fix(anatomy): url() locality skips CSS whitespace only (security audit R3 pass 2)

This commit is contained in:
bchanot committed 2026-10-15 00:40:30 +02:00
1 parent 157c854ed1
commit 23895b9351
6 files changed
+90 -10

No files matched your search

@@ -11,6 +11,7 @@ Migrate the anatomy core to Rust (review R3 scope): SVG analysis, anatomical tre
- `apps/app/tests/unit/canvas.trees.test.ts` is renamed to `.tsx` (it now mounts the canvas): a rename, no test deleted. - `apps/app/tests/unit/canvas.trees.test.ts` is renamed to `.tsx` (it now mounts the canvas): a rename, no test deleted.
- The sanitiser oracle (`oracles/r3/sanitiser.json`) was regenerated as corpus files were added; the TypeScript it runs (`packages/pdf/src`, `packages/anatomy/src`) is unchanged on the branch (criterion 4), so every regeneration is the same TypeScript's verdict. `highlight.json` gained `edgeDocument` only; its cases and sanitised trees are unchanged since d1aa35e (criterion 4). - The sanitiser oracle (`oracles/r3/sanitiser.json`) was regenerated as corpus files were added; the TypeScript it runs (`packages/pdf/src`, `packages/anatomy/src`) is unchanged on the branch (criterion 4), so every regeneration is the same TypeScript's verdict. `highlight.json` gained `edgeDocument` only; its cases and sanitised trees are unchanged since d1aa35e (criterion 4).
- Pass 2 (after the security audit, PASS + 1 MEDIUM + 2 LOW): the prescan searches each closer after its whole opener (`<!-->` desync), `href` locality on ASCII-trimmed text only, root `transform`/`filter`/`mask`/`clip-path` removed; 5 corpus files added (41), the wasm corpus has 49 cases; counts below updated. - Pass 2 (after the security audit, PASS + 1 MEDIUM + 2 LOW): the prescan searches each closer after its whole opener (`<!-->` desync), `href` locality on ASCII-trimmed text only, root `transform`/`filter`/`mask`/`clip-path` removed; 5 corpus files added (41), the wasm corpus has 49 cases; counts below updated.
- Pass 3 (security pass 2 on the fix: PASS + LOW): `url(` skips CSS whitespace only; corpus 42 files, wasm corpus 50 cases.
- Visual screenshots are uncommitted by design (plan O6); their result is recorded in RUST.md §9.3. - Visual screenshots are uncommitted by design (plan O6); their result is recorded in RUST.md §9.3.
## ACCEPTANCE CRITERIA ## ACCEPTANCE CRITERIA
@@ -40,10 +41,10 @@ Migrate the anatomy core to Rust (review R3 scope): SVG analysis, anatomical tre
EXPECT: gzip under 600000 bytes / exit 0 / app 352, core-wasm 29, pdf 140, anatomy 131, all passed EXPECT: gzip under 600000 bytes / exit 0 / app 352, core-wasm 29, pdf 140, anatomy 131, all passed
9. The hostile corpus runs through the shipped wasm under the stated budgets, and the canvas composes within its own. 9. The hostile corpus runs through the shipped wasm under the stated budgets, and the canvas composes within its own.
CHECK: grep -E "R3 hostile corpus|R3 sheet handle" /tmp/r3-verify-check.log CHECK: grep -E "R3 hostile corpus|R3 sheet handle" /tmp/r3-verify-check.log
EXPECT: a hostile-corpus line with 49 cases, slowest under 150 ms, largest growth under 32 MiB; a sheet-handle line with a warm head median under 15 ms EXPECT: a hostile-corpus line with 50 cases, slowest under 150 ms, largest growth under 32 MiB; a sheet-handle line with a warm head median under 15 ms
10. The hostile corpus covers the owner's classes (scripts, events, external and data references, DTD/entities, depth, element count, `<use>` count, cycles and expansion, malformed, namespaces, CSS, size). 10. The hostile corpus covers the owner's classes (scripts, events, external and data references, DTD/entities, depth, element count, `<use>` count, cycles and expansion, malformed, namespaces, CSS, size).
