Terminal::~Terminal() called qDeleteAll(m_conductors_list) on the live
member. Each Conductor destructor calls removeConductor() on both of its
terminals, and that removes the conductor from the same list qDeleteAll
is iterating. Mutating a QList while iterating it is undefined
behaviour; with two or more conductors on one terminal (terminal strips,
bridged terminals) it can skip a delete or delete one conductor twice,
which leaves another conductor's terminal1/terminal2 pointing at freed
memory.
The pattern dates from a00404bc9 (2021), which replaced a foreach loop
(iterating an implicit copy) with a direct qDeleteAll. It went unnoticed
until the deterministic sort keys added to Diagram::toXml() in #844
started reading pos() on both terminals of every conductor on every
save, including the periodic backup, which turned the stale pointer into
an EXC_BAD_ACCESS in QGraphicsItem::pos() while deleting an element.
Copy the list first and delete from the copy, restoring the pre-2021
behaviour. An isolated regression test (a hub terminal with 2 to 8
conductors, under AddressSanitizer) did not trigger the failure with the
old code, so none is included; the crash analysis and that attempt are
recorded in jp2images/qelectrotech-source-mirror#1.
QString::toDouble() reports a successful conversion for "nan"/"inf"/
"-inf" -- confirmed directly -- so the existing conv_ok checks in
Element::valideXml() and Terminal::valideXml() never caught a
non-finite x/y. A NaN-positioned element reaching the scene can hang
QGraphicsScene::addItem() forever: an existing conductor's itemChange()
runs a collision test (calculateTextItemPosition() ->
QGraphicsItem::collidesWithPath()) whose underlying QPathClipper spins
without terminating when fed a NaN-valued QPainterPath, since NaN
breaks the ordering comparisons the clipping algorithm's termination
depends on (#781). Short of that, a non-finite value that doesn't
happen to trigger a collision test simply gets written straight back
out on save with nothing to stop it (#782).
Element::valideXml() and Terminal::valideXml() now also check
qIsFinite() on the parsed x/y, rejecting the whole item the same way a
missing attribute already does. DynamicElementTextItem::fromXml() has
no such reject-the-item gate (void return, no caller check), so its x/y
are clamped to 0 instead -- the same fallback the attribute lookup
already uses when x/y is missing entirely.
Verified against both original findings' exact repro steps:
- #781: Habitat-Unifilaire.qet with x="nan" on one element -- hung
(SIGTERM'd by a 25s timeout) on an unfixed build, resaves cleanly
(exit 0) on this one.
- #782: grafcet.qet with y="nan" on a dynamic_elmt_text -- the value
passed straight through to the resaved file on an unfixed build;
clamped to 0 on this one. The specific field is now stable (y="0" on
two consecutive resaves) where it read "nan" both times before.
(grafcet.qet has an unrelated, already-known, unmerged fix
(PR #779) for element/terminal-order non-determinism, so a whole-file
diff across resaves still differs for reasons unconnected to this
change -- checked the specific once-NaN field in isolation instead.)
- Also checked -inf on affuteuse_250h.qet: same rejection, same result.
Full qet-dbcheck.py sweep of the unmutated example corpus (23
projects), 0 regressions.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
The conductor table keyed on Terminal::uuid(), which comes from the catalog
.elmt definition and is empty for every element authored before that field
existed. A conductor was dropped unless *both* its terminals had one, so the
tables this slice adds were empty on almost every project in existence:
examples corpus conductor rows in the database
industrial.qet 0 of 671
affuteuse_250h.qet 0 of 263
tremie_vibrante.qet 0 of 77
741.qet 0 of 67
Across the 23 example projects, 16 of the 20 that contain conductors have
zero terminal uuids -- 2366 of 3002 conductors -- and overall coverage is
7.3%. Meanwhile --export-cables, already on master, lists all 671 conductors
of industrial.qet from the document. A feature that only works on newly
authored elements is not one users can rely on.
