hugo/resources/resource_transformers/js/options.go

462 lines
12 KiB
Go
Raw Permalink Normal View History

// Copyright 2020 The Hugo Authors. All rights reserved.
//
// Licensed under the Apache License, Version 2.0 (the "License");
// you may not use this file except in compliance with the License.
// You may obtain a copy of the License at
// http://www.apache.org/licenses/LICENSE-2.0
//
// Unless required by applicable law or agreed to in writing, software
// distributed under the License is distributed on an "AS IS" BASIS,
// WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
// See the License for the specific language governing permissions and
// limitations under the License.
package js
import (
"encoding/json"
"fmt"
"os"
"path/filepath"
"strings"
"github.com/gohugoio/hugo/common/maps"
all: Rework page store, add a dynacache, improve partial rebuilds, and some general spring cleaning There are some breaking changes in this commit, see #11455. Closes #11455 Closes #11549 This fixes a set of bugs (see issue list) and it is also paying some technical debt accumulated over the years. We now build with Staticcheck enabled in the CI build. The performance should be about the same as before for regular sized Hugo sites, but it should perform and scale much better to larger data sets, as objects that uses lots of memory (e.g. rendered Markdown, big JSON files read into maps with transform.Unmarshal etc.) will now get automatically garbage collected if needed. Performance on partial rebuilds when running the server in fast render mode should be the same, but the change detection should be much more accurate. A list of the notable new features: * A new dependency tracker that covers (almost) all of Hugo's API and is used to do fine grained partial rebuilds when running the server. * A new and simpler tree document store which allows fast lookups and prefix-walking in all dimensions (e.g. language) concurrently. * You can now configure an upper memory limit allowing for much larger data sets and/or running on lower specced PCs. We have lifted the "no resources in sub folders" restriction for branch bundles (e.g. sections). Memory Limit * Hugos will, by default, set aside a quarter of the total system memory, but you can set this via the OS environment variable HUGO_MEMORYLIMIT (in gigabytes). This is backed by a partitioned LRU cache used throughout Hugo. A cache that gets dynamically resized in low memory situations, allowing Go's Garbage Collector to free the memory. New Dependency Tracker: Hugo has had a rule based coarse grained approach to server rebuilds that has worked mostly pretty well, but there have been some surprises (e.g. stale content). This is now revamped with a new dependency tracker that can quickly calculate the delta given a changed resource (e.g. a content file, template, JS file etc.). This handles transitive relations, e.g. $page -> js.Build -> JS import, or $page1.Content -> render hook -> site.GetPage -> $page2.Title, or $page1.Content -> shortcode -> partial -> site.RegularPages -> $page2.Content -> shortcode ..., and should also handle changes to aggregated values (e.g. site.Lastmod) effectively. This covers all of Hugo's API with 2 known exceptions (a list that may not be fully exhaustive): Changes to files loaded with template func os.ReadFile may not be handled correctly. We recommend loading resources with resources.Get Changes to Hugo objects (e.g. Page) passed in the template context to lang.Translate may not be detected correctly. We recommend having simple i18n templates without too much data context passed in other than simple types such as strings and numbers. Note that the cachebuster configuration (when A changes then rebuild B) works well with the above, but we recommend that you revise that configuration, as it in most situations should not be needed. One example where it is still needed is with TailwindCSS and using changes to hugo_stats.json to trigger new CSS rebuilds. Document Store: Previously, a little simplified, we split the document store (where we store pages and resources) in a tree per language. This worked pretty well, but the structure made some operations harder than they needed to be. We have now restructured it into one Radix tree for all languages. Internally the language is considered to be a dimension of that tree, and the tree can be viewed in all dimensions concurrently. This makes some operations re. language simpler (e.g. finding translations is just a slice range), but the idea is that it should also be relatively inexpensive to add more dimensions if needed (e.g. role). Fixes #10169 Fixes #10364 Fixes #10482 Fixes #10630 Fixes #10656 Fixes #10694 Fixes #10918 Fixes #11262 Fixes #11439 Fixes #11453 Fixes #11457 Fixes #11466 Fixes #11540 Fixes #11551 Fixes #11556 Fixes #11654 Fixes #11661 Fixes #11663 Fixes #11664 Fixes #11669 Fixes #11671 Fixes #11807 Fixes #11808 Fixes #11809 Fixes #11815 Fixes #11840 Fixes #11853 Fixes #11860 Fixes #11883 Fixes #11904 Fixes #7388 Fixes #7425 Fixes #7436 Fixes #7544 Fixes #7882 Fixes #7960 Fixes #8255 Fixes #8307 Fixes #8863 Fixes #8927 Fixes #9192 Fixes #9324
2023-12-24 13:11:05 -05:00
"github.com/gohugoio/hugo/common/paths"
"github.com/gohugoio/hugo/identity"
"github.com/spf13/afero"
"github.com/evanw/esbuild/pkg/api"
"github.com/gohugoio/hugo/hugofs"
"github.com/gohugoio/hugo/media"
"github.com/mitchellh/mapstructure"
)
const (
nsImportHugo = "ns-hugo"
nsParams = "ns-params"
stdinImporter = "<stdin>"
)
// Options esbuild configuration
type Options struct {
// If not set, the source path will be used as the base target path.
