summaryrefslogtreecommitdiffstats
path: root/src/corelib/serialization
Commit message (Collapse)AuthorAgeFilesLines
* QDataStream: remove description from Version enum values' docsAhmad Samir2025-10-301-44/+44
| | | | | | | | | | | | | | | | | | | It's redundant; the repeated values clearly show that the version doesn't necessarily change with each Qt release: Constant Value QDataStream::Qt_6_5 Qt_6_0 QDataStream::Qt_6_6 21 QDataStream::Qt_6_7 22 QDataStream::Qt_6_8 Qt_6_7 This also makes updating the enum values simpler, and avoids bugs like QTBUG-116886. Fixes: QTBUG-116886 Pick-to: 6.10 6.8 6.5 Change-Id: Ie89c29417e5c901f693f0fcc1906fde46f39d27e Reviewed-by: Kai Köhne <kai.koehne@qt.io> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QCborValue: rewrite nextUtf32Character to avoid narrowing char32_t → char16_tMarc Mutz2025-10-191-12/+12
| | | | | | | | | | | | | | | | | | | | | | | | Ought to fix Clang 21 -Wcharacter-conversion in this function. Instead of storing the initial *ptr in r.c (a char32_t) and then narrowing it back to char16_t for the call to QChar::surrogateToUcs4(), store it in a separate variable and construct an R only in the return statement. Also move the len = 1 return from the end of the function to the (inverted) if (c < 0x80) near the start of the function. This is less assmebler-esque, so might execute slower (though it's all values, so the optimizer should have no problem rewirting this to the old code, if that was faster), but is much easier to read, which is not the sneezed at in a security-critical component. Amends d4c7da9a07dc1434692fe08a61ba22c794574c4f. Pick-to: 6.10 6.8 6.5 Change-Id: I1ce610fa80e3d874106c9c5b3a1d50bf99f39732 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com> Reviewed-by: Ahmad Samir <a.samirh78@gmail.com>
* QJsonParseError: fix clang -Wshorten-64-to-32 warningsAhmad Samir2025-10-162-2/+6
| | | | | | | | | Change it to qint64 in Qt7, as requested in code review, so as to make it easier to extend it in the future. Pick-to: 6.10 6.8 6.5 Change-Id: If893d719dd0457fac46fdd37ff95ce95222b1858 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Mark remaining files in src/corelib/serialization as security-insignificantMarc Mutz2025-10-152-0/+2
| | | | | | | | | | | | | | | | | | | | | | | | | | The .gitignore file is obviously insignificant. The file format allows comments, so mark it. Invent a new reason string for that. The shell script is also insignificant. While it runs qlalr for the security-critical QtXmlStream classes, as a build-tool, it's not security-critical itself, and, unlike e.g. util/normalize/main.cpp, it doesn't itself contain code that ends up being compiled. Invent a new reason string for that. This completes the review of src/corelib/serialization: $ for i in $(find src/corelib/serialization -type f); do if ! grep -qE '^ *(//|#) *Qt[ -]Security +score:' "$i"; then echo "$i" fi done <nothing> QUIP: 23 Fixes: QTBUG-135194 Pick-to: 6.10 6.8 Change-Id: Id5d18244fe0e9d18b8891500a3a946ac530671a4 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* Mark the remaining cbor/json headers as security-significantMarc Mutz2025-10-133-0/+3
| | | | | | | | | | | | | | | They only contain declarations; qcborstream.h only includes other headers without interfering with their contents, so even though the components are security-critical, these headers get the header-decls-only get-out-of-jail-free card. Amends cd4813a3c304a6d9d5d8096d91a8fc297818d624. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.8 Change-Id: I67bf7f8c1fdc5e6881df37207236eb3c294535ab Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* Mark qcborcommon_p.h as security-criticalMarc Mutz2025-10-131-0/+1
| | | | | | | | | | | | | | | This might be a bit controversial, but this header configures the API of TinyCBor for use in Qt, so it's not just declarations. TinyCBor, as third-party component, is listed as security-critical in its qt_attribution.json, so I opted to mark the glue code as critial, too. Amends cd4813a3c304a6d9d5d8096d91a8fc297818d624. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.8 Change-Id: I0042a8677eff716eb1924dfc211c740d164029cd Reviewed-by: Matthias Rauter <matthias.rauter@qt.io>
