527 lines
14 KiB
C++
527 lines
14 KiB
C++
|
// Copyright (c) 1994, 1995 James Clark
|
||
|
// See the file COPYING for copying permission.
|
||
|
|
||
|
#include "splib.h"
|
||
|
#include "Parser.h"
|
||
|
#include "MessageArg.h"
|
||
|
#include "token.h"
|
||
|
#include "macros.h"
|
||
|
#include "ParserMessages.h"
|
||
|
|
||
|
#ifdef SP_NAMESPACE
|
||
|
namespace SP_NAMESPACE {
|
||
|
#endif
|
||
|
|
||
|
Boolean Parser::parseAttributeSpec(Mode mode,
|
||
|
AttributeList &atts,
|
||
|
Boolean &netEnabling,
|
||
|
Ptr<AttributeDefinitionList> &newAttDef)
|
||
|
|
||
|
{
|
||
|
unsigned specLength = 0;
|
||
|
AttributeParameter::Type curParm;
|
||
|
|
||
|
if (!parseAttributeParameter(mode, 0, curParm, netEnabling))
|
||
|
return 0;
|
||
|
while (curParm != AttributeParameter::end) {
|
||
|
switch (curParm) {
|
||
|
case AttributeParameter::name:
|
||
|
{
|
||
|
Text text;
|
||
|
text.addChars(currentInput()->currentTokenStart(),
|
||
|
currentInput()->currentTokenLength(),
|
||
|
currentLocation());
|
||
|
size_t nameMarkupIndex;
|
||
|
if (currentMarkup())
|
||
|
nameMarkupIndex = currentMarkup()->size() - 1;
|
||
|
text.subst(*syntax().generalSubstTable(), syntax().space());
|
||
|
if (!parseAttributeParameter(mode == piPasMode ? asMode : mode, 1, curParm, netEnabling))
|
||
|
return 0;
|
||
|
if (curParm == AttributeParameter::vi) {
|
||
|
specLength += text.size() + syntax().normsep();
|
||
|
if (!parseAttributeValueSpec(mode == piPasMode ? asMode : mode, text.string(), atts,
|
||
|
specLength, newAttDef))
|
||
|
return 0;
|
||
|
// setup for next attribute
|
||
|
if (!parseAttributeParameter(mode, 0, curParm, netEnabling))
|
||
|
return 0;
|
||
|
}
|
||
|
else {
|
||
|
if (currentMarkup())
|
||
|
currentMarkup()->changeToAttributeValue(nameMarkupIndex);
|
||
|
if (!handleAttributeNameToken(text, atts, specLength))
|
||
|
return 0;
|
||
|
}
|
||
|
}
|
||
|
break;
|
||
|
case AttributeParameter::nameToken:
|
||
|
{
|
||
|
Text text;
|
||
|
text.addChars(currentInput()->currentTokenStart(),
|
||
|
currentInput()->currentTokenLength(),
|
||
|
currentLocation());
|
||
|
text.subst(*syntax().generalSubstTable(), syntax().space());
|
||
|
if (!handleAttributeNameToken(text, atts, specLength))
|
||
|
return 0;
|
||
|
if (!parseAttributeParameter(mode, 0, curParm, netEnabling))
|
||
|
return 0;
|
||
|
}
|
||
|
break;
|
||
|
case AttributeParameter::recoverUnquoted:
|
||
|
{
|
||
|
if (!atts.recoverUnquoted(currentToken(), currentLocation(), *this)) {
|
||
|
// Don't treat it as an unquoted attribute value.
