diff options
author | Lars Knoll <lars.knoll@qt.io> | 2022-05-31 11:58:50 +0200 |
---|---|---|
committer | Qt Cherry-pick Bot <cherrypick_bot@qt-project.org> | 2022-06-25 15:26:08 +0000 |
commit | 1197280748b017cbf0c6234c7cb226eca73c00eb (patch) | |
tree | d4d16080891cb9263afd766d17d977436f31728a /src | |
parent | 24e7671e2014a9ec4b99a5abebeed0f0fb75bf48 (diff) |
Add QStringDecoder::decoderForHtml()
Now that QStringConverter can handle non UTF encodings through ICU,
add a way to get a decoder for arbitrary HTML code.
Opposed to QStringConverter::encodingForHtml(), this method will
try to create a valid string decoder also for non unicode codecs.
Change-Id: I343584da1b114396c744f482d9b433c9cedcc511
Reviewed-by: Fabian Kosmale <fabian.kosmale@qt.io>
Reviewed-by: Lars Knoll <lars.knoll@qt.io>
Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
(cherry picked from commit 9c1f3b6d4d5a5fe59396062c6f68cc1201665c62)
Reviewed-by: Qt Cherry-pick Bot <cherrypick_bot@qt-project.org>
Diffstat (limited to 'src')
-rw-r--r-- | src/corelib/kernel/qmimedata.cpp | 7 | ||||
-rw-r--r-- | src/corelib/text/qstringconverter.cpp | 64 | ||||
-rw-r--r-- | src/corelib/text/qstringconverter.h | 3 | ||||
-rw-r--r-- | src/widgets/widgets/qtextbrowser.cpp | 9 |
4 files changed, 60 insertions, 23 deletions
diff --git a/src/corelib/kernel/qmimedata.cpp b/src/corelib/kernel/qmimedata.cpp index 661a30b167..02a86bff37 100644 --- a/src/corelib/kernel/qmimedata.cpp +++ b/src/corelib/kernel/qmimedata.cpp @@ -127,10 +127,9 @@ QVariant QMimeDataPrivate::retrieveTypedData(const QString &format, QMetaType ty if (ba.isNull()) return QVariant(); if (format == "text/html"_L1) { - auto encoding = QStringConverter::encodingForHtml(ba); - if (encoding) { - QStringDecoder toUtf16(*encoding); - return QString(toUtf16(ba)); + QStringDecoder decoder = QStringDecoder::decoderForHtml(ba); + if (decoder.isValid()) { + return QString(decoder(ba)); } // fall back to utf8 } diff --git a/src/corelib/text/qstringconverter.cpp b/src/corelib/text/qstringconverter.cpp index ba13194310..38a434f353 100644 --- a/src/corelib/text/qstringconverter.cpp +++ b/src/corelib/text/qstringconverter.cpp @@ -2049,20 +2049,8 @@ QStringConverter::encodingForData(QByteArrayView data, char16_t expectedFirstCha return std::nullopt; } -/*! - Tries to determine the encoding of the HTML in \a data by looking at leading byte - order marks or a charset specifier in the HTML meta tag. If the optional is empty, - the encoding specified is not supported by QStringConverter. If no encoding is - detected, the method returns Utf8. -*/ -std::optional<QStringConverter::Encoding> QStringConverter::encodingForHtml(QByteArrayView data) +static QByteArray parseHtmlMetaForEncoding(QByteArrayView data) { - // determine charset - auto encoding = encodingForData(data); - if (encoding) - // trust the initial BOM - return encoding; - static constexpr auto metaSearcher = qMakeStaticByteArrayMatcher("meta "); static constexpr auto charsetSearcher = qMakeStaticByteArrayMatcher("charset="); @@ -2089,15 +2077,63 @@ std::optional<QStringConverter::Encoding> QStringConverter::encodingForHtml(QByt if (name == "unicode") // QTBUG-41998, ICU will return UTF-16. name = QByteArrayLiteral("UTF-8"); if (!name.isEmpty()) - return encodingForName(name); + return name; } } } } + return QByteArray(); +} + +/*! + Tries to determine the encoding of the HTML in \a data by looking at leading byte + order marks or a charset specifier in the HTML meta tag. If the optional is empty, + the encoding specified is not supported by QStringConverter. If no encoding is + detected, the method returns Utf8. + + \sa QStringDecoder::decoderForHtml() +*/ +std::optional<QStringConverter::Encoding> QStringConverter::encodingForHtml(QByteArrayView data) +{ + // determine charset + std::optional<QStringConverter::Encoding> encoding = encodingForData(data); + if (encoding) + // trust the initial BOM + return encoding; + + QByteArray encodingTag = parseHtmlMetaForEncoding(data); + if (!encodingTag.isEmpty()) + return encodingForName(encodingTag); + return Utf8; } /*! + Tries to determine the encoding of the HTML in \a data by looking at leading byte + order marks or a charset specifier in the HTML meta tag and returns a QStringDecoder + matching the encoding. If the returned decoder is not valid, + the encoding specified is not supported by QStringConverter. If no encoding is + detected, the method returns a decoder for Utf8. + + \sa isValid() +*/ +QStringDecoder QStringDecoder::decoderForHtml(QByteArrayView data) +{ + // determine charset + std::optional<QStringConverter::Encoding> encoding = encodingForData(data); + if (encoding) + // trust the initial BOM + return QStringDecoder(encoding.value()); + + QByteArray encodingTag = parseHtmlMetaForEncoding(data); + if (!encodingTag.isEmpty()) + return QStringDecoder(encodingTag); + + return QStringDecoder(Utf8); +} + + +/*! Returns the canonical name for encoding \a e. */ const char *QStringConverter::nameForEncoding(QStringConverter::Encoding e) diff --git a/src/corelib/text/qstringconverter.h b/src/corelib/text/qstringconverter.h index c1645f0938..6664beccb3 100644 --- a/src/corelib/text/qstringconverter.h +++ b/src/corelib/text/qstringconverter.h @@ -140,6 +140,9 @@ public: } return iface->toUtf16(out, ba, &state); } + + Q_CORE_EXPORT static QStringDecoder decoderForHtml(QByteArrayView data); + private: QString decodeAsString(QByteArrayView in) { diff --git a/src/widgets/widgets/qtextbrowser.cpp b/src/widgets/widgets/qtextbrowser.cpp index 0273331c40..2bab2efc61 100644 --- a/src/widgets/widgets/qtextbrowser.cpp +++ b/src/widgets/widgets/qtextbrowser.cpp @@ -288,12 +288,11 @@ void QTextBrowserPrivate::setSource(const QUrl &url, QTextDocument::ResourceType } else if (data.userType() == QMetaType::QByteArray) { QByteArray ba = data.toByteArray(); if (type == QTextDocument::HtmlResource) { - auto encoding = QStringConverter::encodingForHtml(ba); - if (!encoding) + auto decoder = QStringDecoder::decoderForHtml(ba); + if (!decoder.isValid()) // fall back to utf8 - encoding = QStringDecoder::Utf8; - QStringDecoder toUtf16(*encoding); - txt = toUtf16(ba); + decoder = QStringDecoder(QStringDecoder::Utf8); + txt = decoder(ba); } else { txt = QString::fromUtf8(ba); } |