hugo/tpl/collections/collections.go
Bjørn Erik Pedersen 7285e74090
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
2024-01-27 16:28:14 +01:00

779 lines
18 KiB
Go

// Copyright 2019 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 collections provides template functions for manipulating collections
// such as arrays, maps, and slices.
package collections
import (
"context"
"errors"
"fmt"
"math/rand"
"net/url"
"reflect"
"strings"
"time"
"github.com/gohugoio/hugo/common/collections"
"github.com/gohugoio/hugo/common/hugo"
"github.com/gohugoio/hugo/common/maps"
"github.com/gohugoio/hugo/common/types"
"github.com/gohugoio/hugo/deps"
"github.com/gohugoio/hugo/langs"
"github.com/gohugoio/hugo/tpl/compare"
"github.com/spf13/cast"
)
// New returns a new instance of the collections-namespaced template functions.
func New(deps *deps.Deps) *Namespace {
language := deps.Conf.Language()
if language == nil {
panic("language must be set")
}
loc := langs.GetLocation(language)
return &Namespace{
loc: loc,
sortComp: compare.New(loc, true),
deps: deps,
}
}
// Namespace provides template functions for the "collections" namespace.
type Namespace struct {
loc *time.Location
sortComp *compare.Namespace
deps *deps.Deps
}
// After returns all the items after the first n items in list l.
func (ns *Namespace) After(n any, l any) (any, error) {
if n == nil || l == nil {
return nil, errors.New("both limit and seq must be provided")
}
nv, err := cast.ToIntE(n)
if err != nil {
return nil, err
}
if nv < 0 {
return nil, errors.New("sequence bounds out of range [" + cast.ToString(nv) + ":]")
}
lv := reflect.ValueOf(l)
lv, isNil := indirect(lv)
if isNil {
return nil, errors.New("can't iterate over a nil value")
}
switch lv.Kind() {
case reflect.Array, reflect.Slice, reflect.String:
// okay
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l).Type().String())
}
if nv >= lv.Len() {
return lv.Slice(0, 0).Interface(), nil
}
return lv.Slice(nv, lv.Len()).Interface(), nil
}
// Delimit takes a given list l and returns a string delimited by sep.
// If last is passed to the function, it will be used as the final delimiter.
func (ns *Namespace) Delimit(ctx context.Context, l, sep any, last ...any) (string, error) {
d, err := cast.ToStringE(sep)
if err != nil {
return "", err
}
var dLast *string
if len(last) > 0 {
l := last[0]
dStr, err := cast.ToStringE(l)
if err != nil {
dLast = nil
} else {
dLast = &dStr
}
}
lv := reflect.ValueOf(l)
lv, isNil := indirect(lv)
if isNil {
return "", errors.New("can't iterate over a nil value")
}
var str string
switch lv.Kind() {
case reflect.Map:
sortSeq, err := ns.Sort(ctx, l)
if err != nil {
return "", err
}
lv = reflect.ValueOf(sortSeq)
fallthrough
case reflect.Array, reflect.Slice, reflect.String:
for i := 0; i < lv.Len(); i++ {
val := lv.Index(i).Interface()
valStr, err := cast.ToStringE(val)
if err != nil {
continue
}
switch {
case i == lv.Len()-2 && dLast != nil:
str += valStr + *dLast
case i == lv.Len()-1:
str += valStr
default:
str += valStr + d
}
}
default:
return "", fmt.Errorf("can't iterate over %T", l)
}
return str, nil
}
// Dictionary creates a new map from the given parameters by
// treating values as key-value pairs. The number of values must be even.