* Mark qcbor{map,array}.cpp as security-criticalMarc Mutz2025-10-132-0/+2
| | | | | | | | | | | | | QJsonArray/Map are, too, because the whole resp. component is, and the implementation in those two files cannot be said to be trivial. Amends cd4813a3c304a6d9d5d8096d91a8fc297818d624. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.8 Change-Id: I08dfc1dc1f917832f0cd24d3b575095c5e930a4d Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* Add missing nullptr guard to QTextStreamPrivate::setupDevice(QIODevice*)Christian Heimlich2025-10-101-6/+8
| | | | | | | | | | | | Amends commit e3c290e1947515992821e6bf97d74d65c9254271, which missed this. The guard is necessary to prevent a connection attempt with a nullptr when QTextStream::setDevice(QIODevice*) is called with a nullptr as a means to dissociate it from its current device without attaching it to a new one. Pick-to: 6.10 Change-Id: I51d417fc9d3549b60c1f9e3e5b35e26b2d8fb1a9 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Doc: Remove duplicate "see also" linksDavid Boddie2025-09-263-9/+5
| | | | | | | | | | Also make certain links specific to avoid ambiguity. Pick-to: 6.9 6.10 Change-Id: I77499a1c9c4f4356e2f9957092c16c6572a13569 Task-number: QTBUG-137048 Reviewed-by: Paul Wicking <paul.wicking@qt.io> Reviewed-by: Marc Mutz <marc.mutz@qt.io>
* Move all Q_DECL_(PURE|CONST)_FUNCTION markers to the frontMårten Nordheim2025-09-195-14/+14
| | | | | | | | | While testing defining Q_DECL_PURE_FUNCTION for MSVC I discovered it cannot handle having the marker trailing, while GCC/clang can cope with either. Change-Id: Ic9dd088996c191263e2f2a43c2f6a16533bdb9c5 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QJsonObject: fix -Werror,-Wshorten-64-to-32 on 32-bit platformsMarc Mutz2025-09-091-2/+2
| | | | | | | | | | | | | | The `index` member is a quint64 bitfield, so on 32-bit platforms, where qsizetype is just int, the return expression, indeed, narrows, causing warnings, which may break the build if -Werror is in effect (either for the Qt build itself, or, seeing as this is a public header, for user projects). Pick-to: 6.10 6.9 6.8 6.5 Task-number: QTBUG-139951 Change-Id: I06602ce85234207cf90e1acbbbba4d89998ac399 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com> Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QDataStream: fix -Werror,-Wshorten-64-to-32 on 32-bit ClangMarc Mutz2025-09-051-3/+3
| | | | | | | | | | | | | | | | | | On 32-bit platforms qsizetype is int, so assigning a qint64 to it narrows. It's benign, because we perform the overflow check in the next line, but that doesn't help -Werror. Fix by making the narrowing explicit. As a drive-by, mark `n` const. Amends fd48ce0b73c74dafd5db27bc1f2752ef665df7ef. Reported-by: Karim Pinter <karim.pinter@qt.io> Pick-to: 6.10 6.9 6.8 Change-Id: I14d95c54c501df14a0b76d11262f22605fbb583b Reviewed-by: Øystein Heskestad <oystein.heskestad@qt.io> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Use QUtf8Functions::nextUcs4FromUtf8() in the JSON parserMarc Mutz2025-08-291-6/+6
| | | | | | | | | | | | Can't say this makes the function much simpler, and I didn't find a Coverity issue about this use, but be consistent and use the new function here, too. Besides, if we port Parser to qchar8_t, then it will be simpler, and we might be able to remove the scanUtf8 function completely. Pick-to: 6.10 6.9 6.8 Change-Id: Ib6fa7aac14b942b4002163dccc77513a71378ea1 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QUtf8Functions: wrap fromUtf8() in a simpler APIMarc Mutz2025-08-271-5/+11
