Split sanitizer functions and fine-tune some tests (#31192) (#31200)

Backport #31192 by wxiaoguang

Co-authored-by: wxiaoguang <wxiaoguang@gmail.com>

modules/markup/html_test.go

@@ -156,13 +156,18 @@ func TestRender_links(t *testing.T) {
 		assert.NoError(t, err)
 		assert.Equal(t, strings.TrimSpace(expected), strings.TrimSpace(buffer))
 	}
-	// Text that should be turned into URL
 
-	defaultCustom := setting.Markdown.CustomURLSchemes
+	oldCustomURLSchemes := setting.Markdown.CustomURLSchemes
+	markup.ResetDefaultSanitizerForTesting()
+	defer func() {
+		setting.Markdown.CustomURLSchemes = oldCustomURLSchemes
+		markup.ResetDefaultSanitizerForTesting()
+		markup.CustomLinkURLSchemes(oldCustomURLSchemes)
+	}()
 	setting.Markdown.CustomURLSchemes = []string{"ftp", "magnet"}
-	markup.InitializeSanitizer()
 	markup.CustomLinkURLSchemes(setting.Markdown.CustomURLSchemes)
 
+	// Text that should be turned into URL
 	test(
 		"https://www.example.com",
 		`<p><a href="https://www.example.com" rel="nofollow">https://www.example.com</a></p>`)
@@ -246,11 +251,6 @@ func TestRender_links(t *testing.T) {
 	test(
 		"ftps://gitea.com",
 		`<p>ftps://gitea.com</p>`)
-
-	// Restore previous settings
-	setting.Markdown.CustomURLSchemes = defaultCustom
-	markup.InitializeSanitizer()
-	markup.CustomLinkURLSchemes(setting.Markdown.CustomURLSchemes)
 }
 
 func TestRender_email(t *testing.T) {

modules/markup/renderer.go

@@ -46,7 +46,6 @@ func Init(ph *ProcessorHelper) {
 		DefaultProcessorHelper = *ph
 	}
 
-	NewSanitizer()
 	if len(setting.Markdown.CustomURLSchemes) > 0 {
 		CustomLinkURLSchemes(setting.Markdown.CustomURLSchemes)
 	}

modules/markup/sanitizer.go

@@ -5,13 +5,9 @@
 package markup
 
 import (
-	"io"
-	"net/url"
 	"regexp"
 	"sync"
 
-	"code.gitea.io/gitea/modules/setting"
-
 	"github.com/microcosm-cc/bluemonday"
 )
 
@@ -21,211 +17,35 @@ type Sanitizer struct {
 	defaultPolicy     *bluemonday.Policy
 	descriptionPolicy *bluemonday.Policy
 	rendererPolicies  map[string]*bluemonday.Policy
-	init              sync.Once
+	allowAllRegex     *regexp.Regexp
 }
 
 var (
-	sanitizer     = &Sanitizer{}
+	defaultSanitizer     *Sanitizer
-	allowAllRegex = regexp.MustCompile(".+")
+	defaultSanitizerOnce sync.Once
 )
 
-// NewSanitizer initializes sanitizer with allowed attributes based on settings.
+func GetDefaultSanitizer() *Sanitizer {
-// Multiple calls to this function will only create one instance of Sanitizer during
+	defaultSanitizerOnce.Do(func() {
-// entire application lifecycle.
+		defaultSanitizer = &Sanitizer{
-func NewSanitizer() {
+			rendererPolicies: map[string]*bluemonday.Policy{},
-	sanitizer.init.Do(func() {
+			allowAllRegex:    regexp.MustCompile(".+"),
-		InitializeSanitizer()
-	})
 		}
-
-// InitializeSanitizer (re)initializes the current sanitizer to account for changes in settings
-func InitializeSanitizer() {
-	sanitizer.rendererPolicies = map[string]*bluemonday.Policy{}
-	sanitizer.defaultPolicy = createDefaultPolicy()
-	sanitizer.descriptionPolicy = createRepoDescriptionPolicy()
-
 		for name, renderer := range renderers {
 			sanitizerRules := renderer.SanitizerRules()
 			if len(sanitizerRules) > 0 {
-			policy := createDefaultPolicy()
+				policy := defaultSanitizer.createDefaultPolicy()
-			addSanitizerRules(policy, sanitizerRules)
+				defaultSanitizer.addSanitizerRules(policy, sanitizerRules)
-			sanitizer.rendererPolicies[name] = policy
+				defaultSanitizer.rendererPolicies[name] = policy
 			}
 		}
+		defaultSanitizer.defaultPolicy = defaultSanitizer.createDefaultPolicy()
+		defaultSanitizer.descriptionPolicy = defaultSanitizer.createRepoDescriptionPolicy()
+	})
+	return defaultSanitizer
 }
 
