Files
obscura/crates/obscura-render/tests/layout_test.rs
T
SGavrl 23594b1f96 render: percentage padding, box-shadow, external-sprite svg, gradient text
percentage padding/margin: top/bottom percentages resolve against the
containing-block width (the responsive aspect-ratio-box trick, e.g.
padding-top:56.25%), resolved to px in the top-down layout pass so card
media boxes reserve space instead of collapsing.
box-shadow: parse and paint an outset shadow (rounded, offset, spread,
soft-edge approximation) so cards/buttons/menus have depth.
external-sprite svg: resolve <use href="url#id"> to an external sprite
file, inject the referenced symbol so inline-svg icon systems rasterize.
background-clip:text: paint transparent-with-gradient text (the common
gradient hero/button label) instead of leaving it invisible.
2026-07-11 21:02:57 +00:00

153 lines
7.2 KiB
Rust

//! Regression test: a Hacker-News-shaped nested `<table>` layout must come out
//! right using only the general engine (UA defaults + real CSS cascade), with
//! no per-site hardcoded selectors. This guards against reintroducing the
//! site-specific hacks that used to live in obscura-render for this exact markup.
use obscura_dom::tree_sink::parse_html;
use obscura_render::layout_dom;
const HN_HTML: &str = r##"
<table border="0" cellpadding="0" cellspacing="0" width="85%" bgcolor="#f6f6ef">
<tr>
<td bgcolor="#ff6600">
<table border="0" cellpadding="0" cellspacing="0" width="100%" style="padding:2px">
<tr>
<td style="width:18px;padding-right:4px"><a href="https://news.ycombinator.com"><img src="y18.svg" width="18" height="18"></a></td>
<td style="line-height:12pt; height:10px;"><span class="pagetop"><b class="hnname"><a href="news">Hacker News</a></b> <a href="newest">new</a> | <a href="front">past</a></span></td>
<td style="text-align:right;padding-right:4px;"><span class="pagetop"><a href="login?goto=news">login</a></span></td>
</tr>
</table>
</td>
</tr>
<tr>
<td>
<table border="0" cellpadding="0" cellspacing="0">
<tr class="athing submission" id="48761229">
<td align="right" valign="top" class="title"><span class="rank">24.</span></td>
<td valign="top" class="votelinks"><center><a id="up_48761229" href="vote?id=48761229"><div class="votearrow" title="upvote"></div></a></center></td>
<td class="title"><span class="titleline"><a href="https://www.zachtronics.com/exapunks/">Exapunks (2018)</a></span></td>
</tr>
</table>
</td>
</tr>
</table>
"##;
/// Top-left of the tightest laid-out element box whose text contains
/// `needle`. Text geometry is no longer a per-word list: a pure-text
/// container collapses to a single cosmic-text inline formatting context
/// (see `inline`), and even in the word-split path the wrapping `<a>`/`<span>`
/// are flattened into their block, so the smallest element rect enclosing the
/// text is the meaningful, mode-independent anchor. Picking the smallest-area
/// match skips the giant ancestor tables that also "contain" the text.
fn find_by_text(
tree: &obscura_dom::tree::DomTree,
layout: &obscura_render::DomLayout,
needle: &str,
) -> Option<(f32, f32)> {
let mut best: Option<(f32, obscura_render::Rect)> = None;
for (id, rect) in &layout.rects {
if tree.text_content(*id).contains(needle) {
let area = rect.width * rect.height;
if best.as_ref().map(|(a, _)| area < *a).unwrap_or(true) {
best = Some((area, *rect));
}
}
}
best.map(|(_, r)| (r.x, r.y))
}
#[test]
fn hn_shaped_table_lays_out_without_site_hardcoding() {
let tree = parse_html(HN_HTML);
let layout = layout_dom(&tree, (1000.0, 1000.0));
// Every element got a rect; the tree isn't just being dropped.