// The keys can be string slices, which will create the needed nested structure.
func (ns *Namespace) Dictionary(values ...any) (map[string]any, error) {
if len(values)%2 != 0 {
return nil, errors.New("invalid dictionary call")
}
root := make(map[string]any)
for i := 0; i < len(values); i += 2 {
dict := root
var key string
switch v := values[i].(type) {
case string:
key = v
case []string:
for i := 0; i < len(v)-1; i++ {
key = v[i]
var m map[string]any
v, found := dict[key]
if found {
m = v.(map[string]any)
} else {
m = make(map[string]any)
dict[key] = m
}
dict = m
}
key = v[len(v)-1]
default:
return nil, errors.New("invalid dictionary key")
}
dict[key] = values[i+1]
}
return root, nil
}
// EchoParam returns the value in the collection c with key k if is set; otherwise, it returns an
// empty string.
// Deprecated: Use the index function instead.
func (ns *Namespace) EchoParam(c, k any) any {
hugo.Deprecate("collections.EchoParam", "Use the index function instead.", "v0.120.0")
av, isNil := indirect(reflect.ValueOf(c))
if isNil {
return ""
}
var avv reflect.Value
switch av.Kind() {
case reflect.Array, reflect.Slice:
index, ok := k.(int)
if ok && av.Len() > index {
avv = av.Index(index)
}
case reflect.Map:
kv := reflect.ValueOf(k)
if kv.Type().AssignableTo(av.Type().Key()) {
avv = av.MapIndex(kv)
}
}
avv, isNil = indirect(avv)
if isNil {
return ""
}
if avv.IsValid() {
switch avv.Kind() {
case reflect.Int, reflect.Int8, reflect.Int16, reflect.Int32, reflect.Int64:
return avv.Int()
case reflect.Uint, reflect.Uint8, reflect.Uint16, reflect.Uint32, reflect.Uint64:
return avv.Uint()
case reflect.Float32, reflect.Float64:
return avv.Float()
case reflect.String:
return avv.String()
case reflect.Bool:
return avv.Bool()
}
}
return ""
}
// First returns the first limit items in list l.
func (ns *Namespace) First(limit any, l any) (any, error) {
if limit == nil || l == nil {
return nil, errors.New("both limit and seq must be provided")
}
limitv, err := cast.ToIntE(limit)
if err != nil {
return nil, err
}
if limitv < 0 {
return nil, errors.New("sequence length must be non-negative")
}
lv := reflect.ValueOf(l)
lv, isNil := indirect(lv)
if isNil {
return nil, errors.New("can't iterate over a nil value")
}
switch lv.Kind() {
case reflect.Array, reflect.Slice, reflect.String:
// okay
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l).Type().String())
}
if limitv > lv.Len() {
limitv = lv.Len()
}
return lv.Slice(0, limitv).Interface(), nil
}
// In returns whether v is in the list l. l may be an array or slice.
func (ns *Namespace) In(l any, v any) (bool, error) {
if l == nil || v == nil {
return false, nil
}
lv := reflect.ValueOf(l)
vv := reflect.ValueOf(v)
vvk := normalize(vv)
switch lv.Kind() {
case reflect.Array, reflect.Slice:
for i := 0; i < lv.Len(); i++ {
lvv, isNil := indirectInterface(lv.Index(i))
if isNil {
continue
}
lvvk := normalize(lvv)
if lvvk == vvk {
return true, nil
}
}
}
ss, err := cast.ToStringE(l)
if err != nil {
return false, nil
}
su, err := cast.ToStringE(v)
if err != nil {
return false, nil
}
return strings.Contains(ss, su), nil
}
// Intersect returns the common elements in the given sets, l1 and l2. l1 and
// l2 must be of the same type and may be either arrays or slices.
