Skip to main content

max / makeover-layout

112.6 KB · 2663 lines History Blame Raw
1 //! The renderer-agnostic half of the make-family design system.
2 //!
3 //! <!-- wiki: makeover-layout -->
4 //!
5 //! `makeover` answers *what colour*, and varies by theme. `makeover-geometry`
6 //! answers *how much space*, and varies by density and surface. This crate
7 //! answers *what the thing is*, and varies by nothing.
8 //!
9 //! # The deferral rule
10 //!
11 //! A description names intents and relationships, never values. Say
12 //! [`Fill::Raised`], never `#D9DDF4`. Say `Gap::Peer`, never `6px`. What is
13 //! left once colour and spacing are deferred is **composition**: which edges
14 //! are lit, what inverts on press, what nests in what.
15 //!
16 //! The constraint that shapes all of it: a renderer that can only paint
17 //! rectangles has to be able to express the result. egui has no
18 //! `box-shadow: inset` and one stroke per widget with no per-side control; a
19 //! terminal has box-drawing characters and one cell of resolution, and cannot
20 //! draw a two-tone lit edge at all. A description that assumes per-side edges
21 //! is a CSS description wearing a neutral name. So this crate names the
22 //! *intent* — this region is a well — and each renderer chooses an expression
23 //! it can actually produce, including dropping half of one.
24 //!
25 //! # Scope
26 //!
27 //! Depth came first: the bevel and the surfaces it shapes. That much was
28 //! settled the hard way — the vocabulary here was read off audiofiles'
29 //! `ui::theme` and `ui::widgets`, which are the only implementation written
30 //! by a consumer with no CSS, then checked against both webview apps. All
31 //! three agreed once Balanced Breakfast's fills were corrected.
32 //!
33 //! 0.2.0 adds the rest of the description, each member drawn the same way,
34 //! from what the three apps already hand-write rather than from a taxonomy:
35 //!
36 //! - Components. [`Token`] (badge against chip), [`Notice`] (toast against
37 //! banner), [`RowPart`], [`Heading`], [`Selector`], [`Readiness`], and
38 //! [`Tone`], which is the one intent family they share.
39 //! - Schemas. [`Field`] for forms and [`Column`] for lists and tables.
40 //! - Structure. [`Region`] for the parts of a screen, [`Arrangement`] for how
41 //! a screen is put together.
42 //!
43 //! **Validation** was absent on purpose here, on the grounds that neither app
44 //! had a shared story. That reasoning is retired — see 0.11.0 below, which is
45 //! where the constraints arrived and why the argument did not survive contact
46 //! with what the apps were measured to do.
47 //!
48 //! 0.3.0 closes a gap the first real adoption found, which is what adopting
49 //! against goingson first was for. [`Selector`] described only the *chosen*
50 //! option, so an unchosen one fell through to [`Depth::Flat`] and no renderer
51 //! drew it; goingson's tab strip recesses its unchosen tabs by hand and could
52 //! not delete the line, because being recessed is *why* the chosen tab reads as
53 //! coming forward. So [`Selector::unchosen`] joins `chosen`, and saying it
54 //! needed [`Fill::Sunken`] and [`Depth::Sunken`]: a surface set back by colour
55 //! with no edge, which is neither a well nor level-with.
56 //!
57 //! 0.7.0 adds [`State`], the interaction axis, closing the gap that adopting
58 //! against three apps rather than one made visible. The description named
59 //! rest and, through [`Depth::pressed`], pressed. It named neither focus nor
60 //! disabled, so `makeover-webview` emitted a hover rule and stopped, and each
61 //! consumer completed the primitive from outside by out-specifying a rule it
62 //! did not own: 19 such rules in goingson, 21 in the MNW server, a further set
63 //! in Balanced Breakfast, and three focus rings that do not match. The axis is
64 //! deliberately two members wide, because hover and pressed belong where they
65 //! already are. [`State`]'s own docs carry that argument.
66 //!
67 //! 0.8.0 finishes [`Field`], which described a field well enough to label it and
68 //! not well enough to draw it. Writing `makeover-webview`'s form emitter found
69 //! three things missing and the renderer supplied all three from outside: the
70 //! current value, a select's options, and the placeholder. Two of those move
71 //! here and one does not.
72 //!
73 //! - [`Field::placeholder`] is user-facing text sitting beside `label` and
74 //! `hint`. There was never a reading on which it was renderer state; it was
75 //! outside only because adding a field to a published struct is breaking.
76 //! - [`Field::options`] moves because every renderer needs them and each was
77 //! going to invent its own shape. [`Choice`] is the shape `makeover-webview`
78 //! already arrived at, taken as-is rather than redesigned.
79 //! - The current value stays renderer-side and is not coming here. It is the
80 //! one of the three that is genuinely state: a webview reads it out of the
81 //! DOM, an immediate-mode renderer holds a `&mut` to the app's own field, and
82 //! a description that carried it would be a form model.
83 //!
84 //! 0.9.0 opens [`RowPart`], which was the last closed enum in the vocabulary,
85 //! and adds [`RowPart::Tokens`]. Both halves come from the same finding, made
86 //! by the first two real screens described through the router rather than by
87 //! reading a stylesheet.
88 //!
89 //! A goingson project card carries two trailing badges, a type and a toned
90 //! status; a contact card carries a primary email *and* a strip of tags. `Meta`
91 //! is one slot and one string, so both ports joined their facts with a
92 //! separator and lost what the second one was: a status reads as text where it
93 //! used to read as colour. [`Token`] already says exactly the right thing — a
94 //! small labelled thing with a kind, a tone and an optional action — and could
95 //! only ever be a node in its own right, never inside a row.
96 //!
97 //! So the missing thing was permission rather than a concept. `Tokens` is that
98 //! permission, and `#[non_exhaustive]` arrives with it so the next member is not
99 //! a lockstep event across three renderers. The pairing is the point: this
100 //! enum's own consumer in `makeover-webview` carried a comment predicting it
101 //! would stop compiling one day, which is a lockstep break written down and
102 //! waited for rather than prevented.
103 //!
104 //! Balanced Breakfast was checked before the member was added, because one
105 //! consumer wanting something is not evidence. It packs a count and two icon
106 //! buttons into the same single `Meta` slot while leaving `Actions` empty, so
107 //! the slot was already straining under a second consumer for a different
108 //! reason.
109 //!
110 //! 0.10.0 adds [`Meter`], a proportion carried as a pair rather than as a
111 //! percentage. Its own docs carry the argument; the short form is that the
112 //! percentage shape had already been tried in goingson and had already needed a
113 //! companion flag to recover what rounding and clamping threw away.
114 //!
115 //! 0.11.0 is four members from the quasi proving ground, batched into one
116 //! release because pre-1.0 a minor is breaking and a cascade is nine repos.
117 //! Three findings that arrived with them turned out not to belong here at all:
118 //! this crate has no notion of an action, a route or a destination, so anything
119 //! asking what a control *calls* was never the vocabulary's to say.
120 //!
121 //! - [`Figure`], a value with a caption. goingson had five of them across five
122 //! screens with five class vocabularies for the one shape, which is the
123 //! divergence this crate exists to end, sitting in plain sight and counted for
124 //! the first time.
125 //! - [`RowPart::Proportion`], so a [`Meter`] can sit in a row. `Meter` reached
126 //! two of its seven sites at 0.10.0 and the other five are row-shaped. Exactly
127 //! [`RowPart::Tokens`]'s problem with a different payload, and it takes
128 //! `Tokens`' answer: the part carries the description of a bar, not a node.
129 //! - [`Field::max_length`], [`Field::min`] and [`Field::max`], joining
130 //! [`Field::required`], which had been sitting here as the sole constraint
131 //! while the header above claimed there were none. The set stops before
132 //! `pattern`, which fails the renderer test and is one site in one app.
133 //! - [`FieldKind::File`]. Every host has an honest answer — a native picker, an
134 //! `<input type="file">`, a path prompt, an argument — and it carries no
135 //! accepted-types list because `accept` appears at zero sites in either app.
136 //!
137 //! The evidence rule changed under these, and it is worth recording because four
138 //! earlier decisions were made under the old one. The two-app test said a shape
139 //! earns a word once a second app wants it. It is backwards: a rule that
140 //! withholds a word until a second app has duplicated the code guarantees the
141 //! duplication, and app three writes it a third time. The bar is now generic
142 //! against bespoke — is this furniture any app would have, or is it this app's
143 //! own? Bespoke keeps [`Region::Bespoke`], which already carries a completion
144 //! heatmap and is the right answer for a calendar nobody will build twice.
145 //!
146 //! 0.12.0 is two more from the same proving ground, and the same sorting
147 //! happened first: six findings came out of a measurement of goingson's whole
148 //! frontend, and four of them turned out to be asking what a control *calls*,
149 //! which this crate cannot say. The two that were really here:
150 //!
151 //! - [`Readiness`] grows from two states to four. It named `Ready` and
152 //! `Pending` and stopped, so a screen whose list came back empty had nothing
153 //! to say about it; goingson draws an empty state at 27 sites and Balanced
154 //! Breakfast at 9. `Empty` and `Failed` are the same axis rather than a new
155 //! member beside it, because a region shows one of the four and never two.
156 //! `#[non_exhaustive]` arrives with them, the pairing [`RowPart`] made at
157 //! 0.9.0 and for the same reason.
158 //! - [`Column::sortable`], [`Column::sorted`] and [`Sort`]. The one finding in
159 //! the set that completes a member rather than adding one: `Column` shipped
160 //! with a width and a priority and could not say that a table is ordered by a
161 //! column, so a described table could draw no caret and offer no reordering.
162 //!
163 //! 0.14.0 is two additive members on two `#[non_exhaustive]` enums, released
164 //! together because publishing twice for that is waste and the cascade below
165 //! this crate is nine repos.
166 //!
167 //! - [`Depth::Overlay`]. The enum could say raised, well, sunken and flat, and
168 //! could not say that a surface sits *over* the page. Every renderer already
169 //! had the surface — `makeover-tui`'s `Palette::overlay`,
170 //! `makeover-immediate`'s `Palette::elevation`, `makeover-webview`'s
171 //! `--elevation-overlay` — and none of them could be reached from a
172 //! description. buckets_of_money has 16 modals waiting on it.
173 //! - [`CellPart`], which is [`RowPart`] for tables. A row's parts have carried
174 //! their own content intent since 0.2.0, so `.row-actions` inherits rather
175 //! than taking a text colour; a table cell had no such vocabulary and
176 //! `makeover-webview` emitted one undifferentiated `.cell`, so a button in a
177 //! cell was painted as text. The four members are the four things quasi's
178 //! `Cell` was measured to hold, and the count is in that crate's history
179 //! rather than assumed here.
180 //!
181 //! 0.15.0 adds [`FieldKind::Date`] and [`FieldKind::DateTime`], on the argument
182 //! [`FieldKind::Email`] was admitted on: a webview emits a different `type=`,
183 //! which is a native picker, the platform's validation and a different keyboard
184 //! on a touch device. Described as text with a "YYYY-MM-DD" hint, all three are
185 //! lost.
186 //!
187 //! Two members and not one or five, from a count rather than from symmetry: 13
188 //! sites of `date` and 13 of `datetime-local` across the MNW server and
189 //! goingson, and zero of `time`, `month` or `week`. The wire format each takes
190 //! is named here as [`DATE_FORMAT`] and [`DATETIME_FORMAT`], because a host
191 //! left to pick its own would disagree with a server silently, and
192 //! [`FieldKind::temporal`] is the pair asked about once rather than at each
193 //! renderer. `FieldKind`'s own comment claiming `radio` was the last HTML input
194 //! type missing was already false when 0.8.1 wrote it; these are what it was
195 //! missing.
196 //!
197 //! What each of the 0.12.0 findings deliberately leaves out is the address — what pressing a
198 //! header calls, and where an empty state's "Add your first project" button
199 //! goes. That is the boundary this crate is defined by, and four findings moved
200 //! across it rather than being answered here.
201 //!
202 //! # Where the description stops
203 //!
204 //! The bespoke widgets, a day-plan timeline and a kanban board and a calendar,
205 //! are not describable here and will not become describable. A description
206 //! expressive enough to produce a timeline is a widget library wearing a
207 //! description's name. Generate the boring 80% so the bespoke 20% gets the
208 //! attention.
209 //!
210 //! [`Region::Bespoke`] is how that limit is stated rather than hidden. The
211 //! description names the *place* and the app owns the contents, so a screen
212 //! containing a timeline is still a whole screen and still routable. Without
213 //! it, the four goingson screens that make the app worth using would need a
214 //! second, undescribed path beside the router, and two paths is how a
215 //! vocabulary starts drifting from its app again.
216
217 #![forbid(unsafe_code)]
218
219 /// A colour intent this crate refers to but never resolves.
220 ///
221 /// The string is the token name `makeover` publishes, so a renderer can look
222 /// it up without this crate knowing what colour came back.
223 pub trait Intent {
224 /// The `makeover` intent token this resolves against.
225 fn token(self) -> &'static str;
226 }
227
228 /// Which way the light falls across a two-tone edge.
229 ///
230 /// The whole content of a bevel, once colour and thickness are deferred. The
231 /// light is always assumed to come from the top left: every consumer measured
232 /// agreed on that and none of them ever varied it, so it is an invariant here
233 /// rather than a parameter.
234 ///
235 /// # The two corners that belong to both edges
236 ///
237 /// Top-right and bottom-left are where the lit run meets the shaded one, and
238 /// the description's claim is that they belong to *both*. How a renderer says
239 /// that is its own business, because the answer is bounded by resolution and
240 /// not by taste:
241 ///
242 /// - A terminal cell is roughly 8x17 device pixels, so giving the whole corner
243 /// to one tone thickens that edge by a cell and reads as one run overrunning
244 /// the other. A half-cell glyph divides the cell already, so `makeover-tui`
245 /// splits it and recovers real information. Its box-drawing fallback cannot:
246 /// a single stroke has no half to give, so there both corners go to dark.
247 /// - A pixel bevel is a one-point stroke by default, which makes the corner a
248 /// one-point square. There is nothing to divide — a diagonal seam across one
249 /// point is sub-pixel, and antialiasing renders it as the blend a mitred join
250 /// already produces. So `makeover-immediate` mitres and is *not* diverging;
251 /// it is the same rule at a resolution where the split degenerates.
252 ///
253 /// Stated here so the difference reads as a decision rather than as drift. A
254 /// renderer with room to divide the corner should; one without should mitre or
255 /// pick the shaded tone, and neither is a bug.