// Note that the target path's extension may change if the target MIME type
// is different, e.g. when the source is TypeScript.
TargetPath string
// Whether to minify to output.
Minify bool
// Whether to write mapfiles
SourceMap string
// The language target.
// One of: es2015, es2016, es2017, es2018, es2019, es2020 or esnext.
// Default is esnext.
Target string
// The output format.
// One of: iife, cjs, esm
// Default is to esm.
Format string
// External dependencies, e.g. "react".
Externals []string
// This option allows you to automatically replace a global variable with an import from another file.
// The filenames must be relative to /assets.
// See https://esbuild.github.io/api/#inject
Inject []string
// User defined symbols.
Defines map[string]any
// Maps a component import to another.
Shims map[string]string
// User defined params. Will be marshaled to JSON and available as "@params", e.g.
// import * as params from '@params';
Params any
// What to use instead of React.createElement.
JSXFactory string
// What to use instead of React.Fragment.
JSXFragment string
// What to do about JSX syntax.
// See https://esbuild.github.io/api/#jsx
JSX string
// Which library to use to automatically import JSX helper functions from. Only works if JSX is set to automatic.
// See https://esbuild.github.io/api/#jsx-import-source
JSXImportSource string
2020-11-03 03:05:09 -05:00
// There is/was a bug in WebKit with severe performance issue with the tracking
// of TDZ checks in JavaScriptCore.
//
// Enabling this flag removes the TDZ and `const` assignment checks and
// may improve performance of larger JS codebases until the WebKit fix
// is in widespread use.
//
// See https://bugs.webkit.org/show_bug.cgi?id=199866
// Deprecated: This no longer have any effect and will be removed.
// TODO(bep) remove. See https://github.com/evanw/esbuild/commit/869e8117b499ca1dbfc5b3021938a53ffe934dba
2020-11-03 03:05:09 -05:00
AvoidTDZ bool
mediaType media.Type
outDir string
contents string
sourceDir string
resolveDir string
tsConfig string
}
func decodeOptions(m map[string]any) (Options, error) {
var opts Options
if err := mapstructure.WeakDecode(m, &opts); err != nil {
return opts, err
}
if opts.TargetPath != "" {
all: Rework page store, add a dynacache, improve partial rebuilds, and some general spring cleaning There are some breaking changes in this commit, see #11455. Closes #11455 Closes #11549 This fixes a set of bugs (see issue list) and it is also paying some technical debt accumulated over the years. We now build with Staticcheck enabled in the CI build. The performance should be about the same as before for regular sized Hugo sites, but it should perform and scale much better to larger data sets, as objects that uses lots of memory (e.g. rendered Markdown, big JSON files read into maps with transform.Unmarshal etc.) will now get automatically garbage collected if needed. Performance on partial rebuilds when running the server in fast render mode should be the same, but the change detection should be much more accurate. A list of the notable new features: * A new dependency tracker that covers (almost) all of Hugo's API and is used to do fine grained partial rebuilds when running the server. * A new and simpler tree document store which allows fast lookups and prefix-walking in all dimensions (e.g. language) concurrently. * You can now configure an upper memory limit allowing for much larger data sets and/or running on lower specced PCs. We have lifted the "no resources in sub folders" restriction for branch bundles (e.g. sections). Memory Limit * Hugos will, by default, set aside a quarter of the total system memory, but you can set this