Files
lanework/Kanban/UI/Card/CardWindowView.swift
T
rzen c27cc93ec1 The card window's title learns Edit mode — a sibling session rides the body's own doors
CardWindowView's header is now the title, live: a static Text in Preview, an editable
single-line TextField in Edit, both reading a new CardTitleEditSession's buffer rather
than the card's own snapshot value — the same "buffer outranks the snapshot" reason the
body surface already reads bodySession.text instead of card.body.

CardTitleEditSession is CardBodyEditSession's shape one field over: the one-isDirty write
gate, dirty-buffer-wins on adopt(diskTitle:), the ~700ms injectable debounce, flush() /
flushOrThrow() for DirtyBufferGuard, and beginEditSession()/endEditSession() with a
session-coalesced undo step (one per session, never per debounced tick). It rides the
body's own begin/end/flush doors rather than opening a second session boundary — title is
only ever editable while the body column is in Edit mode — because the two write through
different WriteOperations with different validation and merging them would conflate two
unrelated frontmatter keys behind one buffer.

BoardStore.commitCardTitle(inCard:title:) reuses the same private setTitle helper and the
same .rename WriteOperation the board's own inline rename commits through, so trimming,
empty-removes-the-key, unchanged-writes-nothing and banner enrichment are one code path,
not a re-implementation. It resolves through cardBodyTarget (spans lanes and the trash),
not boardItem (board only), because a card window's title field stays live through the
same dismissal-into-trash flush the body already gets — the one deliberate divergence
from the board's own rename, which treats a trashed target as vanished.
registerTitleEdit(inCard:priorTitle:newTitle🔛) mirrors registerBodyEdit, anchored by
card identity and the already-reserved ExpectedField.title, folding into the same
one-coarse-step-per-window-close undo model.

Every place the body's Edit buffer flushes, the title's now does too: mode exit
(bodyPresentation.flushEdits/beginEdits), window close and the dismissal path
(CardWindowSession.endSession()), raw-source entry (configureRawSource), the close-time
DirtyBufferGuard modal (attemptSave tries body then title), and the fast-path close gate
(closeAfterFlushing() now checks title.isDirty alongside body.isDirty).
holdsUnsavedContent and settlement carry the title too.

Tests: CardTitleEditSessionTests.swift mirrors CardBodyEditSessionTests.swift (write
gates, normalization and newline-stripping, dirty-buffer-wins, debounce, undo-step
coalescing). CardTitleWriteTests.swift covers commitCardTitle/registerTitleEdit:
byte-identity no-op, empty-removes-key, vanished, the trashed-card-is-still-writable
divergence, a readable-but-uneditable target refusing and bannering, and a read-only
board suspending quietly. CardSessionUndoTests.swift gains coverage that a title edit
folds into the coarse close step alongside a body edit and registers on the window's own
stack, never the board's. RawSourceTests.swift's hand-wired rig picks up the title
session configureRawSource now also flushes.

3148 KanbanTests pass, 0 failures.

Claude-Session: https://claude.ai/code/session_014PtZdPwqZuqEDLc6wZMtEy
2026-08-09 10:37:55 -04:00

550 lines
32 KiB
Swift

import SwiftUI
import UniformTypeIdentifiers
// MARK: - CardWindowView
/// The card window's content: **three componentized panes** — a wide body pane leading, the comments
/// pane in the middle when it is shown, and the narrow attributes sidebar trailing
/// (05-card-window.md ▸ Composition, re-composed 2026-07-29).
///
/// ### The composition arranges; the panes do not know about each other
///
/// > each an independent component with its own scroll, arranged by the window's layout rather than
/// > wired to each other; componentization is the rule, so the comments pane mounts beside the body
/// > or below it (the layout option) without either pane knowing which.
///
/// That is enforced here by there being nothing to enforce: `CardCommentsPane` takes no layout
/// parameter and the body column takes none either. This view puts one of them in a frame; the
/// arithmetic behind the frame is `CommentsMount`, which is pure and therefore checkable.
///
/// The sidebar is unchanged by any of it — it is a third pane, it has always been fixed-width when
/// shown, and the resize flex still goes to the body and never to the sidebar. **The comments column
/// is no longer in that second category** (ruled 2026-08-09): its divider is user-draggable in the
/// beside mount, so the flex the body and the comments pane split is now the user's to move, within
/// the two floors `CardWindowMetrics.clampedCommentsColumnWidth` keeps either side from crossing.