256 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
257 pub enum Bevel {
258 /// Lit from the top left: light on top and left, dark on bottom and right.
259 Raised,
260 /// The same edge inverted, which is also the pressed state of anything
261 /// that draws itself [`Bevel::Raised`].
262 Inset,
263 }
264
265 impl Bevel {
266 /// The edge intents, as `(top_left, bottom_right)`.
267 ///
268 /// Split out from any painting because the inversion *is* the idea, and
269 /// it is the one part every renderer implements identically.
270 #[must_use]
271 pub const fn edges(self) -> (Edge, Edge) {
272 match self {
273 Self::Raised => (Edge::Light, Edge::Dark),
274 Self::Inset => (Edge::Dark, Edge::Light),
275 }
276 }
277
278 /// Pressing inverts. A raised control reads as inset while held.
279 ///
280 /// Stated here rather than left to each consumer because a cascade can
281 /// carry a pressed state and an immediate-mode renderer cannot: audiofiles
282 /// resolves this per call site, eighteen times.
283 #[must_use]
284 pub const fn pressed(self) -> Self {
285 match self {
286 Self::Raised => Self::Inset,
287 Self::Inset => Self::Raised,
288 }
289 }
290 }
291
292 /// One side of a bevel, named by the intent it takes.
293 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
294 pub enum Edge {
295 /// The lit side.
296 Light,
297 /// The shadowed side.
298 Dark,
299 }
300
301 impl Intent for Edge {
302 fn token(self) -> &'static str {
303 match self {
304 Self::Light => "bevel-light",
305 Self::Dark => "bevel-dark",
306 }
307 }
308 }
309
310 /// A surface intent a region is filled with.
311 ///
312 /// `#[non_exhaustive]`, so a renderer must carry a wildcard arm and a new
313 /// member is additive rather than breaking. Added 0.4.0, after [`Sunken`]
314 /// (an additive member, 0.3.0) hard-broke `makeover-tui` and
315 /// `makeover-immediate` at compile time and left neither able to move until
316 /// both published. The vocabulary exists to grow and the renderers exist to
317 /// disagree about how much of it they answer, so growth must not be a
318 /// lockstep event. The renderer's wildcard is not a hole: [`Fill`] is
319 /// resolved through a fallible lookup, and a missing intent is answered with
320 /// structure rather than with a substituted colour.
321 ///
322 /// [`Sunken`]: Fill::Sunken
323 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
324 #[non_exhaustive]
325 pub enum Fill {
326 /// The page behind everything.
327 Page,
328 /// A surface lifted off the page: cards, controls, menus, toasts.
329 Raised,
330 /// A surface floating above the page rather than resting on it.
331 Overlay,
332 /// The inside of a well.
333 Well,
334 /// A surface set back from the one it sits on, by colour and nothing else.
335 ///
336 /// Not a well. A well is a hole with an edge, and the two are authored in
337 /// opposite directions: `makeover` derives `surface-well` by inverting
338 /// against the theme's own content colour, while `surface-sunken` is
339 /// authored and free to sit darker than raised (goingson's does). Naming
340 /// only the well left the recessed-with-no-edge surface unsayable, which is
341 /// what an unchosen tab is: it recedes so the chosen one can come forward,
342 /// and it carries no bevel of its own.
343 ///
344 /// Added 0.3.0, from goingson's tab strip, which hand-writes exactly this
345 /// and could not delete the line because no member described it.
346 Sunken,
347 }
348
349 // No `fallback` here, deliberately. An earlier cut had `Fill::Well` fall back
350 // to `Fill::Page` so a consumer on makeover 2.2.0, which has no `surface-well`,
351 // had something to paint. makeover-tui found that wrong within a day: page is
352 // the surface a well is usually cut into, so on a terminal that substitution
353 // produces exactly the invisibility it was meant to prevent, and the right
354 // answer there is a drawn edge rather than a different colour.
355 //
356 // Substituting one intent for another is renderer policy. The description says
357 // what the region is and stops.
358
359 impl Intent for Fill {
360 fn token(self) -> &'static str {
361 match self {
362 Self::Page => "surface-page",
363 Self::Raised => "surface-raised",
364 Self::Overlay => "surface-overlay",
365 Self::Well => "surface-well",
366 Self::Sunken => "surface-sunken",
367 }
368 }
369 }
370
371 /// How a region sits relative to the surface behind it.
372 ///
373 /// Fill and bevel are named together because naming them apart is what let
374 /// them disagree. Every consumer measured had at least one region carrying a
375 /// raised bevel over a recessed fill: audiofiles fixed it in `raised_frame`
376 /// and recorded the bug in its doc comment, and Balanced Breakfast still had
377 /// twelve of them a year later. A single name for the pair makes that
378 /// unrepresentable.
379 /// `#[non_exhaustive]` for the same reason as [`Fill`], and in the same
380 /// release: a depth this renderer has no drawing for should cost it a
381 /// wildcard arm, not a compile error and a wait on someone else's publish.
382 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
383 #[non_exhaustive]
384 pub enum Depth {
385 /// Level with its surroundings. No edge.
386 Flat,
387 /// A card laid on the panel it sits in.
388 Raised,
389 /// A hole in the panel, with content down inside it. For anything the
390 /// user looks *into*: a table body, a tag tree, a text field.
391 Well,
392 /// Set back from what it sits on, by colour alone. No edge.
393 ///
394 /// The one member carrying a fill without a bevel, so a renderer cannot
395 /// assume the two arrive together. That is deliberate and it is still the
396 /// pairing rule: both halves come off the same `Depth`, so they cannot
397 /// disagree, and here one half is legitimately absent.
398 ///
399 /// Distinct from [`Depth::Flat`], which has no fill either and inherits.
400 /// Recessed and level-with are different claims, and only one of them
401 /// needs a colour.
402 Sunken,
403 /// A surface sitting *over* the page rather than in it. A modal, a popover,
404 /// a menu.
405 ///
406 /// Takes elevation and no bevel: a surface overlaying the page is lifted
407 /// off it, and a surface in the page is cut into it. That is the same
408 /// pairing rule the rest of the enum holds, applied to the one case where
409 /// the separation is not an edge at all — the lift and the scrim behind it
410 /// are already saying where the surface is.
411 ///
412 /// Every renderer had the surface before it had this variant.
413 /// `makeover-tui` carries `Palette::overlay`, `makeover-immediate` gained
414 /// `Palette::elevation` at 0.10.0, and `makeover-webview` emits
415 /// `--elevation-overlay`. What was missing was the route from a description
416 /// to any of them, which is why this is one variant rather than a feature.
417 Overlay,
418 }
419
420 impl Depth {
421 /// The edge this depth is drawn with, if it has one.
422 #[must_use]
423 pub const fn bevel(self) -> Option<Bevel> {
424 match self {
425 // Sunken joins Flat here, for the opposite reason: Flat has no edge
426 // because nothing separates it from its surroundings, and Sunken has
427 // none because its colour is already doing the separating.
428 Self::Flat | Self::Sunken => None,
429 // A third reason to have no edge, which is why it gets its own arm
430 // rather than joining the two above: an overlay is separated by the
431 // lift and by the scrim behind it, so an edge would be a second
432 // answer to a question already answered.
433 Self::Overlay => None,
434 Self::Raised => Some(Bevel::Raised),
435 Self::Well => Some(Bevel::Inset),
436 }
437 }
438
439 /// The surface this depth is filled with.
440 ///
441 /// [`Depth::Flat`] has no fill of its own: it inherits whatever it sits on,
442 /// which is the difference between level-with and painted-the-same-colour.
443 #[must_use]
444 pub const fn fill(self) -> Option<Fill> {
445 match self {
446 Self::Flat => None,
447 Self::Raised => Some(Fill::Raised),
448 Self::Well => Some(Fill::Well),
449 Self::Sunken => Some(Fill::Sunken),
450 Self::Overlay => Some(Fill::Overlay),
451 }
452 }
453
454 /// Pressing a raised region reads as a well, and nothing else moves.
455 ///
456 /// [`Depth::Overlay`] is untouched along with the rest: an overlay is a
457 /// surface, not a control, so there is nothing there to press.
458 #[must_use]
459 pub const fn pressed(self) -> Self {
460 match self {
461 Self::Raised => Self::Well,
462 other => other,
463 }
464 }
465 }
466
467 /// An interaction state a region can be in, beside whatever [`Depth`] it is.
468 ///
469 /// Orthogonal to depth on purpose. A disabled button is still [`Depth::Raised`]
470 /// and a disabled field is still a [`Depth::Well`], so folding either member
471 /// into `Depth` would make [`Depth::bevel`] and [`Depth::fill`] answer for
472 /// something that is not a depth, and would leave disabled-button and
473 /// disabled-field sharing one variant that cannot tell them apart.
474 ///
475 /// # Why hover and pressed are not members
476 ///
477 /// The line is whether every renderer has the state to express, not whether CSS
478 /// does. Hover is renderer policy and `makeover-webview` says so in its own
479 /// header: a terminal and an immediate-mode painter have no pointer hovering
480 /// over anything, and pressed already arrives through [`Bevel::pressed`] and
481 /// [`Depth::pressed`], where it belongs, because pressing is a depth inversion
482 /// rather than a separate condition.
483 ///
484 /// Focus and disabled are different in kind. A TUI has a focused widget and a
485 /// greyed-out one; so does egui. Both were unsayable here, so all three webview
486 /// consumers supplied them from outside the primitive by out-specifying rules
487 /// they did not own: goingson alone carries 19 of them, and the MNW server
488 /// another 21. That is the divergence this crate exists to end, arriving one
489 /// layer down.
490 ///
491 /// # The principle this encodes
492 ///
493 /// A primitive owns every state it implies. A renderer that emits a hover rule
494 /// for a thing owes disabled, focus and the capability answer for that same
495 /// thing, because anything less exports the completion work to N consumers who
496 /// will each do it differently.
497 ///
498 /// `#[non_exhaustive]` for the reason [`Fill`] and [`Depth`] carry it: growth
499 /// must not be a lockstep event across the three renderers.
500 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
501 #[non_exhaustive]
502 pub enum State {
503 /// Keyboard focus, as distinct from the pointer having landed on something.
504 ///
505 /// One ring, not one per primitive. Where the ring sits is [`Depth`]'s
506 /// question and not a per-component choice: a well takes it inside its own
507 /// edge and a raised surface takes it outside. That is one decision with
508 /// two renderings rather than one decision per component, which is how the
509 /// three apps ended up with three rings.
510 Focus,
511 /// Present, visible, and not answering.
512 ///
513 /// Not the same as absent, and deliberately not a [`Fill`]: a disabled
514 /// control keeps the surface it always had and stops responding, so what
515 /// changes is its content and its interactivity rather than what it is.
516 Disabled,
517 }
518
519 impl State {
520 /// Whether a region in this state stops answering the pointer.
521 ///
522 /// Stated in the description rather than left to each renderer, on the same
523 /// reasoning as [`Bevel::pressed`]: a cascade carries it for free and an
524 /// immediate-mode renderer resolves it per call site, so leaving it unsaid
525 /// means resolving it once per consumer and disagreeing.
526 #[must_use]
527 pub const fn suppresses_interaction(self) -> bool {
528 match self {
529 Self::Disabled => true,
530 Self::Focus => false,
531 }
532 }
533 }
534
535 impl Intent for State {
536 fn token(self) -> &'static str {
537 match self {
538 // Already derived by `makeover` from `action.primary`, and unused
539 // until now for the same reason `hover-surface` was: nothing
540 // emitted the rule that would consume it.
541 Self::Focus => "focus-ring",
542 // Reusing the muted content intent rather than minting a
543 // `disabled` colour. Disabled is a reduction and not a status, and
544 // `makeover-webview`'s progress rules already record the reading
545 // that `content-muted` is what disabled looks like.
546 Self::Disabled => "content-muted",
547 }
548 }
549 }
550
551 /// What a region is saying, when it is saying something.
552 ///
553 /// The one intent family shared by badges, notices and nothing else. Kept
554 /// separate from [`Fill`] because a surface is where a thing sits and a tone is
555 /// what it means, and the three apps agree on the four statuses:
556 /// `info_banner` / `warning_banner` in audiofiles, `.toast-info` /
557 /// `.toast-success` / `.toast-error` in goingson, `.toast.success` /
558 /// `.toast.error` in Balanced Breakfast.
559 ///
560 /// The per-tag palette (`category-one` through `category-six`) is deliberately
561 /// not here. Which colour a *particular* tag takes is app domain, and both
562 /// webview apps already carry it as a `data-color` attribute.
563 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
564 pub enum Tone {
565 /// No status. Reads as ordinary de-emphasised content.
566 Neutral,
567 /// Something worth knowing and nothing to do about it.
568 Info,
569 /// Something finished and it worked.
570 Success,
571 /// Something the user should look at before continuing.
572 Warning,
573 /// Something broken, or something about to be destroyed.
574 Danger,
575 }
576
577 impl Intent for Tone {
578 fn token(self) -> &'static str {
579 match self {
580 // Neutral has no status token of its own. It takes the muted
581 // content intent, which is what both webview apps already spell as
582 // `data-color="muted"`.
583 Self::Neutral => "content-muted",
584 Self::Info => "info",
585 Self::Success => "success",
586 Self::Warning => "warning",
587 Self::Danger => "danger",
588 }
589 }
590 }
591
592 /// A small labelled thing that sits inside something else.
593 ///
594 /// Two members, because the three apps drew three taxonomies and only one line
595 /// runs through all of them: does it answer a click. audiofiles has
596 /// `classification_badge` (a label) against `tag_chip`, `tag_chip_removable`
597 /// and `selectable_tag` (all of which do). Balanced Breakfast has `.tag` and
598 /// `.badge` against `.tag-chip`. goingson is the one that has to move: its
599 /// `.tag` and `.badge` are a single CSS rule, so every call site has to be read
600 /// to decide which of the two it always was.
601 ///
602 /// The evidence that a chip is a real concept rather than a badge with a
603 /// cursor: audiofiles inverts its bevel on press and Balanced Breakfast latches
604 /// `.tag-chip.active` with the inset bevel. Two independent arrivals at "a chip
605 /// holds itself down", which is exactly what [`Depth::pressed`] already says.
606 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
607 pub enum Token {
608 /// Non-interactive status or count. Answers no click.
609 Badge,
610 /// An interactive or removable token. Answers a click, and latches if it
611 /// stands for a filter that is either on or off.