via the OS environment variable HUGO_MEMORYLIMIT (in gigabytes). This is backed by a partitioned LRU cache used throughout Hugo. A cache that gets dynamically resized in low memory situations, allowing Go's Garbage Collector to free the memory. New Dependency Tracker: Hugo has had a rule based coarse grained approach to server rebuilds that has worked mostly pretty well, but there have been some surprises (e.g. stale content). This is now revamped with a new dependency tracker that can quickly calculate the delta given a changed resource (e.g. a content file, template, JS file etc.). This handles transitive relations, e.g. $page -> js.Build -> JS import, or $page1.Content -> render hook -> site.GetPage -> $page2.Title, or $page1.Content -> shortcode -> partial -> site.RegularPages -> $page2.Content -> shortcode ..., and should also handle changes to aggregated values (e.g. site.Lastmod) effectively. This covers all of Hugo's API with 2 known exceptions (a list that may not be fully exhaustive): Changes to files loaded with template func os.ReadFile may not be handled correctly. We recommend loading resources with resources.Get Changes to Hugo objects (e.g. Page) passed in the template context to lang.Translate may not be detected correctly. We recommend having simple i18n templates without too much data context passed in other than simple types such as strings and numbers. Note that the cachebuster configuration (when A changes then rebuild B) works well with the above, but we recommend that you revise that configuration, as it in most situations should not be needed. One example where it is still needed is with TailwindCSS and using changes to hugo_stats.json to trigger new CSS rebuilds. Document Store: Previously, a little simplified, we split the document store (where we store pages and resources) in a tree per language. This worked pretty well, but the structure made some operations harder than they needed to be. We have now restructured it into one Radix tree for all languages. Internally the language is considered to be a dimension of that tree, and the tree can be viewed in all dimensions concurrently. This makes some operations re. language simpler (e.g. finding translations is just a slice range), but the idea is that it should also be relatively inexpensive to add more dimensions if needed (e.g. role). Fixes #10169 Fixes #10364 Fixes #10482 Fixes #10630 Fixes #10656 Fixes #10694 Fixes #10918 Fixes #11262 Fixes #11439 Fixes #11453 Fixes #11457 Fixes #11466 Fixes #11540 Fixes #11551 Fixes #11556 Fixes #11654 Fixes #11661 Fixes #11663 Fixes #11664 Fixes #11669 Fixes #11671 Fixes #11807 Fixes #11808 Fixes #11809 Fixes #11815 Fixes #11840 Fixes #11853 Fixes #11860 Fixes #11883 Fixes #11904 Fixes #7388 Fixes #7425 Fixes #7436 Fixes #7544 Fixes #7882 Fixes #7960 Fixes #8255 Fixes #8307 Fixes #8863 Fixes #8927 Fixes #9192 Fixes #9324
2023-12-24 13:11:05 -05:00
opts.TargetPath = paths.ToSlashTrimLeading(opts.TargetPath)
}
opts.Target = strings.ToLower(opts.Target)
opts.Format = strings.ToLower(opts.Format)
return opts, nil
}
var extensionToLoaderMap = map[string]api.Loader{
".js": api.LoaderJS,
".mjs": api.LoaderJS,
".cjs": api.LoaderJS,
".jsx": api.LoaderJSX,
".ts": api.LoaderTS,
".tsx": api.LoaderTSX,
".css": api.LoaderCSS,
".json": api.LoaderJSON,
".txt": api.LoaderText,
}
func loaderFromFilename(filename string) api.Loader {
l, found := extensionToLoaderMap[filepath.Ext(filename)]
if found {
return l
}
return api.LoaderJS
}
func resolveComponentInAssets(fs afero.Fs, impPath string) *hugofs.FileMeta {
findFirst := func(base string) *hugofs.FileMeta {
// This is the most common sub-set of ESBuild's default extensions.