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | Coverity had an issue with how we called this function to get the next character from a UTF-8 sequence. In particular, it complained about passing the address of a char32_t object to fromUtf8(), which "treats it as an array", as Coverity notes. The C++ standard says this is fine¹, but callers of this function, specifically, are almost by definition security-critical, so we shouldn't leave Coverity issues unfixed in them. The fix is to use an array of one element instead: char32_t[1]. We have already applied this fix in qstring.cpp (4eb9e0d3eedfc1b6de968308167af01b19f6ffe7) and qurlrecode.cpp (7a32a2238f52217bc4f0dc4c9620a2a2d350a1ca), but there are more cases in QXmlStream and QStringConverter, some of which this patch fixes. In order to simplify the task, package up the fix in a small function for easier reuse and so that the proverbial maintenance programmer is less likely to undo these fixes again. Incidentally, this function makes those callers that can use it much more readable than before. Many other calls (incl. the one in qurlrecode.cpp and some others in qstringconverter.cpp) cannot be ported over, as they are using the return value of fromUtf8(), or have already advanced the input pointer themselves when they call fromUtf8(). Amends 94c62e322264e2e7d61193ae74ba8556a330385c, ea0a08c898fed9cfd8d8eb16613e352740d3eb02 and b977ae371a753a82e1d0bb32c5b62099da663721. ¹ https://eel.is/c++draft/basic.compound#3.sentence-11 Pick-to: 6.10 6.9 6.8 Coverity-Id: 378348 Coverity-Id: 403740 Coverity-Id: 403748 Coverity-Id: 459945 Change-Id: I4957ea2ee7b6e0e9c361eb2bcd4351708762a6e9 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QJsonDocument/Value: fix integer truncation in fromJson(QByteArray)Marc Mutz2025-08-254-6/+6
| | | | | | | | | | | | | | | | | | | | | | | | | The old code passed the QByteArray's size() to the QJsonPrivate::Parser ctor's length argument, which, however, was still 'int', truncating. Making matters worse, the Parser ctor only used the length to calculate an end iterator. Fix by passing by QUtf8StringView. This also indicates that the Parser expects UTF-8 input (and not Latin1, e.g.). As a drive-by, make the Parser ctor explicit. [ChangeLog][QtCore][QJsonDocument/QJsonValue] Fixed a bug on 64-bit platforms where fromJson(QByteArray) could report one of the Unterminated errors for valid input whose size merely exceeded INT_MAX (2GiB). Amends the port of QByteArray to qsizetype (Qt 6.0, didn't check exact commit). Pick-to: 6.10 6.9 6.8 6.5 Task-number: QTBUG-103541 Change-Id: I3e4f7c098a98e8840e15ff6cc63bee2121118129 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Doc: Remove spurious charactersDavid Boddie2025-08-201-1/+0
| | | | | | Pick-to: 6.10 Change-Id: Ibfd907bf8ea80063cb896d2ba36cfd2a2ebfa241 Reviewed-by: Paul Wicking <paul.wicking@qt.io>
* QXmlStream: don't use Structured BindingsMarc Mutz2025-08-201-4/+4
| | | | | | | | | | | | | | | | | | | | | Decoder::operator() already returns a proper struct with named members, so decomposing that using SB can do no good, except introducing a dependency on the order of the struct, which is counter-productive. Fix by receiving the result in a proper variable and using the struct's member names. Amends 53622aca2ad0d13bd16d8307dc28f915c8878b75. The idea may have been to avoid churning the surrounding code, but follow-up c08766abf2267f4eeda159c5688309d22d319722 touched all lines using the result, anyway, so we don't need SB anymore to keep the old names. Pick-to: 6.10 Change-Id: I2890754c2c2189cc2fa395f4500ca6c3cb115d7d Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QXmlStreamWriter: replace Q_ASSERT() with static_assert()Marc Mutz2025-08-191-1/+1
| | | | | | | | | | | The condition is a compile-time constant expression, so we can check it at compile-time. No run-time check needed. Amends 53622aca2ad0d13bd16d8307dc28f915c8878b75. Pick-to: 6.10 Change-Id: Ia73dd99c3670a279ec7f740eecd623d51ae647d4 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QXmlStream: silence clang.exe warningTim Blechmann2025-08-061-1/+1