612 Chip {
613 /// Whether it carries its own remove affordance.
614 removable: bool,
615 },
616 }
617
618 impl Token {
619 /// Whether this answers a click.
620 ///
621 /// The whole difference between the two members, and the reason a renderer
622 /// with no hover (a touch surface, a terminal) can still tell them apart.
623 #[must_use]
624 pub const fn interactive(self) -> bool {
625 matches!(self, Self::Chip { .. })
626 }
627
628 /// How it sits, given whether it is currently latched down.
629 ///
630 /// A badge is flat: it is a label, and giving it an edge would say it can
631 /// be pressed. A chip is raised, and inset while latched.
632 #[must_use]
633 pub const fn depth(self, latched: bool) -> Depth {
634 match self {
635 Self::Badge => Depth::Flat,
636 Self::Chip { .. } if latched => Depth::Well,
637 Self::Chip { .. } => Depth::Raised,
638 }
639 }
640 }
641
642 /// Something the app is telling the user, unprompted.
643 ///
644 /// Two concepts, not one with a placement. They differ in more than where they
645 /// sit: a toast is transient, stacked and self-dismissing, and a banner is
646 /// persistent, in flow, one per region, and dismissed by fixing the condition
647 /// it reports. Folding them into one member with a placement parameter would
648 /// make lifetime, stacking and dismissal all placement-dependent, which is the
649 /// description leaking renderer policy.
650 ///
651 /// All three apps have banners: `info_banner` and `warning_banner` in
652 /// audiofiles, five of them in goingson (sync, sync-result, vacation-day,
653 /// timer-active, past-review), `.update-banner` in Balanced Breakfast. The two
654 /// webview apps also have toasts. So neither member is speculative, and no app
655 /// gains a concept it lacks except audiofiles, whose renderer may legitimately
656 /// decline to draw a toast at all.
657 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
658 pub enum Notice {
659 /// Transient, stacked, dismisses itself.
660 Toast,
661 /// Persistent, in flow, one per region, dismissed by fixing the cause.
662 Banner,
663 }
664
665 impl Notice {
666 /// Whether it goes away on its own.
667 #[must_use]
668 pub const fn transient(self) -> bool {
669 matches!(self, Self::Toast)
670 }
671
672 /// How it sits.
673 ///
674 /// A toast floats above the page rather than resting on it, which is
675 /// [`Fill::Overlay`]'s whole reason to exist. A banner is a card in the
676 /// flow. Both are raised, and they are raised off different things.
677 #[must_use]
678 pub const fn fill(self) -> Fill {
679 match self {
680 Self::Toast => Fill::Overlay,
681 Self::Banner => Fill::Raised,
682 }
683 }
684 }
685
686 /// The parts of a list row.
687 ///
688 /// Four to begin with, taken from Balanced Breakfast, which was the only
689 /// consumer that had all of them (`row-primary`, `row-secondary`, `row-meta`,
690 /// `row-actions`). audiofiles has two and no slot structure at all, so it gains
691 /// meta and actions as real work rather than a rename; goingson moves off
692 /// `task-row` / `task-cell`.
693 ///
694 /// [`Tokens`](Self::Tokens) joined at 0.9.0, and `#[non_exhaustive]` with it.
695 /// See the crate header for why the two arrived together.
696 ///
697 /// # Meta against Tokens
698 ///
699 /// The line is whether the thing has its own standing. `Meta` is one short
700 /// trailing fact about the row, written as text: a count, a size, a date.
701 /// `Tokens` is a set of small labelled things, each of which can be toned and
702 /// can answer a click. "3 files" is meta. A status badge that is amber, and a
703 /// tag you can click to filter by, are tokens.
704 ///
705 /// Keeping them apart is what a single widened slot would have foreclosed. A
706 /// renderer can right-align one string and cannot usefully do the same to a
707 /// strip of chips, and a fact that is not clickable should not be drawn as
708 /// though it were.
709 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
710 #[non_exhaustive]
711 pub enum RowPart {
712 /// The thing itself. What the row is called.
713 Primary,
714 /// Supporting text under the primary.
715 Secondary,
716 /// A short trailing fact: a count, a size, a date.
717 Meta,
718 /// Controls that act on this row.
719 Actions,
720 /// Small labelled things belonging to the row: badges, chips, tags.
721 ///
722 /// Each carries its own [`Token`] kind and [`Tone`], so a renderer with no
723 /// colour still has the kind to work with, and one with no chips still has
724 /// the label. That is the constrained-consumer test this vocabulary exists
725 /// to pass, and it is why the tone lives on the token rather than on the
726 /// part.
727 Tokens,
728 /// How much of a set the row's thing has done: a [`Meter`] in the row.
729 ///
730 /// Added 0.11.0, `da5666ae`, and it is [`Tokens`](Self::Tokens)'s problem
731 /// again with a different payload. [`Meter`] arrived at 0.10.0 and closed
732 /// two of the seven sites that asked for it; the other five sit in rows, and
733 /// a row holds no nodes by the ruling that a row part may not carry an
734 /// arbitrary node — the door through which a description becomes a
735 /// templating language. So the part carries the *description of a bar*
736 /// rather than a node, exactly as `Tokens` carries tags rather than nodes.
737 ///
738 /// Without it a row flattens the proportion into [`Meta`](Self::Meta) as
739 /// "3/7 subtasks", which keeps both numbers and loses the reading, the same
740 /// way a toned status badge read as prose before `Tokens`.
741 Proportion,
742 }
743
744 impl RowPart {
745 /// The content intent the part takes.
746 #[must_use]
747 pub const fn intent(self) -> &'static str {
748 match self {
749 Self::Primary => "content",
750 Self::Secondary => "content-secondary",
751 Self::Meta => "content-muted",
752 // Actions carry controls rather than text, so they inherit.
753 Self::Actions => "content",
754 // So do tokens: each one carries its own tone, and a part-level
755 // intent underneath it would fight the token that sits on it.
756 Self::Tokens => "content",
757 // And so does a proportion, for the same reason: the meter carries
758 // the tone, and it is about the ratio rather than about the row.
759 Self::Proportion => "content",
760 }
761 }
762 }
763
764 /// How far down the heading tree a title sits.
765 ///
766 /// Three, and only the three that are actually headings. The bands those used
767 /// to be filed with (goingson's `.page-header`, Balanced Breakfast's `.header`
768 /// and `.detail-header`) are arrangement, not type, and live at
769 /// [`Region::Band`]. One of them contains no text at all.
770 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
771 pub enum Heading {
772 /// Names the whole screen. One per screen.
773 Page,
774 /// Names a block within the screen.
775 Section,
776 /// Names a sub-block inside an already-named section.
777 Subsection,
778 }
779
780 impl Heading {
781 /// Whether a rule follows the heading.
782 ///
783 /// audiofiles' `section_header` draws a separator and its
784 /// `subsection_label` deliberately does not, which is the only thing
785 /// distinguishing the two once weight and colour are deferred.
786 #[must_use]
787 pub const fn separated(self) -> bool {
788 matches!(self, Self::Section)
789 }
790 }
791
792 /// A control that picks between things.
793 ///
794 /// Three, because three distinct behaviours are in play and collapsing any two
795 /// loses something. A segmented control picks a value; a tab picks a pane; a
796 /// toggle picks nothing and simply holds itself on or off.
797 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
798 pub enum Selector {
799 /// Exactly one of N, and the options abut.
800 Segmented,
801 /// Independent on or off, on its own.
802 Toggle,
803 /// Navigation between panes. The folder semantic.
804 Tabs,
805 }
806
807 impl Selector {
808 /// How the chosen option sits.
809 ///
810 /// Held in for a segmented control and a toggle, which is the same shape
811 /// pressing produces and the whole economy of the idiom: one appearance,
812 /// two reasons to wear it. A tab is the exception, because the selected
813 /// folder tab comes *forward* to join the pane it opens.
814 #[must_use]
815 pub const fn chosen(self) -> Depth {
816 match self {
817 Self::Segmented | Self::Toggle => Depth::Well,
818 Self::Tabs => Depth::Raised,
819 }
820 }
821
822 /// How the options that were *not* picked sit.
823 ///
824 /// Added 0.3.0. Describing only [`Selector::chosen`] left the unchosen
825 /// option falling through to [`Depth::Flat`], which says it is level with
826 /// the strip it sits in, and no renderer emitted anything for it. That is
827 /// wrong in both directions and goingson proved it: its unchosen tabs are
828 /// recessed by hand, and being recessed is *why* the chosen one reads as
829 /// coming forward. Against a flat strip, a raised chosen tab is a bevel
830 /// drawn on the strip's own colour, which is a much weaker folder effect
831 /// than the contrast the idiom is named after.
832 ///
833 /// Each member is the inverse of its chosen state, which is the whole
834 /// content of "picked" once colour is deferred:
835 ///
836 /// - Tabs recede, so the chosen one comes forward.
837 /// - A segment and a toggle stand up, so the chosen one is held in.
838 #[must_use]
839 pub const fn unchosen(self) -> Depth {
840 match self {
841 Self::Tabs => Depth::Sunken,
842 Self::Segmented | Self::Toggle => Depth::Raised,
843 }
844 }
845
846 /// Whether the options touch.
847 ///
848 /// The gap is the entire difference between a segmented control and a row
849 /// of buttons that happen to sit near each other, which is what audiofiles'
850 /// `segmented_control` says in its own comment and why it zeroes the
851 /// spacing by hand.
852 #[must_use]
853 pub const fn abutting(self) -> bool {
854 matches!(self, Self::Segmented | Self::Tabs)
855 }
856 }
857
858 /// What is in a region right now.
859 ///
860 /// The state, not the shimmer. Whether pending paints a skeleton, a spinner or
861 /// nothing at all is renderer policy, the same class of decision that got
862 /// `Fill::fallback` deleted from this crate. goingson and Balanced Breakfast
863 /// each grew a skeleton with differently-named parts; both keep them, as the
864 /// webview renderer's expression of [`Readiness::Pending`]. audiofiles has none
865 /// and needs none, because an immediate-mode renderer simply repaints.
866 ///
867 /// # Four states and not two, as of 0.12.0
868 ///
869 /// `703f4cd2`. It named `Ready` and `Pending` and stopped, so a described screen
870 /// whose list came back empty had to render an empty region or invent its own
871 /// placeholder text, and neither says what it is. goingson draws one at 27 sites
872 /// across 12 files and Balanced Breakfast at 9, with a class family that had
873 /// already drifted into `empty-state`, `empty-state--error`, `error-state` and
874 /// six more.
875 ///
876 /// The four are one axis because they are mutually exclusive: a region shows its
877 /// content, or a sign that it is coming, or a sign that there is none, or a sign
878 /// that it broke. Never two. That is the test for one enum against several
879 /// fields, and it is why this grew rather than a new member arriving beside it.
880 ///
881 /// # What is not here
882 ///
883 /// **The message.** "No projects yet" is content, and this names a state. It
884 /// lives with whatever holds the region — in quasi's case a `Slot` — alongside
885 /// the action that leads out of the emptiness, since an address is the one thing
886 /// this crate never names.
887 ///
888 /// **How much room it gets.** goingson's `--compact`, `--dashboard` and
889 /// `--padded` are the same state at three sizes, and a size is
890 /// `makeover-geometry`'s question. Naming them here would be this crate stating
891 /// values again.
892 ///
893 /// **The icon.** Presentation, and each host has its own answer or none.
894 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
895 #[non_exhaustive]
896 pub enum Readiness {
897 /// The content is here.
898 Ready,
899 /// The content is on its way.
900 Pending,
901 /// The content arrived and there is none of it.
902 ///
903 /// Not a failure. An empty list is the normal state of a new install, and a
904 /// renderer that drew it in a danger tone would be reporting a fault where
905 /// there is none.
906 Empty,
907 /// The content did not arrive.
908 Failed,
909 }
910
911 impl Readiness {
912 /// Whether the region draws its own content, or something standing in for
913 /// it.
914 ///
915 /// The question every renderer asks first, so it is answered once here
916 /// rather than by a `matches!` in each. A state added later is a stand-in
917 /// until proven otherwise: falling back to drawing content that may not be
918 /// there is the worse of the two mistakes.
919 #[must_use]
920 pub const fn shows_content(self) -> bool {
921 matches!(self, Self::Ready)
922 }
923
924 /// What the state means, for a renderer choosing a colour.
925 ///
926 /// Derived rather than carried, which is the opposite of [`Meter`] and
927 /// [`Figure`], and the difference is worth stating: a proportion's meaning
928 /// depends on what is being counted and only the app knows it, while
929 /// "nothing here yet" and "this broke" mean the same thing in every app that
930 /// will ever have them.
931 #[must_use]
932 pub const fn tone(self) -> Tone {
933 match self {
934 Self::Failed => Tone::Danger,
935 _ => Tone::Neutral,
936 }
937 }
938 }
939
940 /// How much of a set is done.
941 ///
942 /// Added 0.10.0. Nine sites across the two webview apps drew a bar and nothing
943 /// here named one, so every described screen concatenated the two numbers into
944 /// its heading text instead: "Subtasks 3/7", "Time Tracking 45m tracked / 30m
945 /// est, over". Every fact survives that and the reading does not, which is the
946 /// same loss `RowPart::Tokens` closed when a toned status badge became prose.
947 ///
948 /// # Why a pair and not a percentage
949 ///
950 /// Both numbers, not the percentage the apps compute from them. The percentage
951 /// was the obvious shape and it had already been tried: goingson's
952 /// `Task::time_progress` divides, rounds, and then clamps to 100, which throws
953 /// away the one case the bar exists to show — 45 minutes tracked against a
954 /// 30-minute estimate. It carries a separate `is_over_estimate` boolean beside
955 /// it to recover the fact the clamp dropped. A pair keeps the over-run without a
956 /// companion flag, and [`percent`](Meter::percent) is still one call away for a
957 /// renderer that wants it.
958 ///
959 /// The pair is also what the apps already have at every site. All seven
960 /// determinate bars write the ratio into the accessible layer and never the
961 /// percentage: `title="3/7 subtasks"`, `aria-label="3 of 7 subtasks completed"`,
962 /// a milestone's own `3/7` span. Given 43 nothing can recover "3 of 7", so a
963 /// percentage member would have made [`label`](Meter::label) mandatory at every
964 /// call site, which is the concatenated text this member removes, moved one
965 /// layer down.
