mozilla
/
gecko-dev
зеркало из https://github.com/mozilla/gecko-dev.git
1
0
Форкнуть 0
Код Задачи Пакеты Проекты Релизы Вики Активность
gecko-dev/editor/libeditor/WSRunObject.h

1523 строки
63 KiB
C++
Исходник Ответственный История

/* -*- Mode: C++; tab-width: 2; indent-tabs-mode: nil; c-basic-offset: 2 -*- */
/* This Source Code Form is subject to the terms of the Mozilla Public
* License, v. 2.0. If a copy of the MPL was not distributed with this
* file, You can obtain one at http://mozilla.org/MPL/2.0/. */
#ifndef WSRunObject_h
#define WSRunObject_h
#include "HTMLEditUtils.h"
#include "mozilla/Assertions.h"
#include "mozilla/EditAction.h"
#include "mozilla/EditorBase.h"
#include "mozilla/EditorDOMPoint.h" // for EditorDOMPoint
#include "mozilla/HTMLEditor.h"
#include "mozilla/Maybe.h"
#include "mozilla/Result.h"
#include "mozilla/dom/Element.h"
#include "mozilla/dom/HTMLBRElement.h"
#include "mozilla/dom/Text.h"
#include "nsCOMPtr.h"
#include "nsIContent.h"
namespace mozilla {
namespace dom {
class Element;
}
class WSRunScanner;
/**
* WSScanResult is result of ScanNextVisibleNodeOrBlockBoundaryFrom(),
* ScanPreviousVisibleNodeOrBlockBoundaryFrom(), and their static wrapper
* methods. This will have information of found visible content (and its
* position) or reached block element or topmost editable content at the
* start of scanner.
*/
class MOZ_STACK_CLASS WSScanResult final {
private:
enum class WSType : uint8_t {
NotInitialized,
// Could be the DOM tree is broken as like crash tests.
UnexpectedError,
// The run is maybe collapsible white-spaces at start of a hard line.
LeadingWhiteSpaces,
// The run is maybe collapsible white-spaces at end of a hard line.
TrailingWhiteSpaces,
// Normal (perhaps, meaning visible) white-spaces.
NormalWhiteSpaces,
// Normal text, not white-spaces.
NormalText,
// Special content such as `<img>`, etc.
SpecialContent,
// <br> element.
BRElement,
// Other block's boundary (child block of current block, maybe).
OtherBlockBoundary,
// Current block's boundary.
CurrentBlockBoundary,
};
friend class WSRunScanner; // Because of WSType.
public:
WSScanResult() = delete;
MOZ_NEVER_INLINE_DEBUG WSScanResult(nsIContent* aContent, WSType aReason)
: mContent(aContent), mReason(aReason) {
AssertIfInvalidData();
}
MOZ_NEVER_INLINE_DEBUG WSScanResult(const EditorDOMPoint& aPoint,
WSType aReason)
: mContent(aPoint.GetContainerAsContent()),
mOffset(Some(aPoint.Offset())),
mReason(aReason) {
AssertIfInvalidData();
}
MOZ_NEVER_INLINE_DEBUG void AssertIfInvalidData() const {
#ifdef DEBUG
MOZ_ASSERT(
mReason == WSType::UnexpectedError || mReason == WSType::NormalText ||
mReason == WSType::NormalWhiteSpaces || mReason == WSType::BRElement ||
mReason == WSType::SpecialContent ||
mReason == WSType::CurrentBlockBoundary ||
mReason == WSType::OtherBlockBoundary);
MOZ_ASSERT_IF(mReason == WSType::UnexpectedError, !mContent);
MOZ_ASSERT_IF(
mReason == WSType::NormalText || mReason == WSType::NormalWhiteSpaces,
mContent && mContent->IsText());
MOZ_ASSERT_IF(mReason == WSType::BRElement,
mContent && mContent->IsHTMLElement(nsGkAtoms::br));
MOZ_ASSERT_IF(
mReason == WSType::SpecialContent,
mContent && ((mContent->IsText() && !mContent->IsEditable()) ||
(!mContent->IsHTMLElement(nsGkAtoms::br) &&
!HTMLEditUtils::IsBlockElement(*mContent))));
MOZ_ASSERT_IF(mReason == WSType::OtherBlockBoundary,
mContent && HTMLEditUtils::IsBlockElement(*mContent));
// If mReason is WSType::CurrentBlockBoundary, mContent can be any content.
// In most cases, it's current block element which is editable. However, if
// there is no editable block parent, this is topmost editable inline
// content. Additionally, if there is no editable content, this is the
// container start of scanner and is not editable.
MOZ_ASSERT_IF(
mReason == WSType::CurrentBlockBoundary,
!mContent || !mContent->GetParentElement() ||
HTMLEditUtils::IsBlockElement(*mContent) ||
HTMLEditUtils::IsBlockElement(*mContent->GetParentElement()) ||
!mContent->GetParentElement()->IsEditable());
#endif // #ifdef DEBUG
}
bool Failed() const {
return mReason == WSType::NotInitialized ||
mReason == WSType::UnexpectedError;
}
/**
* GetContent() returns found visible and editable content/element.
* See MOZ_ASSERT_IF()s in AssertIfInvalidData() for the detail.
*/
nsIContent* GetContent() const { return mContent; }
/**
* The following accessors makes it easier to understand each callers.
*/
MOZ_NEVER_INLINE_DEBUG dom::Element* ElementPtr() const {
MOZ_DIAGNOSTIC_ASSERT(mContent->IsElement());
return mContent->AsElement();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* BRElementPtr() const {
MOZ_DIAGNOSTIC_ASSERT(mContent->IsHTMLElement(nsGkAtoms::br));
return static_cast<dom::HTMLBRElement*>(mContent.get());
}
MOZ_NEVER_INLINE_DEBUG dom::Text* TextPtr() const {
MOZ_DIAGNOSTIC_ASSERT(mContent->IsText());
return mContent->AsText();
}
/**
* Returns true if found or reached content is ediable.
*/
bool IsContentEditable() const { return mContent && mContent->IsEditable(); }
/**
* Offset() returns meaningful value only when InNormalWhiteSpacesOrText()
* returns true or the scanner reached to start or end of its scanning
* range and that is same as start or end container which are specified
* when the scanner is initialized. If it's result of scanning backward,
* this offset means before the found point. Otherwise, i.e., scanning
* forward, this offset means after the found point.
*/
MOZ_NEVER_INLINE_DEBUG uint32_t Offset() const {
NS_ASSERTION(mOffset.isSome(), "Retrieved non-meaningful offset");
return mOffset.valueOr(0);
}
/**
* Point() and RawPoint() return the position in found visible node or
* reached block boundary. So, they return meaningful point only when
* Offset() returns meaningful value.
*/
MOZ_NEVER_INLINE_DEBUG EditorDOMPoint Point() const {
NS_ASSERTION(mOffset.isSome(), "Retrieved non-meaningful point");
return EditorDOMPoint(mContent, mOffset.valueOr(0));
}
MOZ_NEVER_INLINE_DEBUG EditorRawDOMPoint RawPoint() const {
NS_ASSERTION(mOffset.isSome(), "Retrieved non-meaningful raw point");
return EditorRawDOMPoint(mContent, mOffset.valueOr(0));
}
/**
* PointAtContent() and RawPointAtContent() return the position of found
* visible content or reached block element.
*/
MOZ_NEVER_INLINE_DEBUG EditorDOMPoint PointAtContent() const {
MOZ_ASSERT(mContent);
return EditorDOMPoint(mContent);
}
MOZ_NEVER_INLINE_DEBUG EditorRawDOMPoint RawPointAtContent() const {
MOZ_ASSERT(mContent);
return EditorRawDOMPoint(mContent);
}
/**
* PointAfterContent() and RawPointAfterContent() retrun the position after
* found visible content or reached block element.
