avmplus::XMLParser Class Reference

#include <XMLParser16.h>

List of all members.

Public Types
enum	{   kNoError = 0, kEndOfDocument = -1, kUnterminatedCDataSection = -2, kUnterminatedXMLDeclaration = -3,   kUnterminatedDocTypeDeclaration = -4, kUnterminatedComment = -5, kMalformedElement = -6, kOutOfMemory = -7,   kUnterminatedAttributeValue = -8, kUnterminatedElement = -9, kElementNeverBegun = -10, kUnterminatedProcessingInstruction = -11 }
Public Member Functions
	XMLParser (AvmCore *core)
	~XMLParser ()
void	parse (Stringp source, bool ignoreWhite=false)
int	getNext (XMLTag &tag)
bool	getCondenseWhite () const
void	setCondenseWhite (bool flag)
Public Attributes
AvmCore *	core
Private Member Functions
Stringp	unescape (Stringp buffer, const wchar *start, int len, bool bIntern)
void	condenseWhitespace (Stringp text)
Private Attributes
Stringp	m_source
const wchar *	m_ptr
bool	m_ignoreWhite
bool	m_condenseWhite

---

Detailed Description

XMLParser is a XML parser which takes 16-bit wide characters as input. The parser operates in "pull" fashion, returning a single tag or text node on each call to the GetNext method.

This XML parser is used to support E4X in AVM+.

Definition at line 101 of file XMLParser16.h.

---

Member Enumeration Documentation

anonymous enum

Enumerator: kNoError  kEndOfDocument  kUnterminatedCDataSection  kUnterminatedXMLDeclaration  kUnterminatedDocTypeDeclaration  kUnterminatedComment  kMalformedElement  kOutOfMemory  kUnterminatedAttributeValue  kUnterminatedElement  kElementNeverBegun  kUnterminatedProcessingInstruction  Definition at line 119 of file XMLParser16.h. { kNoError = 0, kEndOfDocument = -1, kUnterminatedCDataSection = -2, kUnterminatedXMLDeclaration = -3, kUnterminatedDocTypeDeclaration = -4, kUnterminatedComment = -5, kMalformedElement = -6, kOutOfMemory = -7, kUnterminatedAttributeValue = -8, kUnterminatedElement = -9, kElementNeverBegun = -10, kUnterminatedProcessingInstruction = -11 };

---

Constructor & Destructor Documentation

avmplus::XMLParser::XMLParser (  AvmCore *  core  )

Definition at line 468 of file XMLParser16.cpp. References avmplus::Hashtable::add(), avmplus::AvmCore::constant(), core, MMgc::GCRoot::GetGC(), avmplus::AvmCore::intToAtom(), and avmplus::AvmCore::xmlEntities. { this->core = core; if (!core->xmlEntities) { // Lazy creation of the XML entities table. core->xmlEntities = new (core->GetGC()) Hashtable(core->GetGC()); const char *entities = "&amp\0\"quot\0'apos\0<lt\0>gt\0\xA0nbsp\0"; while (*entities) { core->xmlEntities->add(core->constant(entities+1), (void*)core->intToAtom(*entities)); while (*entities++) { // do nothing } } } }

avmplus::XMLParser::~XMLParser (   )  [inline]

Definition at line 105 of file XMLParser16.h. References core, m_condenseWhite, m_ignoreWhite, m_ptr, m_source, and NULL. { core = NULL; m_source = NULL; m_ptr = NULL; m_ignoreWhite = false; m_condenseWhite = false; }

---

Member Function Documentation

void avmplus::XMLParser::condenseWhitespace (  Stringp  text  )  [private]

Definition at line 73 of file XMLParser16.cpp. References AvmAssert, avmplus::String::isInterned(), avmplus::String::isSpace(), avmplus::String::length(), avmplus::String::lockBuffer(), and avmplus::String::unlockBuffer(). Referenced by getNext(). { AvmAssert (!text->isInterned()); wchar *str = text->lockBuffer(); int len = text->length(); wchar *dst = str; wchar *src = str; bool leadingWhite = true; wchar *lastChar = 0; while (len--) { if (String::isSpace(*src)) { if (!leadingWhite) { *dst++ = *src; } src++; } else { leadingWhite = false; // first non-space char, no more lastChar = dst; *dst++ = *src++; } } if (lastChar) lastChar[1] = 0; *dst = 0; text->unlockBuffer((int)((lastChar ? (lastChar + 1) : dst)-str)); }

bool avmplus::XMLParser::getCondenseWhite (   )  const [inline]

