fix(ui): expand flat threads, reach the first message, localise the date

Four faults from the first hand test, two of them behavioural.

An expander that opened onto nothing. setThreadMessages kept only nodes with
depth > 0, and notmuch_thread_get_toplevel_messages returns every message at
depth 0 when a thread carries no usable In-Reply-To, so a flat thread
contributed no children while its card still advertised the count. Measured in
the user's database: of 396 inbox threads three are flat, one of them nine
messages long, and every two-message thread of that kind was affected, which is
exactly why the fault looked like "the expander only works with more than one
reply". The rule is now position, not depth: every message except the first,
which is the root card itself. That is also the correct rule rather than a
workaround, since the row under the root is the second message however notmuch
chose to nest it.

The thread's first message was unreachable. Selecting a root card loaded the
whole thread, so the pane showed every message with only the last expanded, and
no row in the list offered the first one: the reply rows are messages two
onward. The root card now renders its own message, which is what the card
already claims to be. It keeps its thread id, unlike the message-row path, so
mark-read and the tag-change repaint still work; that is asserted, because
clearing it is the obvious way to write this and silently disables both. Before
the replies are loaded the model has no first message to name and the whole
thread stays the honest answer.

Dates ignored the locale. One hardcoded "yyyy-MM-dd hh:mm" produced a US-looking
format on an Italian desktop; QLocale::system() now formats it, and the width
reserved for the date comes from the same function so a longer locale cannot
clip.

The expander was a bare number on the card's own background. It is a pill now,
carrying "3 replies" (and "1 reply", singular), sized from the label actually
drawn and measured in both glyph states so it does not resize under the pointer
on click. Its fill is blended from Text toward Base rather than taken from
QPalette::Button, which is #2b2b2b against a Base of #2b2b2b on the user's
theme: byte identical, so the pill was invisible. A theme may make any two roles
equal; a blend is defined against the surface it sits on and cannot collide with
it. Checked by rendering both a dark and a light palette and looking.

1 files changed, 75 insertions, 0 deletions

diff --git a/tests/test_cardlayout.cpp b/tests/test_cardlayout.cpp
index c38f728..48bba25 100644
--- a/tests/test_cardlayout.cpp
+++ b/tests/test_cardlayout.cpp
@@ -19,6 +19,7 @@
 #include "cardlayout.h"
 
 #include <QFont>
+#include <QLocale>
 #include <QTest>
 
 class TestCardLayout : public QObject
@@ -32,10 +33,12 @@ private slots:
     void indentStopsAtTheCap();
     void expanderSitsOnTheSecondLine();
     void expanderIsEmptyWithoutReplies();
+    void theExpanderReadsAsAPillWithAWord();
     void dateIsFlushRight();
     void threadCardCarriesAnAccentBar();
     void replyCardCarriesNoAccentBar();
     void theDateFitsWhenTheCardIsBold();
+    void theDateFollowsTheSystemLocale();
 };
 
 namespace {
@@ -183,6 +186,42 @@ void TestCardLayout::expanderIsEmptyWithoutReplies()
     QVERIFY(card.expanderRect.isEmpty());
 }
 
+void TestCardLayout::theExpanderReadsAsAPillWithAWord()
+{
+    // A bare "3" beside the subject reads as an unexplained number and gives
+    // no hint that it can be clicked. The label carries the word, and the rect
+    // carries padding for the pill drawn behind it.
+    QCOMPARE(CardLayout::expanderLabel(3, false),
+             QStringLiteral("\u25b8 3 replies"));
+    QCOMPARE(CardLayout::expanderLabel(3, true),
+             QStringLiteral("\u25be 3 replies"));
+
+    // Singular, because "1 replies" is the kind of detail that makes an
+    // interface look unfinished.
+    QCOMPARE(CardLayout::expanderLabel(1, false),
+             QStringLiteral("\u25b8 1 reply"));
+
+    const QFont font;
+    const int h = CardLayout::heightFor(font);
+    const CardLayout card =
+        CardLayout::compute(threadInput(), QRect(0, 0, 400, h), font);
+    const QFontMetrics small(CardLayout::smallFont(font));
+
+    // The rect must hold the label AND its padding, or the pill's background
+    // is narrower than the text sitting on it.
+    QVERIFY2(card.expanderRect.width()
+                 >= small.horizontalAdvance(CardLayout::expanderLabel(3, false))
+                        + CardLayout::kPillPaddingX * 2,
+             "the expander rect is too narrow for its own label and padding");
+
+    // And it must NOT change width when the card opens: a pill that resized on
+    // click would shift the subject's elision under the pointer.
+    CardLayout::Input open = threadInput();
+    const CardLayout expanded =
+        CardLayout::compute(open, QRect(0, 0, 400, h), font);
+    QCOMPARE(expanded.expanderRect.width(), card.expanderRect.width());
+}
+
 void TestCardLayout::dateIsFlushRight()
 {
     const QFont font;
@@ -232,6 +271,42 @@ void TestCardLayout::replyCardCarriesNoAccentBar()
     QCOMPARE(reply.spines.size(), 1);
 }
 
+void TestCardLayout::theDateFollowsTheSystemLocale()
+{
+    const QDateTime when(QDate(2025, 8, 10), QTime(6, 26));
+
+    // The system locale's own rendering, whatever it is. Asserting a specific
+    // string would only restate the hardcoded pattern this replaced, and would
+    // fail on any machine but the one that wrote it.
+    QCOMPARE(CardLayout::formatDate(when),
+             QLocale::system().toString(when, QLocale::ShortFormat));
+
+    // The specific fault: an ISO-looking pattern on a desktop that does not
+    // use one. Guarded so this test says nothing on a locale that genuinely
+    // formats that way.
+    if (QLocale::system().toString(when, QLocale::ShortFormat)
+            != QStringLiteral("2025-08-10 06:26")) {
+        QVERIFY2(CardLayout::formatDate(when)
+                     != QStringLiteral("2025-08-10 06:26"),
+                 "the date is hardcoded to yyyy-MM-dd hh:mm rather than "
+                 "following the desktop's locale");
+    }
+
+    // And the reserved width has to follow the same formatter, or a locale
+    // whose dates are longer clips them exactly as the bold font did.
+    QFont font;
+    const int h = CardLayout::heightFor(font);
+    const CardLayout card =
+        CardLayout::compute(threadInput(), QRect(0, 0, 400, h), font);
+    QFont bold = font;
+    bold.setBold(true);
+    QVERIFY2(card.dateRect.width()
+                 >= QFontMetrics(bold).horizontalAdvance(
+                        CardLayout::formatDate(when)),
+             "the reserved date width is narrower than this locale's own "
+             "formatting of a date");
+}
+
 void TestCardLayout::theDateFitsWhenTheCardIsBold()
 {
     // An UNREAD card draws BOLD, and bold is wider. The layout is computed from