| | | | | | | | StreamEOF is an enum, but compared against uint instances, emitting signed/unsigned comparison warnings when building with clang.exe using the MSVC ABI. Using a constexpr variable of the correct type to fix the warning. Pick-to: 6.10 Change-Id: I1fd29e0bcc46fc8a75c7024605ba68f68df82af4 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QXmlStreamWriter: use QASV in raiseError()Ivan Solovev2025-07-282-5/+5
| | | | | | | | | | | Amends 53622aca2ad0d13bd16d8307dc28f915c8878b75. Found in 6.10 API review. Pick-to: 6.10 Change-Id: Ib53d4306f6345befd39d1273ae44997b41163b78 Reviewed-by: Marc Mutz <marc.mutz@qt.io> Reviewed-by: Magdalena Stojek <magdalena.stojek@qt.io>
* QTextStream: remove unneeded includemocsMarc Mutz2025-07-211-4/+0
| | | | | | | | | | There is no QObject subclass in qtextstream_p.h anymore. Amends e3c290e1947515992821e6bf97d74d65c9254271. Pick-to: 6.10 Change-Id: If0d3030cbbc3306cd71dfcb793c89379611749e7 Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
* QTextStreamPrivate: fix a particularly nasty Bool TrapMarc Mutz2025-07-212-15/+17
| | | | | | | | | | | | | | | | | | | | | | | | | | | | The bool number = false parameter of putString() overloads is not only a classical Bool Trap (ie. doesn't give any hint to the reader of a call what 'true' or 'false' means), it also, in combination with the (ptr, n, bool=false) and (view, bool=false) overloading may cause a (ptr, n) argument list to accidentally match (view, number) instead: if decltype(ptr) isn't (const) QChar*, but one of the views accepts the type in it's (Char*) ctor, the call might be solved as (View(ptr), bool number(n)). This is not acceptable for a security-critical component. To fix, introduce a strong enum, PutStringMode { String, Number } to replace the bool. Only after this change it is safe to remove the (unused) (ptr, n, bool) overload, for the above-mentioned reason, so do it now. The alternative was to split putString() into putString() and putNumber(), but there's already a putNumber(qulonglong, bool negative), so I opted for an enum argument instead. Fixes: QTBUG-138520 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I78d1a9132d5ef41e33b9462d48d2f41401e1f688 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com> Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QTextStream: remove QDeviceClosedNotifierThiago Macieira2025-07-202-55/+36
| | | | | | | | | Because we have (have had) the new-style connections which allow us to bypass the need for a receiver QObject. Pick-to: 6.10 Change-Id: I6cc5368310bb902d99d6fffd67d98bd67463c99c Reviewed-by: Marc Mutz <marc.mutz@qt.io>
* QTextStream: prefer QStringView overload of Private::putString()Marc Mutz2025-07-191-2/+2
| | | | | | | | | | | | | This call is safer than the (ptr, n, bool=false) one, which may accidentally resolve to (view, bool) if decltype(ptr) is "wrong". This is in preparation of removing said putString() overload (and the bool parameter). Task-number: QTBUG-138520 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I93ff89322468558f7da723ae47a1d67852cc59e6 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QTextStream: delay QByteArray → QString conversion in op<<()Marc Mutz2025-07-191-1/+1
| | | | | | | | | | | | | ... until putString(QUtf8StringView), which, at some point, may be optimized to not create a QString. This way, we centralize the conversion in a single place. The data comes from a QByteArray, so is always a valid range. Task-number: QTBUG-138520 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: Ie6db2b1a254b737af286d7eb4e531829410b92c7 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QTextStream: Extract Method writeImpl() from write() overloadsMarc Mutz2025-07-192-17/+14
| | | | | | | | | | They're all the same, except the type of the thing being append()ed to one of the two QStrings, so make the actual implementation a template. Task-number: QTBUG-138520 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I0a3041984b690cc0febac932b53bd6ec601e8780 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* Prefer QTextStreamPrivate::write(QStringView) over (ptr, n) [2/2]: restMarc Mutz2025-07-192-5/+4
