Element::editProperty() gave PropertiesEditorDialog no size of its own,
so the dialog fell back to its sizeHint, which is sized for the compact
general-purpose editors it usually hosts. A PLC master instead shows a
six-column IO table that grows horizontally, and the dialog came out
too narrow to read those columns in.
For a master whose type is PLC, resize the dialog to three times its
natural width before exec(), keeping the natural height. The width is
clamped to the available screen so it cannot run off the display, and
every other element type opens exactly as before.
The IO table of MasterPropertiesWidget -- the panel that opens when a
PLC master placed on a schematic is edited -- set every section to
QHeaderView::Stretch. Stretch spreads the sections evenly over the
widget and disables section dragging altogether, so the column
boundaries were permanently fixed: they could be neither widened nor
narrowed, and only the width of the whole panel had any effect.
Use QHeaderView::Interactive instead, with movable sections and a set
of default widths, so the columns can be dragged as they are everywhere
else in the application. The layout reached this way is written to
QSettings under masterpropertieswidget/plc-table-header-state on every
sectionResized and sectionMoved, and restored the next time the table
is built -- the same header-state trick the free and linked element
trees of this widget already use, except saved automatically rather
than only from the context menu.
sources/ElementsCollection/fileelementcollectionitem.cpp had a German
tr() source string ("Makros") in a project whose source language is
French/English elsewhere. Renamed to "Macros" (identical in both
languages, so no translation catalog change is needed). Translated
three German-language comments in elementscollectionmodel.cpp and
diagramview.cpp to English.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
PR #983 added a hard SQLite3 system dependency to CMakeLists.txt, but
the workflow only installed libqt6sql6-sqlite (Qt's runtime driver
plugin), not the C headers/library find_package(SQLite3) needs. Every
PR based on current master has been failing CI with "Could NOT find
SQLite3" since #983 merged.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Reviving the still-relevant third of #682 (closed 2026-09-18 purely to
clear a review backlog, not on merit). Investigated fresh against
current master rather than merged wholesale -- two of the original
PR's three findings turned out to already be resolved independently:
- Element::valideXml() and Terminal::valideXml() already reject a
non-finite x/y (qIsFinite checks, with comments citing this exact
class of bug) -- added by someone else since #682 was written.
Verified live: a project with x="nan" on an element loads and
exports cleanly on current master, 3.1s, no hang.
- The illegal-XML-control-byte segfault in QDomDocument::setContent()
does not reproduce either. Tested both bytes from the original
report (0x00, 0x0E) against a real Qt6 build: both are now refused
cleanly (XmlParsingFailed, exit 0), no crash. Qt6's QDom parses
differently to the Qt5 one #682 was written and tested against.
What's still genuinely open: QET::attributeIsAReal() itself --
QString::toDouble()'s output parameter reports success for "nan"/
"inf"/"-inf", and this shared helper (26+ call sites across the
codebase, per #682's own count) had no finiteness check independent of
element.cpp/terminal.cpp's own since-added ones. Confirmed several
call sites are not behind either of those two gates -- notably
elementpicturefactory.cpp's line/rect/ellipse/circle/arc parsing for a
symbol's own drawing (a corrupted .elmt, not just a corrupted project
file), which was and remains reachable through this helper alone.
Qt 6.10.2, ctest 13/13.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Reviving #713, closed 2026-09-21 purely to clear a maintainer review
backlog (#630), not on merit; not superseded. The crash #713 was
originally named after (bugtracker #291) was already fixed separately
by 39ac5716c, merged 14 Aug -- confirmed still on master. What's left,
and what this revives, is the one-line follow-up #713 itself narrowed
to after that: m_future.cancel() before the wait.
Without it, ~ElementsCollectionModel()'s wait runs the whole queued
QtConcurrent::map() to completion, so cancelling the open-element
dialog blocks until every remaining item has been processed -- a
visible hang on the button pressed precisely to stop the work.
cancel() drops the not-yet-started items so the wait is short, while
still waiting for whatever item is already in flight (needed so it
can't dereference this object after it's gone).
