Add Page.Contents with scope support

Note that this also adds a new `.ContentWithoutSummary` method, and to do that we had to unify the different summary types:

Both `auto` and `manual` now returns HTML. Before this commit, `auto` would return plain text. This could be considered to be a slightly breaking change, but for the better: Now you can treat the `.Summary` the same without thinking about where it comes from, and if you want plain text, pipe it into `{{ .Summary | plainify }}`.

Fixes #8680
Fixes #12761
Fixes #12778
Fixes #716

common/hugo/hugo.go

@@ -14,6 +14,7 @@
 package hugo
 
 import (
+	"context"
 	"fmt"
 	"html/template"
 	"os"
@@ -29,6 +30,7 @@ import (
 	"github.com/mitchellh/mapstructure"
 
 	"github.com/bep/godartsass/v2"
+	"github.com/gohugoio/hugo/common/hcontext"
 	"github.com/gohugoio/hugo/common/hexec"
 	"github.com/gohugoio/hugo/common/loggers"
 	"github.com/gohugoio/hugo/hugofs/files"
@@ -69,6 +71,9 @@ type HugoInfo struct {
 
 	conf ConfigProvider
 	deps []*Dependency
+
+	// Context gives access to some of the context scoped variables.
+	Context Context
 }
 
 // Version returns the current version as a comparable version string.
@@ -127,6 +132,26 @@ func (i HugoInfo) IsMultilingual() bool {
 	return i.conf.IsMultilingual()
 }
 
+type contextKey string
+
+var markupScope = hcontext.NewContextDispatcher[string](contextKey("markupScope"))
+
+type Context struct{}
+
+func (c Context) MarkupScope(ctx context.Context) string {
+	return GetMarkupScope(ctx)
+}
+
+// SetMarkupScope sets the markup scope in the context.
+func SetMarkupScope(ctx context.Context, s string) context.Context {
+	return markupScope.Set(ctx, s)
+}
+
+// GetMarkupScope gets the markup scope from the context.
+func GetMarkupScope(ctx context.Context) string {
+	return markupScope.Get(ctx)
+}
+
 // ConfigProvider represents the config options that are relevant for HugoInfo.
 type ConfigProvider interface {
 	Environment() string

common/hugo/hugo_test.go

@@ -14,6 +14,7 @@
 package hugo
 
 import (
+	"context"
 	"fmt"
 	"testing"
 
@@ -64,6 +65,19 @@ func TestDeprecationLogLevelFromVersion(t *testing.T) {
 	c.Assert(deprecationLogLevelFromVersion(ver.String()), qt.Equals, logg.LevelError)
 }
 
+func TestMarkupScope(t *testing.T) {
+	c := qt.New(t)
+
+	conf := testConfig{environment: "production", workingDir: "/mywork", running: false}
+	info := NewInfo(conf, nil)
+
+	ctx := context.Background()
+
+	ctx = SetMarkupScope(ctx, "foo")
+
+	c.Assert(info.Context.MarkupScope(ctx), qt.Equals, "foo")
+}
+
 type testConfig struct {
 	environment  string
 	running      bool

common/paths/pathparser.go

@@ -153,7 +153,7 @@ func (pp *PathParser) doParse(component, s string, p *Path) (*Path, error) {
 				} else {
 					high = len(p.s)
 				}
-				id := types.LowHigh{Low: i + 1, High: high}
+				id := types.LowHigh[string]{Low: i + 1, High: high}
 				if len(p.identifiers) == 0 {
 					p.identifiers = append(p.identifiers, id)
 				} else if len(p.identifiers) == 1 {
@@ -260,7 +260,7 @@ type Path struct {
 	component  string
 	bundleType PathType
 
-	identifiers []types.LowHigh
+	identifiers []types.LowHigh[string]
 
 	posIdentifierLanguage int
 	disabled              bool

common/types/types.go

@@ -107,12 +107,20 @@ func Unwrapv(v any) any {
 	return v
 }
 
-// LowHigh is typically used to represent a slice boundary.
-type LowHigh struct {
+// LowHigh represents a byte or slice boundary.
+type LowHigh[S ~[]byte | string] struct {
 	Low  int
 	High int
 }
 