CHECK: ls crates/core/tests/fixtures/svg/hostile | wc -l; ls crates/core/tests/fixtures/svg/hostile | tr '\n' ' ' CHECK: ls crates/core/tests/fixtures/svg/hostile | wc -l; ls crates/core/tests/fixtures/svg/hostile | tr '\n' ' '
EXPECT: 41 files including comment-desync-depth, comment-desync-attributes, href-unicode-space, root-presentation, and script, events, script-urls, external-use, data-and-images, external-urls, doctype-entities, malformed, namespaces, css, use-self-cycle, use-mutual-cycle, root-class-style, long-title EXPECT: 42 files including url-unicode-space, comment-desync-depth, comment-desync-attributes, href-unicode-space, root-presentation, and script, events, script-urls, external-use, data-and-images, external-urls, doctype-entities, malformed, namespaces, css, use-self-cycle, use-mutual-cycle, root-class-style, long-title
11. The docs record the phase: RUST.md §9 (proofs, measurements, differences, risks), CHANGELOG, CLAUDE.md map. 11. The docs record the phase: RUST.md §9 (proofs, measurements, differences, risks), CHANGELOG, CLAUDE.md map.
CHECK: grep -c "^## 9. R3" docs/development/RUST.md; grep -c "Rust R3, anatomy core" CHANGELOG.md; grep -c "anatomy/{svg/{prescan,tree,sanitize,refs,number,serialize},layer,compose}" CLAUDE.md CHECK: grep -c "^## 9. R3" docs/development/RUST.md; grep -c "Rust R3, anatomy core" CHANGELOG.md; grep -c "anatomy/{svg/{prescan,tree,sanitize,refs,number,serialize},layer,compose}" CLAUDE.md
EXPECT: 1 / 1 / 1 EXPECT: 1 / 1 / 1
+11 -3
View File
@@ -5,7 +5,7 @@
//! reference grammar, no text outside `title`/`desc`. A refused element goes with its //! reference grammar, no text outside `title`/`desc`. A refused element goes with its
//! subtree, a refused attribute alone; both are named in `removed` (names, never values). //! subtree, a refused attribute alone; both are named in `removed` (names, never values).
use super::number::{is_js_space, js_number_of, js_split_list, js_to_string}; use super::number::{js_number_of, js_split_list, js_to_string};
use super::tree::{Element, Node, SvgRejection, XLINK_NS}; use super::tree::{Element, Node, SvgRejection, XLINK_NS};
const ALLOWED_ELEMENTS: [&str; 28] = [ const ALLOWED_ELEMENTS: [&str; 28] = [
@@ -273,6 +273,11 @@ fn value_acceptable(qname: &str, value: &str) -> bool {
} }
} }
/// CSS whitespace (what the tokenizer skips inside `url(`).
fn is_css_space(c: char) -> bool {
matches!(c, ' ' | '\t' | '\n' | '\r' | '\x0C')
}
/// An `href` as the URL parser reads it: only ASCII whitespace is stripped (the engine /// An `href` as the URL parser reads it: only ASCII whitespace is stripped (the engine
/// trimmed JavaScript spaces, so a no-break space before `#a` passed as local and resolved /// trimmed JavaScript spaces, so a no-break space before `#a` passed as local and resolved
/// as a relative URL). /// as a relative URL).
@@ -305,12 +310,15 @@ fn url_openings(value: &str) -> impl Iterator<Item = usize> + '_ {
}) })
} }
/// After `url(`: `\s*`, an optional quote, `\s*`; returns the offset of the next character. /// After `url(`: whitespace, an optional quote, whitespace; returns the offset of the next
/// character. Only CSS whitespace is skipped, as the browser's CSS tokenizer skips it: the
/// engine skipped JavaScript spaces, so a no-break space before `#a` read as local but the
/// browser resolves it as a relative URL (security audit R3, pass 2).
