void
nsHTMLContentSerializer::AppendToString(const nsAString& aStr,
nsAString& aOutputStr,
PRBool aTranslateEntities,
PRBool aIncrColumn)
{
if (mBodyOnly && !mInBody) {
return;
}
if (aIncrColumn) {
mColPos += aStr.Length();
}
if (aTranslateEntities && !mInCDATA) {
if (mFlags & (nsIDocumentEncoder::OutputEncodeBasicEntities |
nsIDocumentEncoder::OutputEncodeLatin1Entities |
nsIDocumentEncoder::OutputEncodeHTMLEntities |
nsIDocumentEncoder::OutputEncodeW3CEntities)) {
nsIParserService* parserService = nsContentUtils::GetParserService();
if (!parserService) {
NS_ERROR("Can't get parser service");
return;
}
nsReadingIterator<PRUnichar> done_reading;
aStr.EndReading(done_reading);
// for each chunk of |aString|...
PRUint32 advanceLength = 0;
nsReadingIterator<PRUnichar> iter;
const char **entityTable = mInAttribute ? kAttrEntities : kEntities;
for (aStr.BeginReading(iter);
iter != done_reading;
iter.advance(PRInt32(advanceLength))) {
PRUint32 fragmentLength = iter.size_forward();
PRUint32 lengthReplaced = 0; // the number of UTF-16 codepoints
// replaced by a particular entity
const PRUnichar* c = iter.get();
const PRUnichar* fragmentStart = c;
const PRUnichar* fragmentEnd = c + fragmentLength;
const char* entityText = nsnull;
nsCAutoString entityReplacement;
char* fullEntityText = nsnull;
advanceLength = 0;
// for each character in this chunk, check if it
// needs to be replaced
for (; c < fragmentEnd; c++, advanceLength++) {
PRUnichar val = *c;
if (val == kValNBSP) {
entityText = kEntityNBSP;
break;
}
else if ((val <= kGTVal) && (entityTable[val][0] != 0)) {
entityText = entityTable[val];
break;
} else if (val > 127 &&
((val < 256 &&
mFlags & nsIDocumentEncoder::OutputEncodeLatin1Entities) ||
mFlags & nsIDocumentEncoder::OutputEncodeHTMLEntities)) {
parserService->HTMLConvertUnicodeToEntity(val, entityReplacement);
if (!entityReplacement.IsEmpty()) {
entityText = entityReplacement.get();
break;
}
}
else if (val > 127 &&
mFlags & nsIDocumentEncoder::OutputEncodeW3CEntities &&
mEntityConverter) {
if (NS_IS_HIGH_SURROGATE(val) &&
c + 1 < fragmentEnd &&
NS_IS_LOW_SURROGATE(*(c + 1))) {
PRUint32 valUTF32 = SURROGATE_TO_UCS4(val, *(++c));
if (NS_SUCCEEDED(mEntityConverter->ConvertUTF32ToEntity(valUTF32,
nsIEntityConverter::entityW3C, &fullEntityText))) {
lengthReplaced = 2;
break;
}
else {
advanceLength++;
}
}
else if (NS_SUCCEEDED(mEntityConverter->ConvertToEntity(val,
nsIEntityConverter::entityW3C,
&fullEntityText))) {
lengthReplaced = 1;
break;
}
}
}
aOutputStr.Append(fragmentStart, advanceLength);
if (entityText) {
aOutputStr.Append(PRUnichar('&'));
AppendASCIItoUTF16(entityText, aOutputStr);
aOutputStr.Append(PRUnichar(';'));
advanceLength++;
}
// if it comes from nsIEntityConverter, it already has '&' and ';'
else if (fullEntityText) {
AppendASCIItoUTF16(fullEntityText, aOutputStr);
nsMemory::Free(fullEntityText);
advanceLength += lengthReplaced;
}
}
} else {
nsXMLContentSerializer::AppendToString(aStr, aOutputStr, aTranslateEntities, aIncrColumn);
}
return;
}
aOutputStr.Append(aStr);
}
void
CopyASCIItoUTF16(const nsACString& aSource, nsAString& aDest)
{
aDest.Truncate();
AppendASCIItoUTF16(aSource, aDest);
}
nsresult
nsHTMLContentSerializer::EscapeURI(const nsAString& aURI, nsAString& aEscapedURI)
{
// URL escape %xx cannot be used in JS.
// No escaping if the scheme is 'javascript'.
if (IsJavaScript(nsGkAtoms::href, aURI)) {
aEscapedURI = aURI;
return NS_OK;
}
// nsITextToSubURI does charset convert plus uri escape
// This is needed to convert to a document charset which is needed to support existing browsers.
// But we eventually want to use UTF-8 instead of a document charset, then the code would be much simpler.
// See HTML 4.01 spec, "Appendix B.2.1 Non-ASCII characters in URI attribute values"
nsCOMPtr<nsITextToSubURI> textToSubURI;
nsAutoString uri(aURI); // in order to use FindCharInSet()
nsresult rv = NS_OK;
if (!mCharset.IsEmpty() && !IsASCII(uri)) {
textToSubURI = do_GetService(NS_ITEXTTOSUBURI_CONTRACTID, &rv);
NS_ENSURE_SUCCESS(rv, rv);
}
PRInt32 start = 0;
PRInt32 end;
nsAutoString part;
nsXPIDLCString escapedURI;
aEscapedURI.Truncate(0);
// Loop and escape parts by avoiding escaping reserved characters (and '%', '#' ).
while ((end = uri.FindCharInSet("%#;/?:@&=+$,", start)) != -1) {
part = Substring(aURI, start, (end-start));
if (textToSubURI && !IsASCII(part)) {
rv = textToSubURI->ConvertAndEscape(mCharset.get(), part.get(), getter_Copies(escapedURI));
NS_ENSURE_SUCCESS(rv, rv);
}
else {
escapedURI.Adopt(nsEscape(NS_ConvertUTF16toUTF8(part).get(), url_Path));
}
AppendASCIItoUTF16(escapedURI, aEscapedURI);
// Append a reserved character without escaping.
part = Substring(aURI, end, 1);
aEscapedURI.Append(part);
start = end + 1;
}
if (start < (PRInt32) aURI.Length()) {
// Escape the remaining part.
part = Substring(aURI, start, aURI.Length()-start);
if (textToSubURI) {
rv = textToSubURI->ConvertAndEscape(mCharset.get(), part.get(), getter_Copies(escapedURI));
NS_ENSURE_SUCCESS(rv, rv);
}
else {
escapedURI.Adopt(nsEscape(NS_ConvertUTF16toUTF8(part).get(), url_Path));
}
AppendASCIItoUTF16(escapedURI, aEscapedURI);
}
return rv;
}
void
nsString::AppendWithConversion( const nsACString& aData )
{
AppendASCIItoUTF16(aData, *this);
}