|
||
|
currentInput()->endToken(1);
|
||
|
if (!atts.handleAsUnterminated(*this))
|
||
|
message(ParserMessages::attributeSpecCharacter,
|
||
|
StringMessageArg(currentToken()));
|
||
|
return 0;
|
||
|
}
|
||
|
if (!parseAttributeParameter(mode, 0, curParm, netEnabling))
|
||
|
return 0;
|
||
|
}
|
||
|
break;
|
||
|
default:
|
||
|
CANNOT_HAPPEN();
|
||
|
}
|
||
|
}
|
||
|
atts.finish(*this);
|
||
|
if (specLength > syntax().attsplen())
|
||
|
message(ParserMessages::attsplen,
|
||
|
NumberMessageArg(syntax().attsplen()),
|
||
|
NumberMessageArg(specLength));
|
||
|
return 1;
|
||
|
}
|
||
|
|
||
|
Boolean Parser::handleAttributeNameToken(Text &text,
|
||
|
AttributeList &atts,
|
||
|
unsigned &specLength)
|
||
|
{
|
||
|
unsigned index;
|
||
|
if (!atts.tokenIndex(text.string(), index)) {
|
||
|
if (atts.handleAsUnterminated(*this))
|
||
|
return 0;
|
||
|
atts.noteInvalidSpec();
|
||
|
message(ParserMessages::noSuchAttributeToken,
|
||
|
StringMessageArg(text.string()));
|
||
|
}
|
||
|
else if (sd().www() && !atts.tokenIndexUnique(text.string(), index)) {
|
||
|
atts.noteInvalidSpec();
|
||
|
message(ParserMessages::attributeTokenNotUnique,
|
||
|
StringMessageArg(text.string()));
|
||
|
}
|
||
|
else {
|
||
|
if (!sd().attributeOmitName())
|
||
|
message(ParserMessages::attributeNameShorttag);
|
||
|
else if (options().warnMissingAttributeName)
|
||
|
message(ParserMessages::missingAttributeName);
|
||
|
atts.setSpec(index, *this);
|
||
|
atts.setValueToken(index, text, *this, specLength);
|
||
|
}
|
||
|
return 1;
|
||
|
}
|
||
|
|
||
|
Boolean Parser::parseAttributeValueSpec(Mode mode,
|
||
|
const StringC &name,
|
||
|
AttributeList &atts,
|
||
|
unsigned &specLength,
|
||
|
Ptr<AttributeDefinitionList> &newAttDef)
|
||
|
{
|
||
|
Markup *markup = currentMarkup();
|
||
|
Token token = getToken(mode);
|
||
|
if (token == tokenS) {
|
||
|
if (markup) {
|
||
|
do {
|
||
|
markup->addS(currentChar());
|
||
|
token = getToken(mode);
|
||
|
} while (token == tokenS);
|
||
|
}
|
||
|
else {
|
||
|
do {
|
||
|
token = getToken(mode);
|
||
|
} while (token == tokenS);
|
||
|
}
|
||
|
}
|
||
|
unsigned index;
|
||
|
if (!atts.attributeIndex(name, index)) {
|
||
|
if (newAttDef.isNull())
|
||
|
newAttDef = new AttributeDefinitionList(atts.def());
|
||
|
AttributeDefinition *newDef = 0;
|
||
|
if (!inInstance()) {
|
||
|
// We are parsing a data attribute specification
|
||
|
Ptr<Notation> notation;
|
||
|
Dtd::NotationIter notationIter(currentDtdNonConst().notationIter());
|
||
|
for (;;) {
|
||
|
notation = notationIter.next();
|
||
|
if (notation.isNull()
|
||
|
|| atts.def() == notation->attributeDef())
|
||
|
break;
|
||
|
}
|
||
|
ASSERT(!notation.isNull());
|
||
|
if (!notation->defined()) {
|
||
|
Notation *nt =
|
||
|