*/
MOZ_NEVER_INLINE_DEBUG EditorDOMPoint PointAfterContent() const {
MOZ_ASSERT(mContent);
return mContent ? EditorDOMPoint::After(mContent) : EditorDOMPoint();
}
MOZ_NEVER_INLINE_DEBUG EditorRawDOMPoint RawPointAfterContent() const {
MOZ_ASSERT(mContent);
return mContent ? EditorRawDOMPoint::After(mContent) : EditorRawDOMPoint();
}
/**
* The scanner reached <img> or something which is inline and is not a
* container.
*/
bool ReachedSpecialContent() const {
return mReason == WSType::SpecialContent;
}
/**
* The point is in normal white-spaces or text.
*/
bool InNormalWhiteSpacesOrText() const {
return mReason == WSType::NormalWhiteSpaces ||
mReason == WSType::NormalText;
}
/**
* The point is in normal white-spaces.
*/
bool InNormalWhiteSpaces() const {
return mReason == WSType::NormalWhiteSpaces;
}
/**
* The point is in normal text.
*/
bool InNormalText() const { return mReason == WSType::NormalText; }
/**
* The scanner reached a <br> element.
*/
bool ReachedBRElement() const { return mReason == WSType::BRElement; }
/**
* The scanner reached a <hr> element.
*/
bool ReachedHRElement() const {
return mContent && mContent->IsHTMLElement(nsGkAtoms::hr);
}
/**
* The scanner reached current block boundary or other block element.
*/
bool ReachedBlockBoundary() const {
return mReason == WSType::CurrentBlockBoundary ||
mReason == WSType::OtherBlockBoundary;
}
/**
* The scanner reached current block element boundary.
*/
bool ReachedCurrentBlockBoundary() const {
return mReason == WSType::CurrentBlockBoundary;
}
/**
* The scanner reached other block element.
*/
bool ReachedOtherBlockElement() const {
return mReason == WSType::OtherBlockBoundary;
}
/**
* The scanner reached something non-text node.
*/
bool ReachedSomething() const { return !InNormalWhiteSpacesOrText(); }
private:
nsCOMPtr<nsIContent> mContent;
Maybe<uint32_t> mOffset;
WSType mReason;
};
class WhiteSpaceVisibilityKeeper;
class MOZ_STACK_CLASS WSRunScanner final {
public:
using WSType = WSScanResult::WSType;
template <typename EditorDOMPointType>
WSRunScanner(const HTMLEditor& aHTMLEditor,
const EditorDOMPointType& aScanStartPoint)
: mScanStartPoint(aScanStartPoint),
mEditingHost(aHTMLEditor.GetActiveEditingHost()),
mHTMLEditor(&aHTMLEditor),
mTextFragmentDataAtStart(mScanStartPoint, mEditingHost) {}
// ScanNextVisibleNodeOrBlockBoundaryForwardFrom() returns the first visible
// node after aPoint. If there is no visible nodes after aPoint, returns
// topmost editable inline ancestor at end of current block. See comments
// around WSScanResult for the detail.
template <typename PT, typename CT>
WSScanResult ScanNextVisibleNodeOrBlockBoundaryFrom(
const EditorDOMPointBase<PT, CT>& aPoint) const;
template <typename PT, typename CT>
static WSScanResult ScanNextVisibleNodeOrBlockBoundary(
const HTMLEditor& aHTMLEditor, const EditorDOMPointBase<PT, CT>& aPoint) {
return WSRunScanner(aHTMLEditor, aPoint)
.ScanNextVisibleNodeOrBlockBoundaryFrom(aPoint);
}
// ScanPreviousVisibleNodeOrBlockBoundaryFrom() returns the first visible node
// before aPoint. If there is no visible nodes before aPoint, returns topmost
// editable inline ancestor at start of current block. See comments around
// WSScanResult for the detail.
template <typename PT, typename CT>
WSScanResult ScanPreviousVisibleNodeOrBlockBoundaryFrom(
const EditorDOMPointBase<PT, CT>& aPoint) const;
template <typename PT, typename CT>
static WSScanResult ScanPreviousVisibleNodeOrBlockBoundary(
const HTMLEditor& aHTMLEditor, const EditorDOMPointBase<PT, CT>& aPoint) {
return WSRunScanner(aHTMLEditor, aPoint)
.ScanPreviousVisibleNodeOrBlockBoundaryFrom(aPoint);
}
/**
* GetInclusiveNextEditableCharPoint() returns a point in a text node which
* is at current editable character or next editable character if aPoint
* does not points an editable character.
*/
template <typename PT, typename CT>
static EditorDOMPointInText GetInclusiveNextEditableCharPoint(
const HTMLEditor& aHTMLEditor, const EditorDOMPointBase<PT, CT>& aPoint) {
if (aPoint.IsInTextNode() && !aPoint.IsEndOfContainer() &&
HTMLEditUtils::IsSimplyEditableNode(*aPoint.ContainerAsText())) {
return EditorDOMPointInText(aPoint.ContainerAsText(), aPoint.Offset());
}
WSRunScanner scanner(aHTMLEditor, aPoint);
return scanner.GetInclusiveNextEditableCharPoint(aPoint);
}
/**
* GetPreviousEditableCharPoint() returns a point in a text node which
* is at previous editable character.
*/
template <typename PT, typename CT>
static EditorDOMPointInText GetPreviousEditableCharPoint(
const HTMLEditor& aHTMLEditor, const EditorDOMPointBase<PT, CT>& aPoint) {
if (aPoint.IsInTextNode() && !aPoint.IsStartOfContainer() &&
HTMLEditUtils::IsSimplyEditableNode(*aPoint.ContainerAsText())) {
return EditorDOMPointInText(aPoint.ContainerAsText(),
aPoint.Offset() - 1);
}
WSRunScanner scanner(aHTMLEditor, aPoint);
return scanner.GetPreviousEditableCharPoint(aPoint);
}
/**
* Scan aTextNode from end or start to find last or first visible things.
* I.e., this returns a point immediately before or after invisible
* white-spaces of aTextNode if aTextNode ends or begins with some invisible
* white-spaces.
* Note that the result may not be in different text node if aTextNode has
* only invisible white-spaces and there is previous or next text node.
*/
template <typename EditorDOMPointType>
static EditorDOMPointType GetAfterLastVisiblePoint(
dom::Text& aTextNode, const dom::Element* aAncestorLimiter);
template <typename EditorDOMPointType>
static EditorDOMPointType GetFirstVisiblePoint(
dom::Text& aTextNode, const dom::Element* aAncestorLimiter);
/**
* GetRangeInTextNodesToForwardDeleteFrom() returns the range to remove
* text when caret is at aPoint.
*/
static Result<EditorDOMRangeInTexts, nsresult>
GetRangeInTextNodesToForwardDeleteFrom(const HTMLEditor& aHTMLEditor,
const EditorDOMPoint& aPoint);
/**
* GetRangeInTextNodesToBackspaceFrom() returns the range to remove text
* when caret is at aPoint.
*/
static Result<EditorDOMRangeInTexts, nsresult>
GetRangeInTextNodesToBackspaceFrom(const HTMLEditor& aHTMLEditor,
const EditorDOMPoint& aPoint);
/**
* GetRangesForDeletingAtomicContent() returns the range to delete
* aAtomicContent. If it's followed by invisible white-spaces, they will
* be included into the range.
*/
static EditorDOMRange GetRangesForDeletingAtomicContent(
const HTMLEditor& aHTMLEditor, const nsIContent& aAtomicContent);
/**
* GetRangeForDeleteBlockElementBoundaries() returns a range starting from end
* of aLeftBlockElement to start of aRightBlockElement and extend invisible
* white-spaces around them.
*
* @param aHTMLEditor The HTML editor.
* @param aLeftBlockElement The block element which will be joined with
* aRightBlockElement.
* @param aRightBlockElement The block element which will be joined with
* aLeftBlockElement. This must be an element
* after aLeftBlockElement.
