hugo/common/paths/pathparser.go

578 lines
13 KiB
Go
Raw Normal View History

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
// Copyright 2024 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 paths
import (
"path"
"path/filepath"
"runtime"
"strings"
"sync"
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/types"
"github.com/gohugoio/hugo/hugofs/files"
"github.com/gohugoio/hugo/identity"
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
)
var defaultPathParser PathParser
// PathParser parses a path into a Path.
type PathParser struct {
// Maps the language code to its index in the languages/sites slice.
LanguageIndex map[string]int
// Reports whether the given language is disabled.
IsLangDisabled func(string) bool
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
}
// Parse parses component c with path s into Path using the default path parser.
func Parse(c, s string) *Path {
return defaultPathParser.Parse(c, s)
}
// NormalizePathString returns a normalized path string using the very basic Hugo rules.
func NormalizePathStringBasic(s string) string {
// All lower case.
s = strings.ToLower(s)
// Replace spaces with hyphens.
s = strings.ReplaceAll(s, " ", "-")
return s
}
// ParseIdentity parses component c with path s into a StringIdentity.
func (pp *PathParser) ParseIdentity(c, s string) identity.StringIdentity {
p := pp.parsePooled(c, s)
defer putPath(p)
return identity.StringIdentity(p.IdentifierBase())
}
// ParseBaseAndBaseNameNoIdentifier parses component c with path s into a base and a base name without any identifier.
func (pp *PathParser) ParseBaseAndBaseNameNoIdentifier(c, s string) (string, string) {
p := pp.parsePooled(c, s)
defer putPath(p)
return p.Base(), p.BaseNameNoIdentifier()
}
func (pp *PathParser) parsePooled(c, s string) *Path {
s = NormalizePathStringBasic(s)
p := getPath()
p.component = c
p, err := pp.doParse(c, s, p)
if err != nil {
panic(err)
}
return p
}
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
// Parse parses component c with path s into Path using Hugo's content path rules.
func (pp *PathParser) Parse(c, s string) *Path {
p, err := pp.parse(c, s)
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
if err != nil {
panic(err)
}
return p
}
func (pp *PathParser) newPath(component string) *Path {
return &Path{
component: component,
posContainerLow: -1,
posContainerHigh: -1,
posSectionHigh: -1,
posIdentifierLanguage: -1,
}
}
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 (pp *PathParser) parse(component, s string) (*Path, error) {
ss := NormalizePathStringBasic(s)
p, err := pp.doParse(component, ss, pp.newPath(component))
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
if err != nil {
return nil, err
}
if s != ss {
var err error
// Preserve the original case for titles etc.
p.unnormalized, err = pp.doParse(component, s, pp.newPath(component))
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
if err != nil {
return nil, err
}
} else {
p.unnormalized = p
}
return p, nil
}
func (pp *PathParser) doParse(component, s string, p *Path) (*Path, error) {
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
hasLang := pp.LanguageIndex != nil
hasLang = hasLang && (component == files.ComponentFolderContent || component == files.ComponentFolderLayouts)
if runtime.GOOS == "windows" {
s = path.Clean(filepath.ToSlash(s))
if s == "." {
s = ""
}
}
if s == "" {
s = "/"
}
// Leading slash, no trailing slash.
if !strings.HasPrefix(s, "/") {
s = "/" + s
}
if s != "/" && s[len(s)-1] == '/' {
s = s[:len(s)-1]
}
p.s = s
slashCount := 0
for i := len(s) - 1; i >= 0; i-- {
c := s[i]
switch c {
case '.':
if p.posContainerHigh == -1 {
var high int
if len(p.identifiers) > 0 {
high = p.identifiers[len(p.identifiers)-1].Low - 1
} else {
high = len(p.s)
}
id := types.LowHigh{Low: i + 1, High: high}
if len(p.identifiers) == 0 {
p.identifiers = append(p.identifiers, id)
} else if len(p.identifiers) == 1 {
// Check for a valid language.