///
/// **Whether it shows at all is its own bit** (View ▸ Show Sidebar, extended for the toolbar-toggle
/// card): app-wide, persisted `showSidebar` below, mirrored by the card window's toolbar
/// (`CardToolbar`). Hidden, the pane and its divider are simply absent from the `HStack` and the body
/// column takes the width back through its own `.infinity` frame — no third state, no collapsed
/// sliver, the comments pane's own "costs the window no width while it is not mounted" rule applied
/// to the sidebar's visibility instead of the comments pane's mount.
///
/// ### What this milestone builds, and what it deliberately does not
///
/// The *shell*: the two columns, their scrolling, the sidebar's fixed width, and the header's
/// renderings of what the loader already knows — the card's title and its created/modified line —
/// over the body column's two live surfaces (Preview and Edit, `CardBodySurface`), with the
/// raw-source outlet swapping the pair of columns out entirely when it is active.
///
/// **The title is an editable field in Edit mode, a static rendering in Preview** (m6-card-body,
/// widened): it rides the body column's own mode rather than opening a second editor of its own, so
/// "Edit mode" means both surfaces at once, and its own session (`titleSession`, `CardTitleEditSession`)
/// begins and ends with the body's — see that type's doc comment for the whole of what that buys.
///
/// The placeholders that remain are structural rather than apologetic: the sidebar's inventory and its
/// order are settled (05 ▸ The attributes sidebar), so the shell states them and the sections fill in
/// underneath without the composition moving.
///
/// ### The width rule, in one line
///
/// The sidebar has a fixed width from `CardWindowMetrics`; the body column takes `.infinity`. That
/// is the whole of "the window's resize flex goes to the body, never the sidebar" — no split view
/// between body and sidebar, no stored divider position there, nothing for a drag to disagree with.
///
/// **The comments column keeps its own, separate width rule** (added 2026-08-09,
/// `CommentsColumnDivider`): a real divider, a drag, and a stored width — the exception this
/// sentence used to have none of, and still does not have between the body and the *sidebar*.
///
/// ### Why the title no longer scrolls with the body
///
/// The body surface is a hosted `NSScrollView` (`CardBodySurface`), because ⌘F's find bar lives in
/// one — "Edit ▸ Find (⌘F) is find-in-text here … the standard find bar" (05 ▸ Preview). A scroll
/// view inside a scroll view is a scroll view that fights, so the column's header — the title and
/// its created/modified line — sits above the body's scroller rather than inside it. 05 fixes the
/// column's *order* ("Body column, top to bottom") and the columns' independent scrolling, and both
/// still hold; which of the two things scrolls the title away was never settled, and pinning the
/// card's name over its own body is the better reading of a window whose subtitle already follows it.
struct CardWindowView: View {
let card: Card
/// The board this card belongs to.
///
/// The one place in this window a whole store is handed to a view rather than a narrow seam, and
/// the Style section is why: it hosts the **shared** style editor, whose API is store-shaped by
/// design (it reads the target set's current values and writes through the one `applyStyle`
/// bracket every anchor shares). Routing it through a closure of this window's own would be a
/// second card-styling path to keep in step with the first — exactly what "one component, one
/// behavior" forbids. Delete moved off the sidebar's own Actions section and onto the card
/// window's toolbar (`CardToolbar`, `CardWindowActions`) — still the store's own move-to-trash,
/// still "exactly the ⌫ delete", just reached through the host's toolbar wiring rather than this
/// view.
let store: BoardStore
/// The app-wide quick-style recents the embedded editor feeds (03-board-ui.md ▸ Styling ▸
/// Controls) — app state, not board state, which is why it arrives beside the store rather than
/// on it.
let recents: StyleRecents
/// The card's folder on disk — what relative images and links in the body resolve against
/// (05 ▸ Preview). `nil` only where a caller has no board root to build it from.
let cardFolder: URL?