| | | | | | | | | | | | | | | | | Replace the remaining calls to write(const QChar *, qsizetype) with calls that use QStringView. Requires adapting a QChar[2] (not null-terminate, and QStringView constructor scan for one). Chose to make it a char16_t literal instead, and make it constexpr as a drive-by. This removes the last caller of the write(p, n) overload, so remove it. Task-number: QTBUG-138520 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: Iddfadc7cd7837a541e5840247f8812398807bcfe Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* Doc: Fix linking in QDS::operator bool()Kai Köhne2025-07-181-1/+1
| | | | | | | | | Fixes corelib/serialization/qdatastream.cpp:1419: (qdoc) warning: Can't link to 'OK' Pick-to: 6.10 Change-Id: I81a32031accbac35e730c452afeb9be453419180 Reviewed-by: Tatiana Borisova <tatiana.borisova@qt.io>
* QTextStream: cope with multi-code-point signsMarc Mutz2025-07-161-4/+21
| | | | | | | | | | | | | | | | | | | | | | | | | | | | QTextStreamPrivate::putString(), in numeric mode, and with accounting-style alignment, parses the string to check whether it starts with the current locale's negativeSign() or positiveSign(), which, since Qt 6.0, are QStrings, when in Qt 5 they were mere QChars (which was wrong). The old code still assumed Qt 5 times and merely compared the front() of the string (the first character) with the locale's QString. This works for locales where the plus/minus signs are just one UTF-16 code point, but not for those that require surrogate pairs or are just plain more than one Unicode code point long, like ar_EG (Arabic as written in Egypt). Fix by using startsWith() instead of front() ==. [ChangeLog][QtCore][QTextStream] Fixed QTextStream::FieldAlignment::AlignAccountingStyle for locales that have negativeSign/positiveSign (-/+) that take more than one UTF-16 code point (e.g. ar (Arabic)). Fixes: QTBUG-138484 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I6120460cb2ea8ce201bca5ba404cdaea442b0cb6 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com> Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QTextStream: Extract Method putStringImpl()Marc Mutz2025-07-162-21/+17
| | | | | | | | | | | | The putString() implementations for QStringView and QLatin1StringView are now almost identical, so replace them with a template implementation, and save ourselves some duplicate effort fixing bugs in both places hereafter. Task-number: QTBUG-138484 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I23051be146970348b195d5b1dea8779a11047018 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QTextStream: swap roles of Private::putString(QStringView) and (ptr, size)Marc Mutz2025-07-152-13/+10
| | | | | | | | | | | | Implement the (ptr, size) overload in terms of the QStringView one. This makes the QStringView and QLatin1StringView overloads almost token-for-token identical, allowing to Extract Method on them in the next step, in preparation of fixing QTBUG-138484. Task-number: QTBUG-138484 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I75abfd29ff98ba79a92555631f4ac7f231571a32 Reviewed-by: Edward Welbourne <edward.welbourne@qt.io>
* QTextStream: use write(QChar) moreMarc Mutz2025-07-142-4/+4
| | | | | | | | | | | | | | | | | | | | Replace write(&ch, 1) calls with write(ch) ones. The QChar overload, added in 31c6d9f04ba80e6776ae01cb0588e6bd228f7184, is safer, because it has less preconditions. With QTextStream being security-critical, we should target to use safer functions to do the same work, even if it introduces a bit of git history churn. The call from a different TU (qdebug.cpp) means we need to remove the lying inline keywords from both declaration and definition of write(QChar), because GCC -Werrors on them. As a drive-by, remove a use of QLatin1Char, in preparation of moving from QChar to char16_t, eventually. Pick-to: 6.10 6.9 6.8 Change-Id: I97b4c8c68a4ee0d4b821c4d2ec4a9dd6f44dceb0 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Add move SMFs to QCbor{Map,Array}Marc Mutz2025-07-046-0/+50