assert!(layout.rects.len() > 10, "expected many laid-out elements, got {}", layout.rects.len());
// "Hacker News" (a bold link in the top bar) sits above the first
// headline row ("Exapunks (2018)"): normal top-to-bottom flow, not a
// hardcoded absolute position.
let (brand_x, brand_y) = find_by_text(&tree, &layout, "Hacker News").expect("brand text laid out");
let (_, headline_y) = find_by_text(&tree, &layout, "Exapunks (2018)").expect("headline text laid out");
assert!(
headline_y > brand_y,
"headline should be below the header bar: brand.y={} headline.y={}",
brand_y,
headline_y
);
// Within the header row, "login" (right-aligned cell) sits to the right
// of "Hacker News" (left cell) — plain flex/table layout, not a magic
// per-class x-offset.
let (login_x, _) = find_by_text(&tree, &layout, "login").expect("login text laid out");
assert!(
login_x > brand_x,
"login cell should be right of the brand: brand.x={} login.x={}",
brand_x,
login_x
);
}
#[test]
fn relative_units_resolve_against_viewport_and_font_size() {
// 50vw of a 1000px viewport = 500px; 10em at the default 16px = 160px.
// Both were previously mis-resolved (vw kept as raw px, em hardcoded to 16
// regardless of context), so this guards the deferred-resolution pass.
let html = r##"<div style="width:50vw;height:10em"></div>"##;
let tree = parse_html(html);
let layout = layout_dom(&tree, (1000.0, 800.0));
let hit = layout.rects.values().any(|r| (r.width - 500.0).abs() < 1.0 && (r.height - 160.0).abs() < 1.0);
assert!(hit, "expected a 500x160 box from 50vw/10em, rects: {:?}", layout.rects.values().map(|r| (r.width, r.height)).collect::<Vec<_>>());
}
#[test]
fn percentage_padding_top_reserves_aspect_ratio_box() {
// The responsive aspect-ratio trick: an empty box with padding-top:56.25%
// inside a 1000px-wide block reserves a 16:9 area (~562px tall), the room a
// `position:absolute; inset:0` media child fills. Percentage padding
// resolves against the containing block WIDTH on every side, so the box
// gains real height instead of collapsing to zero.
let html = r##"<div style="width:1000px"><div style="padding-top:56.25%"></div></div>"##;
let tree = parse_html(html);
let layout = layout_dom(&tree, (1200.0, 800.0));
let hit = layout
.rects
.values()
.any(|r| (r.width - 1000.0).abs() < 1.0 && (r.height - 562.5).abs() < 2.0);
assert!(
hit,
"expected a ~1000x562 aspect-ratio box from padding-top:56.25%, rects: {:?}",
layout.rects.values().map(|r| (r.width, r.height)).collect::<Vec<_>>()
);
}
#[test]
fn long_text_run_wraps_across_multiple_lines() {
// A long single text node with no inline elements breaking it up must
// wrap within a narrow container instead of overflowing on one line. The
// container's height is the mode-independent proof: several wrapped lines
// make it much taller than one line, whether text is shaped by cosmic-text
// (paint) or split into word boxes (layout-only).
let html = r##"<div style="width:100px">This sentence has plenty of words to wrap across several lines</div>"##;
let tree = parse_html(html);
let layout = layout_dom(&tree, (1000.0, 1000.0));
// Tightest element enclosing the text: the 100px div itself.
let mut div_rect: Option<obscura_render::Rect> = None;
for (id, rect) in &layout.rects {
if tree.text_content(*id).contains("several lines") {
if div_rect.as_ref().map(|d| rect.width * rect.height < d.width * d.height).unwrap_or(true) {
div_rect = Some(*rect);
}
}
}
let div_rect = div_rect.expect("text container laid out");
assert!(div_rect.width <= 101.0, "container should hold its 100px width, got {}", div_rect.width);
assert!(
div_rect.height > 60.0,
"text should wrap onto several lines in a 100px-wide box (tall container), got height {}",
div_rect.height
);
}