/// This window's body-column state: which mode it is in, and the find-bar hook.
let bodyPresentation: CardBodyPresentation
/// This window's Edit buffer. It holds the text **both** surfaces show: the editor writes into
/// it, Preview renders it, and `adopt(diskBody:)` below is where the snapshot gets a say —
/// which is exactly the point at which dirty-buffer-wins is decided.
let bodySession: CardBodyEditSession
/// This window's title field — `bodySession`'s sibling, over the header's single line rather than
/// the body span (m6-card-body, `CardTitleEditSession`). Read in both modes for `bodySession`'s
/// own reason: a flushed buffer is ahead of the snapshot by a reload, so it is the truer of the
/// two whether the header is showing static text or the editable field.
let titleSession: CardTitleEditSession
/// This window's raw-source outlet. While it is active the two columns are gone entirely — see
/// `body`.
let rawSource: CardRawSourceSession
/// Whether a checkbox may write — `false` under the board's read-only lock.
let isEditable: Bool
/// This window's attachments section: the listing, the selection, and the two writes it starts
/// (05 ▸ Attachments).
let attachments: CardAttachments
/// This window's comments pane: the thread, the composer's draft buffer, and the one open inline
/// edit session (05 ▸ The comments column). It lives on the window's *session* so the close flush
/// can reach it, which is why it arrives here rather than being made here.
let comments: CardComments
/// This window's thumbnail memory, held by the host so it outlives a snapshot.
let thumbnails: AttachmentThumbnailCache
/// **This window's undo stack** (13-native-undo.md ▸ Rules ▸ two levels). It arrives for exactly
/// one consumer — the Style section, the one sidebar anchor that *writes* — because a gesture
/// issued in this window registers on this window's stack. It lives on the window's session so
/// the close can fold it, which is why it arrives here rather than being made here.
let undo: CardWindowUndo
/// The whole-window file drop (05 ▸ Attachments: "the drop surface remains the **whole
/// window**"). `nil` only where a caller has no store to import through.
let fileDrop: CardWindowDropDelegate?
/// Commits a checkbox flip: the marker's byte offset in the body, and the state the user saw.
let onToggleTask: (Int, Bool) -> Void
/// **View ▸ Show Comments** and **View ▸ Comments Beside Body** — app-wide, persisted, read here
/// rather than passed in (05 ▸ The comments column; ▸ Composition).
///
/// `@AppStorage` because the two bits genuinely are app-wide: every open card window obeys the
/// same pair, so a window that took them as parameters would need something above it keeping
/// every window in step with a value that has exactly one instance. It is also what makes the
/// menu rows' checkmarks and these panes provably the same bit (`ShowCommentsCommand`).
@AppStorage(AppPreferences.showCommentsKey) private var showComments = true
@AppStorage(AppPreferences.commentsBesideBodyKey) private var commentsBesideBody = true
/// **View ▸ Show Sidebar** — the trailing attributes sidebar's own visibility bit, `showComments`'s
/// neighbour and shape exactly (05-card-window.md ▸ Composition, extended for the toolbar toggle):
/// app-wide, persisted, read here rather than passed in so this pane and the toolbar item that
/// mirrors it (`CardToolbar`) and the View-menu row that mirrors it too (`ShowSidebarCommand`) are
/// provably the same bit.
@AppStorage(AppPreferences.showCardSidebarKey) private var showSidebar = true
/// Reduce Motion, read from the environment because this is a view — `Motion`'s own split between
/// callers that have one and callers (a menu command, a toolbar item) that reach `Motion
/// .prefersReducedMotion` instead.
@Environment(\.accessibilityReduceMotion) private var reduceMotion
/// **The comments column's user-set width**, in points — the divider's memory (05-card-window.md
/// ▸ The comments column, extended 2026-08-09). `@AppStorage` for the same reason its two
/// neighbours above are: every open card window's divider answers to one figure. `0` is the
/// "never dragged" reading (`AppPreferences.commentsColumnWidthKey`'s doc) — a real drag can
/// never land there, since `CardWindowMetrics.clampedCommentsColumnWidth`'s floor is always
/// positive.
@AppStorage(AppPreferences.commentsColumnWidthKey) private var storedCommentsColumnWidth: Double = 0
/// **The divider's in-flight width**, live only while a drag is running — `nil` the rest of the
/// time. `LaneResizeSession.liveWidth`'s same split applied to a much smaller session: the drag
/// tracks the cursor in memory on every tick (`CommentsColumnDivider.onChange`), and only the
/// *release* writes `storedCommentsColumnWidth` (`.onCommit`), so a flick that fires dozens of
/// events writes `UserDefaults` — and every sibling card window's divider — exactly once.
@State private var commentsColumnWidthOverride: CGFloat?