func (ns *Namespace) Intersect(l1, l2 any) (any, error) {
if l1 == nil || l2 == nil {
return make([]any, 0), nil
}
var ins *intersector
l1v := reflect.ValueOf(l1)
l2v := reflect.ValueOf(l2)
switch l1v.Kind() {
case reflect.Array, reflect.Slice:
ins = &intersector{r: reflect.MakeSlice(l1v.Type(), 0, 0), seen: make(map[any]bool)}
switch l2v.Kind() {
case reflect.Array, reflect.Slice:
for i := 0; i < l1v.Len(); i++ {
l1vv := l1v.Index(i)
if !l1vv.Type().Comparable() {
return make([]any, 0), errors.New("intersect does not support slices or arrays of uncomparable types")
}
for j := 0; j < l2v.Len(); j++ {
l2vv := l2v.Index(j)
if !l2vv.Type().Comparable() {
return make([]any, 0), errors.New("intersect does not support slices or arrays of uncomparable types")
}
ins.handleValuePair(l1vv, l2vv)
}
}
return ins.r.Interface(), nil
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l2).Type().String())
}
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l1).Type().String())
}
}
// Group groups a set of items by the given key.
// This is currently only supported for Pages.
func (ns *Namespace) Group(key any, items any) (any, error) {
if key == nil {
return nil, errors.New("nil is not a valid key to group by")
}
if g, ok := items.(collections.Grouper); ok {
return g.Group(key, items)
}
in := newSliceElement(items)
if g, ok := in.(collections.Grouper); ok {
return g.Group(key, items)
}
return nil, fmt.Errorf("grouping not supported for type %T %T", items, in)
}
// IsSet returns whether a given array, channel, slice, or map in c has the given key
// defined.
func (ns *Namespace) IsSet(c any, key any) (bool, error) {
av := reflect.ValueOf(c)
kv := reflect.ValueOf(key)
switch av.Kind() {
case reflect.Array, reflect.Chan, reflect.Slice:
k, err := cast.ToIntE(key)
if err != nil {
return false, fmt.Errorf("isset unable to use key of type %T as index", key)
}
if av.Len() > k {
return true, nil
}
case reflect.Map:
if kv.Type() == av.Type().Key() {
return av.MapIndex(kv).IsValid(), nil
}
default:
ns.deps.Log.Warnf("calling IsSet with unsupported type %q (%T) will always return false.\n", av.Kind(), c)
}
return false, nil
}
// Last returns the last limit items in the list l.
func (ns *Namespace) Last(limit any, l any) (any, error) {
if limit == nil || l == nil {
return nil, errors.New("both limit and seq must be provided")
}
limitv, err := cast.ToIntE(limit)
if err != nil {
return nil, err
}
if limitv < 0 {
return nil, errors.New("sequence length must be non-negative")
}
seqv := reflect.ValueOf(l)
seqv, isNil := indirect(seqv)
if isNil {
return nil, errors.New("can't iterate over a nil value")
}
switch seqv.Kind() {
case reflect.Array, reflect.Slice, reflect.String:
// okay
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l).Type().String())
}
if limitv > seqv.Len() {
limitv = seqv.Len()
}
return seqv.Slice(seqv.Len()-limitv, seqv.Len()).Interface(), nil
}
// Querify encodes the given params in URL-encoded form ("bar=baz&foo=quux") sorted by key.
func (ns *Namespace) Querify(params ...any) (string, error) {
qs := url.Values{}
if len(params) == 1 {
switch v := params[0].(type) {
case []string:
if len(v)%2 != 0 {
return "", errors.New("invalid query")
}
for i := 0; i < len(v); i += 2 {
qs.Add(v[i], v[i+1])
}
return qs.Encode(), nil
case []any:
params = v
default:
return "", errors.New("query keys must be strings")
}
}
if len(params)%2 != 0 {
return "", errors.New("invalid query")
}
for i := 0; i < len(params); i += 2 {
switch v := params[i].(type) {
case string:
qs.Add(v, fmt.Sprintf("%v", params[i+1]))
default:
return "", errors.New("query keys must be strings")
}
}
return qs.Encode(), nil
}
// Reverse creates a copy of the list l and reverses it.
func (ns *Namespace) Reverse(l any) (any, error) {
if l == nil {
return nil, nil
}
v := reflect.ValueOf(l)
switch v.Kind() {
case reflect.Slice:
default:
return nil, errors.New("argument must be a slice")
}
sliceCopy := reflect.MakeSlice(v.Type(), v.Len(), v.Len())
for i := v.Len() - 1; i >= 0; i-- {
element := sliceCopy.Index(i)
element.Set(v.Index(v.Len() - 1 - i))
}
return sliceCopy.Interface(), nil
}
// Seq creates a sequence of integers from args. It's named and used as GNU's seq.