lookupCreateNotation(syntax().rniReservedName(Syntax::rIMPLICIT));
|
||
|
ConstPtr<AttributeDefinitionList> common = nt->attributeDef();
|
||
|
if (!common.isNull() && common->attributeIndex(name, index)) {
|
||
|
newDef = common->def(index)->copy();
|
||
|
newDef->setSpecified(1);
|
||
|
}
|
||
|
}
|
||
|
if (!newDef) {
|
||
|
Notation *nt =
|
||
|
lookupCreateNotation(syntax().rniReservedName(Syntax::rALL));
|
||
|
ConstPtr<AttributeDefinitionList> common = nt->attributeDef();
|
||
|
if (!common.isNull() && common->attributeIndex(name, index)) {
|
||
|
newDef = common->def(index)->copy();
|
||
|
newDef->setSpecified(0);
|
||
|
}
|
||
|
}
|
||
|
}
|
||
|
if (!newDef) {
|
||
|
if (!implydefAttlist())
|
||
|
message(ParserMessages::noSuchAttribute, StringMessageArg(name));
|
||
|
newDef = new ImpliedAttributeDefinition(name,
|
||
|
new CdataDeclaredValue);
|
||
|
}
|
||
|
newAttDef->append(newDef);
|
||
|
atts.changeDef(newAttDef);
|
||
|
index = atts.size() - 1;
|
||
|
}
|
||
|
atts.setSpec(index, *this);
|
||
|
Text text;
|
||
|
switch (token) {
|
||
|
case tokenUnrecognized:
|
||
|
if (reportNonSgmlCharacter())
|
||
|
return 0;
|
||
|
// fall through
|
||
|
case tokenEtago:
|
||
|
case tokenStago:
|
||
|
case tokenNestc:
|
||
|
message(ParserMessages::unquotedAttributeValue);
|
||
|
extendUnquotedAttributeValue();
|
||
|
if (markup)
|
||
|
markup->addAttributeValue(currentInput());
|
||
|
text.addChars(currentInput()->currentTokenStart(),
|
||
|
currentInput()->currentTokenLength(),
|
||
|
currentLocation());
|
||
|
break;
|
||
|
case tokenEe:
|
||
|
if (mode != piPasMode) {
|
||
|
message(ParserMessages::attributeSpecEntityEnd);
|
||
|
return 0;
|
||
|
}
|
||
|
case tokenTagc:
|
||
|
case tokenDsc:
|
||
|
case tokenVi:
|
||
|
message(ParserMessages::attributeValueExpected);
|
||
|
return 0;
|
||
|
case tokenNameStart:
|
||
|
case tokenDigit:
|
||
|
case tokenLcUcNmchar:
|
||
|
if (!sd().attributeValueNotLiteral())
|
||
|
message(ParserMessages::attributeValueShorttag);
|
||
|
else if (options().warnAttributeValueNotLiteral)
|
||
|
message(ParserMessages::attributeValueNotLiteral);
|
||
|
extendNameToken(syntax().litlen() >= syntax().normsep()
|
||
|
? syntax().litlen() - syntax().normsep()
|
||
|
: 0,
|
||
|
ParserMessages::attributeValueLength);
|
||
|
if (markup)
|
||
|
markup->addAttributeValue(currentInput());
|
||
|
text.addChars(currentInput()->currentTokenStart(),
|
||
|
currentInput()->currentTokenLength(),
|
||
|
currentLocation());
|
||
|
break;
|
||
|
case tokenLit:
|
||
|
case tokenLita:
|
||
|
Boolean lita;
|
||
|
lita = (token == tokenLita);
|
||
|
if (!(atts.tokenized(index)
|
||
|
? parseTokenizedAttributeValueLiteral(lita, text)
|
||
|
: parseAttributeValueLiteral(lita, text)))
|
||
|
return 0;
|
||
|
if (markup)
|
||