966 ///
967 /// # What this is not
968 ///
969 /// The progress of an *operation*. Two of the nine sites are that — goingson's
970 /// focus timer, Balanced Breakfast's feed fetch — and they get nothing here, on
971 /// purpose. Both are imperative controllers over a live handle, driven by a tick
972 /// or an event stream, and a description is built once and dropped. Holding one
973 /// would mean growing a way to update a description between renders, which is a
974 /// different feature. [`Readiness::Pending`] and a [`Notice::Toast`] carry the
975 /// honest part.
976 ///
977 /// The two cases are distinguishable in the markup rather than by taste: every
978 /// determinate bar in both apps carries a tone, and neither operation bar
979 /// carries one. Two codebases drew that line the same way without coordinating.
980 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
981 pub struct Meter<'a> {
982 /// How much is done. May exceed [`total`](Self::total), and that is the
983 /// case worth drawing.
984 pub done: u32,
985 /// How much there is to do. Zero means there is no set, not that the set is
986 /// complete.
987 pub total: u32,
988 /// What the proportion means right now.
989 ///
990 /// Carried rather than derived, because no renderer can work it out. The
991 /// same 90% is [`Tone::Success`] on a subtask rollup and [`Tone::Danger`] on
992 /// a time estimate, and goingson picks between them from `is_over_estimate`,
993 /// a fact about the data and not about the number.
994 pub tone: Tone,
995 /// What is being counted, if the bar says so: "subtasks", "tasks".
996 ///
997 /// The noun, not the ratio. A renderer builds "3 of 7 subtasks" from this
998 /// and the two numbers; handing it the assembled string would put the
999 /// sentence order in the description, where a terminal at one line and a
1000 /// tooltip want different ones.
1001 pub label: Option<&'a str>,
1002 }
1003
1004 impl<'a> Meter<'a> {
1005 /// A proportion with no tone and no label.
1006 #[must_use]
1007 pub const fn new(done: u32, total: u32) -> Self {
1008 Self {
1009 done,
1010 total,
1011 tone: Tone::Neutral,
1012 label: None,
1013 }
1014 }
1015
1016 /// What the proportion means.
1017 #[must_use]
1018 pub const fn tone(mut self, tone: Tone) -> Self {
1019 self.tone = tone;
1020 self
1021 }
1022
1023 /// What is being counted.
1024 #[must_use]
1025 pub const fn label(mut self, label: &'a str) -> Self {
1026 self.label = Some(label);
1027 self
1028 }
1029
1030 /// How full the bar is, 0 to 100, clamped.
1031 ///
1032 /// For drawing, which is the only thing a clamped number is good for. Ask
1033 /// [`overflowing`](Self::overflowing) before reporting it as a fact, or this
1034 /// is `time_progress`'s bug again with the clamp moved.
1035 ///
1036 /// An empty set reads as 0. Nothing is done, because there is nothing to do
1037 /// and no bar to fill; the apps guard on the count before drawing at all.
1038 #[must_use]
1039 pub const fn percent(&self) -> u8 {
1040 if self.total == 0 {
1041 return 0;
1042 }
1043 let scaled = (self.done as u64 * 100) / self.total as u64;
1044 if scaled > 100 { 100 } else { scaled as u8 }
1045 }
1046
1047 /// Whether more is done than there was to do.
1048 ///
1049 /// The fact [`percent`](Self::percent) destroys, kept reachable so a
1050 /// renderer can mark the over-run rather than drawing a full bar and
1051 /// implying it landed exactly.
1052 #[must_use]
1053 pub const fn overflowing(&self) -> bool {
1054 self.done > self.total
1055 }
1056
1057 /// Whether there is a set at all.
1058 ///
1059 /// A meter over nothing is sayable on purpose, for the same reason a field
1060 /// with no options is: it is what an app with an unloaded count actually
1061 /// has, and a renderer that shows an empty bar says so on screen rather than
1062 /// dividing by zero.
1063 #[must_use]
1064 pub const fn is_empty(&self) -> bool {
1065 self.total == 0
1066 }
1067 }
1068
1069 /// One figure with a caption: a number and what it counts.
1070 ///
1071 /// The dashboard shape. A large value over a small caption, several of them in a
1072 /// strip: a current streak, a completion rate, a total. Added 0.11.0,
1073 /// `93c6a174`, after goingson turned out to have five of them across five
1074 /// screens with five class vocabularies for the one shape — `task-overview-stat`,
1075 /// `stat-box`, `month-stat-item`, `contact-summary-stat`, `sync-stat`. Four put
1076 /// the value above the caption and one inverts it, which is drift inside the
1077 /// shape rather than a second shape.
1078 ///
1079 /// # Why the value is text
1080 ///
1081 /// "17", "84%", "12/30", "3d". A figure is whatever the app computed, already
1082 /// formatted, and the formatting is the app's because only it knows whether the
1083 /// number is a percentage, a duration or a ratio. This carries none of the
1084 /// arithmetic [`Meter`] carries, and that is the difference between them: a
1085 /// meter is a proportion a renderer draws, and a figure is a fact a renderer
1086 /// sets in type.
1087 ///
1088 /// # Tone is carried, for [`Meter`]'s reason
1089 ///
1090 /// Three of the five sites tone the figure by their own means — `red`/`blue` on
1091 /// the weekly review, a `${type}` class on the monthly one, `sync-stat-warn` on
1092 /// sync. So tone is carried at every site that needs it and derived at none, and
1093 /// no renderer can work out that a streak of zero is worth colouring.
1094 ///
1095 /// # What is not here
1096 ///
1097 /// Whether the figure answers a click. One of the five is a control — sync's
1098 /// "Not Applied: 3" opens the list — and an action is not something this crate
1099 /// can name: nothing here knows what a route is. That belongs beside the figure
1100 /// in whatever layer holds the actions, the same way a row's activation sits
1101 /// beside its parts rather than inside them.
1102 ///
1103 /// The arrangement is not here either. Several figures in a strip is a set, and
1104 /// a renderer given them one at a time cannot tell it is looking at one; the
1105 /// layer that holds the tree is where the set gets said.
1106 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1107 pub struct Figure<'a> {
1108 /// The number, formatted the way the app means it to read.
1109 pub value: &'a str,
1110 /// What it counts. The caption under the value.
1111 pub caption: &'a str,
1112 /// How the value has moved, if the app is tracking that.
1113 ///
1114 /// Added 0.13.0. Text, for [`value`](Self::value)'s reason: only the app
1115 /// knows whether a move reads as `+12.5%`, `+3` or `2x`, and a renderer
1116 /// handed a number would have to guess.
1117 ///
1118 /// This is what [`tone`](Self::tone) was for and had no consumer of. The MNW
1119 /// server has four screens whose stat card is a label, a value and a delta,
1120 /// and the delta is the toned part: the figure itself is an ordinary fact
1121 /// and it is the movement that reads as good or bad. Without this the delta
1122 /// has to be folded into the caption, which loses the tone and reads as a
1123 /// longer caption rather than as a second, smaller line.
1124 pub change: Option<&'a str>,
1125 /// What the figure means right now. [`Tone::Neutral`] is an ordinary fact.
1126 ///
1127 /// Applies to [`change`](Self::change) where there is one, since that is the
1128 /// part that carries the judgement, and to the value where there is not.
1129 pub tone: Tone,
1130 }
1131
1132 impl<'a> Figure<'a> {
1133 /// A figure that is an ordinary fact.
1134 #[must_use]
1135 pub const fn new(value: &'a str, caption: &'a str) -> Self {
1136 Self {
1137 value,
1138 caption,
1139 change: None,
1140 tone: Tone::Neutral,
1141 }
1142 }
1143
1144 /// How the value has moved.
1145 #[must_use]
1146 pub const fn change(mut self, change: &'a str) -> Self {
1147 self.change = Some(change);
1148 self
1149 }
1150
1151 /// What the figure means.
1152 #[must_use]
1153 pub const fn tone(mut self, tone: Tone) -> Self {
1154 self.tone = tone;
1155 self
1156 }
1157 }
1158
1159 /// A named part of a screen.
1160 ///
1161 /// The thing `makeover-geometry` deliberately does not name: it names the space
1162 /// *between* things by relationship, and nothing named the things. Six named
1163 /// members, taken from what the two webview apps actually use, plus
1164 /// [`Region::Bespoke`] for the parts no description should reach. Both apps'
1165 /// `layout.css` currently names exactly two things, `.raised` and `.well`, so
1166 /// this layer is absent rather than divergent, which makes it the cheapest of
1167 /// the schemas to add and the easiest to over-build.
1168 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1169 pub enum Region<'a> {
1170 /// A full-width strip with a title slot and an actions cluster, either of
1171 /// which may be empty. goingson's `.page-header`, Balanced Breakfast's
1172 /// `.header` and `.detail-header` are all this, differing only in which
1173 /// slots they fill.
1174 Band,
1175 /// A persistent column beside the content, holding navigation.
1176 Sidebar,
1177 /// A region of content with its own scroll.
1178 Pane,
1179 /// Two panes side by side, where the left chooses what the right shows.
1180 Split,
1181 /// A set of panes, one visible at a time, with a [`Selector::Tabs`] above.
1182 TabGroup,
1183 /// Content over a scrim, taking input until dismissed.
1184 Modal,
1185 /// A region this crate names the *place* of and nothing else. The app owns
1186 /// what goes in it.
1187 ///
1188 /// The escape hatch, and the thing that keeps the description honest about
1189 /// its own limits. A day-plan timeline, a kanban board, a calendar and the
1190 /// paint interaction over the timeline are not describable here and are not
1191 /// going to become describable: a description expressive enough to produce
1192 /// a timeline is a widget library wearing a description's name.
1193 ///
1194 /// But a screen containing one still has to be a screen. Without this
1195 /// member the description covers only the boring screens, and the four that
1196 /// make goingson worth using would need a second, undescribed path beside
1197 /// the router. Two paths is how the vocabulary starts drifting from the app
1198 /// again, which is the exact failure this crate exists to end.
1199 ///
1200 /// So the description says "a thing called `day-plan` goes here" and stops.
1201 /// The name is opaque: this crate never interprets it, and no renderer is
1202 /// expected to know what it means beyond handing the space over.
1203 Bespoke {
1204 /// What the app calls it. Never interpreted here.
1205 name: &'a str,
1206 },
1207 }
1208
1209 impl Region<'_> {
1210 /// How the region sits on what is behind it.
1211 #[must_use]
1212 pub const fn depth(self) -> Depth {
1213 match self {
1214 Self::Band | Self::Sidebar | Self::Split | Self::TabGroup => Depth::Flat,
1215 // A pane is looked into, the same as a table body or a tag tree.
1216 Self::Pane => Depth::Well,
1217 Self::Modal => Depth::Raised,
1218 // Flat because it inherits: a bespoke region takes the depth of
1219 // whatever frames it. An app that wants its timeline in a well puts
1220 // it in a `Pane`, which composes rather than adding a knob here.
1221 Self::Bespoke { .. } => Depth::Flat,
1222 }
1223 }
1224
1225 /// Whether this crate can say anything about the region's contents.
1226 ///
1227 /// A renderer walks the description and hands every region it understands
1228 /// to the right drawing code. This is how it tells the two apart, and the
1229 /// reason it is a method rather than a `matches!` at each renderer: there
1230 /// is exactly one opaque member and there should stay exactly one.
1231 #[must_use]
1232 pub const fn described(self) -> bool {
1233 !matches!(self, Self::Bespoke { .. })
1234 }
1235 }
1236
1237 /// How a screen is laid out.
1238 ///
1239 /// Two, and the second is not a variant of the first. goingson is list-detail,
1240 /// Balanced Breakfast is sidebar plus content, and neither app has a third.
1241 /// The tab group is a modifier rather than a member, because goingson uses it
1242 /// *inside* the same content region rather than instead of one.
1243 ///
1244 /// This exists at all because the router has to be able to express a screen
1245 /// rather than only a control. Discovering the arrangement layer missing after
1246 /// the renderers exist is a redesign; naming two now is a morning.
1247 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1248 pub enum Arrangement {
1249 /// A list that chooses what the detail beside it shows.
1250 ListDetail {
1251 /// Whether the detail side is a [`Region::TabGroup`].
1252 tabbed: bool,
1253 },
1254 /// Navigation down the side, content filling the rest.
1255 SidebarContent,
1256 }
1257
1258 /// What kind of value a form field takes.
1259 ///
1260 /// The union of the two vocabularies that diverged, which is what triggered
1261 /// this crate. They have since converged on their own: both apps now have a
1262 /// `renderFormField` emitting the same anatomy, and what is left differing is
1263 /// the kind set, the error shape, and whether the return is a string or a node.
1264 ///
1265 /// Validation is deliberately absent. Neither app has a shared story (goingson
1266 /// validates after collecting the form data, with per-field transform hooks;
1267 /// Balanced Breakfast has `required` and nothing else), and a schema that
1268 /// describes fields but not constraints acquires a constraint layer per app,
1269 /// which is exactly how the current divergence started. Naming it absent is a
1270 /// decision; leaving it unmentioned would not be.
1271 /// `#[non_exhaustive]` for the reason [`Fill`] is: renderers match on this and
1272 /// the set keeps growing, so growth must not be a lockstep event. Email, Url
1273 /// and Tel arriving in 0.5.0 is the second growth in two releases.
1274 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1275 #[non_exhaustive]
1276 pub enum FieldKind {
1277 /// A single line of text.
1278 Text,
1279 /// A single line of text that must never be echoed, logged or round-tripped
1280 /// through anything that might persist it.
1281 Secret,
1282 /// A number.
1283 Number,
1284 /// An email address.
1285 ///
1286 /// Distinct from [`Text`](Self::Text) because the distinction is not
1287 /// decoration: a webview renderer emits `type="email"`, which on a touch
1288 /// device changes the keyboard that appears and turns on the platform's own
1289 /// validation. goingson ships to iOS, so collapsing this into text costs a
1290 /// keyboard with no `@` on it.
1291 ///
1292 /// Added 0.5.0, from goingson's contact form.
1293 Email,
1294 /// A URL. Same reasoning as [`Email`](Self::Email).
1295 ///
1296 /// Added 0.5.0, from goingson's contact-social and contact-feed forms.
1297 Url,
1298 /// A telephone number. Same reasoning as [`Email`](Self::Email), and the
1299 /// clearest case of it: the keyboard is a numeric pad rather than letters.
1300 ///
1301 /// Added 0.5.0, from goingson's contact-phone form.
1302 Tel,
1303 /// A calendar day, with no time of day in it.