// We assume that imports of JSON, CSS etc. will be using their full
// name with extension.
for _, ext := range []string{".js", ".ts", ".tsx", ".jsx"} {
if strings.HasSuffix(impPath, ext) {
// Import of foo.js.js need the full name.
continue
}
if fi, err := fs.Stat(base + ext); err == nil {
return fi.(hugofs.FileMetaInfo).Meta()
}
}
// Not found.
return nil
}
var m *hugofs.FileMeta
// We need to check if this is a regular file imported without an extension.
2023-02-18 15:47:35 -05:00
// There may be ambiguous situations where both foo.js and foo/index.js exists.
// This import order is in line with both how Node and ESBuild's native
// import resolver works.
// It may be a regular file imported without an extension, e.g.
// foo or foo/index.
m = findFirst(impPath)
if m != nil {
return m
}
base := filepath.Base(impPath)
if base == "index" {
// try index.esm.js etc.
m = findFirst(impPath + ".esm")
if m != nil {
return m
}
}
// Check the path as is.
fi, err := fs.Stat(impPath)
if err == nil {
if fi.IsDir() {
m = findFirst(filepath.Join(impPath, "index"))
if m == nil {
m = findFirst(filepath.Join(impPath, "index.esm"))
}
} else {
m = fi.(hugofs.FileMetaInfo).Meta()
}
} else if strings.HasSuffix(base, ".js") {
m = findFirst(strings.TrimSuffix(impPath, ".js"))
}
return m
}
all: Rework page store, add a dynacache, improve partial rebuilds, and some general spring cleaning There are some breaking changes in this commit, see #11455. Closes #11455 Closes #11549 This fixes a set of bugs (see issue list) and it is also paying some technical debt accumulated over the years. We now build with Staticcheck enabled in the CI build. The performance should be about the same as before for regular sized Hugo sites, but it should perform and scale much better to larger data sets, as objects that uses lots of memory (e.g. rendered Markdown, big JSON files read into maps with transform.Unmarshal etc.) will now get automatically garbage collected if needed. Performance on partial rebuilds when running the server in fast render mode should be the same, but the change detection should be much more accurate. A list of the notable new features: * A new dependency tracker that covers (almost) all of Hugo's API and is used to do fine grained partial rebuilds when running the server. * A new and simpler tree document store which allows fast lookups and prefix-walking in all dimensions (e.g. language) concurrently. * You can now configure an upper memory limit allowing for much larger data sets and/or running on lower specced PCs. We have lifted the "no resources in sub folders" restriction for branch bundles (e.g. sections). Memory Limit * Hugos will, by default, set aside a quarter of the total system memory, but you can set this via the OS environment variable HUGO_MEMORYLIMIT (in gigabytes). This is backed by a partitioned LRU cache used throughout Hugo. A cache that gets dynamically resized in low memory situations, allowing Go's Garbage Collector to free the memory. New Dependency Tracker: Hugo has had a rule based coarse grained approach to server rebuilds that has worked mostly pretty well, but there have been some surprises (e.g. stale content). This is now revamped with a new dependency tracker that can quickly calculate the delta given a changed resource (e.g. a content file, template, JS file etc.). This handles transitive relations, e.g. $page -> js.Build -> JS import, or $page1.Content -> render hook -> site.GetPage -> $page2.Title, or $page1.Content -> shortcode -> partial -> site.RegularPages -> $page2.Content -> shortcode ..., and should also handle changes to aggregated values (e.g. site.Lastmod) effectively. This covers all of Hugo's API with 2 known exceptions (a list that may not be fully exhaustive): Changes to files loaded with template func os.ReadFile may not be handled correctly. We recommend loading resources with resources.Get Changes to Hugo objects (e.g. Page) passed in the template context to lang.Translate may not be detected correctly. We recommend having simple i18n templates without too much data context passed in other than simple types such as strings and