//
// Examples:
//
// 3 => 1, 2, 3
// 1 2 4 => 1, 3
// -3 => -1, -2, -3
// 1 4 => 1, 2, 3, 4
// 1 -2 => 1, 0, -1, -2
func (ns *Namespace) Seq(args ...any) ([]int, error) {
if len(args) < 1 || len(args) > 3 {
return nil, errors.New("invalid number of arguments to Seq")
}
intArgs := cast.ToIntSlice(args)
if len(intArgs) < 1 || len(intArgs) > 3 {
return nil, errors.New("invalid arguments to Seq")
}
inc := 1
var last int
first := intArgs[0]
if len(intArgs) == 1 {
last = first
if last == 0 {
return []int{}, nil
} else if last > 0 {
first = 1
} else {
first = -1
inc = -1
}
} else if len(intArgs) == 2 {
last = intArgs[1]
if last < first {
inc = -1
}
} else {
inc = intArgs[1]
last = intArgs[2]
if inc == 0 {
return nil, errors.New("'increment' must not be 0")
}
if first < last && inc < 0 {
return nil, errors.New("'increment' must be > 0")
}
if first > last && inc > 0 {
return nil, errors.New("'increment' must be < 0")
}
}
// sanity check
if last < -100000 {
return nil, errors.New("size of result exceeds limit")
}
size := ((last - first) / inc) + 1
// sanity check
if size <= 0 || size > 2000 {
return nil, errors.New("size of result exceeds limit")
}
seq := make([]int, size)
val := first
for i := 0; ; i++ {
seq[i] = val
val += inc
if (inc < 0 && val < last) || (inc > 0 && val > last) {
break
}
}
return seq, nil
}
// Shuffle returns list l in a randomised order.
func (ns *Namespace) Shuffle(l any) (any, error) {
if l == nil {
return nil, errors.New("both count and seq must be provided")
}
lv := reflect.ValueOf(l)
lv, isNil := indirect(lv)
if isNil {
return nil, errors.New("can't iterate over a nil value")
}
switch lv.Kind() {
case reflect.Array, reflect.Slice, reflect.String:
// okay
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l).Type().String())
}
shuffled := reflect.MakeSlice(reflect.TypeOf(l), lv.Len(), lv.Len())
randomIndices := rand.Perm(lv.Len())
for index, value := range randomIndices {
shuffled.Index(value).Set(lv.Index(index))
}
return shuffled.Interface(), nil
}
// Slice returns a slice of all passed arguments.
func (ns *Namespace) Slice(args ...any) any {
if len(args) == 0 {
return args
}
return collections.Slice(args...)
}
type intersector struct {
r reflect.Value
seen map[any]bool
}
func (i *intersector) appendIfNotSeen(v reflect.Value) {
vi := v.Interface()
if !i.seen[vi] {
i.r = reflect.Append(i.r, v)
i.seen[vi] = true
}
}
func (i *intersector) handleValuePair(l1vv, l2vv reflect.Value) {
switch kind := l1vv.Kind(); {
case kind == reflect.String:
l2t, err := toString(l2vv)
if err == nil && l1vv.String() == l2t {
i.appendIfNotSeen(l1vv)
}
case isNumber(kind):
f1, err1 := numberToFloat(l1vv)
f2, err2 := numberToFloat(l2vv)
if err1 == nil && err2 == nil && f1 == f2 {
i.appendIfNotSeen(l1vv)
}
case kind == reflect.Ptr, kind == reflect.Struct:
if l1vv.Interface() == l2vv.Interface() {
i.appendIfNotSeen(l1vv)
}
case kind == reflect.Interface:
i.handleValuePair(reflect.ValueOf(l1vv.Interface()), l2vv)
}
}
// Union returns the union of the given sets, l1 and l2. l1 and
// l2 must be of the same type and may be either arrays or slices.