1304 ///
1305 /// [`Email`](Self::Email)'s argument, and it carries further: a webview
1306 /// emits `type="date"`, which is a native picker, the platform's own
1307 /// validation, and on a touch device the date keyboard. Described as
1308 /// [`Text`](Self::Text) with a hint reading "YYYY-MM-DD", all three are
1309 /// lost and the hint is doing the platform's job in prose.
1310 ///
1311 /// The membership test passes on every host without stretching: a webview
1312 /// and a Tauri app emit the input, egui has a date picker, a terminal
1313 /// prompts for a day and can validate it, a CLI takes an argument.
1314 ///
1315 /// # The value is ISO 8601, `YYYY-MM-DD`
1316 ///
1317 /// Named here rather than left to each host, because a host that picks
1318 /// differently sends a server something it parses differently, and the
1319 /// failure is silent and per-host. It is `<input type="date">`'s own wire
1320 /// format, so the webview renderer owes nothing to honour it and the other
1321 /// hosts have one spelling to meet. [`DATE_FORMAT`] is the constant, and a
1322 /// test asserts this doc and that constant agree.
1323 ///
1324 /// Added 0.15.0, from the MNW server's git access-token expiry
1325 /// (`user_ssh_keys_tab.html`) and six further sites across the server and
1326 /// goingson.
1327 Date,
1328 /// A calendar day and a time of day together.
1329 ///
1330 /// Apart from [`Date`](Self::Date) because the question is different rather
1331 /// than more precise: "which day does this expire" and "at what moment does
1332 /// this publish" are asked by different screens and answered by different
1333 /// controls. A webview emits `type="datetime-local"` for one and
1334 /// `type="date"` for the other, and a host that collapsed them would ask
1335 /// half the tree for a precision it does not want.
1336 ///
1337 /// Both arrived together on measurement rather than on symmetry: 13 sites
1338 /// of each across the MNW server and goingson, and **zero** of `time`,
1339 /// `month` or `week`, which is why those are not here. A member added for a
1340 /// case nobody has is a member designed against nothing, which is
1341 /// [`File`](Self::File)'s reasoning about `accept` applied to a whole
1342 /// member.
1343 ///
1344 /// # The value is `YYYY-MM-DDTHH:MM`, local, with no zone
1345 ///
1346 /// `<input type="datetime-local">`'s own format, and the "local" is the
1347 /// load-bearing half: the value carries no offset and no `Z`, so the moment
1348 /// it names is only fixed once something supplies a zone. That is the app's
1349 /// business and not the description's. Seconds are absent, which is the
1350 /// browser's own default and is left as the rule rather than restated as a
1351 /// constraint. [`DATETIME_FORMAT`] is the constant.
1352 ///
1353 /// [`Field::min`] and [`Field::max`] already take "the host's own spelling
1354 /// of a bound", so a floor of *not in the past* needs nothing new here: it
1355 /// is a string in this same format.
1356 ///
1357 /// Added 0.15.0, from goingson's snooze picker and day planner and the MNW
1358 /// server's publish-at fields.
1359 DateTime,
1360 /// Several lines of text.
1361 Textarea,
1362 /// One of a fixed set, offered behind a control that shows one at a time.
1363 Select,
1364 /// One of a fixed set, with every option on screen at once.
1365 ///
1366 /// Not a presentation of [`Select`](Self::Select), which is the reading to
1367 /// resist: what differs is a property of the *question*. A choice that is
1368 /// consequential or irreversible has to be readable without opening
1369 /// anything, because a closed control shows one option and hides the rest,
1370 /// and the one it shows is whichever was current before the user had read
1371 /// the alternatives. audiofiles asks whether a library copies samples into
1372 /// its store or references them where they lie — which cannot be changed
1373 /// afterwards — and had already promoted that out of a checkbox by hand,
1374 /// with a comment giving this reason, before the description could say it.
1375 ///
1376 /// It was described here at 0.8.1 as "the one HTML input type this enum was
1377 /// missing", which was not true then and is not true now: `file` arrived at
1378 /// 0.11.0 and `date` and `datetime-local` at 0.15.0. Everything here is
1379 /// still an `<input type=...>`, a `<select>` or a `<textarea>`, and the way
1380 /// this enum grows is by a site being measured rather than by a list being
1381 /// completed, so "the last one" is not a claim it should make again.
1382 ///
1383 /// Added 0.8.1, from audiofiles' Add Library form.
1384 Radio,
1385 /// On or off.
1386 Checkbox,
1387 /// A file the user picks from wherever the host keeps files.
1388 ///
1389 /// Added 0.11.0, `844b5ae0`, from goingson's project-dashboard attachments
1390 /// column. It was filed as a router finding — a control whose destination is
1391 /// a host capability rather than an address — and splitting it is what made
1392 /// it two answers instead of one member satisfying neither. *Opening* a file
1393 /// is a one-way handoff and needs no new API. *Picking* one returns a value
1394 /// into a write, which is a form concern, which is this.
1395 ///
1396 /// The membership test passes on every host and not by a stretch: a Tauri
1397 /// app opens a native picker, a server renders `<input type="file">`, a
1398 /// terminal prompts for a path, a CLI takes an argument. That is closer to
1399 /// [`Email`](Self::Email), which exists because it changes the keyboard,
1400 /// than to anything bespoke.
1401 ///
1402 /// It carries no accepted-types list and no multiple flag, and that is
1403 /// measured rather than deferred: `accept` appears at zero sites in either
1404 /// app. A member added for a case nobody has is a member designed against
1405 /// nothing.
1406 File,
1407 /// Carried through the form and never shown.
1408 Hidden,
1409 }
1410
1411 /// The wire format a [`FieldKind::Date`] value takes: ISO 8601, `YYYY-MM-DD`.
1412 ///
1413 /// A constant rather than a sentence in a doc comment, because the reason to
1414 /// name the format at all is that a host picking its own would fail silently
1415 /// against a server parsing another. A host that cannot emit the native control
1416 /// still has one spelling to meet, and can say which one it meant.
1417 pub const DATE_FORMAT: &str = "%Y-%m-%d";
1418
1419 /// The wire format a [`FieldKind::DateTime`] value takes: `YYYY-MM-DDTHH:MM`,
1420 /// local, carrying no zone and no seconds.
1421 ///
1422 /// [`DATE_FORMAT`]'s sibling and there for its reason. The absent zone is a
1423 /// property of the value rather than an omission: the moment is not fixed until
1424 /// something outside the description supplies one.
1425 pub const DATETIME_FORMAT: &str = "%Y-%m-%dT%H:%M";
1426
1427 impl FieldKind {
1428 /// Whether the value the kind takes is a moment rather than a string.
1429 ///
1430 /// Named once here for the reason [`offers_options`](Self::offers_options)
1431 /// is: two kinds answer yes, and a host that has to parse or format a value
1432 /// needs to ask without spelling the pair out at each renderer. A third
1433 /// temporal kind should land here and nowhere else.
1434 ///
1435 /// The format each one takes is [`DATE_FORMAT`] and [`DATETIME_FORMAT`].
1436 #[must_use]
1437 pub const fn temporal(self) -> bool {
1438 matches!(self, Self::Date | Self::DateTime)
1439 }
1440
1441 /// Whether the field is drawn at all.
1442 #[must_use]
1443 pub const fn visible(self) -> bool {
1444 !matches!(self, Self::Hidden)
1445 }
1446
1447 /// Whether the value must be kept out of logs and diagnostics.
1448 #[must_use]
1449 pub const fn confidential(self) -> bool {
1450 matches!(self, Self::Secret)
1451 }
1452
1453 /// Where the field's own label sits.
1454 ///
1455 /// A checkbox labels itself on the right of the box; everything else takes
1456 /// a label above. Both webview apps already do this and both special-case
1457 /// it inline, which is the tell that it belongs in the description.
1458 ///
1459 /// A [`Radio`](Self::Radio) is not one of them, and the near-miss is worth
1460 /// naming: its *options* each label themselves, but the field still asks a
1461 /// question above them, so the group takes a label like everything else.
1462 #[must_use]
1463 pub const fn labels_itself(self) -> bool {
1464 matches!(self, Self::Checkbox)
1465 }
1466
1467 /// Whether the kind reads [`Field::options`].
1468 ///
1469 /// Two kinds do, so the pair is named once here rather than spelled out at
1470 /// each renderer and again in [`Field::options`]' own doc, where "every
1471 /// kind but `Select`" was true for exactly one release. A third
1472 /// option-taking kind should land here and nowhere else.
1473 #[must_use]
1474 pub const fn offers_options(self) -> bool {
1475 matches!(self, Self::Select | Self::Radio)
1476 }
1477 }
1478
1479 /// One option offered by a field [`FieldKind::offers_options`] accepts.
1480 ///
1481 /// Two strings, because the submitted value and the read label are different
1482 /// facts and every renderer that has tried to collapse them has had to
1483 /// un-collapse them later. `makeover-webview` invented this shape writing its
1484 /// form emitter and it is taken here unchanged; moving it down rather than
1485 /// re-deriving it is the point, since the second and third renderers were each
1486 /// going to arrive at a near-miss of it.
1487 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1488 pub struct Choice<'a> {
1489 /// What is submitted.
1490 pub value: &'a str,
1491 /// What is read.
1492 pub label: &'a str,
1493 }
1494
1495 impl<'a> Choice<'a> {
1496 /// An option whose submitted value is also its label.
1497 #[must_use]
1498 pub const fn plain(value: &'a str) -> Self {
1499 Self {
1500 value,
1501 label: value,
1502 }
1503 }
1504 }
1505
1506 /// One field of a form.
1507 ///
1508 /// Borrowed rather than owned: a description is built, read once by a renderer,
1509 /// and dropped. Nothing here outlives the screen it describes.
1510 ///
1511 /// # What it carries, and what it does not
1512 ///
1513 /// Stated here so the next renderer does not re-ask, which is what the first
1514 /// two both did. It carries everything a renderer needs to *draw* the field:
1515 /// its kind, what it is called, what it is asked for, its standing help, what
1516 /// is wrong with it now, whether it is compulsory, whether it hides behind a
1517 /// disclosure, its ghost text, and the options it offers.
1518 ///
1519 /// It does not carry the **current value**, and it is not going to. That is the
1520 /// one thing here that is genuinely renderer state: a webview reads it back out
1521 /// of the DOM, an immediate-mode renderer holds a `&mut` to the app's own field
1522 /// and writes through it, and a terminal keeps an edit buffer. A description
1523 /// that carried the value would have to carry a way to write it back, at which
1524 /// point it is a form model and no longer a description.
1525 ///
1526 /// **Constraints** are here and enforcement is not, which is one line rather
1527 /// than two. [`required`], [`max_length`], [`min`] and [`max`] are facts about
1528 /// the *question*, so a renderer can emit its host's idiom for each — an HTML
1529 /// attribute, a marked label, a clamped spinner — and the platform helps the
1530 /// user before anything is submitted. Deciding that a value is wrong stays with
1531 /// whoever validated, and [`error`] is that decision arriving back.
1532 ///
1533 /// The set stops before `pattern`, and stops there on both tests at once. A
1534 /// regex has an honest answer in a webview and none anywhere else: egui would
1535 /// have to run it per keystroke and decide what a half-typed value means, which
1536 /// is enforcement wearing description's clothes. And it is one site in goingson
1537 /// and none in Balanced Breakfast, against 8 and 1 for `maxlength`. Measured
1538 /// 2026-08-09, `2cbad3e2`.
1539 ///
1540 /// [`error`]: Field::error
1541 /// [`required`]: Field::required
1542 /// [`max_length`]: Field::max_length
1543 /// [`min`]: Field::min
1544 /// [`max`]: Field::max
1545 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1546 pub struct Field<'a> {
1547 /// What kind of value it takes.
1548 pub kind: FieldKind,
1549 /// The name the value is submitted under.
1550 pub name: &'a str,
1551 /// What the user is asked for.
1552 pub label: &'a str,
1553 /// Standing help, shown whether or not anything is wrong.
1554 pub hint: Option<&'a str>,
1555 /// What is currently wrong with the value.
1556 pub error: Option<&'a str>,
1557 /// Ghost text shown while the field is empty.
1558 ///
1559 /// User-facing text, and it sits with `label` and `hint` rather than with
1560 /// the value because it is a property of the *question* and not of the
1561 /// answer. It lived renderer-side in `makeover-webview` until 0.8.0 for one
1562 /// reason and it was not a reading on where it belonged: adding a field to
1563 /// a published struct is a breaking change.
1564 ///
1565 /// Not a substitute for a label. A field labelled only by its placeholder
1566 /// loses its label the moment anything is typed, and no renderer here can
1567 /// make that not happen, so the description keeps both.
1568 pub placeholder: Option<&'a str>,
1569 /// The options offered, in the order they are offered.
1570 ///
1571 /// Empty for every kind [`FieldKind::offers_options`] rejects. A field
1572 /// described with no options is sayable on purpose: it is what an app with
1573 /// an unfinished-loading option list actually has, and a renderer showing
1574 /// an empty control says so on screen rather than in a log.
1575 ///
1576 /// Which option is *current* is not here. That is the value, and the value
1577 /// is renderer state.
1578 pub options: &'a [Choice<'a>],
1579 /// Whether the form refuses to submit without it.
1580 pub required: bool,
1581 /// The longest the value may be, in characters.
1582 ///
1583 /// Added 0.11.0 with [`min`](Self::min) and [`max`](Self::max), joining
1584 /// [`required`](Self::required), which had been the only constraint here
1585 /// since before the crate wrote down that it carried none.
1586 pub max_length: Option<u32>,
1587 /// The lowest value accepted, as the host would write it.
1588 ///
1589 /// Text rather than a number, because the bound is only a number for some
1590 /// of the kinds that take one. goingson's own sites are `min="1"` on a
1591 /// duration and `min="2026-08-09T14:30"` on a datetime, and a numeric member
1592 /// could say the first and not the second. The [`kind`](Self::kind) already
1593 /// says how to read it, the same way it does for the value.
1594 pub min: Option<&'a str>,
1595 /// The highest value accepted, as the host would write it. See
1596 /// [`min`](Self::min).
1597 pub max: Option<&'a str>,
1598 /// Whether the field lives behind a "more options" disclosure.
1599 pub extended: bool,
1600 }
1601
1602 impl<'a> Field<'a> {
1603 /// A plain required-nothing field of the given kind.