Definition at line 136 of file XMLParser16.h. References m_condenseWhite. { return m_condenseWhite; }

int avmplus::XMLParser::getNext (  XMLTag &  tag  )

Definition at line 105 of file XMLParser16.cpp. References condenseWhitespace(), core, MMgc::GCRoot::GetGC(), avmplus::String::isSpace(), avmplus::XMLTag::kCDataSection, avmplus::XMLTag::kComment, avmplus::XMLTag::kDocTypeDeclaration, avmplus::XMLTag::kElementType, kEndOfDocument, kMalformedElement, kNoError, avmplus::XMLTag::kProcessingInstruction, avmplus::XMLTag::kTextNodeType, kUnterminatedCDataSection, kUnterminatedComment, kUnterminatedDocTypeDeclaration, kUnterminatedProcessingInstruction, kUnterminatedXMLDeclaration, avmplus::XMLTag::kXMLDeclaration, m_condenseWhite, m_ignoreWhite, m_ptr, m_source, avmplus::XMLTag::nodeType, NULL, avmplus::XMLTag::reset(), avmplus::stripPrefix(), avmplus::XMLTag::text, and unescape(). Referenced by avmplus::XMLObject::XMLObject(). { tag.reset(); // If there's nothing left, exit. if (!m_ptr || !*m_ptr) { return XMLParser::kEndOfDocument; } // R41 // If the ignore whitespace flag is on, don't produce // all-whitespace text nodes. if (m_ignoreWhite) { const wchar *ptr = m_ptr; while (String::isSpace(*ptr)) { ptr++; } if (*ptr == '<' || !*ptr) { // If we reached the end of the document, // or we reached a tag, skip all the // whitesapce, because it would turn into // an empty text node. m_ptr = ptr; } // If there's nothing left, exit. // But only do it for Flash 6 because we want // to exactly preserve Flash 5 behavior. if (!*m_ptr) { return XMLParser::kEndOfDocument; } } // end R41 // If it starts with <, it's an XML element. // If it doesn't, it must be a text element. if (*m_ptr != '<') { // Treat it as text. Scan up to the next < or until EOF. const wchar *start = m_ptr; while (*m_ptr && *m_ptr != '<') { m_ptr++; } tag.text = unescape(m_source, start, (int)(m_ptr-start), false); // Condense whitespace if desired if (m_ignoreWhite && m_condenseWhite) { condenseWhitespace(tag.text); } tag.nodeType = XMLTag::kTextNodeType; return XMLParser::kNoError; } // Is this a <?xml> declaration? wchar *temp; if ((temp = stripPrefix(m_ptr, "<?xml ")) != NULL) { // Scan forward for "?>" const wchar *start = m_ptr; m_ptr = temp; while (*m_ptr) { if (m_ptr[0] == '?' && m_ptr[1] == '>') { // We have the end of the XML declaration // !!@ changed to not return <?...?> parts tag.text = new (core->GetGC()) String(start + 2, (int)(m_ptr - start - 2)); m_ptr += 2; tag.nodeType = XMLTag::kXMLDeclaration; return XMLParser::kNoError; } else { m_ptr++; } } return XMLParser::kUnterminatedXMLDeclaration; } // Is this a <!DOCTYPE> declaration? if ((temp = stripPrefix(m_ptr, "<!DOCTYPE")) != NULL) { // Scan forward for '>'. const wchar *start = m_ptr; m_ptr = temp; int depth = 0; while (*m_ptr) { if (*m_ptr == '<') { depth++; } if (*m_ptr == '>') { if (!depth) { // We've reached the end of the DOCTYPE. m_ptr++; tag.text = new (core->GetGC()) String(start, (int)(m_ptr-start)); tag.nodeType = XMLTag::kDocTypeDeclaration; return XMLParser::kNoError; } depth--; } m_ptr++; } return XMLParser::kUnterminatedDocTypeDeclaration; } // Is this a CDATA section? wchar *cdata; if ((cdata = stripPrefix(m_ptr, "<![CDATA[")) != NULL) { // Scan forward for "]]>" m_ptr = cdata; while (*m_ptr) { if (m_ptr[0] == ']' && m_ptr[1] == ']' && m_ptr[2] == '>') { // We have the end of the CDATA section. tag.text = new (core->GetGC()) String(cdata, (int)(m_ptr-cdata)); tag.nodeType = XMLTag::kCDataSection; m_ptr += 3; return XMLParser::kNoError; } m_ptr++; } return XMLParser::kUnterminatedCDataSection; } // Is this a processing instruction? wchar *pi; if ((pi = stripPrefix(m_ptr, "<?")) != NULL) { // Scan forward for "?