fn skip_url_prefix(value: &str, at: usize) -> usize { fn skip_url_prefix(value: &str, at: usize) -> usize {
let rest = &value[at..]; let rest = &value[at..];
let mut chars = rest.char_indices().peekable(); let mut chars = rest.char_indices().peekable();
let skip_space = |chars: &mut std::iter::Peekable<std::str::CharIndices<'_>>| { let skip_space = |chars: &mut std::iter::Peekable<std::str::CharIndices<'_>>| {
while chars.peek().is_some_and(|&(_, c)| is_js_space(c)) { while chars.peek().is_some_and(|&(_, c)| is_css_space(c)) {
chars.next(); chars.next();
} }
}; };
@@ -0,0 +1,2 @@
<svg xmlns="http://www.w3.org/2000/svg"><filter id="f"/><path d="M0 0h1" fill="url( #f)"/><path d="M0 0h1" style="fill:url( #f)"/><path d="M0 0h1" filter="url( #f)"/><path d="M0 0h1" filter="url(
#f)"/></svg>

After

Width:  |  Height:  |  Size: 211 B

@@ -3960,6 +3960,62 @@
"unbound-prefix.svg": { "unbound-prefix.svg": {
"rejected": "malformed" "rejected": "malformed"
}, },
"url-unicode-space.svg": {
"removed": [],
"tree": {
"ns": "http://www.w3.org/2000/svg",
"name": "svg",
"attrs": {
"xmlns": "http://www.w3.org/2000/svg"
},
"children": [
{
"ns": "http://www.w3.org/2000/svg",
"name": "filter",
"attrs": {
"id": "f"
},
"children": []
},
{
"ns": "http://www.w3.org/2000/svg",
"name": "path",
"attrs": {
"d": "M0 0h1",
"fill": "url( #f)"
},
"children": []
},
{
"ns": "http://www.w3.org/2000/svg",
"name": "path",
"attrs": {
"d": "M0 0h1",
"style": "fill:url( #f)"
},
"children": []
},
{
"ns": "http://www.w3.org/2000/svg",
"name": "path",
"attrs": {
"d": "M0 0h1",
"filter": "url( #f)"
},
"children": []
},
{
"ns": "http://www.w3.org/2000/svg",
"name": "path",
"attrs": {
"d": "M0 0h1",
"filter": "url( #f)"
},
"children": []
}
]
}
},
"use-mutual-cycle.svg": { "use-mutual-cycle.svg": {
"removed": [], "removed": [],
"tree": { "tree": {
+4
View File
@@ -42,6 +42,10 @@ fn declared(file: &str) -> Option<Value> {
// stripped, nothing else); the TS trimmed JavaScript spaces (security audit R3). // stripped, nothing else); the TS trimmed JavaScript spaces (security audit R3).
"href-unicode-space.svg" => json!({ "removed": ["use@href", "use@xlink:href"], "href-unicode-space.svg" => json!({ "removed": ["use@href", "use@xlink:href"],
"strip_unicode_spaced_refs": true }), "strip_unicode_spaced_refs": true }),
// C3 references: `url(` skips CSS whitespace only, as the browser's tokenizer does
// (security audit R3, pass 2): a Unicode space before `#` is not local.
"url-unicode-space.svg" => json!({ "removed": ["path@fill", "path@style"],
"strip": [["path", "fill"], ["path", "style"]] }),
// C3 root: nothing that paints outside the box on the inline root (security audit R3). // C3 root: nothing that paints outside the box on the inline root (security audit R3).
"root-presentation.svg" => { "root-presentation.svg" => {
json!({ "removed": ["svg@transform", "svg@filter", "svg@mask", "svg@clip-path"] }) json!({ "removed": ["svg@transform", "svg@filter", "svg@mask", "svg@clip-path"] })
+14 -5
View File
@@ -752,7 +752,7 @@ imports it and restyles `data-state` in place as before.
- O3 (`oracles/r3/highlight.json`): 36 highlighted compositions (every - O3 (`oracles/r3/highlight.json`): 36 highlighted compositions (every
observation and view of the three reports, an existing nested `<defs>`, a observation and view of the three reports, an existing nested `<defs>`, a
detail view with its own asset and no viewBox): tree-equal. detail view with its own asset and no viewBox): tree-equal.