s := p.s[id.Low:id.High]
if hasLang {
var disabled bool
_, langFound := pp.LanguageIndex[s]
if !langFound {
disabled = pp.IsLangDisabled != nil && pp.IsLangDisabled(s)
if disabled {
p.disabled = true
langFound = true
}
}
if langFound {
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
p.posIdentifierLanguage = 1
p.identifiers = append(p.identifiers, id)
}
}
}
}
case '/':
slashCount++
if p.posContainerHigh == -1 {
p.posContainerHigh = i + 1
} else if p.posContainerLow == -1 {
p.posContainerLow = i + 1
}
if i > 0 {
p.posSectionHigh = i
}
}
}
isContentComponent := p.component == files.ComponentFolderContent || p.component == files.ComponentFolderArchetypes
isContent := isContentComponent && files.IsContentExt(p.Ext())
if isContent {
id := p.identifiers[len(p.identifiers)-1]
b := p.s[p.posContainerHigh : id.Low-1]
switch b {
case "index":
p.bundleType = PathTypeLeaf
case "_index":
p.bundleType = PathTypeBranch
default:
p.bundleType = PathTypeContentSingle
}
if slashCount == 2 && p.IsLeafBundle() {
p.posSectionHigh = 0
}
}
return p, nil
}
func ModifyPathBundleTypeResource(p *Path) {
if p.IsContent() {
p.bundleType = PathTypeContentResource
} else {
p.bundleType = PathTypeFile
}
}
type PathType int
const (
// A generic resource, e.g. a JSON file.
PathTypeFile PathType = iota
// All below are content files.
// A resource of a content type with front matter.
PathTypeContentResource
// E.g. /blog/my-post.md
PathTypeContentSingle
// All bewlow are bundled content files.
// Leaf bundles, e.g. /blog/my-post/index.md
PathTypeLeaf
// Branch bundles, e.g. /blog/_index.md
PathTypeBranch
)
type Path struct {
// Note: Any additions to this struct should also be added to the pathPool.
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
s string
posContainerLow int
posContainerHigh int
posSectionHigh int
component string
bundleType PathType
identifiers []types.LowHigh
posIdentifierLanguage int
disabled bool
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
trimLeadingSlash bool
unnormalized *Path
}
var pathPool = &sync.Pool{
New: func() any {
p := &Path{}
p.reset()
return p
},
}
func getPath() *Path {
return pathPool.Get().(*Path)
}
func putPath(p *Path) {
p.reset()
pathPool.Put(p)
}
func (p *Path) reset() {
p.s = ""
p.posContainerLow = -1
p.posContainerHigh = -1
p.posSectionHigh = -1
p.component = ""
p.bundleType = 0
p.identifiers = p.identifiers[:0]
p.posIdentifierLanguage = -1
p.disabled = false
p.trimLeadingSlash = false
p.unnormalized = 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
// TrimLeadingSlash returns a copy of the Path with the leading slash removed.
func (p Path) TrimLeadingSlash() *Path {
p.trimLeadingSlash = true
return &p
}
func (p *Path) norm(s string) string {
if p.trimLeadingSlash {
s = strings.TrimPrefix(s, "/")
}
return s
}
// IdentifierBase satifies identity.Identity.
func (p *Path) IdentifierBase() string {
return p.Base()
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
}
// Component returns the component for this path (e.g. "content").
func (p *Path) Component() string {
return p.component
}
// Container returns the base name of the container directory for this path.
func (p *Path) Container() string {
if p.posContainerLow == -1 {
return ""
}
return p.norm(p.s[p.posContainerLow : p.posContainerHigh-1])
}
// ContainerDir returns the container directory for this path.
// For content bundles this will be the parent directory.
func (p *Path) ContainerDir() string {
if p.posContainerLow == -1 || !p.IsBundle() {
return p.Dir()
}
return p.norm(p.s[:p.posContainerLow-1])
}
// Section returns the first path element (section).
func (p *Path) Section() string {
if p.posSectionHigh <= 0 {
return ""
}
return p.norm(p.s[1:p.posSectionHigh])
}
// IsContent returns true if the path is a content file (e.g. mypost.md).
// Note that this will also return true for content files in a bundle.
func (p *Path) IsContent() bool {
return p.BundleType() >= PathTypeContentResource
}
// isContentPage returns true if the path is a content file (e.g. mypost.md),
// but nof if inside a leaf bundle.
func (p *Path) isContentPage() bool {
return p.BundleType() >= PathTypeContentSingle
}
// Name returns the last element of path.
func (p *Path) Name() string {
if p.posContainerHigh > 0 {
return p.s[p.posContainerHigh:]
}
return p.s
}
// Name returns the last element of path withhout any extension.
func (p *Path) NameNoExt() string {
if i := p.identifierIndex(0); i != -1 {
return p.s[p.posContainerHigh : p.identifiers[i].Low-1]
}
return p.s[p.posContainerHigh:]
}
// Name returns the last element of path withhout any language identifier.
func (p *Path) NameNoLang() string {
i := p.identifierIndex(p.posIdentifierLanguage)
if i == -1 {
return p.Name()
}
return p.s[p.posContainerHigh:p.identifiers[i].Low-1] + p.s[p.identifiers[i].High:]
}
// BaseNameNoIdentifier returns the logcical base name for a resource without any idenifier (e.g. no extension).
