Home | History | Annotate | Download | only in parser
      1 /*
      2  * Copyright (C) 2000 Peter Kelly (pmk (at) post.com)
      3  * Copyright (C) 2005, 2006, 2007 Apple Inc. All rights reserved.
      4  * Copyright (C) 2007 Samuel Weinig (sam (at) webkit.org)
      5  * Copyright (C) 2008 Nokia Corporation and/or its subsidiary(-ies)
      6  * Copyright (C) 2008, 2009 Torch Mobile Inc. All rights reserved. (http://www.torchmobile.com/)
      7  *
      8  * This library is free software; you can redistribute it and/or
      9  * modify it under the terms of the GNU Library General Public
     10  * License as published by the Free Software Foundation; either
     11  * version 2 of the License, or (at your option) any later version.
     12  *
     13  * This library is distributed in the hope that it will be useful,
     14  * but WITHOUT ANY WARRANTY; without even the implied warranty of
     15  * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
     16  * Library General Public License for more details.
     17  *
     18  * You should have received a copy of the GNU Library General Public License
     19  * along with this library; see the file COPYING.LIB.  If not, write to
     20  * the Free Software Foundation, Inc., 51 Franklin Street, Fifth Floor,
     21  * Boston, MA 02110-1301, USA.
     22  *
     23  */
     24 
     25 #ifndef XMLDocumentParser_h
     26 #define XMLDocumentParser_h
     27 
     28 #include "core/dom/ParserContentPolicy.h"
     29 #include "core/dom/ScriptableDocumentParser.h"
     30 #include "core/fetch/ResourceClient.h"
     31 #include "core/fetch/ResourcePtr.h"
     32 #include "core/fetch/ScriptResource.h"
     33 #include "core/xml/parser/XMLErrors.h"
     34 #include "platform/heap/Handle.h"
     35 #include "platform/text/SegmentedString.h"
     36 #include "wtf/HashMap.h"
     37 #include "wtf/OwnPtr.h"
     38 #include "wtf/text/CString.h"
     39 #include "wtf/text/StringHash.h"
     40 #include <libxml/tree.h>
     41 
     42 namespace blink {
     43 
     44 class ContainerNode;
     45 class ResourceFetcher;
     46 class DocumentFragment;
     47 class Document;
     48 class Element;
     49 class FrameView;
     50 class Text;
     51 
     52 class XMLParserContext : public RefCounted<XMLParserContext> {
     53 public:
     54     static PassRefPtr<XMLParserContext> createMemoryParser(xmlSAXHandlerPtr, void* userData, const CString& chunk);
     55     static PassRefPtr<XMLParserContext> createStringParser(xmlSAXHandlerPtr, void* userData);
     56     ~XMLParserContext();
     57     xmlParserCtxtPtr context() const { return m_context; }
     58 
     59 private:
     60     XMLParserContext(xmlParserCtxtPtr context)
     61         : m_context(context)
     62     {
     63     }
     64 
     65     xmlParserCtxtPtr m_context;
     66 };
     67 
     68 class XMLDocumentParser FINAL : public ScriptableDocumentParser, public ScriptResourceClient {
     69     WTF_MAKE_FAST_ALLOCATED_WILL_BE_REMOVED;
     70 public:
     71     static PassRefPtrWillBeRawPtr<XMLDocumentParser> create(Document& document, FrameView* view)
     72     {
     73         return adoptRefWillBeNoop(new XMLDocumentParser(document, view));
     74     }
     75     static PassRefPtrWillBeRawPtr<XMLDocumentParser> create(DocumentFragment* fragment, Element* element, ParserContentPolicy parserContentPolicy)
     76     {
     77         return adoptRefWillBeNoop(new XMLDocumentParser(fragment, element, parserContentPolicy));
     78     }
     79     virtual ~XMLDocumentParser();
     80     virtual void trace(Visitor*) OVERRIDE;
     81 
     82     // Exposed for callbacks:
     83     void handleError(XMLErrors::ErrorType, const char* message, TextPosition);
     84 
     85     void setIsXHTMLDocument(bool isXHTML) { m_isXHTMLDocument = isXHTML; }
     86     bool isXHTMLDocument() const { return m_isXHTMLDocument; }
     87 
     88     bool isCurrentlyParsing8BitChunk() { return m_isCurrentlyParsing8BitChunk; }
     89 
     90     static bool parseDocumentFragment(const String&, DocumentFragment*, Element* parent = 0, ParserContentPolicy = AllowScriptingContent);
     91 
     92     // Used by the XMLHttpRequest to check if the responseXML was well formed.