Qt 6.10.2, ctest 13/13. The responsiveness gain itself is reasoned
from QFuture's documented cancel()/waitForFinished() semantics rather
than timed -- same as the original PR's own stated verification.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Reviving #664, closed 2026-09-18 purely to clear a maintainer review
backlog (#630), not on merit; not superseded. Rewritten fresh against
current master rather than merged from the old branch -- that branch
predates the Qt6-only switch and much of dataBase/projectdatabase.cpp's
later rewrite, and the two had diverged too far for a textual merge to
be trustworthy.
projectDataBase::removeElement() only ran DELETE FROM element WHERE
uuid=:uuid. It never touched element_info, even though every element
also has a row there (element_uuid is its PRIMARY KEY, with a FOREIGN
KEY back to element.uuid that isn't enforced by this connection -- no
ON DELETE CASCADE in effect). So deleting an element left its
element_info row orphaned.
Re-adding an element with that same uuid later -- undo of that same
deletion, or a redo replaying it -- goes through addElement(), which
INSERTs into both tables. The element insert succeeds (that row really
was removed). The element_info insert hits the orphaned row's primary
key and fails, silently: the error is logged and swallowed, so the
element re-enters the scene with no element_info row at all, and
nothing later re-syncs it.
removeDiagram() already cascades this cleanup when a whole folio is
removed (a later, unrelated addition) -- confirmed on current master --
but that path never runs for a single element removed on its own,
which is the case this fixes.
Verified on the built binary, not just read: placed an element, deleted
it (Ctrl+A, Delete), undid the deletion (Ctrl+Z). Reverting just this
fix and repeating the identical sequence reproduces the exact reported
error:
Debug: projectDataBase::addElement insert element info error :
QSqlError("1555", "Unable to fetch row",
"UNIQUE constraint failed: element_info.element_uuid")
With the fix, the same sequence produces nothing. Qt 6.10.2, ctest
13/13.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Split from #913's second suggestion. There was no shortcut for the common
"duplicate with offset" convention; the nearest existing feature,
"Collage multiple", is a different workflow (a dialog for repeating a
paste in a grid pattern, not a one-shot duplicate).
Ctrl+D copies the selection and places it immediately, offset by a
configured spacing and direction -- no interactive follow-the-cursor
step, unlike Ctrl+V. The first press (or after the setting is explicitly
reopened) shows DuplicateOffsetDialog: spacing in grid steps, direction
up/down/left/right. Every later press reuses whatever was confirmed then,
silently, so a row of copies is one key held down and tapped, not a
dialog every time -- unattended, repeatable stamping is the actual point
of a duplicate shortcut, which a dialog or an interactive placement step
on every press would defeat. A separate "Configurer la duplication..."
entry reopens the dialog on demand to change the setting later. Cancel
leaves the diagram untouched -- verified, not assumed: qet_diff against
the saved file shows 0 added.
Chaining ("keep tapping to lay out a row") needs no special handling:
QET already reselects whatever a paste just added
(PasteDiagramCommand::redo()), so the next Ctrl+D naturally continues
from the copy just placed rather than the original.
The offset is applied by hand rather than by asking paste()/fromXml() to
place the copy at a target position. Both of those feed the position
through Diagram::snapToGrid(), which reads
QApplication::keyboardModifiers() and rounds to the nearest PIXEL instead
of the grid whenever Ctrl is held -- and Ctrl is always held here, this
action's own shortcut being Ctrl+D. Measured before settling on this:
routing the offset through paste() first produced copies off-grid on both
axes, by an amount that tracked the selection's own bounding-box geometry
rather than being a fixed error -- caught by qet-mcp's qet_elements
against the saved file, not by eye. fromXml() is instead called with no
position argument at all (leaves every item at its source coordinates,
landing the copy on top of the originals -- (0,0) is not a position, this
is "keep the source coordinates"), and the offset is added directly with
setPos(). A plain addition cannot be off by a rounding rule that never
runs.
Conductors are not in the hand-translated set: fromXml() itself does not
reposition them either -- they load after elements are already in their
final place and take their geometry from their terminals, which have
already moved with the elements that own them. Verified this holds: drew
a conductor by hand between two elements (drag, not click-click),
selected both, Ctrl+D, and the new conductor correctly joins the two new
elements via qet_conductors -- not the originals, not a mix.
