1 /*
2  * Copyright (C) 2004, 2005, 2006, 2007, 2008, 2009 Apple Inc. All rights reserved.
3  *
4  * Redistribution and use in source and binary forms, with or without
5  * modification, are permitted provided that the following conditions
6  * are met:
7  * 1. Redistributions of source code must retain the above copyright
8  *    notice, this list of conditions and the following disclaimer.
9  * 2. Redistributions in binary form must reproduce the above copyright
10  *    notice, this list of conditions and the following disclaimer in the
11  *    documentation and/or other materials provided with the distribution.
12  *
13  * THIS SOFTWARE IS PROVIDED BY APPLE COMPUTER, INC. ``AS IS'' AND ANY
14  * EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
15  * IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
16  * PURPOSE ARE DISCLAIMED.  IN NO EVENT SHALL APPLE COMPUTER, INC. OR
17  * CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
18  * EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
19  * PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR
20  * PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY
21  * OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
22  * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
23  * OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
24  */
25 
26 #ifndef MarkupAccumulator_h
27 #define MarkupAccumulator_h
28 
29 #include "PlatformString.h"
30 #include "markup.h"
31 #include <wtf/HashMap.h>
32 #include <wtf/Vector.h>
33 
34 namespace WebCore {
35 
36 class Attribute;
37 class DocumentType;
38 class Element;
39 class Node;
40 class Range;
41 
42 typedef HashMap<AtomicStringImpl*, AtomicStringImpl*> Namespaces;
43 
44 enum EntityMask {
45     EntityAmp = 0x0001,
46     EntityLt = 0x0002,
47     EntityGt = 0x0004,
48     EntityQuot = 0x0008,
49     EntityNbsp = 0x0010,
50 
51     // Non-breaking space needs to be escaped in innerHTML for compatibility reason. See http://trac.webkit.org/changeset/32879
52     // However, we cannot do this in a XML document because it does not have the entity reference defined (See the bug 19215).
53     EntityMaskInCDATA = 0,
54     EntityMaskInPCDATA = EntityAmp | EntityLt | EntityGt,
55     EntityMaskInHTMLPCDATA = EntityMaskInPCDATA | EntityNbsp,
56     EntityMaskInAttributeValue = EntityAmp | EntityLt | EntityGt | EntityQuot,
57     EntityMaskInHTMLAttributeValue = EntityMaskInAttributeValue | EntityNbsp,
58 };
59 
60 struct EntityDescription {
61     UChar entity;
62     const String& reference;
63     EntityMask mask;
64 };
65 
66 // FIXME: Noncopyable?
67 class MarkupAccumulator {
68 public:
69     MarkupAccumulator(Vector<Node*>* nodes, EAbsoluteURLs shouldResolveURLs, const Range* range = 0);
70     virtual ~MarkupAccumulator();
71 
72     String serializeNodes(Node* node, Node* nodeToSkip, EChildrenOnly childrenOnly);
73 
74 protected:
75     virtual void appendString(const String&);
76     void appendStartTag(Node*, Namespaces* = 0);
77     virtual void appendEndTag(Node*);
78     static size_t totalLength(const Vector<String>&);
length()79     size_t length() const { return totalLength(m_succeedingMarkup); }
80     void concatenateMarkup(Vector<UChar>& out);
81     void appendAttributeValue(Vector<UChar>& result, const String& attribute, bool documentIsHTML);
82     virtual void appendCustomAttributes(Vector<UChar>&, Element*, Namespaces*);
83     void appendQuotedURLAttributeValue(Vector<UChar>& result, const String& urlString);
84     void appendNodeValue(Vector<UChar>& out, const Node*, const Range*, EntityMask);
85     bool shouldAddNamespaceElement(const Element*);
86     bool shouldAddNamespaceAttribute(const Attribute&, Namespaces&);
87     void appendNamespace(Vector<UChar>& result, const AtomicString& prefix, const AtomicString& namespaceURI, Namespaces&);
88     EntityMask entityMaskForText(Text*) const;
89     virtual void appendText(Vector<UChar>& out, Text*);
90     void appendComment(Vector<UChar>& out, const String& comment);
91     void appendDocumentType(Vector<UChar>& result, const DocumentType*);
92     void appendProcessingInstruction(Vector<UChar>& out, const String& target, const String& data);
93     virtual void appendElement(Vector<UChar>& out, Element*, Namespaces*);
94     void appendOpenTag(Vector<UChar>& out, Element*, Namespaces*);
95     void appendCloseTag(Vector<UChar>& out, Element*);
96     void appendAttribute(Vector<UChar>& out, Element*, const Attribute&, Namespaces*);
97     void appendCDATASection(Vector<UChar>& out, const String& section);
98     void appendStartMarkup(Vector<UChar>& result, const Node*, Namespaces*);
99     bool shouldSelfClose(const Node*);
100     bool elementCannotHaveEndTag(const Node* node);
101     void appendEndMarkup(Vector<UChar>& result, const Node*);
102 
shouldResolveURLs()103     bool shouldResolveURLs() { return m_shouldResolveURLs == AbsoluteURLs; }
104 
105     Vector<Node*>* const m_nodes;
106     const Range* const m_range;
107 
108 private:
109     void serializeNodesWithNamespaces(Node*, Node* nodeToSkip, EChildrenOnly, const Namespaces*);
110 
111     Vector<String> m_succeedingMarkup;
112     const bool m_shouldResolveURLs;
113 };
114 
115 // FIXME: This method should be integrated with MarkupAccumulator.
116 void appendCharactersReplacingEntities(Vector<UChar>& out, const UChar* content, size_t length, EntityMask entityMask);
117 
118 }
119 
120 #endif
121