* @param aPointContainingTheOtherBlock
* When aRightBlockElement is an ancestor of
* aLeftBlockElement, this must be set and the
* container must be aRightBlockElement.
* When aLeftBlockElement is an ancestor of
* aRightBlockElement, this must be set and the
* container must be aLeftBlockElement.
* Otherwise, must not be set.
*/
static EditorDOMRange GetRangeForDeletingBlockElementBoundaries(
const HTMLEditor& aHTMLEditor, const dom::Element& aLeftBlockElement,
const dom::Element& aRightBlockElement,
const EditorDOMPoint& aPointContainingTheOtherBlock);
/**
* ShrinkRangeIfStartsFromOrEndsAfterAtomicContent() may shrink aRange if it
* starts and/or ends with an atomic content, but the range boundary
* is in adjacent text nodes. Returns true if this modifies the range.
*/
static Result<bool, nsresult> ShrinkRangeIfStartsFromOrEndsAfterAtomicContent(
const HTMLEditor& aHTMLEditor, nsRange& aRange,
const dom::Element* aEditingHost);
/**
* GetRangeContainingInvisibleWhiteSpacesAtRangeBoundaries() returns
* extended range if range boundaries of aRange are in invisible white-spaces.
*/
static EditorDOMRange GetRangeContainingInvisibleWhiteSpacesAtRangeBoundaries(
const HTMLEditor& aHTMLEditor, const EditorDOMRange& aRange);
/**
* GetPrecedingBRElementUnlessVisibleContentFound() scans a `<br>` element
* backward, but stops scanning it if the scanner finds visible character
* or something. In other words, this method ignores only invisible
* white-spaces between `<br>` element and aPoint.
*/
template <typename EditorDOMPointType>
MOZ_NEVER_INLINE_DEBUG static dom::HTMLBRElement*
GetPrecedingBRElementUnlessVisibleContentFound(
const HTMLEditor& aHTMLEditor, const EditorDOMPointType& aPoint) {
MOZ_ASSERT(aPoint.IsSetAndValid());
// XXX This method behaves differently even in similar point.
// If aPoint is in a text node following `<br>` element, reaches the
// `<br>` element when all characters between the `<br>` and
// aPoint are ASCII whitespaces.
// But if aPoint is not in a text node, e.g., at start of an inline
// element which is immediately after a `<br>` element, returns the
// `<br>` element even if there is no invisible white-spaces.
if (aPoint.IsStartOfContainer()) {
return nullptr;
}
// TODO: Scan for end boundary is redundant in this case, we should optimize
// it.
TextFragmentData textFragmentData(aPoint,
aHTMLEditor.GetActiveEditingHost());
return textFragmentData.StartsFromBRElement()
? textFragmentData.StartReasonBRElementPtr()
: nullptr;
}
const EditorDOMPoint& ScanStartRef() const { return mScanStartPoint; }
/**
* GetStartReasonContent() and GetEndReasonContent() return a node which
* was found by scanning from mScanStartPoint backward or forward. If there
* was white-spaces or text from the point, returns the text node. Otherwise,
* returns an element which is explained by the following methods. Note that
* when the reason is WSType::CurrentBlockBoundary, In most cases, it's
* current block element which is editable, but also may be non-element and/or
* non-editable. See MOZ_ASSERT_IF()s in WSScanResult::AssertIfInvalidData()
* for the detail.
*/
nsIContent* GetStartReasonContent() const {
return TextFragmentDataAtStartRef().GetStartReasonContent();
}
nsIContent* GetEndReasonContent() const {
return TextFragmentDataAtStartRef().GetEndReasonContent();
}
bool StartsFromNormalText() const {
return TextFragmentDataAtStartRef().StartsFromNormalText();
}
bool StartsFromSpecialContent() const {
return TextFragmentDataAtStartRef().StartsFromSpecialContent();
}
bool StartsFromBRElement() const {
return TextFragmentDataAtStartRef().StartsFromBRElement();
}
bool StartsFromCurrentBlockBoundary() const {
return TextFragmentDataAtStartRef().StartsFromCurrentBlockBoundary();
}
bool StartsFromOtherBlockElement() const {
return TextFragmentDataAtStartRef().StartsFromOtherBlockElement();
}
bool StartsFromBlockBoundary() const {
return TextFragmentDataAtStartRef().StartsFromBlockBoundary();
}
bool StartsFromHardLineBreak() const {
return TextFragmentDataAtStartRef().StartsFromHardLineBreak();
}
bool EndsByNormalText() const {
return TextFragmentDataAtStartRef().EndsByNormalText();
}
bool EndsBySpecialContent() const {
return TextFragmentDataAtStartRef().EndsBySpecialContent();
}
bool EndsByBRElement() const {
return TextFragmentDataAtStartRef().EndsByBRElement();
}
bool EndsByCurrentBlockBoundary() const {
return TextFragmentDataAtStartRef().EndsByCurrentBlockBoundary();
}
bool EndsByOtherBlockElement() const {
return TextFragmentDataAtStartRef().EndsByOtherBlockElement();
}
bool EndsByBlockBoundary() const {
return TextFragmentDataAtStartRef().EndsByBlockBoundary();
}
MOZ_NEVER_INLINE_DEBUG dom::Element* StartReasonOtherBlockElementPtr() const {
return TextFragmentDataAtStartRef().StartReasonOtherBlockElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* StartReasonBRElementPtr() const {
return TextFragmentDataAtStartRef().StartReasonBRElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::Element* EndReasonOtherBlockElementPtr() const {
return TextFragmentDataAtStartRef().EndReasonOtherBlockElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* EndReasonBRElementPtr() const {
return TextFragmentDataAtStartRef().EndReasonBRElementPtr();
}
/**
* Active editing host when this instance is created.
*/
dom::Element* GetEditingHost() const { return mEditingHost; }
protected:
using EditorType = EditorBase::EditorType;
class TextFragmentData;
// VisibleWhiteSpacesData represents 0 or more visible white-spaces.
class MOZ_STACK_CLASS VisibleWhiteSpacesData final {
public:
bool IsInitialized() const {
return mLeftWSType != WSType::NotInitialized ||
mRightWSType != WSType::NotInitialized;
}
EditorDOMPoint StartRef() const { return mStartPoint; }
EditorDOMPoint EndRef() const { return mEndPoint; }
/**
* Information why the white-spaces start from (i.e., this indicates the
* previous content type of the fragment).
*/
bool StartsFromNormalText() const {
return mLeftWSType == WSType::NormalText;
}
bool StartsFromSpecialContent() const {
return mLeftWSType == WSType::SpecialContent;
}
/**
* Information why the white-spaces end by (i.e., this indicates the
* next content type of the fragment).
*/
bool EndsByNormalText() const { return mRightWSType == WSType::NormalText; }
bool EndsByTrailingWhiteSpaces() const {
return mRightWSType == WSType::TrailingWhiteSpaces;
}
bool EndsBySpecialContent() const {
return mRightWSType == WSType::SpecialContent;
}
bool EndsByBRElement() const { return mRightWSType == WSType::BRElement; }
bool EndsByBlockBoundary() const {
return mRightWSType == WSType::CurrentBlockBoundary ||
mRightWSType == WSType::OtherBlockBoundary;
}
/**
* ComparePoint() compares aPoint with the white-spaces.