// If l1 and l2 aren't of the same type then l1 will be returned.
// If either l1 or l2 is nil then the non-nil list will be returned.
func (ns *Namespace) Union(l1, l2 any) (any, error) {
if l1 == nil && l2 == nil {
return []any{}, nil
} else if l1 == nil && l2 != nil {
return l2, nil
} else if l1 != nil && l2 == nil {
return l1, nil
}
l1v := reflect.ValueOf(l1)
l2v := reflect.ValueOf(l2)
var ins *intersector
switch l1v.Kind() {
case reflect.Array, reflect.Slice:
switch l2v.Kind() {
case reflect.Array, reflect.Slice:
ins = &intersector{r: reflect.MakeSlice(l1v.Type(), 0, 0), seen: make(map[any]bool)}
if l1v.Type() != l2v.Type() &&
l1v.Type().Elem().Kind() != reflect.Interface &&
l2v.Type().Elem().Kind() != reflect.Interface {
return ins.r.Interface(), nil
}
var (
l1vv reflect.Value
isNil bool
)
for i := 0; i < l1v.Len(); i++ {
l1vv, isNil = indirectInterface(l1v.Index(i))
if !l1vv.Type().Comparable() {
return []any{}, errors.New("union does not support slices or arrays of uncomparable types")
}
if !isNil {
ins.appendIfNotSeen(l1vv)
}
}
if !l1vv.IsValid() {
// The first slice may be empty. Pick the first value of the second
// to use as a prototype.
if l2v.Len() > 0 {
l1vv = l2v.Index(0)
}
}
for j := 0; j < l2v.Len(); j++ {
l2vv := l2v.Index(j)
switch kind := l1vv.Kind(); {
case kind == reflect.String:
l2t, err := toString(l2vv)
if err == nil {
ins.appendIfNotSeen(reflect.ValueOf(l2t))
}
case isNumber(kind):
var err error
l2vv, err = convertNumber(l2vv, kind)
if err == nil {
ins.appendIfNotSeen(l2vv)
}
case kind == reflect.Interface, kind == reflect.Struct, kind == reflect.Ptr:
ins.appendIfNotSeen(l2vv)
}
}
return ins.r.Interface(), nil
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l2).Type().String())
}
default:
return nil, errors.New("can't iterate over " + reflect.ValueOf(l1).Type().String())
}
}
// Uniq returns a new list with duplicate elements in the list l removed.
func (ns *Namespace) Uniq(l any) (any, error) {
if l == nil {
return make([]any, 0), nil
}
v := reflect.ValueOf(l)
var slice reflect.Value
switch v.Kind() {
case reflect.Slice:
slice = reflect.MakeSlice(v.Type(), 0, 0)
case reflect.Array:
slice = reflect.MakeSlice(reflect.SliceOf(v.Type().Elem()), 0, 0)
default:
return nil, fmt.Errorf("type %T not supported", l)
}
seen := make(map[any]bool)
for i := 0; i < v.Len(); i++ {
ev, _ := indirectInterface(v.Index(i))
key := normalize(ev)
if _, found := seen[key]; !found {
slice = reflect.Append(slice, ev)
seen[key] = true
}
}
return slice.Interface(), nil
}
// KeyVals creates a key and values wrapper.
func (ns *Namespace) KeyVals(key any, values ...any) (types.KeyValues, error) {
return types.KeyValues{Key: key, Values: values}, nil
}
// NewScratch creates a new Scratch which can be used to store values in a
// thread safe way.
func (ns *Namespace) NewScratch() *maps.Scratch {
return maps.NewScratch()
}