Verified end-to-end on a built binary via qet-mcp, not by eye:
before L2 (303,207) L9 (512,196) -- deliberately off-grid
spacing=2, down (303,227) (512,216) -- +0,+20 exactly
same again, 2nd (303,247) (512,236) -- +0,+20 again, chained
Both elements land exactly the configured offset from their immediate
source regardless of the selection's own alignment. Qt 6.10.2, ctest
11/11.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
A property value that is entirely whitespace -- reported in #973 as a
workaround (setting a title-block custom variable to a single space, the
only way to give it a value other than blank before that bug was fixed in
#989) -- did not survive a save/reload cycle. Two independent causes, both
needed for the round trip to actually work:
1. DiagramContext::toXml() called .trimmed() on every stored value before
writing it, unconditionally. For ordinary content this only strips
accidental leading/trailing whitespace, but for a value that IS
whitespace it collapses the entire thing to "", indistinguishable from
a value that was never set.
2. QDomDocument::setContent(), used to parse the project file, discards a
text node that is entirely whitespace by default. Confirmed in
isolation, outside any QET code: parsing "<a> </a>" with the default
ParseOptions gives QDomElement::text() == ""; adding
ParseOption::PreserveSpacingOnlyNodes gives " ". So even once (1) stops
destroying the value on save, the very next load throws it away again.
Fix (1) only trims when the trimmed result isn't empty, i.e. leaves an
all-whitespace value untouched. Fix (2) adds PreserveSpacingOnlyNodes to
the one setContent() call that parses a project file
(QETProject::readProjectXml()) -- not the other ~19 call sites in the
codebase (clipboard paste, element/macro loading, translations, autonum
context), which read different, narrower documents and are not implicated
in this report.
Blast radius of (2): every place that walks a QDomNode's children already
filters on isElement() (see QET::findInDomElement()), so the extra
whitespace-only text-node siblings this keeps around are inert wherever
current code already expected only elements. The one place it isn't inert
is exactly the bug -- calling .text() on an element whose entire content
is whitespace.
Verified end-to-end, not just at one stage: a single-space title-block
variable now survives two successive --resave cycles unchanged (confirmed
byte-for-byte in the saved XML), and renders as blank space rather than
literal placeholder text or a vanished value. Re-saved all 24 shipped
examples with and without this change and diffed: 23 byte-identical, the
one that differs (schema_indus.qet) differs only in element uuids -- and
resaving it twice with the SAME unpatched binary produces that same kind
of diff, confirming it is pre-existing non-determinism in files that
predate persisted uuids, unrelated to this change. Qt 6.10.2, ctest 11/11.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
BorderTitleBlock::updateDiagramContextForTitleBlock() skips merging a
page-level custom variable into the title block's render context whenever
its value is empty -- added by PR #572 to fix#531, where an empty
page-level value was shadowing a real project-level one of the same name.
But skipping the merge removes the key from the context entirely, and
TitleBlockTemplate::interpreteVariables() only replaces "%name"/"%{name}"
when "name" is an actual key in that context -- anything absent is left as
its own literal placeholder text. Folio Properties auto-adds every one of
a template's custom variables to the Custom tab with an empty value (#271/
#495) precisely so the user only has to fill in what's missing; until they
do, that variable now renders as e.g. "%label1" instead of blank.
Reproduced two ways: a synthetic fixture, and examples/2612_ats_singlephase.qet
itself, which already carries three such auto-added-but-unset properties
("label1", "label2", "label3") and renders all three literally on current
master.
Fix: skip the empty page-level value only when a project-level one already
exists to show through (preserving #531's guarantee); otherwise still merge
it in empty, so the placeholder resolves to blank rather than falling out
of the context altogether.
Verified against the shipped example (--export-png, before/after crop of
the rendered title block): "%label1"/"%label2"/"%label3" now blank. A
variable never added to the Custom tab at all ("%client", also present in
the same example) is unaffected -- nothing was ever configured for it, and
that is a separate, narrower case. Qt 6.10.2, ctest 11/11.
A related but distinct issue -- DiagramContext::toXml() trims a stored
value before saving, so an all-whitespace value is written as empty --
explains a second symptom from the same report (a single-space "workaround"
value vanishing after the project is reopened) but touches every
context-backed property, not just title blocks, and is left for a separate
fix.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Use a bound VACUUM INTO path and remove the stale SQLite
handle declaration. Fix shell continuations in Windows CI and
Debian installation instructions.
#984 switches JavaScript scripting off by default, and five tools here
drive QElectroTech through --run: qet_query, qet_continuity, qet_check,
qet_project_new, qet_edit. Against such a build they all stop working, and
what came back was exit code 3 and a paragraph of French naming a settings
dialog nobody driving an MCP server is looking at.