*/
enum class PointPosition {
BeforeStartOfFragment,
StartOfFragment,
MiddleOfFragment,
EndOfFragment,
AfterEndOfFragment,
NotInSameDOMTree,
};
template <typename EditorDOMPointType>
PointPosition ComparePoint(const EditorDOMPointType& aPoint) const {
MOZ_ASSERT(aPoint.IsSetAndValid());
if (StartRef() == aPoint) {
return PointPosition::StartOfFragment;
}
if (EndRef() == aPoint) {
return PointPosition::EndOfFragment;
}
const bool startIsBeforePoint = StartRef().IsBefore(aPoint);
const bool pointIsBeforeEnd = aPoint.IsBefore(EndRef());
if (startIsBeforePoint && pointIsBeforeEnd) {
return PointPosition::MiddleOfFragment;
}
if (startIsBeforePoint) {
return PointPosition::AfterEndOfFragment;
}
if (pointIsBeforeEnd) {
return PointPosition::BeforeStartOfFragment;
}
return PointPosition::NotInSameDOMTree;
}
private:
// Initializers should be accessible only from `TextFragmentData`.
friend class WSRunScanner::TextFragmentData;
VisibleWhiteSpacesData()
: mLeftWSType(WSType::NotInitialized),
mRightWSType(WSType::NotInitialized) {}
template <typename EditorDOMPointType>
void SetStartPoint(const EditorDOMPointType& aStartPoint) {
mStartPoint = aStartPoint;
}
template <typename EditorDOMPointType>
void SetEndPoint(const EditorDOMPointType& aEndPoint) {
mEndPoint = aEndPoint;
}
void SetStartFrom(WSType aLeftWSType) { mLeftWSType = aLeftWSType; }
void SetStartFromLeadingWhiteSpaces() {
mLeftWSType = WSType::LeadingWhiteSpaces;
}
void SetStartFromNormalWhiteSpaces() {
mLeftWSType = WSType::NormalWhiteSpaces;
}
void SetEndBy(WSType aRightWSType) { mRightWSType = aRightWSType; }
void SetEndByNormalWiteSpaces() {
mRightWSType = WSType::NormalWhiteSpaces;
}
void SetEndByTrailingWhiteSpaces() {
mRightWSType = WSType::TrailingWhiteSpaces;
}
EditorDOMPoint mStartPoint;
EditorDOMPoint mEndPoint;
WSType mLeftWSType, mRightWSType;
};
using PointPosition = VisibleWhiteSpacesData::PointPosition;
/**
* GetInclusiveNextEditableCharPoint() returns aPoint if it points a character
* in an editable text node, or start of next editable text node otherwise.
* FYI: For the performance, this does not check whether given container
* is not after mStart.mReasonContent or not.
*/
template <typename PT, typename CT>
EditorDOMPointInText GetInclusiveNextEditableCharPoint(
const EditorDOMPointBase<PT, CT>& aPoint) const {
return TextFragmentDataAtStartRef().GetInclusiveNextEditableCharPoint(
aPoint);
}
/**
* GetPreviousEditableCharPoint() returns previous editable point in a
* text node. Note that this returns last character point when it meets
* non-empty text node, otherwise, returns a point in an empty text node.
* FYI: For the performance, this does not check whether given container
* is not before mEnd.mReasonContent or not.
*/
template <typename PT, typename CT>
EditorDOMPointInText GetPreviousEditableCharPoint(
const EditorDOMPointBase<PT, CT>& aPoint) const {
return TextFragmentDataAtStartRef().GetPreviousEditableCharPoint(aPoint);
}
/**
* GetEndOfCollapsibleASCIIWhiteSpaces() returns the next visible char
* (meaning a character except ASCII white-spaces) point or end of last text
* node scanning from aPointAtASCIIWhiteSpace.
* Note that this may return different text node from the container of
* aPointAtASCIIWhiteSpace.
*/
EditorDOMPointInText GetEndOfCollapsibleASCIIWhiteSpaces(
const EditorDOMPointInText& aPointAtASCIIWhiteSpace) const {
return TextFragmentDataAtStartRef().GetEndOfCollapsibleASCIIWhiteSpaces(
aPointAtASCIIWhiteSpace);
}
/**
* GetFirstASCIIWhiteSpacePointCollapsedTo() returns the first ASCII
* white-space which aPointAtASCIIWhiteSpace belongs to. In other words,
* the white-space at aPointAtASCIIWhiteSpace should be collapsed into
* the result.
* Note that this may return different text node from the container of
* aPointAtASCIIWhiteSpace.
*/
EditorDOMPointInText GetFirstASCIIWhiteSpacePointCollapsedTo(
const EditorDOMPointInText& aPointAtASCIIWhiteSpace) const {
return TextFragmentDataAtStartRef().GetFirstASCIIWhiteSpacePointCollapsedTo(
aPointAtASCIIWhiteSpace);
}
EditorDOMPointInText GetPreviousCharPointFromPointInText(
const EditorDOMPointInText& aPoint) const;
char16_t GetCharAt(dom::Text* aTextNode, int32_t aOffset) const;
/**
* TextFragmentData stores the information of white-space sequence which
* contains `aPoint` of the constructor.
*/
class MOZ_STACK_CLASS TextFragmentData final {
private:
class NoBreakingSpaceData;
class MOZ_STACK_CLASS BoundaryData final {
public:
using NoBreakingSpaceData =
WSRunScanner::TextFragmentData::NoBreakingSpaceData;
/**
* ScanCollapsibleWhiteSpaceStartFrom() returns start boundary data of
* white-spaces containing aPoint. When aPoint is in a text node and
* points a non-white-space character or the text node is preformatted,
* this returns the data at aPoint.
*
* @param aPoint Scan start point.
* @param aEditableBlockParentOrTopmostEditableInlineContent
* Nearest editable block parent element of
* aPoint if there is. Otherwise, inline editing host.
* @param aEditingHost Active editing host.
* @param aNBSPData Optional. If set, this recodes first and last
* NBSP positions.
*/
template <typename EditorDOMPointType>
static BoundaryData ScanCollapsibleWhiteSpaceStartFrom(
const EditorDOMPointType& aPoint,
const nsIContent& aEditableBlockParentOrTopmostEditableInlineContent,
const dom::Element* aEditingHost, NoBreakingSpaceData* aNBSPData);
/**
* ScanCollapsibleWhiteSpaceEndFrom() returns end boundary data of
* white-spaces containing aPoint. When aPoint is in a text node and
* points a non-white-space character or the text node is preformatted,
* this returns the data at aPoint.
*
* @param aPoint Scan start point.
* @param aEditableBlockParentOrTopmostEditableInlineContent
* Nearest editable block parent element of
* aPoint if there is. Otherwise, inline editing host.
* @param aEditingHost Active editing host.
* @param aNBSPData Optional. If set, this recodes first and last
* NBSP positions.
*/
template <typename EditorDOMPointType>
static BoundaryData ScanCollapsibleWhiteSpaceEndFrom(
const EditorDOMPointType& aPoint,
const nsIContent& aEditableBlockParentOrTopmostEditableInlineContent,
const dom::Element* aEditingHost, NoBreakingSpaceData* aNBSPData);
enum class Preformatted : bool { Yes, No };
BoundaryData()
: mReason(WSType::NotInitialized),
mAcrossPreformattedCharacter(Preformatted::No) {}
template <typename EditorDOMPointType>
BoundaryData(const EditorDOMPointType& aPoint, nsIContent& aReasonContent,
WSType aReason, Preformatted aDidCrossPreformattedCharacter)
: mReasonContent(&aReasonContent),
mPoint(aPoint),
mReason(aReason),
mAcrossPreformattedCharacter(aDidCrossPreformattedCharacter) {}
bool Initialized() const { return mReasonContent && mPoint.IsSet(); }
nsIContent* GetReasonContent() const { return mReasonContent; }
const EditorDOMPoint& PointRef() const { return mPoint; }
WSType RawReason() const { return mReason; }
bool AcrossPreformattedCharacter() const {
return mAcrossPreformattedCharacter == Preformatted::Yes;
}
bool IsNormalText() const { return mReason == WSType::NormalText; }
bool IsSpecialContent() const {
return mReason == WSType::SpecialContent;
}
bool IsBRElement() const { return mReason == WSType::BRElement; }
bool IsCurrentBlockBoundary() const {
return mReason == WSType::CurrentBlockBoundary;
}
bool IsOtherBlockBoundary() const {
return mReason == WSType::OtherBlockBoundary;
}
bool IsBlockBoundary() const {
return mReason == WSType::CurrentBlockBoundary ||
mReason == WSType::OtherBlockBoundary;
}
bool IsHardLineBreak() const {
return mReason == WSType::CurrentBlockBoundary ||
mReason == WSType::OtherBlockBoundary ||
mReason == WSType::BRElement;
}
MOZ_NEVER_INLINE_DEBUG dom::Element* OtherBlockElementPtr() const {
MOZ_DIAGNOSTIC_ASSERT(mReasonContent->IsElement());
return mReasonContent->AsElement();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* BRElementPtr() const {
MOZ_DIAGNOSTIC_ASSERT(mReasonContent->IsHTMLElement(nsGkAtoms::br));
return static_cast<dom::HTMLBRElement*>(mReasonContent.get());
}
private:
/**
* Helper methods of ScanCollapsibleWhiteSpaceStartFrom() and
* ScanCollapsibleWhiteSpaceEndFrom() when they need to scan in a text
* node.