1604 #[must_use]
1605 pub const fn new(kind: FieldKind, name: &'a str, label: &'a str) -> Self {
1606 Self {
1607 kind,
1608 name,
1609 label,
1610 hint: None,
1611 error: None,
1612 placeholder: None,
1613 options: &[],
1614 required: false,
1615 max_length: None,
1616 min: None,
1617 max: None,
1618 extended: false,
1619 }
1620 }
1621
1622 /// A select offering the given options.
1623 ///
1624 /// One of the two kinds under-described by [`Field::new`], so it gets a
1625 /// constructor rather than leaving every call site to remember that a
1626 /// select with an empty `options` renders as an empty select.
1627 #[must_use]
1628 pub const fn select(name: &'a str, label: &'a str, options: &'a [Choice<'a>]) -> Self {
1629 Self::offering(FieldKind::Select, name, label, options)
1630 }
1631
1632 /// A radio group offering the given options.
1633 ///
1634 /// The other. Same hazard as [`select`](Self::select) and a worse one: a
1635 /// radio group with no options draws nothing at all, so a call site that
1636 /// forgot them has an empty rectangle rather than a visibly empty control.
1637 #[must_use]
1638 pub const fn radio(name: &'a str, label: &'a str, options: &'a [Choice<'a>]) -> Self {
1639 Self::offering(FieldKind::Radio, name, label, options)
1640 }
1641
1642 /// The shared body of the two constructors that take options.
1643 ///
1644 /// Private, and keyed on the kind rather than exposed, because the two
1645 /// public names are the point: a call site says which question it is
1646 /// asking, not which flag it is setting.
1647 const fn offering(
1648 kind: FieldKind,
1649 name: &'a str,
1650 label: &'a str,
1651 options: &'a [Choice<'a>],
1652 ) -> Self {
1653 Self {
1654 options,
1655 ..Self::new(kind, name, label)
1656 }
1657 }
1658
1659 /// Whether the field is currently reporting a problem.
1660 ///
1661 /// Read this rather than testing `error.is_some()` at each renderer: the
1662 /// error state has to mark the field's whole group and not only the
1663 /// message, because a renderer with no descendant selectors (egui, a
1664 /// terminal) cannot find the group from the message. goingson already marks
1665 /// the group and Balanced Breakfast does not, so goingson's shape is the
1666 /// one taken here.
1667 #[must_use]
1668 pub const fn invalid(&self) -> bool {
1669 self.error.is_some()
1670 }
1671 }
1672
1673 /// How much room a column asks for.
1674 ///
1675 /// An intent, so the actual floor stays with `makeover-geometry`. goingson's
1676 /// task table spells these as `minmax(200px, 1fr)`, `140px` and content-sized;
1677 /// only the first three words of that survive deferral.
1678 /// `#[non_exhaustive]`, for the reason [`Fill`] and [`FieldKind`] are: a
1679 /// renderer matches on this and a vocabulary that grows must not break every
1680 /// renderer when it does.
1681 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1682 #[non_exhaustive]
1683 pub enum Width {
1684 /// Takes what it needs and no more.
1685 Content,
1686 /// A fixed share, the same at every width.
1687 Fixed,
1688 /// Absorbs whatever is left over.
1689 Fill,
1690 }
1691
1692 /// What a column is worth when there is not room for all of them.
1693 ///
1694 /// Ordered: [`Priority::Optional`] drops first, [`Priority::Essential`] never
1695 /// drops. This replaces addressing columns by position, which is what both
1696 /// webview apps do today and is a live bug rather than only verbosity. goingson
1697 /// hides mobile columns with `nth-child(n+5)` against a seven-column table, so
1698 /// inserting a column silently hides the wrong one.
1699 /// `#[non_exhaustive]`, same reasoning as [`Width`]. Note the ordering is the
1700 /// whole point of the type, so a new tier has to be declared in its place in
1701 /// the sequence rather than appended.
1702 #[derive(Debug, Clone, Copy, PartialEq, Eq, PartialOrd, Ord, Hash)]
1703 #[non_exhaustive]
1704 pub enum Priority {
1705 /// Dropped first.
1706 Optional,
1707 /// Dropped once the optional columns are gone.
1708 Secondary,
1709 /// Never dropped. Without it the row does not identify itself.
1710 Essential,
1711 }
1712
1713 /// One column of a table.
1714 ///
1715 /// Described once. The grid track, the cell order and the drop behaviour are
1716 /// all derived from this, rather than being three hand-written encodings that
1717 /// must agree and are never checked against each other.
1718 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1719 pub struct Column<'a> {
1720 /// The heading, and the name the cell is addressed by.
1721 pub name: &'a str,
1722 /// How much room it asks for.
1723 pub width: Width,
1724 /// What it is worth when room runs out.
1725 pub priority: Priority,
1726 /// Whether the user can reorder the table by this column.
1727 ///
1728 /// `ce620871`. What reordering *calls* is not here — that is an address, and
1729 /// this crate names none — so a host pairs this with the route the way it
1730 /// pairs a row's parts with the row's activation. This says the affordance
1731 /// exists, which is what a renderer needs to draw a header a user can press
1732 /// rather than a heading they cannot.
1733 pub sortable: bool,
1734 /// Which way the table is ordered by this column, if it is.
1735 ///
1736 /// `None` on every column but the one in force. A renderer draws the caret
1737 /// from this and a webview sets `aria-sort`, which is why it is per column
1738 /// rather than a single fact on the table: the host idiom is a property of
1739 /// the header cell.
1740 ///
1741 /// Independent of [`sortable`](Self::sortable) rather than implied by it,
1742 /// because both combinations mean something. A column sorted and not
1743 /// sortable is a list ordered by a key the user cannot change, which is a
1744 /// real thing to describe and a caret worth drawing.
1745 pub sorted: Option<Sort>,
1746 }
1747
1748 /// Which way a column is ordered.
1749 ///
1750 /// Two, because there is no third. "Unsorted" is [`Column::sorted`] being
1751 /// `None`, and folding it in here would be the same absence said twice.
1752 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1753 pub enum Sort {
1754 /// Smallest, earliest or first alphabetically at the top.
1755 Ascending,
1756 /// The other way.
1757 Descending,
1758 }
1759
1760 impl Sort {
1761 /// The other direction, for a header that flips when pressed.
1762 #[must_use]
1763 pub const fn reversed(self) -> Self {
1764 match self {
1765 Self::Ascending => Self::Descending,
1766 Self::Descending => Self::Ascending,
1767 }
1768 }
1769
1770 /// What a webview writes into `aria-sort`.
1771 ///
1772 /// Named here rather than in the webview renderer because a terminal and an
1773 /// immediate-mode painter both want the same two words for a caret's label,
1774 /// and three renderers picking their own is the drift this crate ends.
1775 #[must_use]
1776 pub const fn as_str(self) -> &'static str {
1777 match self {
1778 Self::Ascending => "ascending",
1779 Self::Descending => "descending",
1780 }
1781 }
1782 }
1783
1784 impl<'a> Column<'a> {
1785 /// A column that absorbs slack and drops after the optional ones.
1786 #[must_use]
1787 pub const fn new(name: &'a str) -> Self {
1788 Self {
1789 name,
1790 width: Width::Fill,
1791 priority: Priority::Secondary,
1792 sortable: false,
1793 sorted: None,
1794 }
1795 }
1796
1797 /// Whether this column survives at the given cutoff.
1798 ///
1799 /// A renderer narrows by raising the cutoff, and never by counting
1800 /// positions.
1801 #[must_use]
1802 pub const fn kept_at(&self, cutoff: Priority) -> bool {
1803 (self.priority as u8) >= (cutoff as u8)
1804 }
1805 }
1806
1807 /// What a table cell holds.
1808 ///
1809 /// [`RowPart`] for tables, and it exists for the same reason: a part that
1810 /// carries a control is not text, and a renderer with one class for the whole
1811 /// cell paints it as though it were. `makeover-webview` emitted a single
1812 /// `.cell` until 0.25.0, so a button in a cell inherited the cell's content
1813 /// colour, which is the exact drift [`RowPart::intent`] prevents for rows and
1814 /// prevented for nothing here.
1815 ///
1816 /// Four members, and the count is what quasi's `Cell` was measured to carry:
1817 /// a value, tokens (33 cells across 22 server templates), actions (30 rows
1818 /// carrying a control, 5 beside a value) and a link (35 cells across 18
1819 /// templates). Nothing was added past what something holds.
1820 ///
1821 /// `#[non_exhaustive]` for [`RowPart`]'s reason: growth here must not be a
1822 /// lockstep event across three renderers.
1823 ///
1824 /// # No hover-reveal
1825 ///
1826 /// [`RowPart`] carried a `revealed_on_hover` until 0.13.0 retired it, and this
1827 /// enum never gets one. A cell's actions are shown at rest in every consumer
1828 /// measured, and a member nothing uses is one three renderers owe an answer
1829 /// for.
1830 #[derive(Debug, Clone, Copy, PartialEq, Eq, Hash)]
1831 #[non_exhaustive]
1832 pub enum CellPart {
1833 /// The cell's own text.
1834 Value,
1835 /// Small labelled things in the cell: a status badge, a chip.
1836 Tokens,
1837 /// Controls that act on what the row is about.
1838 Actions,
1839 /// The cell's value, where the value is itself a link.
1840 Link,
1841 }
1842
1843 impl CellPart {
1844 /// The content intent the part takes.
1845 ///
1846 /// One part is text and three are not, so three answer with the intent
1847 /// inheriting already gives. That is [`RowPart::intent`]'s shape with the
1848 /// text side narrower: a cell's secondary and muted readings are the
1849 /// column's business, not the cell's.
1850 #[must_use]
1851 pub const fn intent(self) -> &'static str {
1852 match self {
1853 Self::Value => "content",
1854 // A token carries its own tone, and a part-level intent underneath
1855 // it would fight the token sitting on it.
1856 Self::Tokens => "content",
1857 // Actions carry controls rather than text.
1858 Self::Actions => "content",
1859 // A link takes the action colour from the control it is, rather
1860 // than the cell's text colour from the cell it sits in.
1861 Self::Link => "content",
1862 }
1863 }
1864 }
1865
1866 #[cfg(test)]
1867 mod tests {
1868 use super::*;
1869
1870 #[test]
1871 fn the_four_readiness_states_are_one_axis_and_only_one_shows_content() {
1872 // Mutually exclusive is the test for one enum against several fields: a
1873 // region shows its content, or that it is coming, or that there is none,
1874 // or that it broke. Never two.
1875 assert!(Readiness::Ready.shows_content());
1876 for state in [Readiness::Pending, Readiness::Empty, Readiness::Failed] {
1877 assert!(!state.shows_content());
1878 }
1879 }
1880
1881 #[test]
1882 fn an_empty_region_is_not_a_broken_one() {
1883 // An empty list is the normal state of a new install. Drawing it in a
1884 // danger tone reports a fault where there is none, and this is the one
1885 // place the distinction is carried.
1886 assert_eq!(Readiness::Empty.tone(), Tone::Neutral);
1887 assert_eq!(Readiness::Failed.tone(), Tone::Danger);
1888 assert_eq!(Readiness::Pending.tone(), Tone::Neutral);
1889 }
1890
1891 #[test]
1892 fn a_column_can_be_sorted_without_being_sortable() {
1893 // Both combinations mean something, which is why the two fields are
1894 // independent rather than one implying the other. A list ordered by a
1895 // key the user cannot change is a real thing with a caret worth drawing.
1896 let fixed = Column {
1897 sorted: Some(Sort::Descending),
1898 ..Column::new("Created")
1899 };
1900
1901 assert!(!fixed.sortable);
1902 assert_eq!(fixed.sorted.map(Sort::as_str), Some("descending"));
1903
1904 let offered = Column {
1905 sortable: true,
1906 ..Column::new("Name")
1907 };
1908 assert_eq!(offered.sorted, None);
1909 }
1910
1911 #[test]
1912 fn a_direction_flips_and_says_what_it_is() {
1913 assert_eq!(Sort::Ascending.reversed(), Sort::Descending);
1914 assert_eq!(Sort::Descending.reversed().reversed(), Sort::Descending);
1915 assert_eq!(Sort::Ascending.as_str(), "ascending");
1916 }
1917
1918 #[test]
1919 fn a_figure_carries_its_tone_because_no_renderer_can_derive_it() {
1920 // Three of goingson's five sites tone the figure by their own means, so
1921 // tone is carried at every site that needs it and derived at none. The
1922 // same reasoning `Meter` reached, from a different direction.
1923 let streak = Figure::new("0", "Current Streak").tone(Tone::Warning);
1924 assert_eq!(streak.tone, Tone::Warning);
1925 assert_eq!(Figure::new("17", "Total").tone, Tone::Neutral);
1926 }
1927
1928 #[test]
1929 fn a_figures_change_is_the_toned_part_and_is_absent_by_default() {
1930 // 0.13.0. The MNW server's stat card is a label, a value and a delta,
1931 // across four screens, and the delta is what reads as good or bad. Tone
1932 // had no consumer before this: the figure itself is an ordinary fact.
1933 let views = Figure::new("1,204", "Views")
1934 .change("+12.5%")
1935 .tone(Tone::Success);
1936 assert_eq!(views.change, Some("+12.5%"));
1937 assert_eq!(views.tone, Tone::Success);
1938
1939 // A figure with nothing to compare against says so by having no change,
1940 // rather than by carrying an empty string a renderer has to test for.
1941 assert_eq!(Figure::new("3.1%", "Conversion").change, None);
1942 }
1943
1944 #[test]
1945 fn a_figures_value_is_text_because_only_the_app_knows_what_it_is() {
1946 // "84%", "12/30", "3d". A figure is whatever the app computed, already
1947 // formatted, and that is the line between this and `Meter`: a meter is
1948 // a proportion a renderer draws, a figure is a fact it sets in type.
1949 for value in ["84%", "12/30", "3d"] {
1950 assert_eq!(Figure::new(value, "Rate").value, value);
1951 }
1952 }
1953
1954 #[test]
1955 fn a_proportion_is_a_row_part_and_takes_no_intent_of_its_own() {
1956 // The meter carries the tone, so a part-level intent underneath would
1957 // fight it. Same answer `Tokens` needed, for the same reason.
1958 assert_eq!(RowPart::Proportion.intent(), RowPart::Tokens.intent());
1959 }
1960
1961 #[test]
1962 fn a_file_field_is_drawn_and_offers_no_options() {
1963 // It is a control the user operates, unlike `Hidden`, and it does not
1964 // pick from a list the description carries, unlike `Select`.