Nothing needed building to make them work again -- _run_qet() inherits its
environment, so QET_ENABLE_SCRIPTING=1 in the "env" block of the client's
own configuration already reaches QElectroTech. Verified both ways against
a #984 binary: without it qet_query returns ok=false exit=3, with it
ok=true and the rows.
So this is about saying so. The refusal is now recognised and answered with
an instruction the caller can act on, keyed on QElectroTech naming the
variable with exit 3 as a fallback for a future build that words it
differently. Two older hints fitted the same symptom and were overwriting
it -- "the binary never ran the script ... is it a build with --run
support?" sends the reader to check the one thing that is fine -- so both
now yield to whatever the launch already reported. qet_check builds its
answer fresh rather than layering onto the launch result, so it carries the
reason across explicitly; without that every check read "no result came
back", which is true and tells nobody why.
The server does not set the variable itself, on purpose. A switch a program
turns on for itself is not a switch: whoever configured this server and
pointed it at a QElectroTech binary made that choice, and their interactive
QElectroTech keeps whatever its own setting says. README says this, and the
registration example now shows the env block with both variables in it.
Six tests, faking subprocess.run so they cost no launch. Two are structural
rather than behavioural: one fails if either older hint goes back to
assigning over the specific one, the other reads which tools actually pass
script= to _run_qet and fails if the hint's list of them drifts. Both were
mutation-checked by reintroducing exactly those mistakes.
176 tests pass with QET_BINARY, QET_ELEMENTS, QET_EXAMPLES and
QET_ENABLE_SCRIPTING set.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A <graphics_table>'s <query> is stored in the .qet and executed when the
project loads. SQLite produces rows lazily, so the cost of that query is
not bounded by anything the project contains -- it is bounded by how long
the loop reading the rows is willing to run. A recursive CTE takes one line
to make that forever:
WITH RECURSIVE c(n) AS (SELECT 1 UNION ALL SELECT n+1 FROM c) SELECT n ...
Put that in the <query> of any project's summary table and opening the file
pins a core at 100% and grows ProjectDBModel::m_record until memory runs
out. Measured on examples/industrial.qet with the query swapped, built from
master:
clean --export-bom 3.6 s, 396 rows, exit 0
poisoned --export-bom killed at 90 s, still going, no output
No scripting, no MCP, no flag beyond an ordinary export. Opening the file in
the editor is the same code path.
QetScriptApi::query() has the identical loop, and the script engine's own
30 s interrupt does not reach it: that aborts JavaScript, and this is C++
inside a single call. Left alone it hung a --run for 45 s until the harness
killed it.
Both loops now stop at projectDataBase::MaxResultRows (100000) and say so.
That is a backstop, not a page size: the largest table in the shipped
examples is 396 rows, and a caller that reaches 100000 has been handed
something it should not run to completion. It is not silent either way --
the model logs the offending query text, and qet.query() sets queryError(),
so a truncated result is never mistaken for a complete one.
clean --export-bom 3.6 s, 396 rows, exit 0 (unchanged)
poisoned --export-bom 20.2 s, 396 rows, exit 0, warning names the query
qet.query(recursive CTE) 3.8 s, 100000 rows, queryError() set
Reverting each cap restores the hang, so both checks discriminate.
Related to #983, which fixes a different flaw reachable through the same
stored query. Neither depends on the other.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
A script reaches the whole project and, through the export calls, the
filesystem. That is a capability most people installing an electrical CAD
program never asked for, and leaving it on by default hands it to them
anyway. So QET_HAS_SCRIPTING builds now ship with it switched off.
QetSettings::scriptingEnabled() is the single answer, read by all three
places that need it, with QET_ENABLE_SCRIPTING=1 overriding the stored
value. The override is not decoration: a CI job or a batch run has no
dialog to tick, and a machine whose HOME is created fresh for each run has
nowhere to keep the setting either. It beats a stored "false" on purpose,
so a box unticked once cannot lock a build server out of --run for good.
Only the exact value "1" counts.
--run refuses with exit 3 and a message naming both ways in.
Projet > Exécuter un script... asks once, and turns the setting on if
the answer is yes. Asking beats grey: a disabled menu
entry says something exists and nothing about how to have
it, and this is the pattern people already know from
macro security in office software.