*/
template <typename EditorDOMPointType>
static Maybe<BoundaryData> ScanCollapsibleWhiteSpaceStartInTextNode(
const EditorDOMPointType& aPoint, NoBreakingSpaceData* aNBSPData);
template <typename EditorDOMPointType>
static Maybe<BoundaryData> ScanCollapsibleWhiteSpaceEndInTextNode(
const EditorDOMPointType& aPoint, NoBreakingSpaceData* aNBSPData);
nsCOMPtr<nsIContent> mReasonContent;
EditorDOMPoint mPoint;
// Must be one of WSType::NotInitialized, WSType::NormalText,
// WSType::SpecialContent, WSType::BRElement, WSType::CurrentBlockBoundary
// or WSType::OtherBlockBoundary.
WSType mReason;
// If the point crosses a preformatted character from scanning start
// point, set to "Yes". So, this may NOT equal to the style at mPoint
// nor mReasonContent.
Preformatted mAcrossPreformattedCharacter;
};
class MOZ_STACK_CLASS NoBreakingSpaceData final {
public:
enum class Scanning { Forward, Backward };
void NotifyNBSP(const EditorDOMPointInText& aPoint,
Scanning aScanningDirection) {
MOZ_ASSERT(aPoint.IsSetAndValid());
MOZ_ASSERT(aPoint.IsCharNBSP());
if (!mFirst.IsSet() || aScanningDirection == Scanning::Backward) {
mFirst = aPoint;
}
if (!mLast.IsSet() || aScanningDirection == Scanning::Forward) {
mLast = aPoint;
}
}
const EditorDOMPointInText& FirstPointRef() const { return mFirst; }
const EditorDOMPointInText& LastPointRef() const { return mLast; }
bool FoundNBSP() const {
MOZ_ASSERT(mFirst.IsSet() == mLast.IsSet());
return mFirst.IsSet();
}
private:
EditorDOMPointInText mFirst;
EditorDOMPointInText mLast;
};
public:
TextFragmentData() = delete;
template <typename EditorDOMPointType>
TextFragmentData(const EditorDOMPointType& aPoint,
const dom::Element* aEditingHost);
bool IsInitialized() const {
return mStart.Initialized() && mEnd.Initialized();
}
nsIContent* GetStartReasonContent() const {
return mStart.GetReasonContent();
}
nsIContent* GetEndReasonContent() const { return mEnd.GetReasonContent(); }
bool StartsFromNormalText() const { return mStart.IsNormalText(); }
bool StartsFromSpecialContent() const { return mStart.IsSpecialContent(); }
bool StartsFromBRElement() const { return mStart.IsBRElement(); }
bool StartsFromCurrentBlockBoundary() const {
return mStart.IsCurrentBlockBoundary();
}
bool StartsFromOtherBlockElement() const {
return mStart.IsOtherBlockBoundary();
}
bool StartsFromBlockBoundary() const { return mStart.IsBlockBoundary(); }
bool StartsFromHardLineBreak() const { return mStart.IsHardLineBreak(); }
bool EndsByNormalText() const { return mEnd.IsNormalText(); }
bool EndsBySpecialContent() const { return mEnd.IsSpecialContent(); }
bool EndsByBRElement() const { return mEnd.IsBRElement(); }
bool EndsByCurrentBlockBoundary() const {
return mEnd.IsCurrentBlockBoundary();
}
bool EndsByOtherBlockElement() const { return mEnd.IsOtherBlockBoundary(); }
bool EndsByBlockBoundary() const { return mEnd.IsBlockBoundary(); }
WSType StartRawReason() const { return mStart.RawReason(); }
WSType EndRawReason() const { return mEnd.RawReason(); }
MOZ_NEVER_INLINE_DEBUG dom::Element* StartReasonOtherBlockElementPtr()
const {
return mStart.OtherBlockElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* StartReasonBRElementPtr() const {
return mStart.BRElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::Element* EndReasonOtherBlockElementPtr() const {
return mEnd.OtherBlockElementPtr();
}
MOZ_NEVER_INLINE_DEBUG dom::HTMLBRElement* EndReasonBRElementPtr() const {
return mEnd.BRElementPtr();
}
const EditorDOMPoint& StartRef() const { return mStart.PointRef(); }
const EditorDOMPoint& EndRef() const { return mEnd.PointRef(); }
const EditorDOMPoint& ScanStartRef() const { return mScanStartPoint; }
bool FoundNoBreakingWhiteSpaces() const { return mNBSPData.FoundNBSP(); }
const EditorDOMPointInText& FirstNBSPPointRef() const {
return mNBSPData.FirstPointRef();
}
const EditorDOMPointInText& LastNBSPPointRef() const {
return mNBSPData.LastPointRef();
}
bool IsPreformatted() const { return mIsPreformatted; }
template <typename PT, typename CT>
EditorDOMPointInText GetInclusiveNextEditableCharPoint(
const EditorDOMPointBase<PT, CT>& aPoint) const;
template <typename PT, typename CT>
EditorDOMPointInText GetPreviousEditableCharPoint(
const EditorDOMPointBase<PT, CT>& aPoint) const;
EditorDOMPointInText GetEndOfCollapsibleASCIIWhiteSpaces(
const EditorDOMPointInText& aPointAtASCIIWhiteSpace) const;
EditorDOMPointInText GetFirstASCIIWhiteSpacePointCollapsedTo(
const EditorDOMPointInText& aPointAtASCIIWhiteSpace) const;
/**
* GetNonCollapsedRangeInTexts() returns non-empty range in texts which
* is the largest range in aRange if there is some text nodes.
*/
EditorDOMRangeInTexts GetNonCollapsedRangeInTexts(
const EditorDOMRange& aRange) const;
/**
* InvisibleLeadingWhiteSpaceRangeRef() retruns reference to two DOM points,
* start of the line and first visible point or end of the hard line. When
* this returns non-positioned range or positioned but collapsed range,
* there is no invisible leading white-spaces.
* Note that if there are only invisible white-spaces in a hard line,
* this returns all of the white-spaces.
*/
const EditorDOMRange& InvisibleLeadingWhiteSpaceRangeRef() const;
/**
* InvisibleTrailingWhiteSpaceRangeRef() returns reference to two DOM
* points, first invisible white-space and end of the hard line. When this
* returns non-positioned range or positioned but collapsed range,
* there is no invisible trailing white-spaces.
* Note that if there are only invisible white-spaces in a hard line,
* this returns all of the white-spaces.
*/
const EditorDOMRange& InvisibleTrailingWhiteSpaceRangeRef() const;
/**
* GetNewInvisibleLeadingWhiteSpaceRangeIfSplittingAt() returns new
* invisible leading white-space range which should be removed if
* splitting invisible white-space sequence at aPointToSplit creates
* new invisible leading white-spaces in the new line.
* Note that the result may be collapsed range if the point is around
* invisible white-spaces.
