// SPDX-FileCopyrightText: 2023 James Graham
"));
}
// Linkify any plain text urls
m_dataBuffer = linkifyUrls(m_dataBuffer);
// Apply user style
m_dataBuffer.replace(TextRegex::userPill, QStringLiteral(R"(\1)"));
// Make all media URLs resolvable.
if (room && event) {
QRegularExpressionMatchIterator i = TextRegex::mxcImage.globalMatch(m_dataBuffer);
while (i.hasNext()) {
const QRegularExpressionMatch match = i.next();
const QUrl mediaUrl = room->makeMediaUrl(event->id(), QUrl(QStringLiteral("mxc://") + match.captured(2) + u'/' + match.captured(3)));
m_dataBuffer.replace(match.captured(0),
QStringLiteral("');
}
}
// Strip any disallowed tags/attributes.
QString outputString;
m_nextTokenType = nextTokenType(m_dataBuffer, m_pos, m_nextToken, m_nextTokenType);
while (m_pos < m_dataBuffer.length()) {
next();
QString nextTokenBuffer = m_nextToken;
if (m_nextTokenType == Type::Text || m_nextTokenType == Type::TextCode) {
nextTokenBuffer = escapeHtml(nextTokenBuffer);
} else if (m_nextTokenType == Type::Tag) {
if (!isAllowedTag(getTagType(m_nextToken))) {
nextTokenBuffer = QString();
} else if ((getTagType(m_nextToken) == QStringLiteral("br") && stripNewlines)) {
nextTokenBuffer = u' ';
}
nextTokenBuffer = cleanAttributes(getTagType(m_nextToken), nextTokenBuffer);
}
outputString.append(nextTokenBuffer);
m_nextTokenType = nextTokenType(m_dataBuffer, m_pos, m_nextToken, m_nextTokenType);
}
if (isEdited) {
if (outputString.endsWith(QStringLiteral("
%1
").arg(editString())); } else { outputString.append(editString()); } } /** * Replace tag.
if (nextTokenType == Text) {
int pos = 0;
while (pos < string.size()) {
pos = string.indexOf(u'<', pos);
if (pos == -1) {
pos = string.size();
} else {
const auto tagType = getTagType(string.mid(pos, string.indexOf(u'>', pos) - pos));
if (blockTags.contains(tagType)) {
return pos;
}
}
pos++;
}
return string.size();
}
int tagEndPos = string.indexOf(u'>');
QString tag = string.first(tagEndPos + 1);
QString tagType = getTagType(tag);
// If the start tag is not a block tag there can be only 1 block.
if (!blockTags.contains(tagType)) {
return string.size();
}
const auto closeTag = QStringLiteral("%1>").arg(tagType);
int closeTagPos = string.indexOf(closeTag);
// If the close tag can't be found assume malformed html and process as single block.
if (closeTagPos == -1) {
return string.size();
}
return closeTagPos + closeTag.size();
}
MessageComponent TextHandler::nextBlock(const QString &string,
int nextBlockPos,
Qt::TextFormat inputFormat,
const NeoChatRoom *room,
const Quotient::RoomEvent *event,
bool isEdited)
{
if (string.isEmpty()) {
return {};
}
int tagEndPos = string.indexOf(u'>');
QString tag = string.first(tagEndPos + 1);
QString tagType = getTagType(tag);
const auto messageComponentType = MessageComponentType::typeForTag(tagType);
QVariantMap attributes;
if (messageComponentType == MessageComponentType::Code) {
attributes = getAttributes(QStringLiteral("code"), string.mid(tagEndPos + 1, string.indexOf(u'>', tagEndPos + 1) - tagEndPos));
}
auto content = stripBlockTags(string.first(nextBlockPos), tagType);
setData(content);
switch (messageComponentType) {
case MessageComponentType::Code:
content = unescapeHtml(content);
break;
default:
content = handleRecieveRichText(inputFormat, room, event, false, isEdited);
}
return MessageComponent{messageComponentType, content, attributes};
}
QString TextHandler::stripBlockTags(QString string, const QString &tagType) const
{
if (blockTags.contains(tagType) && tagType != QStringLiteral("ol") && tagType != QStringLiteral("ul") && tagType != QStringLiteral("table")