numbers. Note that the cachebuster configuration (when A changes then rebuild B) works well with the above, but we recommend that you revise that configuration, as it in most situations should not be needed. One example where it is still needed is with TailwindCSS and using changes to hugo_stats.json to trigger new CSS rebuilds. Document Store: Previously, a little simplified, we split the document store (where we store pages and resources) in a tree per language. This worked pretty well, but the structure made some operations harder than they needed to be. We have now restructured it into one Radix tree for all languages. Internally the language is considered to be a dimension of that tree, and the tree can be viewed in all dimensions concurrently. This makes some operations re. language simpler (e.g. finding translations is just a slice range), but the idea is that it should also be relatively inexpensive to add more dimensions if needed (e.g. role). Fixes #10169 Fixes #10364 Fixes #10482 Fixes #10630 Fixes #10656 Fixes #10694 Fixes #10918 Fixes #11262 Fixes #11439 Fixes #11453 Fixes #11457 Fixes #11466 Fixes #11540 Fixes #11551 Fixes #11556 Fixes #11654 Fixes #11661 Fixes #11663 Fixes #11664 Fixes #11669 Fixes #11671 Fixes #11807 Fixes #11808 Fixes #11809 Fixes #11815 Fixes #11840 Fixes #11853 Fixes #11860 Fixes #11883 Fixes #11904 Fixes #7388 Fixes #7425 Fixes #7436 Fixes #7544 Fixes #7882 Fixes #7960 Fixes #8255 Fixes #8307 Fixes #8863 Fixes #8927 Fixes #9192 Fixes #9324
2023-12-24 13:11:05 -05:00
func createBuildPlugins(depsManager identity.Manager, c *Client, opts Options) ([]api.Plugin, error) {
fs := c.rs.Assets
resolveImport := func(args api.OnResolveArgs) (api.OnResolveResult, error) {
impPath := args.Path
if opts.Shims != nil {
override, found := opts.Shims[impPath]
if found {
impPath = override
}
}
isStdin := args.Importer == stdinImporter
var relDir string
if !isStdin {
rel, found := fs.MakePathRelative(args.Importer, true)
if !found {
// Not in any of the /assets folders.
// This is an import from a node_modules, let
// ESBuild resolve this.
return api.OnResolveResult{}, nil
}
all: Rework page store, add a dynacache, improve partial rebuilds, and some general spring cleaning There are some breaking changes in this commit, see #11455. Closes #11455 Closes #11549 This fixes a set of bugs (see issue list) and it is also paying some technical debt accumulated over the years. We now build with Staticcheck enabled in the CI build. The performance should be about the same as before for regular sized Hugo sites, but it should perform and scale much better to larger data sets, as objects that uses lots of memory (e.g. rendered Markdown, big JSON files read into maps with transform.Unmarshal etc.) will now get automatically garbage collected if needed. Performance on partial rebuilds when running the server in fast render mode should be the same, but the change detection should be much more accurate. A list of the notable new features: * A new dependency tracker that covers (almost) all of Hugo's API and is used to do fine grained partial rebuilds when running the server. * A new and simpler tree document store which allows fast lookups and prefix-walking in all dimensions (e.g. language) concurrently. * You can now configure an upper memory limit allowing for much larger data sets and/or running on lower specced PCs. We have lifted the "no resources in sub folders" restriction for branch bundles (e.g. sections). Memory Limit * Hugos will, by default, set aside a quarter of the total system memory, but you can set this via the OS environment variable HUGO_MEMORYLIMIT (in gigabytes). This is backed by a partitioned LRU cache used throughout Hugo. A cache that gets dynamically resized in low memory situations, allowing Go's Garbage Collector to free the memory. New Dependency Tracker: Hugo has had a rule based coarse grained approach to server rebuilds that has worked mostly pretty well, but there have been some surprises (e.g. stale content). This is now revamped with a new dependency tracker that can quickly calculate the delta given a changed resource (e.g. a