+func (l LowHigh[S]) IsZero() bool {
+	return l.Low < 0 || (l.Low == 0 && l.High == 0)
+}
+
+func (l LowHigh[S]) Value(source S) S {
+	return source[l.Low:l.High]
+}
+
 // This is only used for debugging purposes.
 var InvocationCounter atomic.Int64
 

common/types/types_test.go

@@ -27,3 +27,25 @@ func TestKeyValues(t *testing.T) {
 	c.Assert(kv.KeyString(), qt.Equals, "key")
 	c.Assert(kv.Values, qt.DeepEquals, []any{"a1", "a2"})
 }
+
+func TestLowHigh(t *testing.T) {
+	c := qt.New(t)
+
+	lh := LowHigh[string]{
+		Low:  2,
+		High: 10,
+	}
+
+	s := "abcdefghijklmnopqrstuvwxyz"
+	c.Assert(lh.IsZero(), qt.IsFalse)
+	c.Assert(lh.Value(s), qt.Equals, "cdefghij")
+
+	lhb := LowHigh[[]byte]{
+		Low:  2,
+		High: 10,
+	}
+
+	sb := []byte(s)
+	c.Assert(lhb.IsZero(), qt.IsFalse)
+	c.Assert(lhb.Value(sb), qt.DeepEquals, []byte("cdefghij"))
+}

helpers/content.go

@@ -22,7 +22,6 @@ import (
 	"html/template"
 	"strings"
 	"unicode"
-	"unicode/utf8"
 
 	"github.com/gohugoio/hugo/common/hexec"
 	"github.com/gohugoio/hugo/common/loggers"
@@ -165,75 +164,6 @@ func TotalWords(s string) int {
 	return n
 }
 
-// TruncateWordsByRune truncates words by runes.
-func (c *ContentSpec) TruncateWordsByRune(in []string) (string, bool) {
-	words := make([]string, len(in))
-	copy(words, in)
-
-	count := 0
-	for index, word := range words {
-		if count >= c.Cfg.SummaryLength() {
-			return strings.Join(words[:index], " "), true
-		}
-		runeCount := utf8.RuneCountInString(word)
-		if len(word) == runeCount {
-			count++
-		} else if count+runeCount < c.Cfg.SummaryLength() {
-			count += runeCount
-		} else {
-			for ri := range word {
-				if count >= c.Cfg.SummaryLength() {
-					truncatedWords := append(words[:index], word[:ri])
-					return strings.Join(truncatedWords, " "), true
-				}
-				count++
-			}
-		}
-	}
-
-	return strings.Join(words, " "), false
-}
-
-// TruncateWordsToWholeSentence takes content and truncates to whole sentence
-// limited by max number of words. It also returns whether it is truncated.
-func (c *ContentSpec) TruncateWordsToWholeSentence(s string) (string, bool) {
-	var (
-		wordCount     = 0
-		lastWordIndex = -1
-	)
-
-	for i, r := range s {
-		if unicode.IsSpace(r) {
-			wordCount++
-			lastWordIndex = i
-
-			if wordCount >= c.Cfg.SummaryLength() {
-				break
-			}
-
-		}
-	}
-
-	if lastWordIndex == -1 {
-		return s, false
-	}
-
-	endIndex := -1
-
-	for j, r := range s[lastWordIndex:] {
-		if isEndOfSentence(r) {
-			endIndex = j + lastWordIndex + utf8.RuneLen(r)
-			break
-		}
-	}
-
-	if endIndex == -1 {
-		return s, false
-	}
-
-	return strings.TrimSpace(s[:endIndex]), endIndex < len(s)
-}
-
 // TrimShortHTML removes the outer tags from HTML input where (a) the opening
 // tag is present only once with the input, and (b) the opening and closing
 // tags wrap the input after white space removal.
@@ -256,7 +186,3 @@ func (c *ContentSpec) TrimShortHTML(input []byte, markup string) []byte {
 	}
 	return input
 }
-
-func isEndOfSentence(r rune) bool {
-	return r == '.' || r == '?' || r == '!' || r == '"' || r == '\n'
-}

helpers/content_test.go

@@ -20,7 +20,6 @@ import (
 	"testing"
 
 	qt "github.com/frankban/quicktest"
-	"github.com/gohugoio/hugo/config"
 	"github.com/gohugoio/hugo/helpers"
 )
 
@@ -66,84 +65,6 @@ func TestBytesToHTML(t *testing.T) {
 	c.Assert(helpers.BytesToHTML([]byte("dobedobedo")), qt.Equals, template.HTML("dobedobedo"))
 }
 