|
markup->addLiteral(text);
|
||
|
break;
|
||
|
default:
|
||
|
CANNOT_HAPPEN();
|
||
|
}
|
||
|
return atts.setValue(index, text, *this, specLength);
|
||
|
}
|
||
|
|
||
|
|
||
|
Boolean Parser::parseAttributeParameter(Mode mode,
|
||
|
Boolean allowVi,
|
||
|
AttributeParameter::Type &result,
|
||
|
Boolean &netEnabling)
|
||
|
{
|
||
|
Token token = getToken(mode);
|
||
|
Markup *markup = currentMarkup();
|
||
|
if (mode == piPasMode) {
|
||
|
for (;;) {
|
||
|
switch (token) {
|
||
|
case tokenCom:
|
||
|
if (!parseComment(comMode))
|
||
|
return 0;
|
||
|
if (options().warnPsComment)
|
||
|
message(ParserMessages::psComment);
|
||
|
// fall through
|
||
|
case tokenS:
|
||
|
token = getToken(mode);
|
||
|
continue;
|
||
|
default:
|
||
|
break;
|
||
|
}
|
||
|
break;
|
||
|
}
|
||
|
}
|
||
|
else if (markup) {
|
||
|
while (token == tokenS) {
|
||
|
markup->addS(currentChar());
|
||
|
token = getToken(mode);
|
||
|
}
|
||
|
}
|
||
|
else {
|
||
|
while (token == tokenS)
|
||
|
token = getToken(mode);
|
||
|
}
|
||
|
switch (token) {
|
||
|
case tokenUnrecognized:
|
||
|
if (reportNonSgmlCharacter())
|
||
|
return 0;
|
||
|
extendUnquotedAttributeValue();
|
||
|
result = AttributeParameter::recoverUnquoted;
|
||
|
break;
|
||
|
case tokenEe:
|
||
|
if (mode != piPasMode) {
|
||
|
message(ParserMessages::attributeSpecEntityEnd);
|
||
|
return 0;
|
||
|
}
|
||
|
result = AttributeParameter::end;
|
||
|
break;
|
||
|
case tokenEtago:
|
||
|
case tokenStago:
|
||
|
if (!sd().startTagUnclosed())
|
||
|
message(ParserMessages::unclosedStartTagShorttag);
|
||
|
result = AttributeParameter::end;
|
||
|
currentInput()->ungetToken();
|
||
|
netEnabling = 0;
|
||
|
break;
|
||
|
case tokenNestc:
|
||
|
if (markup)
|
||
|
markup->addDelim(Syntax::dNESTC);
|
||
|
switch (sd().startTagNetEnable()) {
|
||
|
case Sd::netEnableNo:
|
||
|
message(ParserMessages::netEnablingStartTagShorttag);
|
||
|
break;
|
||
|
case Sd::netEnableImmednet:
|
||
|
if (getToken(econnetMode) != tokenNet)
|
||
|
message(ParserMessages::nestcWithoutNet);
|
||
|
currentInput()->ungetToken();
|
||
|
break;
|
||
|
case Sd::netEnableAll:
|
||
|
break;
|
||
|
}
|
||
|
netEnabling = 1;
|
||
|
result = AttributeParameter::end;
|
||
|
break;
|
||
|
case tokenTagc:
|
||
|
if (markup)
|
||
|
markup->addDelim(Syntax::dTAGC);
|
||
|
netEnabling = 0;
|
||
|
result = AttributeParameter::end;
|
||
|
break;
|
||
|
case tokenDsc:
|
||
|
if (markup)
|
||
|
markup->addDelim(Syntax::dDSC);
|
||
|
result = AttributeParameter::end;
|
||
|
break;
|
||
|
case tokenNameStart:
|
||
|
extendNameToken(syntax().namelen(), ParserMessages::nameTokenLength);
|
||
|
if (markup)
|
||
|
markup->addName(currentInput());
|
||
|
result = AttributeParameter::name;
|
||
|
break;
|
||
|
case tokenDigit:
|
||
|
case tokenLcUcNmchar:
|
||
|
extendNameToken(syntax().namelen(), ParserMessages::nameTokenLength);
|
||
|
if (markup)
|
||
|
markup->addName(currentInput());
|
||
|
result = AttributeParameter::nameToken;
|
||
|
break;
|
||
|
case tokenLit:
|
||
|
case tokenLita:
|
||
|
message(allowVi
|
||
|
? ParserMessages::attributeSpecLiteral
|
||
|
: ParserMessages::attributeSpecNameTokenExpected);
|
||
|
return 0;
|
||
|
case tokenVi:
|
||
|
if (!allowVi) {
|
||
|
message(ParserMessages::attributeSpecNameTokenExpected);
|
||
|
return 0;
|
||
|
}
|
||
|
if (markup)
|
||
|
markup->addDelim(Syntax::dVI);
|
||
|
result = AttributeParameter::vi;
|
||
|
break;
|
||
|
default:
|
||
|
CANNOT_HAPPEN();
|
||
|
}
|
||
|
return 1;
|
||
|
}
|
||
|
|
||
|
void Parser::extendUnquotedAttributeValue()
|
||
|
{
|
||
|
InputSource *in = currentInput();
|
||
|
size_t length = in->currentTokenLength();
|
||
|
const Syntax &syn = syntax();
|
||
|
for (;;) {
|
||
|
Xchar c = in->tokenChar(messenger());
|
||
|
if (syn.isS(c)
|
||
|
|| !syn.isSgmlChar(c)
|
||
|
|| c == InputSource::eE
|
||
|
|| c == syn.delimGeneral(Syntax::dTAGC)[0])
|
||
|
break;
|
||
|
length++;
|
||
|
}
|
||
|
in->endToken(length);
|
||
|
}
|
||
|
|
||
|
Boolean Parser::parseAttributeValueLiteral(Boolean lita, Text &text)
|
||
|
{
|
||
|
size_t maxLength = (syntax().litlen() > syntax().normsep()
|
||
|
? syntax().litlen() - syntax().normsep()
|
||
|
: 0);
|
||
|
if (parseLiteral(lita ? alitaMode : alitMode, aliteMode,
|
||
|
maxLength,
|
||
|
ParserMessages::attributeValueLength,
|
||
|
literalNonSgml
|
||
|
| (wantMarkup() ? unsigned(literalDelimInfo) : 0),
|
||
|
text)) {
|
||
|
if (text.size() == 0
|
||
|
&& syntax().normsep() > syntax().litlen())
|
||
|
message(ParserMessages::attributeValueLengthNeg,
|
||
|
NumberMessageArg(syntax().normsep() - syntax().litlen()));
|
||
|
return 1;
|
||
|
}
|
||
|
else
|
||
|
return 0;
|
||
|
}
|
||
|
|
||
|
Boolean Parser::parseTokenizedAttributeValueLiteral(Boolean lita, Text &text)
|
||
|
{
|
||
|
size_t maxLength = (syntax().litlen() > syntax().normsep()
|
||
|
? syntax().litlen() - syntax().normsep()
|
||
|
: 0);
|
||
|
if (parseLiteral(lita ? talitaMode : talitMode, taliteMode,
|
||
|
maxLength,
|
||
|
ParserMessages::tokenizedAttributeValueLength,
|
||
|
literalSingleSpace
|
||
|
| (wantMarkup() ? unsigned(literalDelimInfo) : 0),
|
||
|
text)) {
|
||
|
if (text.size() == 0
|
||
|
&& syntax().normsep() > syntax().litlen())
|
||
|
message(ParserMessages::tokenizedAttributeValueLengthNeg,
|
||
|
NumberMessageArg(syntax().normsep() - syntax().litlen()));
|
||
|
return 1;
|
||
|
}
|
||
|
else
|
||
|
return 0;
|
||
|
}
|
||
|
|
||
|
|
||
|
Boolean Parser::skipAttributeSpec()
|
||
|
{
|
||
|
AttributeParameter::Type parm;
|
||
|
Boolean netEnabling;