1965 assert!(FieldKind::File.visible());
1966 assert!(!FieldKind::File.offers_options());
1967 assert!(!FieldKind::File.confidential());
1968 }
1969
1970 #[test]
1971 fn a_constraint_is_a_fact_about_the_question_and_not_a_verdict() {
1972 // The whole model: the description carries the rule, the renderer emits
1973 // its host's idiom, and `error` is what arrives back when someone
1974 // validated. Nothing here decides a value is wrong.
1975 let field = Field {
1976 max_length: Some(100),
1977 min: Some("1"),
1978 max: Some("240"),
1979 required: true,
1980 ..Field::new(FieldKind::Number, "minutes", "Minutes")
1981 };
1982 assert!(!field.invalid());
1983
1984 // A bound is text because it is only a number for some of the kinds
1985 // that take one. goingson has both shapes live.
1986 let when = Field {
1987 min: Some("2026-08-09T14:30"),
1988 ..Field::new(FieldKind::Text, "starts", "Starts")
1989 };
1990 assert_eq!(when.min, Some("2026-08-09T14:30"));
1991 }
1992
1993 #[test]
1994 fn a_meter_keeps_the_over_run_the_percentage_throws_away() {
1995 // The whole reason this is a pair. goingson's `Task::time_progress`
1996 // clamps to 100 and then carries `is_over_estimate` beside it to say
1997 // what the clamp dropped; a meter says both from one fact.
1998 let over = Meter::new(45, 30);
1999 assert_eq!(over.percent(), 100);
2000 assert!(over.overflowing());
2001
2002 let exact = Meter::new(30, 30);
2003 assert_eq!(exact.percent(), over.percent());
2004 assert!(!exact.overflowing());
2005 }
2006
2007 #[test]
2008 fn an_empty_set_does_not_divide_by_zero() {
2009 // Sayable on purpose, so it has to be answerable. A meter over an
2010 // unloaded count is what an app actually has for a frame.
2011 let none = Meter::new(0, 0);
2012 assert_eq!(none.percent(), 0);
2013 assert!(none.is_empty());
2014 assert!(!none.overflowing());
2015 }
2016
2017 #[test]
2018 fn the_ratio_survives_where_a_percentage_would_not() {
2019 // Given 43 nothing can recover "3 of 7", which is why the numbers are
2020 // carried and the label names only the noun.
2021 let m = Meter::new(3, 7).label("subtasks");
2022 assert_eq!(m.percent(), 42);
2023 assert_eq!((m.done, m.total), (3, 7));
2024 assert_eq!(m.label, Some("subtasks"));
2025 }
2026
2027 #[test]
2028 fn tone_is_carried_because_no_renderer_can_derive_it() {
2029 // The same fullness means opposite things on two of goingson's bars,
2030 // and only the app knows which.
2031 let subtasks = Meter::new(9, 10).tone(Tone::Success);
2032 let estimate = Meter::new(9, 10).tone(Tone::Danger);
2033 assert_eq!(subtasks.percent(), estimate.percent());
2034 assert_ne!(subtasks.tone, estimate.tone);
2035 // Untoned by default: a bar says nothing about status until something
2036 // says so, the same way a row is not selectable until told.
2037 assert_eq!(Meter::new(9, 10).tone, Tone::Neutral);
2038 }
2039
2040 #[test]
2041 fn a_meter_does_not_overflow_on_large_counts() {
2042 // done * 100 in u32 would wrap somewhere past 42 million. Counts that
2043 // size are not tasks, but a description layer that silently reports 3%
2044 // for a full bar is worse than one that is slow.
2045 let big = Meter::new(u32::MAX, u32::MAX);
2046 assert_eq!(big.percent(), 100);
2047 assert!(!big.overflowing());
2048 }
2049
2050 #[test]
2051 fn inset_is_raised_with_the_light_moved() {
2052 let (rl, rd) = Bevel::Raised.edges();
2053 let (il, id) = Bevel::Inset.edges();
2054 assert_eq!((rl, rd), (Edge::Light, Edge::Dark));
2055 assert_eq!((il, id), (rd, rl));
2056 }
2057
2058 #[test]
2059 fn pressing_twice_is_a_no_op() {
2060 for b in [Bevel::Raised, Bevel::Inset] {
2061 assert_eq!(b.pressed().pressed(), b);
2062 }
2063 }
2064
2065 #[test]
2066 fn a_raised_region_is_never_filled_with_a_recessed_surface() {
2067 // The bug this vocabulary exists to make unrepresentable.
2068 assert_eq!(Depth::Raised.fill(), Some(Fill::Raised));
2069 assert_eq!(Depth::Raised.bevel(), Some(Bevel::Raised));
2070 assert_eq!(Depth::Well.bevel(), Some(Bevel::Inset));
2071 assert_ne!(Depth::Well.fill(), Depth::Raised.fill());
2072 }
2073
2074 #[test]
2075 fn state_is_orthogonal_to_depth() {
2076 // The reason State is its own axis and not a Depth member: a disabled
2077 // button and a disabled field are both disabled and are not the same
2078 // shape, which one shared variant could not have said.
2079 assert_eq!(Depth::Raised.fill(), Some(Fill::Raised));
2080 assert_eq!(Depth::Well.fill(), Some(Fill::Well));
2081 assert!(State::Disabled.suppresses_interaction());
2082 }
2083
2084 #[test]
2085 fn only_disabled_stops_answering() {
2086 // Focus is a thing you can still click. Getting this backwards is how
2087 // a focus ring ends up on something inert.
2088 assert!(!State::Focus.suppresses_interaction());
2089 assert!(State::Disabled.suppresses_interaction());
2090 }
2091
2092 #[test]
2093 fn both_states_resolve_against_intents_makeover_already_derives() {
2094 // Neither needs a new token, so this costs no `makeover` release.
2095 assert_eq!(State::Focus.token(), "focus-ring");
2096 assert_eq!(State::Disabled.token(), "content-muted");
2097 }
2098
2099 #[test]
2100 fn flat_has_neither_edge_nor_fill() {
2101 assert_eq!(Depth::Flat.bevel(), None);
2102 assert_eq!(Depth::Flat.fill(), None);
2103 }
2104
2105 #[test]
2106 fn sunken_is_recessed_by_colour_with_no_edge() {
2107 // The one member carrying a fill without a bevel. A renderer that
2108 // assumes the two arrive together drops the fill silently, which is
2109 // exactly what makeover-webview did before 0.3.0.
2110 assert_eq!(Depth::Sunken.fill(), Some(Fill::Sunken));
2111 assert_eq!(Depth::Sunken.bevel(), None);
2112 }
2113
2114 #[test]
2115 fn sunken_and_flat_are_different_claims() {
2116 // Both edgeless, and only one of them needs a colour. Collapsing them
2117 // is what left an unchosen tab unsayable.
2118 assert_eq!(Depth::Flat.bevel(), Depth::Sunken.bevel());
2119 assert_ne!(Depth::Flat.fill(), Depth::Sunken.fill());
2120 }
2121
2122 #[test]
2123 fn a_sunken_surface_is_not_a_well() {
2124 // Authored in opposite directions: makeover derives surface-well by
2125 // inverting against the theme's content colour, while surface-sunken is
2126 // authored and may sit darker than raised.
2127 assert_ne!(Fill::Sunken, Fill::Well);
2128 assert_eq!(Fill::Sunken.token(), "surface-sunken");
2129 assert_eq!(Fill::Well.token(), "surface-well");
2130 }
2131
2132 #[test]
2133 fn every_selector_describes_both_of_its_states() {
2134 // The gap 0.3.0 closed. Before it, only `chosen` existed and the
2135 // unchosen option fell through to Flat at every renderer.
2136 for s in [Selector::Tabs, Selector::Segmented, Selector::Toggle] {
2137 assert_ne!(
2138 s.chosen(),
2139 s.unchosen(),
2140 "{s:?} cannot tell picked from unpicked"
2141 );
2142 }
2143 }
2144
2145 #[test]
2146 fn only_a_tab_inverts_the_other_way() {
2147 // Tabs recede so the chosen one comes forward; a segment and a toggle
2148 // stand up so the chosen one is held in. That inversion is the whole
2149 // content of "picked" once colour is deferred, and it is why the three
2150 // are not one member with a flag.
2151 assert_eq!(Selector::Tabs.unchosen(), Depth::Sunken);
2152 assert_eq!(Selector::Tabs.chosen(), Depth::Raised);
2153
2154 for s in [Selector::Segmented, Selector::Toggle] {
2155 assert_eq!(s.unchosen(), Depth::Raised);
2156 assert_eq!(s.chosen(), Depth::Well);
2157 // Held in is what pressing produces: one appearance, two reasons.
2158 assert_eq!(s.unchosen().pressed(), s.chosen());
2159 }
2160 }
2161
2162 #[test]
2163 fn pressing_a_card_makes_a_well() {
2164 assert_eq!(Depth::Raised.pressed(), Depth::Well);
2165 assert_eq!(
2166 Depth::Raised.pressed().bevel(),
2167 Depth::Raised.bevel().map(Bevel::pressed)
2168 );
2169 // Only raised regions respond to being pressed.
2170 assert_eq!(Depth::Flat.pressed(), Depth::Flat);
2171 assert_eq!(Depth::Well.pressed(), Depth::Well);
2172 // An overlay is a surface, not a control.
2173 assert_eq!(Depth::Overlay.pressed(), Depth::Overlay);
2174 }
2175
2176 #[test]
2177 fn an_overlay_is_lifted_rather_than_edged() {
2178 // The wave-2 rule: a surface over the page takes elevation, a surface
2179 // in the page takes a bevel. Both halves come off the one Depth, so
2180 // they cannot disagree.
2181 assert_eq!(Depth::Overlay.fill(), Some(Fill::Overlay));
2182 assert_eq!(Depth::Overlay.bevel(), None);
2183
2184 // Three depths have no bevel and they are not the same claim. Flat has
2185 // nothing to separate from, Sunken's colour is doing the separating,
2186 // and an overlay is separated by the lift.
2187 assert_ne!(Depth::Overlay.fill(), Depth::Sunken.fill());
2188 assert_ne!(Depth::Overlay.fill(), Depth::Flat.fill());
2189 }
2190
2191 #[test]
2192 fn intents_name_makeover_tokens_and_nothing_else() {
2193 assert_eq!(Edge::Light.token(), "bevel-light");
2194 assert_eq!(Edge::Dark.token(), "bevel-dark");
2195 assert_eq!(Fill::Raised.token(), "surface-raised");
2196 assert_eq!(Fill::Well.token(), "surface-well");
2197 // No value ever leaves this crate.
2198 for t in [
2199 Edge::Light.token(),
2200 Edge::Dark.token(),
2201 Tone::Danger.token(),
2202 Tone::Neutral.token(),
2203 State::Focus.token(),
2204 State::Disabled.token(),
2205 ] {
2206 assert!(!t.starts_with('#'), "{t} looks like a value");
2207 assert!(
2208 !t.chars().next().unwrap().is_ascii_digit(),
2209 "{t} is a value"
2210 );
2211 }
2212 }
2213
2214 #[test]
2215 fn a_badge_cannot_be_pressed_and_a_chip_latches() {
2216 // The one line that runs through all three apps' taxonomies.
2217 assert!(!Token::Badge.interactive());
2218 assert!(Token::Chip { removable: false }.interactive());
2219 assert!(Token::Chip { removable: true }.interactive());
2220
2221 // A badge is a label, so giving it an edge would lie about it.
2222 assert_eq!(Token::Badge.depth(false), Depth::Flat);
2223 assert_eq!(Token::Badge.depth(true), Depth::Flat);
2224
2225 // A latched chip wears the same shape a pressed one does.
2226 let chip = Token::Chip { removable: false };
2227 assert_eq!(chip.depth(false), Depth::Raised);
2228 assert_eq!(chip.depth(true), Depth::Raised.pressed());
2229 }
2230
2231 #[test]
2232 fn a_toast_and_a_banner_differ_in_more_than_placement() {
2233 assert!(Notice::Toast.transient());
2234 assert!(!Notice::Banner.transient());
2235 // A toast floats above the page; a banner rests in the flow.
2236 assert_eq!(Notice::Toast.fill(), Fill::Overlay);
2237 assert_eq!(Notice::Banner.fill(), Fill::Raised);
2238 }
2239
2240 #[test]
2241 fn emphasis_falls_off_down_the_row() {
2242 // `revealed_on_hover` was asserted here until 0.13.0 retired it. It said
2243 // a row's actions stay hidden until hover, which stopped being true when
2244 // makeover-webview 0.23.0 showed them at rest, and nothing had consumed
2245 // it for a release either way.
2246 assert_eq!(RowPart::Primary.intent(), "content");
2247 assert_eq!(RowPart::Secondary.intent(), "content-secondary");
2248 assert_eq!(RowPart::Meta.intent(), "content-muted");
2249 }
2250
2251 #[test]
2252 fn a_token_part_carries_no_intent_of_its_own() {
2253 // Each token carries its own tone, so a part-level intent underneath
2254 // would fight the thing sitting on it. Same reasoning as actions, which
2255 // is why they answer alike.
2256 assert_eq!(RowPart::Tokens.intent(), RowPart::Actions.intent());
2257 assert_eq!(RowPart::Tokens.intent(), "content");
2258 }
2259
2260 #[test]
2261 fn the_two_temporal_kinds_are_the_two_that_name_a_moment() {
2262 // The pair is named once so a host with parsing to do asks here rather
2263 // than spelling it out, which is `offers_options`' reason.
2264 assert!(FieldKind::Date.temporal());
2265 assert!(FieldKind::DateTime.temporal());
2266
2267 for kind in [
2268 FieldKind::Text,
2269 FieldKind::Secret,
2270 FieldKind::Number,
2271 FieldKind::Email,
2272 FieldKind::Url,
2273 FieldKind::Tel,
2274 FieldKind::Textarea,
2275 FieldKind::Select,
2276 FieldKind::Radio,
2277 FieldKind::Checkbox,
2278 FieldKind::File,
2279 FieldKind::Hidden,
2280 ] {
2281 assert!(!kind.temporal(), "{kind:?}");
2282 }
2283 }
2284
2285 #[test]
2286 fn a_date_carries_no_time_and_a_datetime_carries_no_zone() {
2287 // The formats are the whole reason the members are worth naming apart
2288 // from text, so the doc comments and the constants have to agree. A
2289 // host reading one and meeting the other is the silent failure.
