This is a follow up the bug fix in #15189. With the example given in that issue, the `css.Build` would be performed twice, which was unfortunate.
This commit fixes that by resuming the transformation from the last executed transformation.
See #15189
This allows @import statements to be resolved in a set of user provided resources (e.g. from resources.FromString or css.ChromaStyles) before the assets filesystem.
The option also applies to css.PostCSS via the shared import inlining, and css.Sass requires the dartsass transpiler. The import context is part of the transformation cache key.
Fixes#15103
Co-authored-by: Claude Fable 5 <noreply@anthropic.com>
The old implementation copied the incoming bytes into a fixed-size buffer
one byte at a time, and for every byte ran bytes.Contains for every
pattern over the whole buffer. That's O(n·patterns) scans of the buffer,
which showed up badly on large output (a full rendered page).
Scan each not-yet-matched pattern once per Write with bytes.Contains over
the chunk itself, in place, rather than per byte. The only carried state
is a small boundary window (the last maxPatternLen-1 bytes) joined with
the head of the next chunk, so a pattern straddling a Write boundary is
still detected. The chunk is never copied, so the extra allocation is
bounded by the longest pattern and independent of the output size. Once
all patterns have matched we mark done and drop the buffer.
patternLen summed the pattern lengths (to size the old buffer); the
boundary window only needs the longest pattern, so it's renamed
maxPatternLen and returns the max.
```bash
│ benchcmp.bench │ fix-hasbytewriter.bench │
│ sec/op │ sec/op vs base │
HasBytesWriter-10 2473.095µ ± ∞ ¹ 6.114µ ± ∞ ¹ -99.75% (p=0.029 n=4)
¹ need >= 6 samples for confidence interval at level 0.95
│ benchcmp.bench │ fix-hasbytewriter.bench │
│ B/op │ B/op vs base │
HasBytesWriter-10 48.00 ± ∞ ¹ 128.00 ± ∞ ¹ +166.67% (p=0.029 n=4)
¹ need >= 6 samples for confidence interval at level 0.95
│ benchcmp.bench │ fix-hasbytewriter.bench │
│ allocs/op │ allocs/op vs base │
```
And then rethrow the ERR_ACCESS_DENIED if we cannot recover.
There's more details in #15041, but this error has been seen on Netlify with the CJS because of how Netlify has their node_modules cache folder set up.
Fixes#15041
The encode/decode is implemented in a WebAssembly module built from a
small C wrapper around libavif. Bundled libraries (statically linked,
compiled with the WASI SDK):
* libavif v1.4.1 (container + codec glue)
* libaom v3.14.1 (AV1 encoder + decoder)
* dav1d 1.5.3 (AV1 decoder)
* libyuv (Chromium pin) for color conversion
* parson for JSON message passing across the wasm boundary
HDR handling on the encoder:
* SDR images are written as BT.709 / sRGB / BT.601 (8-bit).
* 10-bit and up are written as BT.2020 primaries with PQ (SMPTE
ST 2084) transfer and BT.2020-NCL matrix coefficients, signalled
via CICP.
* Adobe-style SDR+gainmap inputs (e.g. Lightroom HDR exports) are
baked into a single true-HDR image in BT.2020/PQ at 10-bit, with
the CLLI (Content Light Level Information) box carried through so
HDR-capable clients can tone-map correctly.
Limitations:
* Animated input (animated WebP/GIF) is collapsed to its first frame
when re-encoded as AVIF; animated AVIF output is not yet supported.
Fixes#7837
When a module provides a config key whose value is a slice (e.g.
cascade or permalinks), and the site config declares the same key as a
map with only a merge strategy marker (_merge = 'deep'), the types do
not match and Params.merge silently dropped the module's value, leaving
the site with no effective cascade or permalink config from the module.
Fix Params.merge so that when the destination value is an empty Params
(IsZero — only the _merge key is present) and the source value is a
non-Params type, the source value is used provided the user-declared
merge strategy is not 'none'. This honours the explicit _merge
directive regardless of the surrounding shallow-merge context.
Closes#13869
Co-authored-by: Claude Sonnet 4.6 <noreply@anthropic.com>
Node's ESM resolver does not consult NODE_PATH (unlike CJS require), so
an ESM postcss.config.js shipped by a Hugo theme fails when loaded from
the module cache: bare imports like `import x from "postcss-import"`
have no node_modules to walk up to.
Install a synchronous resolver hook (module.registerHooks) via
--import=data:... on every Node invocation. On ERR_MODULE_NOT_FOUND for
a bare specifier it resolves the package from each NODE_PATH entry via
createRequire().resolve(). No-op for relative, absolute, URL-scheme and
non-MODULE_NOT_FOUND failures. Synchronous hooks run on the main thread,
so no --allow-worker is needed under the Node permission model.
Fixes#13987
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
When CSS imports assets via the file loader (fonts, images), the emitted
URLs were relative to the CSS output directory. That broke when the CSS
was inlined into HTML, since browsers then resolved the URLs against the
page rather than the CSS file.
Set esbuild's PublicPath to the CSS output directory joined with the
site base path so URLs work whether the CSS is published as a file or
inlined.
Fixes#14849
Some Linux setups trigger detect-libc's spawnSync('getconf') fallback
when process.report does not expose glibcVersionRuntime, breaking
tailwindcss under the Node permission model. Add AllowChildProcess
mirroring AllowAddons/AllowWorker, default to ["tailwindcss"], and
emit --allow-child-process accordingly.
Fixes#14824
Allow CSS variables to be grouped under sub-paths and imported via
@import "hugo:vars/mobile" (or @use for Dart Sass), so callers can pass
nested dicts like:
{{ dict "primary-color" "blue" "mobile" (dict "primary-color" "red") }}
Top-level "hugo:vars" now skips nested map entries instead of emitting
garbage for them.
Fixes#14705
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Add security.node.permissions config to run Node tools (PostCSS, Babel,
TailwindCSS) under Node's permission model, restricting file system access
to the working directory by default.
The binary resolution is simplified to node_modules/.bin → PATH (npx removed).
For both locations, the actual JS entry point is resolved via symlinks (macOS/Linux)
or by parsing npm wrapper scripts (Windows .cmd), then executed as
"node --permission --allow-fs-read=<path> --allow-fs-write=<path> <script>".
Users can opt out by removing "node" from security.exec.allow.
Closes#7287
Bjørn Erik Pedersen