| | | | | | | | | | | | | | | | Coverity (rightfully) complained that QCborMap is lacking move SMFs. Add them, also for QCborArray. For move-assignment, use pure swap, which matches what QCborValue and the QJson classes do. Coverity-Id: 218756 Pick-to: 6.10 Task-number: QTBUG-53264 Change-Id: I9327b884e6e1ac08254552dfcf64b006f9a963a2 Reviewed-by: Ahmad Samir <a.samirh78@gmail.com> Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QXmlStream: fix incorrect use of realloc()Marc Mutz2025-07-011-4/+6
| | | | | | | | | | | | | | | | | One must never store the result of realloc() back into the first realloc() argument, because, if relloc() fails, one still has to free() the old pointer, but one has just overwritten its value with nullptr, leaking memory. To fix, first check, and then assign. Found in manual review of realloc() uses. Amends the start of the public history. Pick-to: 6.10 6.9 6.8 6.5 Change-Id: Id6d369116d4e01fb9cb1d6db3627b9956d4b6201 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QXmlStreamSimpleStack: Q_CHECK_PTR() the correct pointerMarc Mutz2025-07-011-1/+1
| | | | | | | | | | | | | | | | | | | | | | Coverity complains that the QXmlStreamPrivateTagStack default ctor dereferences nullptr. I haven't understood, yet, why it thinks that, but manual inspection revealed a Q_CHECK_PTR() in QXmlStreamSimpleStack::reserve(), called from push(), fingered by Coverity, that comes too late: If Q_CHECK_PTR() is configured to throw std::bad_alloc, then we'd leak the previous buffer if realloc() failed, because we had already overwritten the old value. Fix by moving the Q_CHECK_PTR to after the realloc() and before the assignment of the realloc() result to the data member. Fingers cross that this fixes Coverity's complaint, too. Amends 403343039d07812c0beee9260b291f86e14d8ac4. Coverity-Id: 474174 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: Ib3564c34626950b927aa54df50abeecc42e8e674 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QXmlStreamReader: fix a missing std::move()Marc Mutz2025-07-011-1/+1
| | | | | | | | | | | | | | | | Coverity (rightfully) complains that the assignment uses the copy- and not the move-assignment operator. Fix by adding the missing move(). Amends 927798f5deb0bd95fc8ce5be9fb0cfceef94a46b. Coverity-Id: 481486 Pick-to: 6.10 Change-Id: I9044847abb88adaaa55f49e7578290d11ddb8655 Reviewed-by: Magdalena Stojek <magdalena.stojek@qt.io> Reviewed-by: Ivan Solovev <ivan.solovev@qt.io> Reviewed-by: Giuseppe D'Angelo <giuseppe.dangelo@kdab.com>
* Mark QTextStream as security-criticalMarc Mutz2025-06-273-0/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This component processes input external to the process, and the application depends on this component to report errors to detect failure, so this component is security-critical. The headers, however, contain only a miniscule amount of implementation, essentially a few forwarders each and, in the public header, the QTextStreamManipulator, everything else is out-of-line, so I've opted to mark both headers as significant, header, declarations-only. I'll de-inline what makes sense in a follow-up commit. Amends 8df072fc8006510c9b743e8ffedaaf51a876883a. This completes the review of src/corelib/serialization, building on the following prequel commits: - 0b9d4c3a857144bad2937f964101c57ae7efec4c (QDataStream) - 0a7ee06b27c55f10b65b053cabdc893ae8f23893 (QXmlStream{Reader,Writer}) - 046e6c5701da96dd64b613c2bc9cfd245c83b1a5 (QXmlUtils) - cd4813a3c304a6d9d5d8096d91a8fc297818d624 (QJson and QCbor) There are no other implementation files in this subdir at the time of commit. QUIP: 23 Fixes: QTBUG-135194 Pick-to: 6.10 6.9 6.8 Change-Id: If5264e3ad1eca230f48042888b1256ce0a329d8e Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
* Mark QXmlUtils as security-criticalMarc Mutz2025-06-262-0/+2