// For bundles this will be the containing directory's name, e.g. "blog".
func (p *Path) BaseNameNoIdentifier() string {
if p.IsBundle() {
return p.Container()
}
return p.NameNoIdentifier()
}
// NameNoIdentifier returns the last element of path withhout any identifier (e.g. no extension).
func (p *Path) NameNoIdentifier() string {
if len(p.identifiers) > 0 {
return p.s[p.posContainerHigh : p.identifiers[len(p.identifiers)-1].Low-1]
}
return p.s[p.posContainerHigh:]
}
// Dir returns all but the last element of path, typically the path's directory.
func (p *Path) Dir() (d string) {
if p.posContainerHigh > 0 {
d = p.s[:p.posContainerHigh-1]
}
if d == "" {
d = "/"
}
d = p.norm(d)
return
}
// Path returns the full path.
func (p *Path) Path() (d string) {
return p.norm(p.s)
}
// Unnormalized returns the Path with the original case preserved.
func (p *Path) Unnormalized() *Path {
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
return p.unnormalized
}
// PathNoLang returns the Path but with any language identifier removed.
func (p *Path) PathNoLang() string {
return p.base(true, false)
}
// PathNoIdentifier returns the Path but with any identifier (ext, lang) removed.
func (p *Path) PathNoIdentifier() string {
return p.base(false, false)
}
// PathRel returns the path relativeto the given owner.
func (p *Path) PathRel(owner *Path) string {
ob := owner.Base()
if !strings.HasSuffix(ob, "/") {
ob += "/"
}
return strings.TrimPrefix(p.Path(), ob)
}
// BaseRel returns the base path relative to the given owner.
func (p *Path) BaseRel(owner *Path) string {
ob := owner.Base()
if ob == "/" {
ob = ""
}
return p.Base()[len(ob)+1:]
}
// For content files, Base returns the path without any identifiers (extension, language code etc.).
// Any 'index' as the last path element is ignored.
//
// For other files (Resources), any extension is kept.
func (p *Path) Base() string {
return p.base(!p.isContentPage(), p.IsBundle())
}
// BaseNoLeadingSlash returns the base path without the leading slash.
func (p *Path) BaseNoLeadingSlash() string {
return p.Base()[1:]
}
func (p *Path) base(preserveExt, isBundle bool) string {
if len(p.identifiers) == 0 {
return p.norm(p.s)
}
if preserveExt && len(p.identifiers) == 1 {
// Preserve extension.
return p.norm(p.s)
}
id := p.identifiers[len(p.identifiers)-1]
high := id.Low - 1
if isBundle {
high = p.posContainerHigh - 1
}
if high == 0 {
high++
}
if !preserveExt {
return p.norm(p.s[:high])
}
// For txt files etc. we want to preserve the extension.
id = p.identifiers[0]
return p.norm(p.s[:high] + p.s[id.Low-1:id.High])
}
func (p *Path) Ext() string {
return p.identifierAsString(0)
}
func (p *Path) Lang() string {
return p.identifierAsString(1)
}
func (p *Path) Identifier(i int) string {
return p.identifierAsString(i)
}
func (p *Path) Disabled() bool {
return p.disabled
}
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 (p *Path) Identifiers() []string {
ids := make([]string, len(p.identifiers))
for i, id := range p.identifiers {
ids[i] = p.s[id.Low:id.High]
}
return ids
}
func (p *Path) IsHTML() bool {
return files.IsHTML(p.Ext())
}
func (p *Path) BundleType() PathType {
return p.bundleType
}
func (p *Path) IsBundle() bool {
return p.bundleType >= PathTypeLeaf
}
func (p *Path) IsBranchBundle() bool {
return p.bundleType == PathTypeBranch
}
func (p *Path) IsLeafBundle() bool {
return p.bundleType == PathTypeLeaf
}
func (p Path) ForBundleType(t PathType) *Path {
p.bundleType = t
return &p
}
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 (p *Path) identifierAsString(i int) string {
i = p.identifierIndex(i)
if i == -1 {
return ""
}
id := p.identifiers[i]
return p.s[id.Low:id.High]
}
func (p *Path) identifierIndex(i int) int {
if i < 0 || i >= len(p.identifiers) {
return -1
}
return i
}
// HasExt returns true if the Unix styled path has an extension.
func HasExt(p string) bool {
for i := len(p) - 1; i >= 0; i-- {
if p[i] == '.' {
return true
}
if p[i] == '/' {
return false
}
}
return false
}