| 1 |
# makeover-webview |
| 2 |
|
| 3 |
The webview renderer for |
| 4 |
[`makeover-layout`](https://makenot.work/git/max/makeover-layout). Emits CSS. |
| 5 |
|
| 6 |
## The renderer that needs no palette |
| 7 |
|
| 8 |
`makeover-immediate` and `makeover-tui` both take a `Palette`, because egui and |
| 9 |
a terminal need an actual colour before they can put anything on screen. A |
| 10 |
webview does not. `var(--surface-raised)` *is* the late binding, and the browser |
| 11 |
resolves it against whatever the theme layer last wrote onto `:root`. |
| 12 |
|
| 13 |
So this crate emits text naming intents and never learns a colour, which is the |
| 14 |
deferral rule with no adapter in the way. It is also why the webview was always |
| 15 |
the wrong renderer to derive a vocabulary from: it can express anything, so it |
| 16 |
never pushes back. |
| 17 |
|
| 18 |
## Phase A: the stylesheet |
| 19 |
|
| 20 |
No markup here, deliberately. GoingsOn has 145 `innerHTML` sites and Balanced |
| 21 |
Breakfast 175 `createElement` sites, so moving markup is a migration while |
| 22 |
adopting a generated stylesheet is a deletion. The apps keep every line of their |
| 23 |
markup and gain the classes. |
| 24 |
|
| 25 |
```css |
| 26 |
:root { |
| 27 |
--bevel-raised: inset 1px 1px 0 var(--bevel-light), inset -1px -1px 0 var(--bevel-dark); |
| 28 |
--bevel-inset: inset 1px 1px 0 var(--bevel-dark), inset -1px -1px 0 var(--bevel-light); |
| 29 |
} |
| 30 |
|
| 31 |
.raised { background: var(--surface-raised); box-shadow: var(--bevel-raised); } |
| 32 |
.well { background: var(--surface-well, var(--surface-page)); box-shadow: var(--bevel-inset); } |
| 33 |
.card:active { background: var(--surface-well, var(--surface-page)); box-shadow: var(--bevel-inset); } |
| 34 |
``` |
| 35 |
|
| 36 |
Those two custom properties are byte-identical to what both apps already |
| 37 |
hand-write, which is asserted in the tests. Adoption removes duplicated lines |
| 38 |
rather than changing a pixel. |
| 39 |
|
| 40 |
The depth classes are surfaces and nothing more. A raised object that is not |
| 41 |
pressable is spelled `.raised`, and the states belong to the named surfaces: |
| 42 |
`.card` and `.button` are that same depth *and* controls, so they carry hover, |
| 43 |
press, focus and disabled. |
| 44 |
|
| 45 |
`.card:active` is the one thing this renderer gets for free: the cascade carries |
| 46 |
a pressed state that an immediate-mode renderer resolves per call site, eighteen |
| 47 |
of them in audiofiles. |
| 48 |
|
| 49 |
## Phase B: form markup |
| 50 |
|
| 51 |
`form::field_html` renders a `makeover_layout::Field` to the group both apps |
| 52 |
build by hand today: label, control, hint, error, in goingson's shape and class |
| 53 |
names, so adopting it deletes `renderFormField` rather than restyling anything. |
| 54 |
|
| 55 |
```rust |
| 56 |
use makeover_layout::{Field, FieldKind}; |
| 57 |
use makeover_webview::{Emit, form::{Filling, Value, field_html}}; |
| 58 |
|
| 59 |
let field = Field::new(FieldKind::Text, "title", "Title"); |
| 60 |
let html = field_html(&field, &Filling::of(Value::Text("Ship it")), &Emit::default()); |
| 61 |
``` |
| 62 |
|
| 63 |
It emits strings because both apps interpolate fields into larger string-built |
| 64 |
forms, and it escapes them itself. One escaper covers both sinks: goingson needs |
| 65 |
four and a correct choice at 543 call sites only because `textContent` |
| 66 |
serialization will not encode a double quote, which is not a constraint Rust |
| 67 |
has. The single hole is `form::Markup`, which a caller has to name. |
| 68 |
|
| 69 |
Three things the emitter does that neither app does at initial render: |
| 70 |
|
| 71 |
- `aria-invalid="true"` on an invalid control, which is what the generated |
| 72 |
stylesheet keys its danger ring on. goingson sets it from its runtime |
| 73 |
validation path and not from its renderer, so a field rendered already-invalid |
| 74 |
is styled as though nothing were wrong. |
| 75 |
- `aria-describedby` naming the hint as well as the error, so standing help |
| 76 |
survives an error appearing. |
| 77 |
- A secret never carries its value into the markup, on `FieldKind::confidential`. |
| 78 |
|
| 79 |
Rows and tables are the other half of phase B and are not written yet. |
| 80 |
|
| 81 |
## Substitution, three ways |
| 82 |
|
| 83 |
`Fill::Well` has no colour on makeover before 2.3.0, and the three renderers |
| 84 |
answer that differently. That is the evidence that taking `Fill::fallback` out of |
| 85 |
the description was right: |
| 86 |
|
| 87 |
- `makeover-immediate` substitutes the page in Rust. |
| 88 |
- `makeover-tui` refuses to substitute and draws an edge, because a terminal |
| 89 |
quantises the two together. |
| 90 |
- here, CSS already has the mechanism. `var(--surface-well, var(--surface-page))` |
| 91 |
falls back in the browser and nothing in Rust decides anything. |
| 92 |
|
| 93 |
## Status |
| 94 |
|
| 95 |
On crates.io at 0.5.1, which is the CSS half only; the form markup is unreleased. |
| 96 |
Consumers are GoingsOn and Balanced Breakfast, both through makeover-build. |
| 97 |
|
| 98 |
`cargo run --example dump` prints the stylesheet. |
| 99 |
|
| 100 |
## Licence |
| 101 |
|
| 102 |
MIT. |
| 103 |
|