- O4/O5 (`oracles/r3/sanitiser.json`): 41 hostile files - O4/O5 (`oracles/r3/sanitiser.json`): 42 hostile files
(`tests/fixtures/svg/hostile`) and 10 generated cases: equal to the (`tests/fixtures/svg/hostile`) and 10 generated cases: equal to the
TypeScript verdict or a declared one (9.4); the 4 product SVGs (horse TypeScript verdict or a declared one (9.4); the 4 product SVGs (horse
drawing, partner logo, 2 public logos) untouched, tree-equal. Mutation drawing, partner logo, 2 public logos) untouched, tree-equal. Mutation
@@ -762,7 +762,11 @@ imports it and restyles `data-state` in place as before.
searched from inside its own opener: `<!-->` let crafted markup hide 5,000 searched from inside its own opener: `<!-->` let crafted markup hide 5,000
nesting levels or 3,000 attributes, then roxmltree overflowed the stack), nesting levels or 3,000 attributes, then roxmltree overflowed the stack),
fixed with regression cases in cargo and through the shipped wasm; two LOW fixed with regression cases in cargo and through the shipped wasm; two LOW
(Unicode-spaced `href`, root painting attributes) fixed the same way. (Unicode-spaced `href`, root painting attributes) fixed the same way. Pass 2
on the fix: PASS (about 4 million differential documents, prescan against
roxmltree: no undercount, no cap breach, no stack overflow on a 64 KiB stack;
the old desync and its variants answer `too_deep` on the shipped module); its
LOW (the same Unicode-space tolerance inside `url(`) fixed.
- Shipped wasm (`packages/core-wasm/tests/anatomy.test.ts`): the corpus plus - Shipped wasm (`packages/core-wasm/tests/anatomy.test.ts`): the corpus plus
floods (99,000 nesting levels, 20,001 nodes, 150,000 `<` in a comment, 257 floods (99,000 nesting levels, 20,001 nodes, 150,000 `<` in a comment, 257
attributes, 2 MiB + 1, a 12 × 20 `<use>` chain, 1.9 MiB of `>`), each in a attributes, 2 MiB + 1, a 12 × 20 `<use>` chain, 1.9 MiB of `>`), each in a
@@ -792,7 +796,8 @@ imports it and restyles `data-state` in place as before.
canvas box); an `id` outside `[A-Za-z_][\w.:-]*` is removed; an `href` is canvas box); an `id` outside `[A-Za-z_][\w.:-]*` is removed; an `href` is
local only when it reads `#name` once ASCII whitespace is stripped, as the local only when it reads `#name` once ASCII whitespace is stripped, as the
URL parser strips it (the TypeScript also trimmed Unicode spaces, so a URL parser strips it (the TypeScript also trimmed Unicode spaces, so a
no-break space before `#a` passed and resolved as a relative URL). no-break space before `#a` passed and resolved as a relative URL); inside
`url(` only CSS whitespace is skipped before `#`, for the same reason.
4. A reference cycle, an expansion over 50,000 elements, and the floods of 4. A reference cycle, an expansion over 50,000 elements, and the floods of
9.2 step 1 are refused; the TypeScript accepted them (size counted on the 9.2 step 1 are refused; the TypeScript accepted them (size counted on the
raw text, refused subtrees included). raw text, refused subtrees included).
@@ -833,8 +838,12 @@ imports it and restyles `data-state` in place as before.
- The engine (PDF) still sanitises with the TypeScript allowlist and has no - The engine (PDF) still sanitises with the TypeScript allowlist and has no
reference budget: the `<use>` expansion gate of `PDF_IMPORT_HARDENING.md` reference budget: the `<use>` expansion gate of `PDF_IMPORT_HARDENING.md`
§5 is met in the core, closed for the PDF path only when R4 switches it. §5 is met in the core, closed for the PDF path only when R4 switches it.
- roxmltree's recursion: safe behind the prescan depth cap; pin the release - roxmltree's recursion: safe behind the prescan depth cap (differentially
with the iterative parser when it ships. fuzzed, 9.3); pin the release with the iterative parser when it ships.
- References are checked inside one drawing; in the page `#id` resolves against
the whole document (the detail modal imports a second copy of the drawing,
as before R3). Drawings with their own assets (Step 13) should get ids
scoped per view before external uploads are drawn.
- Firefox and WebKit: not run here (no browser build); the owner's LAN run on - Firefox and WebKit: not run here (no browser build); the owner's LAN run on
the phone covers WebKit. the phone covers WebKit.
- The degraded notice inside the canvas (9.4 item 7) and its wording ("le - The degraded notice inside the canvas (9.4 item 7) and its wording ("le