-var benchmarkTruncateString = strings.Repeat("This is a sentence about nothing.", 20)
-
-func BenchmarkTestTruncateWordsToWholeSentence(b *testing.B) {
-	c := newTestContentSpec(nil)
-	b.ResetTimer()
-	for i := 0; i < b.N; i++ {
-		c.TruncateWordsToWholeSentence(benchmarkTruncateString)
-	}
-}
-
-func TestTruncateWordsToWholeSentence(t *testing.T) {
-	type test struct {
-		input, expected string
-		max             int
-		truncated       bool
-	}
-	data := []test{
-		{"a b c", "a b c", 12, false},
-		{"a b c", "a b c", 3, false},
-		{"a", "a", 1, false},
-		{"This is a sentence.", "This is a sentence.", 5, false},
-		{"This is also a sentence!", "This is also a sentence!", 1, false},
-		{"To be. Or not to be. That's the question.", "To be.", 1, true},
-		{" \nThis is not a sentence\nAnd this is another", "This is not a sentence", 4, true},
-		{"", "", 10, false},
-		{"This... is a more difficult test?", "This... is a more difficult test?", 1, false},
-	}
-	for i, d := range data {
-		cfg := config.New()
-		cfg.Set("summaryLength", d.max)
-		c := newTestContentSpec(cfg)
-		output, truncated := c.TruncateWordsToWholeSentence(d.input)
-		if d.expected != output {
-			t.Errorf("Test %d failed. Expected %q got %q", i, d.expected, output)
-		}
-
-		if d.truncated != truncated {
-			t.Errorf("Test %d failed. Expected truncated=%t got %t", i, d.truncated, truncated)
-		}
-	}
-}
-
-func TestTruncateWordsByRune(t *testing.T) {
-	type test struct {
-		input, expected string
-		max             int
-		truncated       bool
-	}
-	data := []test{
-		{"", "", 1, false},
-		{"a b c", "a b c", 12, false},
-		{"a b c", "a b c", 3, false},
-		{"a", "a", 1, false},
-		{"Hello 中国", "", 0, true},
-		{"这是中文，全中文。", "这是中文，", 5, true},
-		{"Hello 中国", "Hello 中", 2, true},
-		{"Hello 中国", "Hello 中国", 3, false},
-		{"Hello中国 Good 好的", "Hello中国 Good 好", 9, true},
-		{"This is a sentence.", "This is", 2, true},
-		{"This is also a sentence!", "This", 1, true},
-		{"To be. Or not to be. That's the question.", "To be. Or not", 4, true},
-		{" \nThis is    not a sentence\n ", "This is not", 3, true},
-	}
-	for i, d := range data {
-		cfg := config.New()
-		cfg.Set("summaryLength", d.max)
-		c := newTestContentSpec(cfg)
-		output, truncated := c.TruncateWordsByRune(strings.Fields(d.input))
-		if d.expected != output {
-			t.Errorf("Test %d failed. Expected %q got %q", i, d.expected, output)
-		}
-
-		if d.truncated != truncated {
-			t.Errorf("Test %d failed. Expected truncated=%t got %t", i, d.truncated, truncated)
-		}
-	}
-}
-
 func TestExtractTOCNormalContent(t *testing.T) {
 	content := []byte("<nav>\n<ul>\nTOC<li><a href=\"#")
 

hugolib/page.go

@@ -61,6 +61,7 @@ var (
 	pageTypesProvider = resource.NewResourceTypesProvider(media.Builtin.OctetType, pageResourceType)
 	nopPageOutput     = &pageOutput{
 		pagePerOutputProviders: nopPagePerOutput,
+		MarkupProvider:         page.NopPage,
 		ContentProvider:        page.NopPage,
 	}
 )
@@ -213,11 +214,8 @@ func (p *pageHeadingsFiltered) page() page.Page {
 
 // For internal use by the related content feature.
 func (p *pageState) ApplyFilterToHeadings(ctx context.Context, fn func(*tableofcontents.Heading) bool) related.Document {
-	r, err := p.m.content.contentToC(ctx, p.pageOutput.pco)
-	if err != nil {
-		panic(err)
-	}
-	headings := r.tableOfContents.Headings.FilterBy(fn)
+	fragments := p.pageOutput.pco.c().Fragments(ctx)
+	headings := fragments.Headings.FilterBy(fn)
 	return &pageHeadingsFiltered{
 		pageState: p,
 		headings:  headings,
@@ -719,6 +717,7 @@ func (p *pageState) shiftToOutputFormat(isRenderingSite bool, idx int) error {
 			})
 			p.pageOutput.contentRenderer = lcp
 			p.pageOutput.ContentProvider = lcp
+			p.pageOutput.MarkupProvider = lcp
 			p.pageOutput.PageRenderProvider = lcp
 			p.pageOutput.TableOfContentsProvider = lcp
 		}