/// The body font's point size, read once per body evaluation: every measurement in this view —
/// the sidebar's width, both gutters, the vertical rhythm — is derived from it, so they scale
/// together when the system text size changes.
private var bodyPointSize: CGFloat { CardWindowMetrics.bodyPointSize }
/// Where the comments pane sits, when it is shown at all.
private var mount: CommentsMount { CommentsMount(besideBody: commentsBesideBody) }
/// The comments column's width before this render's clamp: the live drag's own width while one is
/// running, the persisted figure once the divider has ever been dragged and released, and
/// `CardWindowMetrics.commentsColumnWidth`'s font-derived default before either has ever happened.
/// Unclamped on purpose: clamping needs the beside `HStack`'s live container width, which only the
/// `.beside` render branch has (`clampedCommentsColumnWidth(in:)`), so this stays the one place
/// that decides *which number* to clamp rather than also deciding *how far*.
private var commentsColumnWidth: CGFloat {
if let commentsColumnWidthOverride { return commentsColumnWidthOverride }
return storedCommentsColumnWidth > 0
? CGFloat(storedCommentsColumnWidth)
: CardWindowMetrics.commentsColumnWidth(bodyPointSize: bodyPointSize)
}
/// `commentsColumnWidth` clamped against `containerWidth` — the beside `HStack`'s own width this
/// render (`CardWindowMetrics.clampedCommentsColumnWidth`). Re-derived every render rather than
/// written back into `storedCommentsColumnWidth`: a window too narrow for the persisted figure
/// (a smaller display, say) shows the clamped width without silently shrinking the preference a
/// wider window will want back.
private func clampedCommentsColumnWidth(in containerWidth: CGFloat) -> CGFloat {
CardWindowMetrics.clampedCommentsColumnWidth(
commentsColumnWidth,
bodyPointSize: bodyPointSize,
containerWidth: containerWidth
)
}
/// The two columns — **or the raw-source editor in place of both of them**.
///
/// A swap rather than an overlay, which is 05 ▸ Raw source outlet's own word for it ("swaps the
/// **entire content area — title, body, and sidebar —** for the literal on-disk `index.md`") and
/// what the rule underneath it requires: the same frontmatter is being edited as raw text, so a
/// sidebar still offering to restyle the card, or a title field still writing to `title`, would
/// be two editors racing for one file. Unmounting them is the only version of "they can't fight"
/// that cannot be got wrong later.
///
/// The cost is one thing and it is accepted: the body editor's scroll position and selection do
/// not survive a round trip through source mode, because its text view genuinely goes away. What
/// does survive is the buffer, which is the Edit session's, not the view's — and it was flushed
/// to disk on the way in regardless.
var body: some View {
content
// **The drop surface is the whole window** (05 ▸ Attachments), which is why it hangs
// here — outside the raw-source swap, so a file dropped while the source outlet is open
// still imports — rather than on the attachments section it fills. The body editor lets
// file drags through to this by declining the file types
// (`CardBodyTextView.acceptableDragTypes`); dragged *text* never matches `.fileURL` and
// so is never offered here at all, which is the other half of the payload split.
.modifier(WindowFileDrop(delegate: fileDrop))
}
@ViewBuilder
private var content: some View {
if rawSource.isActive {
// **All three panes**, comments included: "Raw Source still swaps the entire content area
// — all panes, comments included; the raw outlet's rule is unchanged" (05 ▸ The comments
// column). The swap encloses the whole composition below rather than any one pane, which
// is what keeps that true as the composition grows.
CardRawSourceView(session: rawSource, presentation: bodyPresentation)
} else {
HStack(spacing: 0) {
contentPanes
.frame(maxWidth: .infinity, maxHeight: .infinity, alignment: .topLeading)
// **View ▸ Show Sidebar** (`ShowSidebarCommand`, `CardToolbar`): the whole trailing
// column, divider included, comes and goes with the one bit — the body/comments
// columns above take the width back through their own `.infinity` frames, exactly as
// they do while the pane is merely narrow. A structural reflow, the Show Trash
// re-divide's own voice (`Motion.structural`), with the pane itself sliding off the
// trailing edge it lives on (`Motion.cardSidebarTransition`) rather than just
// vanishing.
if showSidebar {
Divider()
sidebar
// Fixed, and the one place it comes from.