| | | | | | | | | | | | | | | | | | QXmlUtils are used by QXmlStream and QDom, both of which are (or ought soon to be) marked as security-critical. The component is clearly a data-parser, too, with the same input as QXmlStream and QDom, so has to be security-critical, too. The header file contains only declarations, so it gets the default score:significant. Amends 8df072fc8006510c9b743e8ffedaaf51a876883a. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.9 6.8 Change-Id: I415486dbba0c748e6af561ea4f098ea42c4b1830 Reviewed-by: Ivan Solovev <ivan.solovev@qt.io>
* Mark QJson and QCbor classes as security-criticalMarc Mutz2025-06-2625-0/+25
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | These components process input external to the process, and the application depends on these components to report errors to detect failure, so these components are security-critical. The writers are, too, because a valid use of the classes is to stream-read something, filter or edit a few nodes, and then stream back, in which case writers see the same input as the readers, only in reverse, and have to build a valid output. Where there's doubt whether a header should be considered critical or not, err on the side of critical. We can always revisit the decision later when we know what semantics we'll attach to this marking. Where a header only contains declarations (and class definitions), mark as significant with reason "header, declarations only". It is probably a good idea to avoid any non-trivial inline code in headers whose .cpp files are security-critical, but moving all inline code out-of-line is a task for a different patch. Amends 8df072fc8006510c9b743e8ffedaaf51a876883a. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.9 6.8 Change-Id: Ie317c4df652430e6e68954c37d553b760836dff8 Reviewed-by: Matthias Rauter <matthias.rauter@qt.io>
* Doc: Document QTextStream::setCodec removal in Qt 6 changesDheerendra Purohit2025-06-251-0/+1
| | | | | | | | | | Add a paragraph about the removal of QTextStream::setCodec() and the introduction of setEncoding() in Qt 6. Pick-to: 6.10 6.9 Fixes: QTBUG-95325 Change-Id: I0785dcb493a5dcab83c85e59cb549a998ee25af0 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* QJsonObject/QCborMap: Add asKeyValueRange()Johannes Grunenberg2025-06-184-0/+270
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | This adds {QJsonObject,QCborMap}::asKeyValueRange() which returns a range over key-value pairs of the object/map (`pair<QAnyStringView, QJsonValueRef>` and `pair<QCborValueConstRef, QCborValue>`). This uses QKeyValueIterator under the hood. QJsonObject's iterator only iterates over the items, so using it in a range-based for loop won't give users access to the key. With `asKeyValueRange` one can iterate over both keys and values and conveniently use structured bindings. QCborMap's iterator already iterates over key-value pairs, so `asKeyValueRange` is provided for API symmetry. In `QKeyValueIterator`, this adds a fourth template parameter `Traits` to support custom `key()` and `value()` functions. This is specifically needed for `QJsonObject`, as its actualy key is a string view, but `key()` returns a `QString`. [ChangeLog][QtCore][QJsonObject] Added asKeyValueRange to iterate with a range-based for loop over key-value pairs with support for structured bindings. [ChangeLog][QtCore][QCborMap] Added asKeyValueRange to iterate with a range-based for loop over key-value pairs with support for structured bindings. Pick-to: 6.10 Change-Id: I68d97fada8b2d7ef7224f1beb5aa685aac3d1b16 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Mark QDataStream as security-criticalMarc Mutz2025-06-172-0/+2
| | | | | | | | | | | | | | | | | QDataStream clearly parses data, both in the .cpp file and the header. The class is used for over the network data communication in QtRemoteObjects, so this obviously needs to be security-critical, because de-facto, users will use this class in such ways, too. Amends 8df072fc8006510c9b743e8ffedaaf51a876883a. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.10 6.9 6.8 Change-Id: I02a14e3c3fa1e008a532dd2e752b667cdc254e8e Reviewed-by: Juha Vuolle <juha.vuolle@qt.io> Reviewed-by: Volker Hilsheimer <volker.hilsheimer@qt.io>