*/
template <typename EditorDOMPointType>
EditorDOMRange GetNewInvisibleLeadingWhiteSpaceRangeIfSplittingAt(
const EditorDOMPointType& aPointToSplit) const {
// If there are invisible trailing white-spaces and some or all of them
// become invisible leading white-spaces in the new line, although we
// don't need to delete them, but for aesthetically and backward
// compatibility, we should remove them.
const EditorDOMRange& trailingWhiteSpaceRange =
InvisibleTrailingWhiteSpaceRangeRef();
// XXX Why don't we check leading white-spaces too?
if (!trailingWhiteSpaceRange.IsPositioned()) {
return trailingWhiteSpaceRange;
}
// If the point is before the trailing white-spaces, the new line won't
// start with leading white-spaces.
if (aPointToSplit.IsBefore(trailingWhiteSpaceRange.StartRef())) {
return EditorDOMRange();
}
// If the point is in the trailing white-spaces, the new line may
// start with some leading white-spaces. Returning collapsed range
// is intentional because the caller may want to know whether the
// point is in trailing white-spaces or not.
if (aPointToSplit.EqualsOrIsBefore(trailingWhiteSpaceRange.EndRef())) {
return EditorDOMRange(trailingWhiteSpaceRange.StartRef(),
aPointToSplit);
}
// Otherwise, if the point is after the trailing white-spaces, it may
// be just outside of the text node. E.g., end of parent element.
// This is possible case but the validation cost is not worthwhile
// due to the runtime cost in the worst case. Therefore, we should just
// return collapsed range at the end of trailing white-spaces. Then,
// callers can know the point is immediately after the trailing
// white-spaces.
return EditorDOMRange(trailingWhiteSpaceRange.EndRef());
}
/**
* GetNewInvisibleTrailingWhiteSpaceRangeIfSplittingAt() returns new
* invisible trailing white-space range which should be removed if
* splitting invisible white-space sequence at aPointToSplit creates
* new invisible trailing white-spaces in the new line.
* Note that the result may be collapsed range if the point is around
* invisible white-spaces.
*/
template <typename EditorDOMPointType>
EditorDOMRange GetNewInvisibleTrailingWhiteSpaceRangeIfSplittingAt(
const EditorDOMPointType& aPointToSplit) const {
// If there are invisible leading white-spaces and some or all of them
// become end of current line, they will become visible. Therefore, we
// need to delete the invisible leading white-spaces before insertion
// point.
const EditorDOMRange& leadingWhiteSpaceRange =
InvisibleLeadingWhiteSpaceRangeRef();
if (!leadingWhiteSpaceRange.IsPositioned()) {
return leadingWhiteSpaceRange;
}
// If the point equals or is after the leading white-spaces, the line
// will end without trailing white-spaces.
if (leadingWhiteSpaceRange.EndRef().IsBefore(aPointToSplit)) {
return EditorDOMRange();
}
// If the point is in the leading white-spaces, the line may
// end with some trailing white-spaces. Returning collapsed range
// is intentional because the caller may want to know whether the
// point is in leading white-spaces or not.
if (leadingWhiteSpaceRange.StartRef().EqualsOrIsBefore(aPointToSplit)) {
return EditorDOMRange(aPointToSplit, leadingWhiteSpaceRange.EndRef());
}
// Otherwise, if the point is before the leading white-spaces, it may
// be just outside of the text node. E.g., start of parent element.
// This is possible case but the validation cost is not worthwhile
// due to the runtime cost in the worst case. Therefore, we should
// just return collapsed range at start of the leading white-spaces.
// Then, callers can know the point is immediately before the leading
// white-spaces.
return EditorDOMRange(leadingWhiteSpaceRange.StartRef());
}
/**
* FollowingContentMayBecomeFirstVisibleContent() returns true if some
* content may be first visible content after removing content after aPoint.
* Note that it's completely broken what this does. Don't use this method
* with new code.
*/
template <typename EditorDOMPointType>
bool FollowingContentMayBecomeFirstVisibleContent(
const EditorDOMPointType& aPoint) const {
MOZ_ASSERT(aPoint.IsSetAndValid());
if (!mStart.IsHardLineBreak()) {
return false;
}
// If the point is before start of text fragment, that means that the
// point may be at the block boundary or inline element boundary.
if (aPoint.EqualsOrIsBefore(mStart.PointRef())) {
return true;
}
// VisibleWhiteSpacesData is marked as start of line only when it
// represents leading white-spaces.
const EditorDOMRange& leadingWhiteSpaceRange =
InvisibleLeadingWhiteSpaceRangeRef();
if (!leadingWhiteSpaceRange.StartRef().IsSet()) {
return false;
}
if (aPoint.EqualsOrIsBefore(leadingWhiteSpaceRange.StartRef())) {
return true;
}
if (!leadingWhiteSpaceRange.EndRef().IsSet()) {
return false;
}
return aPoint.EqualsOrIsBefore(leadingWhiteSpaceRange.EndRef());
}
/**
* PrecedingContentMayBecomeInvisible() returns true if end of preceding
* content is collapsed (when ends with an ASCII white-space).
* Note that it's completely broken what this does. Don't use this method
* with new code.
*/
template <typename EditorDOMPointType>
bool PrecedingContentMayBecomeInvisible(
const EditorDOMPointType& aPoint) const {
MOZ_ASSERT(aPoint.IsSetAndValid());
// If this fragment is ends by block boundary, always the caller needs
// additional check.
if (mEnd.IsBlockBoundary()) {
return true;
}
// If the point is in visible white-spaces and ends with an ASCII
// white-space, it may be collapsed even if it won't be end of line.
const VisibleWhiteSpacesData& visibleWhiteSpaces =
VisibleWhiteSpacesDataRef();
if (!visibleWhiteSpaces.IsInitialized()) {
return false;
}
// XXX Odd case, but keep traditional behavior of `FindNearestRun()`.
if (!visibleWhiteSpaces.StartRef().IsSet()) {
return true;
}
if (!visibleWhiteSpaces.StartRef().EqualsOrIsBefore(aPoint)) {
return false;
}
// XXX Odd case, but keep traditional behavior of `FindNearestRun()`.
if (visibleWhiteSpaces.EndsByTrailingWhiteSpaces()) {
return true;
}
// XXX Must be a bug. This claims that the caller needs additional
// check even when there is no white-spaces.
if (visibleWhiteSpaces.StartRef() == visibleWhiteSpaces.EndRef()) {
return true;
}
return aPoint.IsBefore(visibleWhiteSpaces.EndRef());
}
/**
* GetPreviousNBSPPointIfNeedToReplaceWithASCIIWhiteSpace() may return an
* NBSP point which should be replaced with an ASCII white-space when we're
* inserting text into aPointToInsert. Note that this is a helper method for
* the traditional white-space normalizer. Don't use this with the new
* white-space normalizer.
* Must be called only when VisibleWhiteSpacesDataRef() returns initialized
* instance and previous character of aPointToInsert is in the range.
*/
EditorDOMPointInText GetPreviousNBSPPointIfNeedToReplaceWithASCIIWhiteSpace(
const EditorDOMPoint& aPointToInsert) const;
/**
* GetInclusiveNextNBSPPointIfNeedToReplaceWithASCIIWhiteSpace() may return
* an NBSP point which should be replaced with an ASCII white-space when
* the caller inserts text into aPointToInsert.
* Note that this is a helper method for the traditional white-space
* normalizer. Don't use this with the new white-space normalizer.
* Must be called only when VisibleWhiteSpacesDataRef() returns initialized
* instance, and inclusive next char of aPointToInsert is in the range.
*/
EditorDOMPointInText
GetInclusiveNextNBSPPointIfNeedToReplaceWithASCIIWhiteSpace(
const EditorDOMPoint& aPointToInsert) const;
/**
* GetReplaceRangeDataAtEndOfDeletionRange() and
* GetReplaceRangeDataAtStartOfDeletionRange() return delete range if
* end or start of deleting range splits invisible trailing/leading
* white-spaces and it may become visible, or return replace range if
* end or start of deleting range splits visible white-spaces and it
* causes some ASCII white-spaces become invisible unless replacing
* with an NBSP.