     93     virtual bool wellFormed() const OVERRIDE { return !m_sawError; }
     94 
     95     virtual TextPosition textPosition() const OVERRIDE;
     96 
     97     static bool supportsXMLVersion(const String&);
     98 
     99     class PendingCallback {
    100     public:
    101         virtual ~PendingCallback() { }
    102         virtual void call(XMLDocumentParser*) = 0;
    103     };
    104 
    105 private:
    106     explicit XMLDocumentParser(Document&, FrameView* = 0);
    107     XMLDocumentParser(DocumentFragment*, Element*, ParserContentPolicy);
    108 
    109     // From DocumentParser
    110     virtual void insert(const SegmentedString&) OVERRIDE;
    111     virtual void append(PassRefPtr<StringImpl>) OVERRIDE;
    112     virtual void finish() OVERRIDE;
    113     virtual bool isWaitingForScripts() const OVERRIDE;
    114     virtual void stopParsing() OVERRIDE;
    115     virtual void detach() OVERRIDE;
    116     virtual OrdinalNumber lineNumber() const OVERRIDE;
    117     OrdinalNumber columnNumber() const;
    118 
    119     // from ResourceClient
    120     virtual void notifyFinished(Resource*) OVERRIDE;
    121 
    122     void end();
    123 
    124     void pauseParsing();
    125     void resumeParsing();
    126 
    127     bool appendFragmentSource(const String&);
    128 
    129 public:
    130     // Callbacks from parser SAX
    131     void error(XMLErrors::ErrorType, const char* message, va_list args) WTF_ATTRIBUTE_PRINTF(3, 0);
    132     void startElementNs(const AtomicString& localName, const AtomicString& prefix, const AtomicString& uri, int namespaceCount,
    133         const xmlChar** namespaces, int attributeCount, int defaultedCount, const xmlChar** libxmlAttributes);
    134     void endElementNs();
    135     void characters(const xmlChar* chars, int length);
    136     void processingInstruction(const String& target, const String& data);
    137     void cdataBlock(const String&);
    138     void comment(const String&);
    139     void startDocument(const String& version, const String& encoding, int standalone);
    140     void internalSubset(const String& name, const String& externalID, const String& systemID);
    141     void endDocument();
    142 
    143 private:
    144     void initializeParserContext(const CString& chunk = CString());
    145 
    146     void pushCurrentNode(ContainerNode*);
    147     void popCurrentNode();
    148     void clearCurrentNodeStack();
    149 
    150     void insertErrorMessageBlock();
    151 
    152     void enterText();
    153     void exitText();
    154 
    155     void doWrite(const String&);
    156     void doEnd();
    157 
    158     bool m_hasView;
    159 
    160     SegmentedString m_originalSourceForTransform;
    161 
    162     xmlParserCtxtPtr context() const { return m_context ? m_context->context() : 0; };
    163     RefPtr<XMLParserContext> m_context;
    164     Deque<OwnPtr<PendingCallback> > m_pendingCallbacks;
    165     Vector<xmlChar> m_bufferedText;
    166 
    167     RawPtrWillBeMember<ContainerNode> m_currentNode;
    168     WillBeHeapVector<RawPtrWillBeMember<ContainerNode> > m_currentNodeStack;
    169 
    170     RefPtrWillBeMember<Text> m_leafTextNode;
    171 
    172     bool m_isCurrentlyParsing8BitChunk;
    173     bool m_sawError;
    174     bool m_sawCSS;
    175     bool m_sawXSLTransform;
    176     bool m_sawFirstElement;
    177     bool m_isXHTMLDocument;
    178     bool m_parserPaused;
    179     bool m_requestingScript;
    180     bool m_finishCalled;
    181 
    182     XMLErrors m_xmlErrors;
    183 
    184     ResourcePtr<ScriptResource> m_pendingScript;
    185     RefPtrWillBeMember<Element> m_scriptElement;
    186     TextPosition m_scriptStartPosition;
    187 
    188     bool m_parsingFragment;
    189     AtomicString m_defaultNamespaceURI;
    190 
    191     typedef HashMap<AtomicString, AtomicString> PrefixForNamespaceMap;
    192     PrefixForNamespaceMap m_prefixToNamespaceMap;
    193     SegmentedString m_pendingSrc;
    194 };
    195 
    196 xmlDocPtr xmlDocPtrForString(ResourceFetcher*, const String& source, const String& url);
    197 HashMap<String, String> parseAttributes(const String&, bool& attrsOK);
    198 
    199 } // namespace blink
    200 
    201 #endif // XMLDocumentParser_h
    202