.frame(width: CardWindowMetrics.sidebarWidth(bodyPointSize: bodyPointSize))
.frame(maxHeight: .infinity, alignment: .top)
.background(.background.secondary)
.transition(Motion.cardSidebarTransition(reduced: reduceMotion))
}
}
.animation(Motion.structural(reduced: reduceMotion), value: showSidebar)
}
}
/// The body pane and the comments pane, in whichever of the two mounts is current — or the body
/// pane alone, when Show Comments is off.
///
/// **The thread stays visible through body Edit in either mount**, and needs no rule of its own:
/// Edit swaps the content of the body pane (`CardBodySurface`), which is *inside* the body column
/// here, so nothing about the composition changes when the mode flips. That is the sidebar's own
/// precedent, which 05 names when it states the rule.
@ViewBuilder
private var contentPanes: some View {
if showComments {
switch mount {
case .beside:
// The `GeometryReader` is new here for the divider's sake — a resizable column needs
// the `HStack`'s own live width to clamp against (`clampedCommentsColumnWidth(in:)`),
// exactly the reason the stacked branch below already reaches for one.
GeometryReader { proxy in
HStack(spacing: 0) {
bodyColumn
.frame(maxWidth: .infinity, maxHeight: .infinity, alignment: .topLeading)
CommentsColumnDivider(
bodyPointSize: bodyPointSize,
containerWidth: proxy.size.width,
currentWidth: clampedCommentsColumnWidth(in: proxy.size.width),
onChange: { commentsColumnWidthOverride = $0 },
onCommit: { width in
storedCommentsColumnWidth = Double(width)
// The persisted figure now says the same thing the override does —
// clearing it hands `commentsColumnWidth` back to one source of truth
// rather than two that happen to agree.
commentsColumnWidthOverride = nil
}
)
commentsPane
// **User-resizable since 2026-08-09** — the sidebar's "resize flex
// always goes to the body" (05 ▸ Composition) still holds between the
// body and the *sidebar*; the body and the *comments column* now split
// it by the divider's drag instead, clamped so neither pane's own floor
// gives way (`CardWindowMetrics.clampedCommentsColumnWidth`).
.frame(width: clampedCommentsColumnWidth(in: proxy.size.width))
.frame(maxHeight: .infinity, alignment: .top)
}
}
case .stacked:
// The ≈3:2 split needs a height to divide, and a `GeometryReader` is the only way to
// have one — `layoutPriority` and flexible frames express *preferences*, and this is
// a ratio the design fixes. The body takes its share; the comments pane takes the
// remainder, so the divider between them can never leave a gap or overlap.
GeometryReader { proxy in
VStack(spacing: 0) {
bodyColumn
.frame(height: mount.bodyHeight(in: proxy.size.height))
.frame(maxWidth: .infinity, alignment: .topLeading)
Divider()
commentsPane
.frame(maxWidth: .infinity, maxHeight: .infinity, alignment: .topLeading)
}
}
}
} else {
bodyColumn
.frame(maxWidth: .infinity, maxHeight: .infinity, alignment: .topLeading)
}
}
private var commentsPane: some View {
CardCommentsPane(comments: comments, cardFolder: cardFolder, thumbnails: thumbnails)
}
// MARK: - Body column
/// Title, the quiet created/modified line, then the body — 05's top-to-bottom order.
private var bodyColumn: some View {
VStack(alignment: .leading, spacing: 0) {
VStack(alignment: .leading, spacing: bodyPointSize * 0.5) {
titleRow
if let dateLine {
Text(dateLine)
.font(.caption)
.foregroundStyle(.secondary)
.textSelection(.enabled)
}
}
.frame(maxWidth: .infinity, alignment: .leading)
.padding(.horizontal, CardWindowMetrics.gutter(bodyPointSize: bodyPointSize))
.padding(.top, CardWindowMetrics.gutter(bodyPointSize: bodyPointSize))
CardBodySurface(
// The session's text, never `card.body` directly: a dirty buffer outranks the
// snapshot (05 ▸ Write rules) and a flushed one is ahead of it by a reload, so the
// buffer is the truer of the two in both modes — which is also how Preview shows the
// text that produced it the instant Edit is left.
body: bodySession.text,
mode: bodyPresentation.mode,
cardFolder: cardFolder,
presentation: bodyPresentation,
session: bodySession,
isTaskToggleEnabled: isEditable,
onToggleTask: onToggleTask
)
.frame(maxWidth: .infinity, maxHeight: .infinity)
}
// **Dirty-buffer-wins, applied on every snapshot** (05 ▸ Write rules): the session takes
// disk's word for what the file says, and takes it into the editor only when the buffer has
// nothing unsaved. `initial: true` is also how the buffer is filled at all — a window opens
// by adopting its card's body.