*/
ReplaceRangeData GetReplaceRangeDataAtEndOfDeletionRange(
const TextFragmentData& aTextFragmentDataAtStartToDelete) const;
ReplaceRangeData GetReplaceRangeDataAtStartOfDeletionRange(
const TextFragmentData& aTextFragmentDataAtEndToDelete) const;
/**
* VisibleWhiteSpacesDataRef() returns reference to visible white-spaces
* data. That is zero or more white-spaces which are visible.
* Note that when there is no visible content, it's not initialized.
* Otherwise, even if there is no white-spaces, it's initialized and
* the range is collapsed in such case.
*/
const VisibleWhiteSpacesData& VisibleWhiteSpacesDataRef() const;
private:
/**
* IsPreformattedOrSurrondedByVisibleContent() returns true if the text is
* preformatted or the text fragment is surrounded by visible content.
* When this returns true, all of the text is visible.
*/
bool IsPreformattedOrSurrondedByVisibleContent() const {
return mIsPreformatted ||
((StartsFromNormalText() || StartsFromSpecialContent()) &&
(EndsByNormalText() || EndsBySpecialContent() ||
EndsByBRElement()));
}
EditorDOMPoint mScanStartPoint;
BoundaryData mStart;
BoundaryData mEnd;
NoBreakingSpaceData mNBSPData;
RefPtr<const dom::Element> mEditingHost;
mutable Maybe<EditorDOMRange> mLeadingWhiteSpaceRange;
mutable Maybe<EditorDOMRange> mTrailingWhiteSpaceRange;
mutable Maybe<VisibleWhiteSpacesData> mVisibleWhiteSpacesData;
// XXX Currently we set mIsPreformatted to `WSRunScanner::mPRE` value
// even if some text nodes between mStart and mEnd are different styled
// nodes. This caused some bugs actually, but we now keep traditional
// behavior for now.
bool mIsPreformatted;
};
const TextFragmentData& TextFragmentDataAtStartRef() const {
return mTextFragmentDataAtStart;
}
// The node passed to our constructor.
EditorDOMPoint mScanStartPoint;
// Together, the above represent the point at which we are building up ws
// info.
// The editing host when the instance is created.
RefPtr<dom::Element> mEditingHost;
// Non-owning.
const HTMLEditor* mHTMLEditor;
private:
/**
* ComputeRangeInTextNodesContainingInvisibleWhiteSpaces() returns range
* containing invisible white-spaces if deleting between aStart and aEnd
* causes them become visible.
*
* @param aStart TextFragmentData at start of deleting range.
* This must be initialized with DOM point in a text node.
* @param aEnd TextFragmentData at end of deleting range.
* This must be initialized with DOM point in a text node.
*/
static EditorDOMRangeInTexts
ComputeRangeInTextNodesContainingInvisibleWhiteSpaces(
const TextFragmentData& aStart, const TextFragmentData& aEnd);
TextFragmentData mTextFragmentDataAtStart;
friend class WhiteSpaceVisibilityKeeper;
};
/**
* WhiteSpaceVisibilityKeeper class helps `HTMLEditor` modifying the DOM tree
* with keeps white-space sequence visibility automatically. E.g., invisible
* leading/trailing white-spaces becomes visible, this class members delete
* them. E.g., when splitting visible-white-space sequence, this class may
* replace ASCII white-spaces at split edges with NBSPs.
*/
class WhiteSpaceVisibilityKeeper final {
private:
using AutoTransactionsConserveSelection =
EditorBase::AutoTransactionsConserveSelection;
using EditorType = EditorBase::EditorType;
using PointPosition = WSRunScanner::PointPosition;
using TextFragmentData = WSRunScanner::TextFragmentData;
using VisibleWhiteSpacesData = WSRunScanner::VisibleWhiteSpacesData;
public:
WhiteSpaceVisibilityKeeper() = delete;
explicit WhiteSpaceVisibilityKeeper(
const WhiteSpaceVisibilityKeeper& aOther) = delete;
WhiteSpaceVisibilityKeeper(WhiteSpaceVisibilityKeeper&& aOther) = delete;
/**
* DeleteInvisibleASCIIWhiteSpaces() removes invisible leading white-spaces
* and trailing white-spaces if there are around aPoint.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
DeleteInvisibleASCIIWhiteSpaces(HTMLEditor& aHTMLEditor,
const EditorDOMPoint& aPoint);
// PrepareToDeleteRange fixes up ws before aStartPoint and after aEndPoint in
// preperation for content in that range to be deleted. Note that the nodes
// and offsets are adjusted in response to any dom changes we make while
// adjusting ws.
// example of fixup: trailingws before aStartPoint needs to be removed.
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
PrepareToDeleteRangeAndTrackPoints(HTMLEditor& aHTMLEditor,
EditorDOMPoint* aStartPoint,
EditorDOMPoint* aEndPoint) {
MOZ_ASSERT(aStartPoint->IsSetAndValid());
MOZ_ASSERT(aEndPoint->IsSetAndValid());
AutoTrackDOMPoint trackerStart(aHTMLEditor.RangeUpdaterRef(), aStartPoint);
AutoTrackDOMPoint trackerEnd(aHTMLEditor.RangeUpdaterRef(), aEndPoint);
return WhiteSpaceVisibilityKeeper::PrepareToDeleteRange(
aHTMLEditor, EditorDOMRange(*aStartPoint, *aEndPoint));
}
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult PrepareToDeleteRange(
HTMLEditor& aHTMLEditor, const EditorDOMPoint& aStartPoint,
const EditorDOMPoint& aEndPoint) {
MOZ_ASSERT(aStartPoint.IsSetAndValid());
MOZ_ASSERT(aEndPoint.IsSetAndValid());
return WhiteSpaceVisibilityKeeper::PrepareToDeleteRange(
aHTMLEditor, EditorDOMRange(aStartPoint, aEndPoint));
}
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult PrepareToDeleteRange(
HTMLEditor& aHTMLEditor, const EditorDOMRange& aRange) {
MOZ_ASSERT(aRange.IsPositionedAndValid());
nsresult rv = WhiteSpaceVisibilityKeeper::
MakeSureToKeepVisibleStateOfWhiteSpacesAroundDeletingRange(aHTMLEditor,
aRange);
NS_WARNING_ASSERTION(
NS_SUCCEEDED(rv),
"WhiteSpaceVisibilityKeeper::"
"MakeSureToKeepVisibleStateOfWhiteSpacesAroundDeletingRange() failed");
return rv;
}
// PrepareToSplitAcrossBlocks fixes up ws before and after
// {aSplitNode,aSplitOffset} in preparation for a block parent to be split.
// Note that the aSplitNode and aSplitOffset are adjusted in response to
// any DOM changes we make while adjusting ws. Example of fixup: normalws
// before {aSplitNode,aSplitOffset} needs to end with nbsp.
MOZ_CAN_RUN_SCRIPT static nsresult PrepareToSplitAcrossBlocks(
HTMLEditor& aHTMLEditor, nsCOMPtr<nsINode>* aSplitNode,
int32_t* aSplitOffset);
/**
* MergeFirstLineOfRightBlockElementIntoDescendantLeftBlockElement() merges
* first line in aRightBlockElement into end of aLeftBlockElement which
* is a descendant of aRightBlockElement.
*
* @param aHTMLEditor The HTML editor.
* @param aLeftBlockElement The content will be merged into end of
* this element.
* @param aRightBlockElement The first line in this element will be
* moved to aLeftBlockElement.
* @param aAtRightBlockChild At a child of aRightBlockElement and inclusive
* ancestor of aLeftBlockElement.