Configurer QElectroTech > Général > Projets has the checkbox, for
turning it back off. While the environment forces
scripting on, the box is disabled and says why, and
applyConf() then leaves the stored value alone rather
than quietly overwriting it.
runOnProject() checks as well, after both callers have. It is the one
function that actually evaluates JavaScript, so it is the one place a
future caller cannot forget to ask; the callers check first only to give a
better answer than it can.
Verified on the built binary, all four states, with an isolated HOME:
stored env result
absent - refused, exit 3
true - script runs, exit 0
false - refused, exit 3
false 1 script runs, exit 0
tst_scriptingsetting covers the same matrix hermetically, in its own
QSettings scope, and was mutation-checked: flipping the default to true
turns defaultsToOff() red.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two security reviews of #980 landed on the same gap: every path in a tool
call is chosen by the model, and nothing checked where those paths pointed.
That made the server a read/write primitive for anything the process could
reach -- read any project on the disk, export one somewhere else, overwrite
an unrelated file, embed an arbitrary local image or PDF. The sandboxed HOME
each QElectroTech launch gets isolates settings, not the filesystem.
Data paths are now confined to a workspace: QET_MCP_WORKSPACE (os.pathsep
separated), defaulting to the directory the server was started in, which is
what an MCP host normally sets anyway. QET_MCP_ALLOW_ANY_PATH=1 turns the
check off; it exists so that is a visible choice rather than the default.
Paths are resolved before comparison, so a symlink planted inside the
workspace is judged by where it points -- the case a string-prefix check
gets wrong.
Two arguments are deliberately exempt: "binary" and "elements_dir". Those
are configuration, chosen once by whoever runs the server, and both normally
live in /usr or a build tree. Confining them would reject the ordinary case
while stopping nothing -- they are not where a model gets to point the
server at /etc.
Enforcement sits at the dispatcher, where model-supplied arguments enter,
not inside each tool. Importing the module and calling tool_export() from
Python stays unconfined and is meant to: that is the caller's own code with
the caller's own paths.
Separately, an existing "output" is now refused unless the call passes
"overwrite": true. qet_project_new already worked this way; qet_export,
qet_edit and qet_element_build now match it. Replacing a file is the one
step this server cannot undo.
17 tests cover it, including the symlink escape, the traversal, the
overwrite gate and the operation-level file paths that add_image and
add_pdf_page carry one level down. Two of them compare the policy table
against the tool schemas, because a write tool missing from either list
fails silently in opposite directions. Two more drive a real server process
over stdio, which is the only thing that shows a call is gated rather than
merely gate-able. Mutation-checked: removing the confinement fails 8, and
desynchronising the two lists fails the drift pair.
170 tests pass, with QET_BINARY, QET_ELEMENTS and QET_EXAMPLES set so none
are skipped.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
projectDataBase::isReadOnlySelect() decides whether a query only reads
by looking at its first keyword and rejecting internal semicolons.
SQLite has allowed a CTE prefix in front of a data-modifying statement
since 3.8.3, so
WITH x AS (SELECT 1) DELETE FROM element
begins with WITH, contains no semicolon, passes the check, and deletes
every row. UPDATE and INSERT go through the same way.
This is not only reachable from the custom-query box. ProjectDBModel::
fromXml() reads a <graphics_table>'s saved <query> straight out of the
.qet and fillValue() executes it, so a project file can carry the
statement. Reproduced against a build of this branch's parent, with no
scripting and no CLI flag beyond the export itself: a project whose
stored table query was replaced with the DELETE above exported a bill
of materials of 0 rows instead of 14, exit code 0, nothing logged. A
silently empty or -- with UPDATE -- silently altered BOM is the kind of
output someone orders parts from.
Fixed by asking SQLite about the statement it actually compiled.
sqlite3_prepare_v2() compiles without running, sqlite3_stmt_readonly()
reports on the compiled statement rather than on how it was spelled,
and the prepare tail catches a second statement structurally. The same
project now exports its 14 rows again and logs a reason for the
refusal, while an ordinary WITH ... SELECT in a project file still runs
untouched -- the fix is not "ban CTEs".
isReadOnlySelect() stays in front of it rather than being replaced:
SQLite considers ATTACH, BEGIN and several PRAGMAs read-only too, since
none of them change the contents of the database, so dropping the
statement-type allowlist would have widened what is accepted while
fixing what is executed.