content file, template, JS file etc.). This handles transitive relations, e.g. $page -> js.Build -> JS import, or $page1.Content -> render hook -> site.GetPage -> $page2.Title, or $page1.Content -> shortcode -> partial -> site.RegularPages -> $page2.Content -> shortcode ..., and should also handle changes to aggregated values (e.g. site.Lastmod) effectively. This covers all of Hugo's API with 2 known exceptions (a list that may not be fully exhaustive): Changes to files loaded with template func os.ReadFile may not be handled correctly. We recommend loading resources with resources.Get Changes to Hugo objects (e.g. Page) passed in the template context to lang.Translate may not be detected correctly. We recommend having simple i18n templates without too much data context passed in other than simple types such as strings and numbers. Note that the cachebuster configuration (when A changes then rebuild B) works well with the above, but we recommend that you revise that configuration, as it in most situations should not be needed. One example where it is still needed is with TailwindCSS and using changes to hugo_stats.json to trigger new CSS rebuilds. Document Store: Previously, a little simplified, we split the document store (where we store pages and resources) in a tree per language. This worked pretty well, but the structure made some operations harder than they needed to be. We have now restructured it into one Radix tree for all languages. Internally the language is considered to be a dimension of that tree, and the tree can be viewed in all dimensions concurrently. This makes some operations re. language simpler (e.g. finding translations is just a slice range), but the idea is that it should also be relatively inexpensive to add more dimensions if needed (e.g. role). Fixes #10169 Fixes #10364 Fixes #10482 Fixes #10630 Fixes #10656 Fixes #10694 Fixes #10918 Fixes #11262 Fixes #11439 Fixes #11453 Fixes #11457 Fixes #11466 Fixes #11540 Fixes #11551 Fixes #11556 Fixes #11654 Fixes #11661 Fixes #11663 Fixes #11664 Fixes #11669 Fixes #11671 Fixes #11807 Fixes #11808 Fixes #11809 Fixes #11815 Fixes #11840 Fixes #11853 Fixes #11860 Fixes #11883 Fixes #11904 Fixes #7388 Fixes #7425 Fixes #7436 Fixes #7544 Fixes #7882 Fixes #7960 Fixes #8255 Fixes #8307 Fixes #8863 Fixes #8927 Fixes #9192 Fixes #9324
2023-12-24 13:11:05 -05:00
relDir = filepath.Dir(rel)
} else {
relDir = opts.sourceDir
}
// Imports not starting with a "." is assumed to live relative to /assets.
// Hugo makes no assumptions about the directory structure below /assets.
if relDir != "" && strings.HasPrefix(impPath, ".") {
impPath = filepath.Join(relDir, impPath)
}
m := resolveComponentInAssets(fs.Fs, impPath)
if m != nil {
all: Rework page store, add a dynacache, improve partial rebuilds, and some general spring cleaning There are some breaking changes in this commit, see #11455. Closes #11455 Closes #11549 This fixes a set of bugs (see issue list) and it is also paying some technical debt accumulated over the years. We now build with Staticcheck enabled in the CI build. The performance should be about the same as before for regular sized Hugo sites, but it should perform and scale much better to larger data sets, as objects that uses lots of memory (e.g. rendered Markdown, big JSON files read into maps with transform.Unmarshal etc.) will now get automatically garbage collected if needed. Performance on partial rebuilds when running the server in fast render mode should be the same, but the change detection should be much more accurate. A list of the notable new features: * A new dependency tracker that covers (almost) all of Hugo's API and is used to do fine grained partial rebuilds when running the server. * A new and simpler tree document store which allows fast lookups and prefix-walking in all dimensions (e.g. language) concurrently. * You can now configure an upper memory limit allowing for much larger data sets and/or running on lower specced PCs. We have lifted the "no resources in sub folders" restriction for branch bundles (e.g. sections). Memory Limit * Hugos will, by default, set aside a