fix: generate IDs for HTML headings without id attribute

modules/markup/html_node.go

@@ -6,6 +6,8 @@ package markup
 import (
 	"strings"
 
+	"code.gitea.io/gitea/modules/markup/common"
+
 	"golang.org/x/net/html"
 )
 
@@ -23,16 +25,56 @@ func isAnchorHrefFootnote(s string) bool {
 	return strings.HasPrefix(s, "#fnref:user-content-") || strings.HasPrefix(s, "#fn:user-content-")
 }
 
+// isHeadingTag returns true if the node is a heading tag (h1-h6)
+func isHeadingTag(node *html.Node) bool {
+	return node.Type == html.ElementNode &&
+		len(node.Data) == 2 &&
+		node.Data[0] == 'h' &&
+		node.Data[1] >= '1' && node.Data[1] <= '6'
+}
+
+// getNodeText extracts the text content from a node and its children
+func getNodeText(node *html.Node) string {
+	var text strings.Builder
+	var extractText func(*html.Node)
+	extractText = func(n *html.Node) {
+		if n.Type == html.TextNode {
+			text.WriteString(n.Data)
+		}
+		for c := n.FirstChild; c != nil; c = c.NextSibling {
+			extractText(c)
+		}
+	}
+	extractText(node)
+	return text.String()
+}
+
 func processNodeAttrID(node *html.Node) {
 	// Add user-content- to IDs and "#" links if they don't already have them,
 	// and convert the link href to a relative link to the host root
+	hasID := false
 	for idx, attr := range node.Attr {
 		if attr.Key == "id" {
+			hasID = true
 			if !isAnchorIDUserContent(attr.Val) {
 				node.Attr[idx].Val = "user-content-" + attr.Val
 			}
 		}
 	}
+
+	// For heading tags (h1-h6) without an id attribute, generate one from the text content
+	// This ensures HTML headings like <h1>Title</h1> get proper permalink anchors
+	// matching the behavior of Markdown headings
+	if !hasID && isHeadingTag(node) {
+		text := getNodeText(node)
+		if text != "" {
+			// Use the same CleanValue function used by Markdown heading ID generation
+			cleanedID := string(common.CleanValue([]byte(text)))
+			if cleanedID != "" {
+				node.Attr = append(node.Attr, html.Attribute{Key: "id", Val: "user-content-" + cleanedID})
+			}
+		}
+	}
 }
 
 func processFootnoteNode(ctx *RenderContext, node *html.Node) {

modules/markup/html_node_test.go

@@ -0,0 +1,65 @@
+// Copyright 2024 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package markup
+
+import (
+	"strings"
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestProcessNodeAttrID_HTMLHeadingWithoutID(t *testing.T) {
+	// Test that HTML headings without id get an auto-generated id from their text content
+	testCases := []struct {
+		name     string
+		input    string
+		expected string
+	}{
+		{
+			name:     "h1 without id",
+			input:    `<h1>Heading without ID</h1>`,
+			expected: `<h1 id="user-content-heading-without-id">Heading without ID</h1>`,
+		},
+		{
+			name:     "h2 without id",
+			input:    `<h2>Another Heading</h2>`,
+			expected: `<h2 id="user-content-another-heading">Another Heading</h2>`,
+		},
+		{
+			name:     "h3 without id",
+			input:    `<h3>Third Level</h3>`,
+			expected: `<h3 id="user-content-third-level">Third Level</h3>`,
+		},
+		{
+			name:     "h1 with existing id should keep it",
+			input:    `<h1 id="my-custom-id">Heading with ID</h1>`,
+			expected: `<h1 id="user-content-my-custom-id">Heading with ID</h1>`,
+		},
+		{
+			name:     "h1 with user-content prefix should not double prefix",
+			input:    `<h1 id="user-content-already-prefixed">Already Prefixed</h1>`,
+			expected: `<h1 id="user-content-already-prefixed">Already Prefixed</h1>`,
+		},
+		{
+			name:     "heading with special characters",
+			input:    `<h1>What is Wine Staging?</h1>`,
+			expected: `<h1 id="user-content-what-is-wine-staging">What is Wine Staging?</h1>`,
+		},
+		{
+			name:     "heading with nested elements",
+			input:    `<h2><strong>Bold</strong> and <em>Italic</em></h2>`,
+			expected: `<h2 id="user-content-bold-and-italic"><strong>Bold</strong> and <em>Italic</em></h2>`,
+		},
+	}
+
+	for _, tc := range testCases {
+		t.Run(tc.name, func(t *testing.T) {
+			var result strings.Builder
+			err := PostProcessDefault(NewTestRenderContext(), strings.NewReader(tc.input), &result)
+			assert.NoError(t, err)
+			assert.Equal(t, tc.expected, strings.TrimSpace(result.String()))
+		})
+	}
+}