.onChange(of: card.body, initial: true) { _, body in
bodySession.adopt(diskBody: body)
}
// The title field's own dirty-buffer-wins, `bodySession`'s exact rule one property over — see
// `titleSession`'s doc comment.
.onChange(of: card.title.value, initial: true) { _, title in
titleSession.adopt(diskTitle: title)
}
// **The opening rule, applied once** (05 ▸ Mode grammar): a card opens in Preview unless its
// body is empty, in which case it opens straight into Edit. `openIfNeeded` is what makes it
// "once" — a later reload that empties the file must not drag a reader into Edit.
.task { bodyPresentation.openIfNeeded(body: card.body) }
}
/// The header's title — large and borderless in both readings (m6-card-body). **Edit mode makes it
/// an editable field; Preview keeps it static text** — the same surfaces `CardBodySurface` gives
/// the body one column down, over the smaller session one property to the side.
///
/// Both branches read `titleSession.text`, never `card.title.value` directly, for the reason
/// `titleSession`'s doc comment gives: the buffer outranks the snapshot while it is dirty, and
/// outruns it by exactly one reload the instant a save lands, so reading the snapshot in Preview
/// would show a stale title for that reload's length — precisely what a card window must never do
/// (05-card-window.md ▸ Window).
@ViewBuilder
private var titleRow: some View {
if bodyPresentation.mode == .edit {
TextField(
"Card title",
text: Binding(
get: { titleSession.text },
set: { titleSession.edited($0) }
)
)
.textFieldStyle(.plain)
.lineLimit(1)
.font(.largeTitle)
// Return commits the field rather than exiting Edit mode — the title's own session ends
// with the body's (mode exit, window close), not with this key. `.onSubmit` never sees a
// literal newline in the first place: a plain `TextField` treats Return as a submit, which
// is the other half of "no newlines" beside `CardTitleEditSession.edited(_:)`'s paste
// filter.
.onSubmit { titleSession.flush() }
} else {
// Trimmed for the read-only rendering only — the field above shows `titleSession.text`
// raw, so a mid-typed run of padding never has its cursor position disturbed
// (`CardTitleEditSession.saveNow()` never rewrites the live buffer either, for the same
// reason). Preview only ever renders while nothing is being typed, so trimming purely for
// display is free of that hazard and is what keeps a title committed down to nothing
// (spaces alone) reading as "Untitled" here rather than as a run of blank space.
let displayed = titleSession.text.trimmingCharacters(in: .whitespacesAndNewlines)
Text(displayed.isEmpty ? "Untitled" : displayed)
.font(.largeTitle)
// "Untitled" is a rendering, never a value (03-board-ui.md § Card face) — the same
// secondary treatment the face gives it.
.foregroundStyle(displayed.isEmpty ? .secondary : .primary)
.textSelection(.enabled)
}
}
/// "Created ⟨date⟩ · Modified ⟨date⟩ · by ⟨modified-by⟩", **omitting whichever keys are absent**
/// (05 ▸ Composition) — the whole line disappears when the card carries none of the three.
///
/// The "by" segment renders only with the self-reported provenance stamp present
/// (01-storage-format.md), which is the point of showing it at all: provenance made visible with
/// no commit trail to read it from.
private var dateLine: String? {
var parts: [String] = []
if let created = card.created.value {
parts.append("Created \(Self.dateText(created))")
}
if let modified = card.modified.value {
parts.append("Modified \(Self.dateText(modified))")
}
if let by = card.modifiedBy.value, !by.isEmpty {
parts.append("by \(by)")
}
return parts.isEmpty ? nil : parts.joined(separator: " · ")
}
private static func dateText(_ date: Date) -> String {
date.formatted(date: .abbreviated, time: .shortened)
}
// MARK: - Attributes sidebar
/// The sidebar's sections: **Style, Details, Attachments** — Attachments at the bottom of the
/// stack (reordered 2026-08-09, Pipeline card 8f26b029), Actions gone entirely (retired the same
/// day, Pipeline card bcd3b323): its Delete and Reveal in Finder are now the card window's own
/// toolbar items (`CardToolbar`), reachable from Customize and, for Delete, on by default. This
/// is exactly the `VStack`'s child order, so keyboard Tab order and VoiceOver's reading order
/// follow it for free; ⇧⌘A (`AddAttachmentCommand`) still opens the same file panel regardless of
/// where Attachments sits in the stack, since it reaches the window through the focus system
/// rather than through this view's layout.