quarter of the total system memory, but you can set this via the OS environment variable HUGO_MEMORYLIMIT (in gigabytes). This is backed by a partitioned LRU cache used throughout Hugo. A cache that gets dynamically resized in low memory situations, allowing Go's Garbage Collector to free the memory. New Dependency Tracker: Hugo has had a rule based coarse grained approach to server rebuilds that has worked mostly pretty well, but there have been some surprises (e.g. stale content). This is now revamped with a new dependency tracker that can quickly calculate the delta given a changed resource (e.g. a content file, template, JS file etc.). This handles transitive relations, e.g. $page -> js.Build -> JS import, or $page1.Content -> render hook -> site.GetPage -> $page2.Title, or $page1.Content -> shortcode -> partial -> site.RegularPages -> $page2.Content -> shortcode ..., and should also handle changes to aggregated values (e.g. site.Lastmod) effectively. This covers all of Hugo's API with 2 known exceptions (a list that may not be fully exhaustive): Changes to files loaded with template func os.ReadFile may not be handled correctly. We recommend loading resources with resources.Get Changes to Hugo objects (e.g. Page) passed in the template context to lang.Translate may not be detected correctly. We recommend having simple i18n templates without too much data context passed in other than simple types such as strings and numbers. Note that the cachebuster configuration (when A changes then rebuild B) works well with the above, but we recommend that you revise that configuration, as it in most situations should not be needed. One example where it is still needed is with TailwindCSS and using changes to hugo_stats.json to trigger new CSS rebuilds. Document Store: Previously, a little simplified, we split the document store (where we store pages and resources) in a tree per language. This worked pretty well, but the structure made some operations harder than they needed to be. We have now restructured it into one Radix tree for all languages. Internally the language is considered to be a dimension of that tree, and the tree can be viewed in all dimensions concurrently. This makes some operations re. language simpler (e.g. finding translations is just a slice range), but the idea is that it should also be relatively inexpensive to add more dimensions if needed (e.g. role). Fixes #10169 Fixes #10364 Fixes #10482 Fixes #10630 Fixes #10656 Fixes #10694 Fixes #10918 Fixes #11262 Fixes #11439 Fixes #11453 Fixes #11457 Fixes #11466 Fixes #11540 Fixes #11551 Fixes #11556 Fixes #11654 Fixes #11661 Fixes #11663 Fixes #11664 Fixes #11669 Fixes #11671 Fixes #11807 Fixes #11808 Fixes #11809 Fixes #11815 Fixes #11840 Fixes #11853 Fixes #11860 Fixes #11883 Fixes #11904 Fixes #7388 Fixes #7425 Fixes #7436 Fixes #7544 Fixes #7882 Fixes #7960 Fixes #8255 Fixes #8307 Fixes #8863 Fixes #8927 Fixes #9192 Fixes #9324
2023-12-24 13:11:05 -05:00
depsManager.AddIdentity(m.PathInfo)
// Store the source root so we can create a jsconfig.json
// to help IntelliSense when the build is done.
// This should be a small number of elements, and when
// in server mode, we may get stale entries on renames etc.,
// but that shouldn't matter too much.
c.rs.JSConfigBuilder.AddSourceRoot(m.SourceRoot)
return api.OnResolveResult{Path: m.Filename, Namespace: nsImportHugo}, nil
}
// Fall back to ESBuild's resolve.
return api.OnResolveResult{}, nil
}
importResolver := api.Plugin{
Name: "hugo-import-resolver",
Setup: func(build api.PluginBuild) {
build.OnResolve(api.OnResolveOptions{Filter: `.*`},
func(args api.OnResolveArgs) (api.OnResolveResult, error) {
return resolveImport(args)
})
build.OnLoad(api.OnLoadOptions{Filter: `.*`, Namespace: nsImportHugo},
func(args api.OnLoadArgs) (api.OnLoadResult, error) {
b, err := os.ReadFile(args.Path)
if err != nil {
return api.OnLoadResult{}, fmt.Errorf("failed to read %q: %w", args.Path, err)
}
c := string(b)
return api.OnLoadResult{
// See https://github.com/evanw/esbuild/issues/502
// This allows all modules to resolve dependencies
// in the main project's node_modules.