&& string.startsWith(QLatin1String("<%1").arg(tagType))) {
string.remove(0, string.indexOf(u'>') + 1).remove(string.indexOf(QLatin1String("%1>").arg(tagType)), string.size());
}
if (string.startsWith(QStringLiteral("\n"))) {
string.remove(0, 1);
}
if (string.endsWith(QStringLiteral("\n"))) {
string.remove(string.size() - 1, string.size());
}
if (tagType == QStringLiteral("pre")) {
if (string.startsWith(QStringLiteral("') + 1);
string.remove(string.indexOf(QLatin1String("")), string.size());
}
if (string.endsWith(QStringLiteral("\n"))) {
string.remove(string.size() - 1, string.size());
}
}
if (tagType == QStringLiteral("blockquote")) {
if (string.startsWith(QStringLiteral("
"))) { string.remove(0, string.indexOf(u'>') + 1); string.remove(string.indexOf(QLatin1String("
")), string.size()); } if (!string.startsWith(u'"')) { string.prepend(u'"'); } if (!string.endsWith(u'"')) { string.append(u'"'); } } return string; } QString TextHandler::getTagType(const QString &tagToken) const { if (tagToken.isEmpty() || tagToken.length() < 2) { return QString(); } const int tagTypeStart = tagToken[1] == u'/' ? 2 : 1; const int tagTypeEnd = tagToken.indexOf(TextRegex::endTagType, tagTypeStart); return tagToken.mid(tagTypeStart, tagTypeEnd - tagTypeStart); } bool TextHandler::isCloseTag(const QString &tagToken) const { if (tagToken.isEmpty()) { return false; } return tagToken[1] == u'/'; } QString TextHandler::getAttributeType(const QString &string) { if (!string.contains(u'=')) { return string; } const int equalsPos = string.indexOf(u'='); return string.left(equalsPos); } QString TextHandler::getAttributeData(const QString &string, bool stripQuotes) { if (!string.contains(u'=')) { return QStringLiteral(); } const int equalsPos = string.indexOf(u'='); auto data = string.right(string.length() - equalsPos - 1); if (stripQuotes) { data = TextRegex::attributeData.match(data).captured(1); } return data; } bool TextHandler::isAllowedTag(const QString &type) { return allowedTags.contains(type); } bool TextHandler::isAllowedAttribute(const QString &tag, const QString &attribute) { return allowedAttributes[tag].contains(attribute); } bool TextHandler::isAllowedLink(const QString &link, bool isImg) { const QUrl linkUrl = QUrl(link); if (isImg) { return !linkUrl.isRelative() && linkUrl.scheme() == QStringLiteral("mxc"); } else { return !linkUrl.isRelative() && allowedLinkSchemes.contains(linkUrl.scheme()); } } QString TextHandler::cleanAttributes(const QString &tag, const QString &tagString) { int nextAttributeIndex = tagString.indexOf(u' ', 1); if (nextAttributeIndex != -1) { QString outputString = tagString.left(nextAttributeIndex); QString nextAttribute; int nextSpaceIndex; nextAttributeIndex += 1; while (nextAttributeIndex < tagString.length()) { nextSpaceIndex = tagString.indexOf(TextRegex::endTagType, nextAttributeIndex); if (nextSpaceIndex == -1) { nextSpaceIndex = tagString.length(); } nextAttribute = tagString.mid(nextAttributeIndex, nextSpaceIndex - nextAttributeIndex); if (isAllowedAttribute(tag, getAttributeType(nextAttribute))) { if (tag == QStringLiteral("img") && getAttributeType(nextAttribute) == QStringLiteral("src")) { QString attributeData = TextRegex::attributeData.match(getAttributeData(nextAttribute)).captured(1); if (isAllowedLink(attributeData, true)) { outputString.append(u' ' + nextAttribute); } } else if (tag == u'a' && getAttributeType(nextAttribute) == QStringLiteral("href")) { QString attributeData = TextRegex::attributeData.match(getAttributeData(nextAttribute)).captured(1); if (isAllowedLink(attributeData)) { outputString.append(u' ' + nextAttribute); } } else