///
/// 05-card-window.md ▸ The attributes sidebar still documents the pathfinder's four-section order
/// with Actions at the bottom — an owed amendment, tracked on the two cards' own threads rather
/// than made here.
///
/// One of the two remaining sections is conditional, and the condition is the section's own
/// rather than a rule restated here: **Details** renders nothing when the card carries no unknown
/// frontmatter keys ("shown only when any exist"). Style and Attachments are unconditional, so
/// the composition a user learns on one card is the composition they get on the next.
///
/// The History section that once sat between Details and Actions left with app-managed git
/// (strategy/01-git-excision.md, 2026-08-08); View ▸ History (`FutureCommands.swift`) is the only
/// surviving reservation of that slot, and it anticipates the foreign-change journal successor.
private var sidebar: some View {
ScrollView(.vertical) {
VStack(alignment: .leading, spacing: bodyPointSize * 1.25) {
CardStyleSection(store: store, recents: recents, cardID: card.id, undo: undo)
// The snapshot's own document, not a re-read: the loader parsed this file, unknown
// keys and their order included, and `Card` has carried it since (`BoardModel`).
CardDetailsSection(rows: CardDetails.rows(of: card.document))
CardAttachmentsSection(attachments: attachments, thumbnails: thumbnails)
}
.frame(maxWidth: .infinity, alignment: .leading)
.padding(CardWindowMetrics.gutter(bodyPointSize: bodyPointSize))
}
}
}
// MARK: - View ▸ Show Sidebar
/// View ▸ Show Sidebar (checkmark toggle, **no default chord**) — the trailing attributes sidebar's
/// visibility (05-card-window.md ▸ Composition, extended for the toolbar toggle).
///
/// `ShowCommentsCommand`'s shape and reasons exactly: one bit, app-wide and persisted, the checkmark
/// reading exactly that bit so the row never lies, and it mirrors the toolbar's own Show Sidebar item
/// (`CardToolbar`) the way View ▸ Show Trash mirrors its toolbar twin — a menu command with a second
/// face, never two implementations of one.
///
/// Validation is **scope and nothing else**, `ShowCommentsCommand`'s own posture: with no card window
/// in front there is no focused value and the row disables, and the read-only lock plays no part —
/// showing or hiding a pane is not a mutation. `\.cardAttachments` is read only as the "a card window
/// is frontmost" signal `AddAttachmentCommand` already uses; the sidebar this toggles shows attachments
/// among its other sections, but the row does not otherwise touch that handle.
struct ShowSidebarCommand: View {
@FocusedValue(\.cardAttachments) private var attachments
@AppStorage(AppPreferences.showCardSidebarKey) private var isShown = true
static func isEnabled(_ attachments: CardAttachments?) -> Bool {
attachments != nil
}
/// The toggle's binding, `ShowTrashCommand.isVisible`'s own shape: the setter routes through
/// `AppPreferences.setShowCardSidebar` — the one write path this bit's two faces share — rather
/// than writing `$isShown` directly, which is what makes the animated reflow the toolbar item
/// gets true of this row too.
private var isVisible: Binding<Bool> {
Binding(get: { isShown }, set: { AppPreferences.setShowCardSidebar($0) })
}
var body: some View {
Toggle("Show Sidebar", isOn: isVisible)
.disabled(!Self.isEnabled(attachments))
}
}
// MARK: - The window-wide drop
/// Attaches the whole-window file drop, or nothing at all.
///
/// A modifier rather than an `if` inside `body` because `.onDrop` has to be applied to the *same*
/// view identity in both cases: a window whose store arrives a turn after its view would otherwise
/// re-mount its entire content when the drop target appeared, throwing away the body's scroll
/// position for nothing.
private struct WindowFileDrop: ViewModifier {
let delegate: CardWindowDropDelegate?
func body(content: Content) -> some View {
if let delegate {
// `.fileURL` alone: a text drag never matches, so it is never offered here and falls to
// the Edit editor, where `NSTextView` inserts it at the caret (05 ▸ Attachments).
content.onDrop(of: [.fileURL], delegate: delegate)
} else {
content
}
}
}