2290 assert_eq!(DATE_FORMAT, "%Y-%m-%d");
2291 assert!(!DATE_FORMAT.contains("%H"), "a day carries no hour");
2292
2293 assert_eq!(DATETIME_FORMAT, "%Y-%m-%dT%H:%M");
2294 assert!(
2295 DATETIME_FORMAT.starts_with(DATE_FORMAT),
2296 "a moment starts with the day it is on"
2297 );
2298 // Local, and that is a property of the value rather than an omission.
2299 assert!(!DATETIME_FORMAT.contains("%Z"), "no zone name");
2300 assert!(!DATETIME_FORMAT.ends_with('Z'), "not UTC-stamped");
2301 assert!(!DATETIME_FORMAT.contains("%S"), "no seconds by default");
2302 }
2303
2304 #[test]
2305 fn a_temporal_kind_takes_a_label_above_it_and_offers_no_options() {
2306 // Neither is a checkbox and neither is a fixed set, so both fall where
2307 // text does. Asserted because a new kind lands in three predicates and
2308 // only one of them is the interesting one.
2309 for kind in [FieldKind::Date, FieldKind::DateTime] {
2310 assert!(kind.visible(), "{kind:?}");
2311 assert!(!kind.confidential(), "{kind:?}");
2312 assert!(!kind.labels_itself(), "{kind:?}");
2313 assert!(!kind.offers_options(), "{kind:?}");
2314 }
2315 }
2316
2317 #[test]
2318 fn a_cell_part_names_an_intent_and_only_the_value_is_text() {
2319 // The table half of what RowPart::intent does for rows. A cell holding
2320 // a control and a cell holding text answered alike until 0.14.0, and a
2321 // control in a cell took the cell's text colour.
2322 assert_eq!(CellPart::Value.intent(), "content");
2323
2324 for part in [CellPart::Tokens, CellPart::Actions, CellPart::Link] {
2325 // Each for its own reason -- a token carries its tone, an action is
2326 // a control, a link takes the action colour -- and all three reach
2327 // the intent inheriting already gives.
2328 assert_eq!(part.intent(), CellPart::Value.intent(), "{part:?}");
2329 }
2330 }
2331
2332 #[test]
2333 fn every_cell_part_answers_with_a_token_and_never_a_value() {
2334 for part in [
2335 CellPart::Value,
2336 CellPart::Tokens,
2337 CellPart::Actions,
2338 CellPart::Link,
2339 ] {
2340 let intent = part.intent();
2341 assert!(!intent.is_empty(), "{part:?} names nothing");
2342 assert!(!intent.starts_with('#'), "{part:?} looks like a value");
2343 }
2344 }
2345
2346 #[test]
2347 fn a_separator_is_what_tells_a_section_from_a_subsection() {
2348 assert!(Heading::Section.separated());
2349 assert!(!Heading::Subsection.separated());
2350 assert!(!Heading::Page.separated());
2351 }
2352
2353 #[test]
2354 fn a_chosen_segment_is_held_in_and_a_chosen_tab_comes_forward() {
2355 assert_eq!(Selector::Segmented.chosen(), Depth::Well);
2356 assert_eq!(Selector::Toggle.chosen(), Depth::Well);
2357 // The exception, and the whole folder semantic: the open tab joins its
2358 // pane rather than sinking away from it.
2359 assert_eq!(Selector::Tabs.chosen(), Depth::Raised);
2360
2361 // A held-in segment is indistinguishable from a pressed raised one,
2362 // which is the economy the light model buys over a colour swap.
2363 assert_eq!(Selector::Segmented.chosen(), Depth::Raised.pressed());
2364
2365 // A toggle stands alone; the other two are built out of parts that
2366 // touch.
2367 assert!(Selector::Segmented.abutting());
2368 assert!(Selector::Tabs.abutting());
2369 assert!(!Selector::Toggle.abutting());
2370 }
2371
2372 #[test]
2373 fn a_pane_is_looked_into_and_a_band_is_not() {
2374 assert_eq!(Region::Pane.depth(), Depth::Well);
2375 assert_eq!(Region::Modal.depth(), Depth::Raised);
2376 for r in [
2377 Region::Band,
2378 Region::Sidebar,
2379 Region::Split,
2380 Region::TabGroup,
2381 ] {
2382 assert_eq!(r.depth(), Depth::Flat, "{r:?} should carry no edge");
2383 }
2384 }
2385
2386 #[test]
2387 fn exactly_one_region_is_opaque() {
2388 // The escape hatch is one member and stays one member. If a second
2389 // undescribed region ever appears, the description has started
2390 // conceding rather than deferring.
2391 for r in [
2392 Region::Band,
2393 Region::Sidebar,
2394 Region::Pane,
2395 Region::Split,
2396 Region::TabGroup,
2397 Region::Modal,
2398 ] {
2399 assert!(r.described(), "{r:?} should be describable");
2400 }
2401 assert!(!Region::Bespoke { name: "day-plan" }.described());
2402 }
2403
2404 #[test]
2405 fn a_bespoke_region_inherits_its_depth_rather_than_choosing_one() {
2406 // The app owns the contents, not the placement. An app that wants its
2407 // timeline in a well frames it in a Pane.
2408 assert_eq!(Region::Bespoke { name: "day-plan" }.depth(), Depth::Flat);
2409 assert_eq!(Region::Bespoke { name: "kanban" }.depth(), Depth::Flat);
2410 }
2411
2412 #[test]
2413 fn a_screen_with_a_bespoke_region_is_still_a_whole_screen() {
2414 // The argument the member exists for: goingson's day-plan has to be
2415 // routable, or the description covers only the boring screens and the
2416 // interesting four need a second path beside the router.
2417 let day_plan = [
2418 Region::Band,
2419 Region::Bespoke { name: "day-plan" },
2420 Region::Sidebar,
2421 ];
2422 assert_eq!(day_plan.iter().filter(|r| r.described()).count(), 2);
2423 assert_eq!(day_plan.iter().filter(|r| !r.described()).count(), 1);
2424 }
2425
2426 #[test]
2427 fn a_secret_field_is_marked_as_one_and_a_hidden_field_is_not_drawn() {
2428 let secret = Field::new(FieldKind::Secret, "password", "Password");
2429 assert!(secret.kind.confidential());
2430 assert!(secret.kind.visible());
2431
2432 assert!(!FieldKind::Hidden.visible());
2433 // Nothing else is confidential, or the marker means nothing.
2434 for k in [
2435 FieldKind::Text,
2436 FieldKind::Number,
2437 FieldKind::Textarea,
2438 FieldKind::Select,
2439 FieldKind::Checkbox,
2440 FieldKind::Hidden,
2441 ] {
2442 assert!(!k.confidential(), "{k:?} should not be confidential");
2443 }
2444
2445 // Only a checkbox carries its own label.
2446 assert!(FieldKind::Checkbox.labels_itself());
2447 assert!(!FieldKind::Text.labels_itself());
2448 }
2449
2450 #[test]
2451 fn a_plain_field_offers_nothing_and_a_select_offers_its_options() {
2452 let text = Field::new(FieldKind::Text, "title", "Title");
2453 assert!(text.options.is_empty());
2454 assert_eq!(text.placeholder, None);
2455
2456 let sizes = [Choice::plain("small"), Choice::plain("large")];
2457 let select = Field::select("size", "Size", &sizes);
2458 assert_eq!(select.kind, FieldKind::Select);
2459 assert_eq!(select.options.len(), 2);
2460 }
2461
2462 #[test]
2463 fn a_choice_says_what_submits_and_what_is_read_apart() {
2464 // The whole reason it is two strings. `plain` is the case where they
2465 // coincide, and it is a shorthand rather than the general shape.
2466 let plain = Choice::plain("7");
2467 assert_eq!((plain.value, plain.label), ("7", "7"));
2468
2469 let spelled = Choice {
2470 value: "7",
2471 label: "One week",
2472 };
2473 assert_ne!(spelled.value, spelled.label);
2474 }
2475
2476 #[test]
2477 fn a_radio_asks_the_same_question_as_a_select_and_is_not_the_same_kind() {
2478 // Both offer a fixed set and both read `options`, so the two
2479 // constructors differ in exactly one thing. That one thing is the
2480 // point: a renderer decides whether the alternatives are readable
2481 // without opening anything, and it can only decide that if the
2482 // description said which question was asked.
2483 let styles = [
2484 Choice {
2485 value: "copy",
2486 label: "Copy samples in",
2487 },
2488 Choice {
2489 value: "reference",
2490 label: "Reference in place",
2491 },
2492 ];
2493 let radio = Field::radio("storage", "Storage style", &styles);
2494 let select = Field::select("storage", "Storage style", &styles);
2495
2496 assert_eq!(radio.kind, FieldKind::Radio);
2497 assert_ne!(radio.kind, select.kind);
2498 assert_eq!(radio.options, select.options);
2499 assert_eq!(
2500 Field {
2501 kind: select.kind,
2502 ..radio
2503 },
2504 select
2505 );
2506 }
2507
2508 #[test]
2509 fn exactly_the_option_taking_kinds_say_so() {
2510 // The renderers branch on this rather than on a list of their own, so
2511 // a kind added without a decision here renders its options nowhere.
2512 assert!(FieldKind::Select.offers_options());
2513 assert!(FieldKind::Radio.offers_options());
2514 for kind in [
2515 FieldKind::Text,
2516 FieldKind::Secret,
2517 FieldKind::Number,
2518 FieldKind::Email,
2519 FieldKind::Url,
2520 FieldKind::Tel,
2521 FieldKind::Textarea,
2522 FieldKind::Checkbox,
2523 FieldKind::Hidden,
2524 ] {
2525 assert!(!kind.offers_options(), "{kind:?} does not offer options");
2526 }
2527 }
2528
2529 #[test]
2530 fn a_radio_group_takes_a_label_even_though_its_options_carry_their_own() {
2531 // The near-miss: each option is labelled beside its own button, so a
2532 // renderer could plausibly read the group as self-labelling and drop
2533 // the question. Checkbox is the only kind that does that.
2534 assert!(!FieldKind::Radio.labels_itself());
2535 assert!(FieldKind::Checkbox.labels_itself());
2536 }
2537
2538 #[test]
2539 fn a_select_with_no_options_is_sayable() {
2540 // An app whose option list has not loaded has exactly this. Making it
2541 // unrepresentable would push the state somewhere less visible, and a
2542 // renderer drawing an empty select reports it on screen.
2543 let loading = Field::select("project", "Project", &[]);
2544 assert!(loading.options.is_empty());
2545 }
2546
2547 #[test]
2548 fn the_description_carries_the_question_and_never_the_answer() {
2549 // The line 0.8.0 drew. Placeholder and options are properties of what
2550 // is being asked; the current value is what came back, and no field
2551 // here holds one.
2552 let f = Field {
2553 placeholder: Some("yyyy-mm-dd"),
2554 ..Field::new(FieldKind::Text, "due", "Due")
2555 };
2556 assert_eq!(f.placeholder, Some("yyyy-mm-dd"));
2557 // A placeholder is not a label, and having one does not excuse the
2558 // field from carrying the other.
2559 assert_eq!(f.label, "Due");
2560 }
2561
2562 #[test]
2563 fn a_field_reports_its_own_error_state() {
2564 let mut f = Field::new(FieldKind::Text, "title", "Title");
2565 assert!(!f.invalid());
2566 f.error = Some("Required");
2567 assert!(f.invalid());
2568 }
2569
2570 #[test]
2571 fn columns_drop_by_priority_and_never_by_position() {
2572 let cols = [
2573 Column {
2574 width: Width::Fill,
2575 priority: Priority::Essential,
2576 ..Column::new("Title")
2577 },
2578 Column {
2579 width: Width::Fixed,
2580 priority: Priority::Secondary,
2581 ..Column::new("Due")
2582 },
2583 Column {
2584 width: Width::Fixed,
2585 priority: Priority::Optional,
2586 ..Column::new("Estimate")
2587 },
2588 ];
2589
2590 // Widest: everything survives.
2591 assert_eq!(
2592 cols.iter()
2593 .filter(|c| c.kept_at(Priority::Optional))
2594 .count(),
2595 3
2596 );
2597 // Narrower: the optional column goes first.
2598 let kept: Vec<_> = cols
2599 .iter()
2600 .filter(|c| c.kept_at(Priority::Secondary))
2601 .map(|c| c.name)
2602 .collect();
2603 assert_eq!(kept, ["Title", "Due"]);
2604 // Narrowest: only what identifies the row.
2605 let kept: Vec<_> = cols
2606 .iter()
2607 .filter(|c| c.kept_at(Priority::Essential))
2608 .map(|c| c.name)
2609 .collect();
2610 assert_eq!(kept, ["Title"]);
2611 }
2612
2613 #[test]
2614 fn inserting_a_column_does_not_move_what_gets_dropped() {
2615 // The bug the ordinal form has and this form cannot: goingson hides
2616 // `nth-child(n+5)` against a seven-column table, so a column inserted
2617 // anywhere to the left silently hides a different one.
2618 let before = [
2619 Column::new("Title"),
2620 Column {
2621 width: Width::Fixed,
2622 priority: Priority::Optional,
2623 ..Column::new("Estimate")
2624 },
2625 ];
2626 let after = [
2627 Column::new("Title"),
2628 Column::new("Project"), // inserted
2629 Column {
2630 width: Width::Fixed,
2631 priority: Priority::Optional,
2632 ..Column::new("Estimate")
2633 },
2634 ];
2635
2636 fn dropped<'a>(cols: &[Column<'a>]) -> Vec<&'a str> {
2637 cols.iter()
2638 .filter(|c| !c.kept_at(Priority::Secondary))
2639 .map(|c| c.name)
2640 .collect()
2641 }
2642 assert_eq!(dropped(&before), ["Estimate"]);
2643 assert_eq!(dropped(&after), ["Estimate"]);
2644 }
2645
2646 #[test]
2647 fn an_arrangement_carries_the_tab_group_as_a_modifier() {
2648 // goingson uses the tab group inside the content region rather than
2649 // instead of one, so it is not a third arrangement.
2650 let go = Arrangement::ListDetail { tabbed: true };
2651 let plain = Arrangement::ListDetail { tabbed: false };
2652 assert_ne!(go, plain);
2653 assert_ne!(go, Arrangement::SidebarContent);
2654 }
2655
2656 #[test]
2657 fn readiness_names_the_state_and_not_the_shimmer() {
2658 // Two members and no third. If a skeleton ever appears in this enum,
2659 // the deferral rule has been broken.
2660 assert_ne!(Readiness::Ready, Readiness::Pending);
2661 }
2662 }
2663