Terminal::stableUuid() returns the terminal's own uuid when it has one and
otherwise derives one from its local position and orientation inside its
element. That is not an invented scheme: it is what the project format
already does. TerminalData::fromXml() says so where it parses the field --
"if the attribute not exists, means, the element is created with an older
version of qet. So use the legacy approach to identify terminals" -- and the
legacy approach is the terminal's position. m_pos is read from the definition
and is not touched by moving the element on a folio, so the identity survives
loads, saves and folio moves. Derived values are UUID v5 in a fixed namespace,
so they are reproducible without being stored, and cannot collide with the v4
uuids the element editor generates.
Every project in the corpus now has exactly as many conductor rows as the
document has conductors -- 20 of 20 measured, 0 mismatches. (schema_indus.qet
is excluded: it blocks on a modal dialog at zero CPU under any CLI flag, the
pre-existing hang PR #661 addresses.)
Two things this deliberately does not key on:
- The terminal name. It is not stable: QET rewrites a terminal named "_" as
unnamed, which would have silently changed the identity of 1421 of
industrial.qet's 1790 terminals on their first resave. Measured across the
corpus, dropping it costs nothing -- geometry alone yields exactly the same
three collisions -- and it means renaming a terminal no longer changes what
it is.
- Uniqueness in the face of a definition that declares two terminals at the
same point and orientation. Three cases exist in the whole corpus. They
merge to a single terminal row, which is harmless: two terminals identical
in position and orientation are indistinguishable in every observable
respect, and every conductor on either still resolves to the right element
and terminal name. Both affected projects (industrial, perceuse) return
their full conductor count.
The only conductor still skipped is one whose terminal has no parent element,
which has no identity to key on at all.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
Terminal stores its TerminalData* as member d but never deletes it.
Every Element creation (placing on diagram, loading icon for the
element browser, drag previews) leaks one TerminalData per terminal.
ASan confirmed 112 leaked objects (9856 bytes) in a short session
across four call sites all rooted in Element::parseTerminal.
Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
Extend TerminalData::Type enum with three new semantic values:
- No : Normally Open terminal of a switch (SW) contact
- Nc : Normally Closed terminal of a switch (SW) contact
- Common : Common terminal of a switch (SW) contact
Update typeToString() and typeFromString() accordingly.
Fully backward compatible: existing Generic/Inner/Outer types
are unchanged. Elements without typed terminals fall back
to the previous behavior (first 2 named terminals).
terminal: expose terminalType() as public accessor
Add Terminal::terminalType() returning the TerminalData::Type
of this terminal. This allows crossrefitem and other consumers
to filter terminals by semantic role (No, Nc, Common) without
accessing TerminalData internals directly.
terminaleditor: add No, Nc, Common entries to type combobox
Expose the three new TerminalData types (No, Nc, Common) in
the element editor UI so users can assign a semantic role to
each terminal of a SW contact element.
Also fix a pre-existing bug in updateForm() where m_type_cb
was incorrectly using m_orientation_cb->findData() instead
of m_type_cb->findData(), preventing the type from being
restored correctly when selecting a terminal.
terminaleditor: add No, Nc, Common entries to type combobox
Expose the three new TerminalData types (No, Nc, Common) in
the element editor UI so users can assign a semantic role to
each terminal of a SW contact element.
Also fix a pre-existing bug in updateForm() where m_type_cb
was incorrectly using m_orientation_cb->findData() instead
of m_type_cb->findData(), preventing the type from being
restored correctly when selecting a terminal.
When reading and comparing Qt5-docs and Qt6-docs, I read,
that there are no differences in the functions!
So it is not necessary to have the differentiation
between the Qt-Versions.
Code compiles without errors or warnings for Qt5 and Qt6.
clazy is a compiler plugin which allows clang to understand Qt
semantics. You get more than 50 Qt related compiler warnings, ranging
from unneeded memory allocations to misusage of API, including fix-its
for automatic refactoring.
https://invent.kde.org/sdk/clazy
Jeff Patterson