The check lives in its own translation unit depending on nothing but
QString and SQLite, so tests/qttest/tst_sqlreadonly.cpp can link it
alone and exercise the security property without standing up a
QETProject: 18 assertions covering the three CTE-prefixed writes named
in the review, bare writes, trailing statements, comment-only input
(which compiles to a null statement sqlite3_stmt_readonly() must not be
handed) and a null connection (refused, not waved through). Confirmed
the suite discriminates by deliberately disabling the new check and
watching exactly the nine write-refusal assertions go red while the
accept cases stayed green.
ctest 13/13, qet-coherence-check and qet-pdflink-check clean on the
example corpus.
Reported in PR #980's review thread by @elevatormind and confirmed
against this code by @scorpio810; fixed here on its own because the
flaw is in already-released code and needs none of that branch to
reach.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Two issues on the interactive paste path:
1. Stall: DiagramEventAddPaste's constructor called Diagram::fromXml()
with no database batching, so every addItem() emitted dataBaseUpdated()
and each connected table model re-ran its full SQL query. A typical
paste (~40 elements + ~40 conductors) triggered ~77 rebuilds of the
table models -- measured at ~2.1 s of pure fromXml time on a large
project. Project loading already batches this via
setUpdateBlocked()/blockSignals() (QETProject::readProjectXml); the
paste path now does the same: block during fromXml, one updateDB()
after. Measured fromXml: 2114 ms -> 143 ms.
2. Cursor jump: m_initial_cursor was set to the group origin but the
physical cursor stayed at the Ctrl+V press location, so the first
mouseMoveEvent computed a large delta and the items jumped on first
touch. Warp the cursor to the group origin after placement so the
baseline and the actual cursor position match.
Commit dd0c194a3 (#913) moved the pasted group to the cursor position
at construction time. The desired behaviour is that items appear at
their original XML coordinates (where they were copied from) so the
user starts from the origin. The grid-snapped movement baseline and
the context-menu restoration from that commit are kept.
LinkElementCommand::redo() already had a check meant to catch exactly
this -- two report-linked conductors whose properties disagree -- and
ask the user which to keep via PotentialSelectorDialog. It never
worked: it built ONE combined list from three unrelated fields
(tension_protocol, wire_color, wire_section) and tested that whole
list for string equality, so a tension-protocol value could never
equal a wire-colour value even when every field individually matched
across every conductor. Worse, "wire_color"/"wire_section" are
ConductorProperties::m_wire_color/m_wire_section, a separate free-text
documentation pair that says nothing about how the wire is actually
drawn -- that's "color"/"style" -- so the one field #974 is actually
about was never compared at all.
Fixed by comparing each relevant field (text/num, function, tension
protocol, colour, line style) separately. Downloaded the reporter's
actual project, confirmed the mismatched wire reads color="#0000ff" on
one side of a "Folio suivant"/"Folio precedent" link and
color="#55aa00" on the other, with the link's other four conductors
matching correctly (ruling out a rendering artifact) -- see PR #980's
checkContinuity() extension, which now flags this class of mismatch on
sight.
Extracted the comparison into its own static
reportLinkNeedsPotentialChoice(), for the same reason
ConductorCreator::needsPotentialChoice() already exists as its own
method: a caller with nobody there to answer a modal dialog needs to
check first and decline, and the condition must not drift away from
the one redo() actually applies.
Fixing the comparison surfaced a real, previously-latent hang in this
session's own qet.linkElements(): PotentialSelectorDialog::exec() is a
plain QDialog::exec(), not routed through QET::QetMessageBox, so
headless --run has nobody to answer it. Measured directly -- hung
until killed with the property-comparison fix alone, clean refusal
after adding the guard. linkElements() now calls
reportLinkNeedsPotentialChoice() before constructing the command and
declines with a clear reason, the same choice addConductor() already
makes about ConductorCreator's own equivalent dialog.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
link_elements() now correctly refuses a mismatched report link instead
of allowing it (see the two preceding commits), so the
report_link_mismatch reproduction can no longer be built by linking
two already-differently-coloured conductors -- that path is refused
before it happens. Updated to patch a saved file's XML directly
instead (the same technique test_continuity_detects_a_tampered_
potential already uses), since a file QElectroTech's own edits produce
can no longer end up in this state at all. Adds
test_link_elements_refuses_a_mismatched_report_link to cover the
refusal itself.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Fixing LinkElementCommand's property comparison (previous commit)
means it now correctly detects a report-link colour/style mismatch --
which means it now correctly pops PotentialSelectorDialog for one,
same as the GUI. Under headless --run there is nobody to answer a
plain QDialog::exec(), so this hangs forever; confirmed directly with
a timeout before adding this guard.