>" m_ptr = pi; while (*m_ptr) { if (m_ptr[0] == '?' && m_ptr[1] == '>') { // We have the end of the processing instruction. tag.text = new (core->GetGC()) String(pi, (int)(m_ptr - pi)); tag.nodeType = XMLTag::kProcessingInstruction; m_ptr += 2; return XMLParser::kNoError; } m_ptr++; } return XMLParser::kUnterminatedProcessingInstruction; } // Advance past the "<" m_ptr++; // Is this a comment? Return a comment tag-> const wchar *comment; if (m_ptr[0] == '!' && m_ptr[1] == '-' && m_ptr[2] == '-') { // Skip up to '-->'. m_ptr += 3; comment = m_ptr; while (*m_ptr) { if (m_ptr[0] == '-' && m_ptr[1] == '-' && m_ptr[2] == '>') { tag.text = new (core->GetGC()) String(comment, (int)(m_ptr-comment)); tag.nodeType = XMLTag::kComment; m_ptr += 3; return XMLParser::kNoError; } m_ptr++; } // Got to the end of the buffer without finding a new tag-> return XMLParser::kUnterminatedComment; } // Extract the tag name. Scan up to ">" or whitespace. const wchar *tagStart = m_ptr; while (!String::isSpace(*m_ptr) && *m_ptr != '>') { if (*m_ptr == '/' && *(m_ptr+1) == '>') { // Found close of an empty element. // Exit! break; } if (!*m_ptr) { // Premature end! return XMLParser::kMalformedElement; } m_ptr++; } // Give up if tag name is empty if (m_ptr == tagStart) { return XMLParser::kMalformedElement; } tag.text = unescape(m_source, tagStart, (int)(m_ptr-tagStart), true); tag.nodeType = XMLTag::kElementType; // Extract attributes. for (;;) { if (!*m_ptr) { // Premature end! return XMLParser::kMalformedElement; } // Skip any whitespace. while (String::isSpace(*m_ptr)) { m_ptr++; } if (*m_ptr == '>') { break; } if (*m_ptr == '/' && *(m_ptr+1) == '>') { // Found close of an empty element. // Exit! tag.empty = true; m_ptr++; break; } // Extract the attribute name. const wchar *nameStart = m_ptr; while (!String::isSpace(*m_ptr) && *m_ptr != '=' && *m_ptr != '>') { if (!*m_ptr) { // Premature end! return XMLParser::kMalformedElement; } m_ptr++; } if (m_ptr == nameStart) { // Empty attribute name? return XMLParser::kMalformedElement; } Stringp attributeName = unescape(m_source, nameStart, (int)(m_ptr-nameStart), true); while (String::isSpace(*m_ptr)) { m_ptr++; } if (*m_ptr != '=') { // No '=' sign, no attribute value, error! return XMLParser::kMalformedElement; } else { // Skip over whitespace. while (String::isSpace(*++m_ptr)) ; const wchar *attrStart = m_ptr; // Extract the attribute value. if (*m_ptr != '"' && *m_ptr != '\'') { // Error; no opening quote for attribute value. return XMLParser::kMalformedElement; } wchar delimiter = *m_ptr; // Extract up to the next quote. attrStart++; while (*++m_ptr != delimiter) { if (*m_ptr == '<') { // '<' is not permitted in an attribute value // Changed this from kMalformedElement to kUnterminatedAttributeValue for bug 117058(105422) return XMLParser::kUnterminatedAttributeValue; } if (!*m_ptr) { // If at end of file, // we have an unterminated attribute value on our hands. return XMLParser::kUnterminatedAttributeValue; } } const wchar *attrEnd = m_ptr; m_ptr++; Stringp attributeValue = unescape(m_source, attrStart, (int)(attrEnd-attrStart), false); AvmAssert (attributeName->isInterned()); tag.attributes.add(attributeName); tag.attributes.add(attributeValue); } } // Advance past the end > of this element. if (*m_ptr == '>') { m_ptr++; } return XMLParser::kNoError; }

void avmplus::XMLParser::parse (  Stringp  source, bool  ignoreWhite = false )