* @param aListElementTagName Set some if aRightBlockElement is a list
* element and it'll be merged with another
* list element.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static EditActionResult
MergeFirstLineOfRightBlockElementIntoDescendantLeftBlockElement(
HTMLEditor& aHTMLEditor, dom::Element& aLeftBlockElement,
dom::Element& aRightBlockElement,
const EditorDOMPoint& aAtRightBlockChild,
const Maybe<nsAtom*>& aListElementTagName,
const dom::HTMLBRElement* aPrecedingInvisibleBRElement);
/**
* MergeFirstLineOfRightBlockElementIntoAncestorLeftBlockElement() merges
* first line in aRightBlockElement into end of aLeftBlockElement which
* is an ancestor of aRightBlockElement, then, removes aRightBlockElement
* if it becomes empty.
*
* @param aHTMLEditor The HTML editor.
* @param aLeftBlockElement The content will be merged into end of
* this element.
* @param aRightBlockElement The first line in this element will be
* moved to aLeftBlockElement and maybe
* removed when this becomes empty.
* @param aAtLeftBlockChild At a child of aLeftBlockElement and inclusive
* ancestor of aRightBlockElement.
* @param aLeftContentInBlock The content whose inclusive ancestor is
* aLeftBlockElement.
* @param aListElementTagName Set some if aRightBlockElement is a list
* element and it'll be merged with another
* list element.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static EditActionResult
MergeFirstLineOfRightBlockElementIntoAncestorLeftBlockElement(
HTMLEditor& aHTMLEditor, dom::Element& aLeftBlockElement,
dom::Element& aRightBlockElement, const EditorDOMPoint& aAtLeftBlockChild,
nsIContent& aLeftContentInBlock,
const Maybe<nsAtom*>& aListElementTagName,
const dom::HTMLBRElement* aPrecedingInvisibleBRElement);
/**
* MergeFirstLineOfRightBlockElementIntoLeftBlockElement() merges first
* line in aRightBlockElement into end of aLeftBlockElement and removes
* aRightBlockElement when it has only one line.
*
* @param aHTMLEditor The HTML editor.
* @param aLeftBlockElement The content will be merged into end of
* this element.
* @param aRightBlockElement The first line in this element will be
* moved to aLeftBlockElement and maybe
* removed when this becomes empty.
* @param aListElementTagName Set some if aRightBlockElement is a list
* element and its type needs to be changed.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static EditActionResult
MergeFirstLineOfRightBlockElementIntoLeftBlockElement(
HTMLEditor& aHTMLEditor, dom::Element& aLeftBlockElement,
dom::Element& aRightBlockElement,
const Maybe<nsAtom*>& aListElementTagName,
const dom::HTMLBRElement* aPrecedingInvisibleBRElement);
/**
* InsertBRElement() inserts a <br> node at (before) aPointToInsert and delete
* unnecessary white-spaces around there and/or replaces white-spaces with
* non-breaking spaces. Note that if the point is in a text node, the
* text node will be split and insert new <br> node between the left node
* and the right node.
*
* @param aPointToInsert The point to insert new <br> element. Note that
* it'll be inserted before this point. I.e., the
* point will be the point of new <br>.
* @return The new <br> node. If failed to create new <br>
* node, returns nullptr.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static Result<RefPtr<dom::Element>, nsresult>
InsertBRElement(HTMLEditor& aHTMLEditor,
const EditorDOMPoint& aPointToInsert);
/**
* InsertText() inserts aStringToInsert to aPointToInsert and makes any needed
* adjustments to white-spaces around the insertion point.
*
* @param aStringToInsert The string to insert.
* @param aRangeToBeReplaced The range to be deleted.
* @param aPointAfterInsertedString
* The point after inserted aStringToInsert.
* So, when this method actually inserts string,
* this is set to a point in the text node.
* Otherwise, this may be set to mScanStartPoint.
* @return When this succeeds to insert the string or
* does nothing during composition, returns NS_OK.
* Otherwise, an error code.
*/
template <typename EditorDOMPointType>
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult InsertText(
HTMLEditor& aHTMLEditor, const nsAString& aStringToInsert,
const EditorDOMPointType& aPointToInsert,
EditorRawDOMPoint* aPointAfterInsertedString = nullptr) {
return WhiteSpaceVisibilityKeeper::ReplaceText(
aHTMLEditor, aStringToInsert, EditorDOMRange(aPointToInsert),
aPointAfterInsertedString);
}
/**
* ReplaceText() repaces aRangeToReplace with aStringToInsert and makes any
* needed adjustments to white-spaces around both start of the range and
* end of the range.
*
* @param aStringToInsert The string to insert.
* @param aRangeToBeReplaced The range to be deleted.
* @param aPointAfterInsertedString
* The point after inserted aStringToInsert.
* So, when this method actually inserts string,
* this is set to a point in the text node.
* Otherwise, this may be set to mScanStartPoint.
* @return When this succeeds to insert the string or
* does nothing during composition, returns NS_OK.
* Otherwise, an error code.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult ReplaceText(
HTMLEditor& aHTMLEditor, const nsAString& aStringToInsert,
const EditorDOMRange& aRangeToBeReplaced,
EditorRawDOMPoint* aPointAfterInsertedString = nullptr);
/**
* DeletePreviousWhiteSpace() deletes previous white-space of aPoint.
* This automatically keeps visibility of white-spaces around aPoint.
* E.g., may remove invisible leading white-spaces.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult DeletePreviousWhiteSpace(
HTMLEditor& aHTMLEditor, const EditorDOMPoint& aPoint);
/**
* DeleteInclusiveNextWhiteSpace() delete inclusive next white-space of
* aPoint. This automatically keeps visiblity of white-spaces around aPoint.
* E.g., may remove invisible trailing white-spaces.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
DeleteInclusiveNextWhiteSpace(HTMLEditor& aHTMLEditor,
const EditorDOMPoint& aPoint);
/**
* DeleteContentNodeAndJoinTextNodesAroundIt() deletes aContentToDelete and
* may remove/replace white-spaces around it. Then, if deleting content makes
* 2 text nodes around it are adjacent siblings, this joins them and put
* selection at the joined point.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
DeleteContentNodeAndJoinTextNodesAroundIt(HTMLEditor& aHTMLEditor,
nsIContent& aContentToDelete,
const EditorDOMPoint& aCaretPoint);
/**
* NormalizeVisibleWhiteSpacesAt() tries to normalize visible white-space
* sequence around aPoint.
*/
template <typename EditorDOMPointType>
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
NormalizeVisibleWhiteSpacesAt(HTMLEditor& aHTMLEditor,
const EditorDOMPointType& aPoint);
private:
/**
* MakeSureToKeepVisibleStateOfWhiteSpacesAroundDeletingRange() may delete
* invisible white-spaces for keeping make them invisible and/or may replace
* ASCII white-spaces with NBSPs for making visible white-spaces to keep
* visible.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
MakeSureToKeepVisibleStateOfWhiteSpacesAroundDeletingRange(
HTMLEditor& aHTMLEditor, const EditorDOMRange& aRangeToDelete);
/**
* MakeSureToKeepVisibleWhiteSpacesVisibleAfterSplit() replaces ASCII white-
* spaces which becomes invisible after split with NBSPs.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
MakeSureToKeepVisibleWhiteSpacesVisibleAfterSplit(
HTMLEditor& aHTMLEditor, const EditorDOMPoint& aPointToSplit);
/**
* ReplaceTextAndRemoveEmptyTextNodes() replaces the range between
* aRangeToReplace with aReplaceString simply. Additionally, removes
* empty text nodes in the range.
*
* @param aRangeToReplace Range to replace text.
* @param aReplaceString The new string. Empty string is allowed.
*/
[[nodiscard]] MOZ_CAN_RUN_SCRIPT static nsresult
ReplaceTextAndRemoveEmptyTextNodes(
HTMLEditor& aHTMLEditor, const EditorDOMRangeInTexts& aRangeToReplace,
const nsAString& aReplaceString);
};
} // namespace mozilla
#endif // #ifndef WSRunObject_h
Ссылка в новой задаче Показать git blame Копировать постоянную ссылку