qet.linkElements() now checks LinkElementCommand::
reportLinkNeedsPotentialChoice() before constructing the command and
declines with a clear reason pointing at checkContinuity() and
setConductorProperty(), the same choice addConductor() already makes
about ConductorCreator's own equivalent ambiguous-potential dialog.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
LinkElementCommand::redo() already had a check meant to catch exactly
this -- two report-linked conductors whose properties disagree -- and
ask the user which to keep via PotentialSelectorDialog. It never
worked: it built ONE combined list from three unrelated fields
(tension_protocol, wire_color, wire_section) and tested that whole
list for string equality, so a tension-protocol value could never
equal a wire-colour value even when every field individually matched
across every conductor. Worse, "wire_color"/"wire_section" are
ConductorProperties::m_wire_color/m_wire_section, a separate free-text
documentation pair that says nothing about how the wire is actually
drawn -- that's "color"/"style" -- so the one field #974 is actually
about was never compared at all.
Fixed by comparing each relevant field (text/num, function, tension
protocol, colour, line style) separately. Downloaded the reporter's
actual project, confirmed the mismatched wire reads color="#0000ff" on
one side of a "Folio suivant"/"Folio precedent" link and
color="#55aa00" on the other, with the link's other four conductors
matching correctly (ruling out a rendering artifact) -- see PR #980's
checkContinuity() extension, which now flags this class of mismatch on
sight.
Extracted the comparison into its own static
reportLinkNeedsPotentialChoice(), for the same reason
ConductorCreator::needsPotentialChoice() already exists as its own
method: a caller with nobody there to answer a modal dialog needs to
check first and decline, and the condition must not drift away from
the one redo() actually applies.
Fixing the comparison surfaced a real, previously-latent hang in this
session's own qet.linkElements(): PotentialSelectorDialog::exec() is a
plain QDialog::exec(), not routed through QET::QetMessageBox, so
headless --run has nobody to answer it. Measured directly -- hung
until killed with the property-comparison fix alone, clean refusal
after adding the guard. linkElements() now calls
reportLinkNeedsPotentialChoice() before constructing the command and
declines with a clear reason, the same choice addConductor() already
makes about ConductorCreator's own equivalent dialog.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Brings misc/qet-mcp up to date with checkContinuity()'s new
report_link_mismatch finding: updated qet_continuity's description,
and two new tests reproducing #974 with the shipped 02going_arrow.elmt/
01coming_arrow.elmt pair (no custom fixtures needed -- this element
type already ships).
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
New report_link_mismatch finding (severity "warning", not "error" the
way potential_mismatch is): a next_report/previous_report folio-jump
pair whose conductors disagree on colour, style, num, or any of the
other checked properties. Unlike potential_mismatch, this one is not
proof of external tampering -- LinkElementCommand::isLinkable() only
ever checks type and freedom (see its own doc comment), never conductor
properties, so nothing in QElectroTech copies one side's colour onto
the other when a report link is made or keeps them in sync afterwards.
This is a real, unenforced gap reachable through completely ordinary
use, not a defect a script or the GUI could introduce.
Reproduces qelectrotech/qelectrotech-source-mirror#974 exactly:
downloaded the reporter's actual project, traced the mismatched wire to
a "Folio suivant"/"Folio precedent" link pair, and confirmed via query
that the two sides read color="#0000ff" and color="#55aa00" while the
link's other four conductors (0V/Low/High/Ground) matched -- ruling out
a rendering artifact. Verified fresh with a synthetic reproduction
(tests in misc/qet-mcp) using the shipped 02going_arrow.elmt/
01coming_arrow.elmt pair, giving exactly one finding, not one per
folio-link conductor.
Also fixes a real gap in the existing potential_mismatch check while
here: checked_properties was missing "color" and "style" entirely,
checking only "conductor_color" (ConductorProperties::m_wire_color, a
separate free-text documentation field, typically empty) -- meaning
the same-folio version of this exact bug class would have gone
undetected too.
Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Laurent Trinques