Definition at line 490 of file XMLParser16.cpp. References avmplus::String::c_str(), m_ignoreWhite, and m_source. Referenced by avmplus::XMLObject::XMLObject(). { m_source = source; m_ptr = m_source->c_str(); m_ignoreWhite = ignoreWhite; }

void avmplus::XMLParser::setCondenseWhite (  bool  flag  )  [inline]

Definition at line 137 of file XMLParser16.h. References m_condenseWhite. Referenced by avmplus::XMLObject::XMLObject(). { m_condenseWhite = flag; }

Stringp avmplus::XMLParser::unescape (  Stringp  buffer, const wchar *  start, int  len, bool  bIntern )  [private]

Definition at line 380 of file XMLParser16.cpp. References util::threadpool::i. Referenced by getNext(). { bool bUseSubString = true; for (int i = 0; i < len; i++) { if (startChar[i] == '&') { bUseSubString = false; break; } } if (bUseSubString) { if (bIntern) { return core->internAlloc (startChar, len); } else { MMgc::GC* gc = MMgc::GC::GetGC(text); int start = (int)(startChar - text->c_str()); AvmAssert (start < text->length()); return new (gc) String (text, start, len); } } MMgc::GC* gc = MMgc::GC::GetGC(text); Stringp news = new (gc) String (startChar, len); wchar *buffer = news->lockBuffer(); // Remove XML &#xx; escape entities, and &lt; &gt; &amp; &apos; wchar *dst = buffer; wchar *src = buffer; while (*src) { if (*src == '&') { bool success = false; // Scan forward to the ';' wchar *endPtr = src; while (*endPtr && *endPtr != ';') { endPtr++; } if (*endPtr) { *endPtr = 0; int len = (int)(endPtr-src-1); if (*(src+1) == '#') { // Parse a &#xx; decimal sequence. Or a &#xDD hex sequence double value = MathUtils::parseInt(src+2, len-1); if (MathUtils::isNaN(value)) { if (len > 2 && src[2] == 'x') { // Handle xFF hex encoded tags, too value = MathUtils::parseInt(src+3, len-2, 16); } } if (!MathUtils::isNaN(value)) { *dst++ = (wchar) (int) value; success = true; } } else if (len <= 4) // Our xmlEntities are only 4 characters or less { Atom entityAtom = core->internAlloc(src+1, len)->atom(); Atom result = core->xmlEntities->get(entityAtom); if (result != undefinedAtom) { *dst++ = (wchar)(result>>3); success = true; } } *endPtr = ';'; } if (success) { // If successful, advance past the sequence src = endPtr+1; } else { // Otherwise copy the sequence literally *dst++ = *src++; } } else { *dst++ = *src++; } } *dst = 0; news->unlockBuffer((int)(dst-buffer)); return (bIntern) ? core->internString (news) : news; }

---

Member Data Documentation

AvmCore* avmplus::XMLParser::core

Definition at line 134 of file XMLParser16.h. Referenced by getNext(), XMLParser(), and ~XMLParser().

bool avmplus::XMLParser::m_condenseWhite [private]

Definition at line 146 of file XMLParser16.h. Referenced by getCondenseWhite(), getNext(), setCondenseWhite(), and ~XMLParser().

bool avmplus::XMLParser::m_ignoreWhite [private]

Definition at line 145 of file XMLParser16.h. Referenced by getNext(), parse(), and ~XMLParser().

const wchar* avmplus::XMLParser::m_ptr [private]

Definition at line 143 of file XMLParser16.h. Referenced by getNext(), and ~XMLParser().

Stringp avmplus::XMLParser::m_source [private]

Definition at line 142 of file XMLParser16.h. Referenced by getNext(), parse(), and ~XMLParser().

---

The documentation for this class was generated from the following files:

• tamarin-central/core/XMLParser16.h
• tamarin-central/core/XMLParser16.cpp

---