_NODE ) {
matches = this.firstChild.data.match( /^(\r\n|\r|\n)/ );
if ( matches && matches[1] ) {
// Prepend a newline exactly like the one we saw
this.firstChild.insertData( 0, matches[1] );
}
}
} );
return container;
};
/* Initialization */
ve.dm.converter = new ve.dm.Converter( ve.dm.nodeFactory, ve.dm.annotationFactory );