if (tag == QStringLiteral("code") && getAttributeType(nextAttribute) == QStringLiteral("class")) { if (getAttributeData(nextAttribute).remove(u'"').startsWith(QStringLiteral("language-"))) { outputString.append(u' ' + nextAttribute); } } else { outputString.append(u' ' + nextAttribute); } } nextAttributeIndex = nextSpaceIndex + 1; } outputString += u'>'; return outputString; } return tagString; } QVariantMap TextHandler::getAttributes(const QString &tag, const QString &tagString) { QVariantMap attributes; int nextAttributeIndex = tagString.indexOf(u' ', 1); if (nextAttributeIndex != -1) { QString nextAttribute; int nextSpaceIndex; nextAttributeIndex += 1; while (nextAttributeIndex < tagString.length()) { nextSpaceIndex = tagString.indexOf(TextRegex::endTagType, nextAttributeIndex); if (nextSpaceIndex == -1) { nextSpaceIndex = tagString.length(); } nextAttribute = tagString.mid(nextAttributeIndex, nextSpaceIndex - nextAttributeIndex); if (isAllowedAttribute(tag, getAttributeType(nextAttribute))) { if (tag == QStringLiteral("img") && getAttributeType(nextAttribute) == QStringLiteral("src")) { QString attributeData = TextRegex::attributeData.match(getAttributeData(nextAttribute)).captured(1); if (isAllowedLink(attributeData, true)) { attributes[getAttributeType(nextAttribute)] = getAttributeData(nextAttribute, true); } } else if (tag == u'a' && getAttributeType(nextAttribute) == QStringLiteral("href")) { QString attributeData = TextRegex::attributeData.match(getAttributeData(nextAttribute)).captured(1); if (isAllowedLink(attributeData)) { attributes[getAttributeType(nextAttribute)] = getAttributeData(nextAttribute, true); } } else if (tag == QStringLiteral("code") && getAttributeType(nextAttribute) == QStringLiteral("class")) { if (getAttributeData(nextAttribute).remove(u'"').startsWith(QStringLiteral("language-"))) { attributes[getAttributeType(nextAttribute)] = convertCodeLanguageString(getAttributeData(nextAttribute, true)); } } else { attributes[getAttributeType(nextAttribute)] = getAttributeData(nextAttribute, true); } } nextAttributeIndex = nextSpaceIndex + 1; } } return attributes; } QList")) == stringIn.left(index).count(QStringLiteral(""))) {
auto replacement = QStringLiteral("%1").arg(match.captured(2));
stringIn = stringIn.replace(index, match.captured(0).size(), replacement);
} else {
skip = match.captured().length();
}
}
start = index + skip;
match = {};
}
start = 0;
match = {};
for (int index = 0; index != -1; index = stringIn.indexOf(TextRegex::plainUrl, start, &match)) {
int skip = 0;
if (match.captured(0).size() > 0) {
if (stringIn.left(index).count(QStringLiteral("")) == stringIn.left(index).count(QStringLiteral(""))) {
auto replacement = QStringLiteral("%1").arg(match.captured(1));
stringIn = stringIn.replace(index, match.captured(0).size(), replacement);
skip = replacement.length();
} else {
skip = match.captured().length();
}
}
start = index + skip;
match = {};
}
start = 0;
match = {};
for (int index = 0; index != -1; index = stringIn.indexOf(TextRegex::emailAddress, start, &match)) {
int skip = 0;
if (match.captured(0).size() > 0) {
if (stringIn.left(index).count(QStringLiteral("")) == stringIn.left(index).count(QStringLiteral(""))) {
auto replacement = QStringLiteral("%1").arg(match.captured(2));
stringIn = stringIn.replace(index, match.captured(0).size(), replacement);
skip = replacement.length();
} else {
skip = match.captured().length();
}
}
start = index + skip;
match = {};
}
return stringIn;
}
QString TextHandler::editString() const
{
Kirigami::Platform::PlatformTheme *theme =
static_cast