* QDataStream: add operator bool()Tatiana Borisova2025-06-172-6/+12
| | | | | | | | | | | | | | | | | | - Add operator bool() that checks whether the associated stream has no error status. This operator makes it possible to use streams and functions that return references to streams as loop conditions: `while (stream >> data)`. - Update existing testcases that use or can use the status of stream operations. [ChangeLog][QtCore][QDataStream] Added implicit conversion to bool, returning `status() == Ok`. Found in API-review (added for symmetry with QTextStream). Pick-to: 6.10 Change-Id: I3eb8251f40eba1a8164e088fa10de670564f428e Reviewed-by: Marc Mutz <marc.mutz@qt.io>
* QXmlStreamWriter: Clear errorString when resetting error codeMagdalena Stojek2025-06-161-0/+1
| | | | | | | | | | | | This change ensures that errorString is explicitly cleared when resetting the error code to NoError. Resulted from API review. Pick-to: 6.10 Change-Id: I182575df7a058045ea58424b64063cd6026fb21d Reviewed-by: Ivan Solovev <ivan.solovev@qt.io> Reviewed-by: Marc Mutz <marc.mutz@qt.io>
* Change enum value names in QXmlStreamWriter::Error enum classMagdalena Stojek2025-06-162-27/+27
| | | | | | | | | | | Renamed enum values to avoid redundant naming. Resulted from API review. Pick-to: 6.10 Change-Id: I5213695c02a763d4689c5df97b20c282368b4fbd Reviewed-by: Edward Welbourne <edward.welbourne@qt.io> Reviewed-by: Marc Mutz <marc.mutz@qt.io>
* Rework tst_QDate{,Time} QDataStream test to include all versionsEdward Welbourne2025-06-101-0/+4
| | | | | | | | | | | | | | | | In tst_QDate the test previously had a hard-coded list of versions by name. This, however, needed updates on each new release (or at least new version of QDS), which have been neglected since Qt 5.0. The tst_QDateTime version, meanwhile, only identified versions by their numeric value, not their symbolic names. So borrow the way tst_QCalendar enumerates QCalendar::System, making QDataStream a Q_GADGET and its Version a Q_ENUM to support this, to iterate over all versions algorithmically. Apply to both data stream tests to get a more consistent form for their data tags. Change-Id: I04d1060cfec822cfcdc0bce6d15b636fc165fe36 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Add support for font features and variable axes to QTextCharFormatEskil Abrahamsen Blomfeldt2025-06-061-1/+1
| | | | | | | | | | These can be set on the font directly, but had not been added to QTextCharFormat, so there would be no way to override them by formatting in a rich text document. Fixes: QTBUG-134060 Change-Id: I4494e24cb9b99d84fb376ba895e2461fc3cd054b Reviewed-by: Eirik Aavitsland <eirik.aavitsland@qt.io>
* Bump version to 6.11.0Jani Heikkinen2025-06-052-2/+4
| | | | | Change-Id: I14ec3a08bf704535b9b10e33ec8ee0f9415f2d7c Reviewed-by: Qt Submodule Update Bot <qt_submodule_update_bot@qt-project.org>
* Update QJsonValue(qint64) docs to reflect full precisionEdward Welbourne2025-06-041-3/+8
| | | | | | | | | | | Amends commit 35adb74ddd915831789f0175423660f8e898942e (from 5.15) to reflect the improved precision of 64-bit int when stored as a JSON value. Task-number: QTBUG-28560 Pick-to: 6.10 6.9 6.8 6.5 Change-Id: I07915bc190dd7696e58356143e8857f91e888c67 Reviewed-by: Thiago Macieira <thiago.macieira@intel.com>
* Mark QXmlStream{Reader,Writer} as security-criticalMarc Mutz2025-05-217-0/+8
| | | | | | | | | | | | | | | | If QXmlStream isn't security-criticial, what is? qxmlstream.h contains the definition of the QXmlString work-horse, so it has to be security-critical, too (until we remove it). Amends 8df072fc8006510c9b743e8ffedaaf51a876883a. QUIP: 23 Task-number: QTBUG-135194 Pick-to: 6.9 6.8 Change-Id: Ib366e63fb89aa0b69ad437f6688285b2c390c5c1 Reviewed-by: Ivan Solovev <ivan.solovev@qt.io> Reviewed-by: Kai Köhne <kai.koehne@qt.io>
generated by cgit v1.2.3 (git 2.25.1) at 2025-11-29 20:25:39 +0000