|
||
|
if (!parseAttributeParameter(tagMode, 0, parm, netEnabling))
|
||
|
return 0;
|
||
|
while (parm != AttributeParameter::end) {
|
||
|
if (parm == AttributeParameter::name) {
|
||
|
size_t nameMarkupIndex = 0;
|
||
|
if (currentMarkup())
|
||
|
nameMarkupIndex = currentMarkup()->size() - 1;
|
||
|
if (!parseAttributeParameter(tagMode, 1, parm, netEnabling))
|
||
|
return 0;
|
||
|
if (parm == AttributeParameter::vi) {
|
||
|
Token token = getToken(tagMode);
|
||
|
while (token == tokenS) {
|
||
|
if (currentMarkup())
|
||
|
currentMarkup()->addS(currentChar());
|
||
|
token = getToken(tagMode);
|
||
|
}
|
||
|
switch (token) {
|
||
|
case tokenUnrecognized:
|
||
|
if (!reportNonSgmlCharacter())
|
||
|
message(ParserMessages::attributeSpecCharacter,
|
||
|
StringMessageArg(currentToken()));
|
||
|
return 0;
|
||
|
case tokenEe:
|
||
|
message(ParserMessages::attributeSpecEntityEnd);
|
||
|
return 0;
|
||
|
case tokenEtago:
|
||
|
case tokenStago:
|
||
|
case tokenNestc:
|
||
|
case tokenTagc:
|
||
|
case tokenDsc:
|
||
|
case tokenVi:
|
||
|
message(ParserMessages::attributeValueExpected);
|
||
|
return 0;
|
||
|
case tokenNameStart:
|
||
|
case tokenDigit:
|
||
|
case tokenLcUcNmchar:
|
||
|
if (!sd().attributeValueNotLiteral())
|
||
|
message(ParserMessages::attributeValueShorttag);
|
||
|
extendNameToken(syntax().litlen() >= syntax().normsep()
|
||
|
? syntax().litlen() - syntax().normsep()
|
||
|
: 0,
|
||
|
ParserMessages::attributeValueLength);
|
||
|
if (currentMarkup())
|
||
|
currentMarkup()->addAttributeValue(currentInput());
|
||
|
break;
|
||
|
case tokenLit:
|
||
|
case tokenLita:
|
||
|
{
|
||
|
Text text;
|
||
|
if (!parseLiteral(token == tokenLita ? talitaMode : talitMode,
|
||
|
taliteMode,
|
||
|
syntax().litlen(),
|
||
|
ParserMessages::tokenizedAttributeValueLength,
|
||
|
(currentMarkup() ? literalDelimInfo : 0)
|
||
|
| literalNoProcess,
|
||
|
text))
|
||
|
return 0;
|
||
|
if (currentMarkup())
|
||
|
currentMarkup()->addLiteral(text);
|
||
|
}
|
||
|
break;
|
||
|
default:
|
||
|
CANNOT_HAPPEN();
|
||
|
}
|
||
|
if (!parseAttributeParameter(tagMode, 0, parm, netEnabling))
|
||
|
return 0;
|
||
|
}
|
||
|
else {
|
||
|
if (currentMarkup())
|
||
|
currentMarkup()->changeToAttributeValue(nameMarkupIndex);
|
||
|
if (!sd().attributeOmitName())
|
||
|
message(ParserMessages::attributeNameShorttag);
|
||
|
}
|
||
|
}
|
||
|
else {
|
||
|
// It's a name token.
|
||
|
if (!parseAttributeParameter(tagMode, 0, parm, netEnabling))
|
||
|
return 0;
|
||
|
if (!sd().attributeOmitName())
|
||
|
message(ParserMessages::attributeNameShorttag);
|
||
|
}
|
||
|
}
|
||
|
if (netEnabling)
|
||
|
message(ParserMessages::startTagGroupNet);
|
||
|
return 1;
|
||
|
}
|
||
|
|
||
|
#ifdef SP_NAMESPACE
|
||
|
}
|
||
|
#endif
|