[GITEA] Use restricted sanitizer for repository description

- Currently the repository description uses the same sanitizer as a
normal markdown document. This means that element such as heading and
images are allowed and can be abused.
- Create a minimal restricted sanitizer for the repository description,
which only allows what the postprocessor currently allows, which are
links and emojis.
- Added unit testing.
- Resolves https://codeberg.org/forgejo/forgejo/issues/1202
- Resolves https://codeberg.org/Codeberg/Community/issues/1122

(cherry picked from commit a8afa4cd181d7c31f73d6a8fae4c6a4b9622a425)
(cherry picked from commit 0238587c51e2c749413ca5a63e47590399fe5a2b)
(cherry picked from commit a8c7bbf728326b992e000a3d19c8833610f960c9)
(cherry picked from commit 80e05a8245092b4158c6c970ca0563181b40f2eb)
(cherry picked from commit f5af5050b34891ff16a4ef1f8e3d805fe135238d)
(cherry picked from commit 608f981e551db5f38550b622646cc307fe0566b9)
(cherry picked from commit 659186750201672391223fe4584b292e2ae26c62)
(cherry picked from commit b7e6dedafd84a7f240d7e36ce414659d3ecb7a84)
(cherry picked from commit 1cd196da49cbc352280affe8b391c70c95f64a81)
(cherry picked from commit 4c74fd4a04)

models/repo/repo.go

@@ -576,9 +576,9 @@ func (repo *Repository) DescriptionHTML(ctx context.Context) template.HTML {
 	}, repo.Description)
 	if err != nil {
 		log.Error("Failed to render description for %s (ID: %d): %v", repo.Name, repo.ID, err)
-		return template.HTML(markup.Sanitize(repo.Description))
+		return template.HTML(markup.SanitizeDescription(repo.Description))
 	}
-	return template.HTML(markup.Sanitize(desc))
+	return template.HTML(markup.SanitizeDescription(desc))
 }
 
 // CloneLink represents different types of clone URLs of repository.

modules/markup/sanitizer.go

@@ -19,6 +19,7 @@ import (
 // any modification to the underlying policies once it's been created.
 type Sanitizer struct {
 	defaultPolicy     *bluemonday.Policy
+	descriptionPolicy *bluemonday.Policy
 	rendererPolicies  map[string]*bluemonday.Policy
 	init              sync.Once
 }
@@ -41,6 +42,7 @@ func NewSanitizer() {
 func InitializeSanitizer() {
 	sanitizer.rendererPolicies = map[string]*bluemonday.Policy{}
 	sanitizer.defaultPolicy = createDefaultPolicy()
+	sanitizer.descriptionPolicy = createRepoDescriptionPolicy()
 
 	for name, renderer := range renderers {
 		sanitizerRules := renderer.SanitizerRules()
@@ -161,6 +163,27 @@ func createDefaultPolicy() *bluemonday.Policy {
 	return policy
 }
 
+// createRepoDescriptionPolicy returns a minimal more strict policy that is used for
+// repository descriptions.
+func createRepoDescriptionPolicy() *bluemonday.Policy {
+	policy := bluemonday.NewPolicy()
+
+	// Allow italics and bold.
+	policy.AllowElements("i", "b", "em", "strong")
+
+	// Allow code.
+	policy.AllowElements("code")
+
+	// Allow links
+	policy.AllowAttrs("href", "target", "rel").OnElements("a")
+
+	// Allow classes for emojis
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^emoji$`)).OnElements("img", "span")
+	policy.AllowAttrs("aria-label").OnElements("span")
+
+	return policy
+}
+
 func addSanitizerRules(policy *bluemonday.Policy, rules []setting.MarkupSanitizerRule) {
 	for _, rule := range rules {
 		if rule.AllowDataURIImages {
@@ -176,6 +199,12 @@ func addSanitizerRules(policy *bluemonday.Policy, rules []setting.MarkupSanitize
 	}
 }
 
+// SanitizeDescription sanitizes the HTML generated for a repository description.
+func SanitizeDescription(s string) string {
+	NewSanitizer()
+	return sanitizer.descriptionPolicy.Sanitize(s)
+}
+
 // Sanitize takes a string that contains a HTML fragment or document and applies policy whitelist.
 func Sanitize(s string) string {
 	NewSanitizer()

modules/markup/sanitizer_test.go

@@ -73,6 +73,28 @@ func Test_Sanitizer(t *testing.T) {
 	}
 }
 
+func TestDescriptionSanitizer(t *testing.T) {
+	NewSanitizer()
+
+	testCases := []string{
+		`<h1>Title</h1>`, `Title`,
+		`<img src='img.png' alt='image'>`, ``,
+		`<span class="emoji" aria-label="thumbs up">THUMBS UP</span>`, `<span class="emoji" aria-label="thumbs up">THUMBS UP</span>`,
+		`<span style="color: red">Hello World</span>`, `<span>Hello World</span>`,
+		`<br>`, ``,
+		`<a href="https://example.com" target="_blank" rel="noopener noreferrer">https://example.com</a>`, `<a href="https://example.com" target="_blank" rel="noopener noreferrer">https://example.com</a>`,
+		`<mark>Important!</mark>`, `Important!`,
+		`<details>Click me! <summary>Nothing to see here.</summary></details>`, `Click me! Nothing to see here.`,
+		`<input type="hidden">`, ``,
+		`<b>I</b> have a <i>strong</i> <strong>opinion</strong> about <em>this</em>.`, `<b>I</b> have a <i>strong</i> <strong>opinion</strong> about <em>this</em>.`,
+		`Provides alternative <code>wg(8)</code> tool`, `Provides alternative <code>wg(8)</code> tool`,
+	}
+
+	for i := 0; i < len(testCases); i += 2 {
+		assert.Equal(t, testCases[i+1], SanitizeDescription(testCases[i]))
+	}
+}
+
 func TestSanitizeNonEscape(t *testing.T) {
 	descStr := "<scrİpt>&lt;script&gt;alert(document.domain)&lt;/script&gt;</scrİpt>"