ResolveDir: opts.resolveDir,
Contents: &c,
Loader: loaderFromFilename(args.Path),
}, nil
})
},
}
params := opts.Params
if params == nil {
// This way @params will always resolve to something.
params = make(map[string]any)
}
b, err := json.Marshal(params)
if err != nil {
return nil, fmt.Errorf("failed to marshal params: %w", err)
}
bs := string(b)
paramsPlugin := api.Plugin{
Name: "hugo-params-plugin",
Setup: func(build api.PluginBuild) {
build.OnResolve(api.OnResolveOptions{Filter: `^@params$`},
func(args api.OnResolveArgs) (api.OnResolveResult, error) {
return api.OnResolveResult{
Path: args.Path,
Namespace: nsParams,
}, nil
})
build.OnLoad(api.OnLoadOptions{Filter: `.*`, Namespace: nsParams},
func(args api.OnLoadArgs) (api.OnLoadResult, error) {
return api.OnLoadResult{
Contents: &bs,
Loader: api.LoaderJSON,
}, nil
})
},
}
return []api.Plugin{importResolver, paramsPlugin}, nil
}
func toBuildOptions(opts Options) (buildOptions api.BuildOptions, err error) {
var target api.Target
switch opts.Target {
case "", "esnext":
target = api.ESNext
case "es5":
target = api.ES5
case "es6", "es2015":
target = api.ES2015
case "es2016":
target = api.ES2016
case "es2017":
target = api.ES2017
case "es2018":
target = api.ES2018
case "es2019":
target = api.ES2019
case "es2020":
target = api.ES2020
case "es2021":
target = api.ES2021
case "es2022":
target = api.ES2022
case "es2023":
target = api.ES2023
default:
err = fmt.Errorf("invalid target: %q", opts.Target)
return
}
mediaType := opts.mediaType
if mediaType.IsZero() {
mediaType = media.Builtin.JavascriptType
}
var loader api.Loader
switch mediaType.SubType {
// TODO(bep) ESBuild support a set of other loaders, but I currently fail
// to see the relevance. That may change as we start using this.
case media.Builtin.JavascriptType.SubType:
loader = api.LoaderJS
case media.Builtin.TypeScriptType.SubType:
loader = api.LoaderTS
case media.Builtin.TSXType.SubType:
loader = api.LoaderTSX
case media.Builtin.JSXType.SubType:
loader = api.LoaderJSX
default:
err = fmt.Errorf("unsupported Media Type: %q", opts.mediaType)
return
}
var format api.Format
// One of: iife, cjs, esm
switch opts.Format {
case "", "iife":
format = api.FormatIIFE
case "esm":
format = api.FormatESModule
case "cjs":
format = api.FormatCommonJS
default:
err = fmt.Errorf("unsupported script output format: %q", opts.Format)
return
}
var jsx api.JSX
switch opts.JSX {
case "", "transform":
jsx = api.JSXTransform
case "preserve":
jsx = api.JSXPreserve
case "automatic":
jsx = api.JSXAutomatic
default:
err = fmt.Errorf("unsupported jsx type: %q", opts.JSX)
return
}
var defines map[string]string
if opts.Defines != nil {
defines = maps.ToStringMapString(opts.Defines)
}
// By default we only need to specify outDir and no outFile
outDir := opts.outDir
outFile := ""
var sourceMap api.SourceMap
switch opts.SourceMap {
case "inline":
sourceMap = api.SourceMapInline
case "external":
sourceMap = api.SourceMapExternal
case "":
sourceMap = api.SourceMapNone
default:
err = fmt.Errorf("unsupported sourcemap type: %q", opts.SourceMap)
return
}
buildOptions = api.BuildOptions{
Outfile: outFile,
Bundle: true,
Target: target,
Format: format,
Sourcemap: sourceMap,
MinifyWhitespace: opts.Minify,
MinifyIdentifiers: opts.Minify,
MinifySyntax: opts.Minify,
Outdir: outDir,
Define: defines,
External: opts.Externals,
JSXFactory: opts.JSXFactory,
JSXFragment: opts.JSXFragment,
JSX: jsx,
JSXImportSource: opts.JSXImportSource,
Tsconfig: opts.tsConfig,
// Note: We're not passing Sourcefile to ESBuild.
// This makes ESBuild pass `stdin` as the Importer to the import
// resolver, which is what we need/expect.
Stdin: &api.StdinOptions{
Contents: opts.contents,
ResolveDir: opts.resolveDir,
Loader: loader,
},
}
return
}