-func createDefaultPolicy() *bluemonday.Policy {
+func ResetDefaultSanitizerForTesting() {
-	policy := bluemonday.UGCPolicy()
+	defaultSanitizer = nil
-
+	defaultSanitizerOnce = sync.Once{}
-	// For JS code copy and Mermaid loading state
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-block( is-loading)?$`)).OnElements("pre")
-
-	// For code preview
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-preview-[-\w]+( file-content)?$`)).Globally()
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-num$`)).OnElements("td")
-	policy.AllowAttrs("data-line-number").OnElements("span")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-code chroma$`)).OnElements("td")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-inner$`)).OnElements("div")
-
-	// For code preview (unicode escape)
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^file-view( unicode-escaped)?$`)).OnElements("table")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-escape$`)).OnElements("td")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^toggle-escape-button btn interact-bg$`)).OnElements("a") // don't use button, button might submit a form
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(ambiguous-code-point|escaped-code-point|broken-code-point)$`)).OnElements("span")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^char$`)).OnElements("span")
-	policy.AllowAttrs("data-tooltip-content", "data-escaped").OnElements("span")
-
-	// For color preview
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^color-preview$`)).OnElements("span")
-
-	// For attention
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-header attention-\w+$`)).OnElements("blockquote")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-\w+$`)).OnElements("strong")
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-icon attention-\w+ svg octicon-[\w-]+$`)).OnElements("svg")
-	policy.AllowAttrs("viewBox", "width", "height", "aria-hidden").OnElements("svg")
-	policy.AllowAttrs("fill-rule", "d").OnElements("path")
-
-	// For Chroma markdown plugin
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(chroma )?language-[\w-]+( display)?( is-loading)?$`)).OnElements("code")
-
-	// Checkboxes
-	policy.AllowAttrs("type").Matching(regexp.MustCompile(`^checkbox$`)).OnElements("input")
-	policy.AllowAttrs("checked", "disabled", "data-source-position").OnElements("input")
-
-	// Custom URL-Schemes
-	if len(setting.Markdown.CustomURLSchemes) > 0 {
-		policy.AllowURLSchemes(setting.Markdown.CustomURLSchemes...)
-	} else {
-		policy.AllowURLSchemesMatching(allowAllRegex)
-
-		// Even if every scheme is allowed, these three are blocked for security reasons
-		disallowScheme := func(*url.URL) bool {
-			return false
-		}
-		policy.AllowURLSchemeWithCustomPolicy("javascript", disallowScheme)
-		policy.AllowURLSchemeWithCustomPolicy("vbscript", disallowScheme)
-		policy.AllowURLSchemeWithCustomPolicy("data", disallowScheme)
-	}
-
-	// Allow classes for anchors
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`ref-issue( ref-external-issue)?`)).OnElements("a")
-
-	// Allow classes for task lists
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`task-list-item`)).OnElements("li")
-
-	// Allow classes for org mode list item status.
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(unchecked|checked|indeterminate)$`)).OnElements("li")
-
-	// Allow icons
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^icon(\s+[\p{L}\p{N}_-]+)+$`)).OnElements("i")
-
-	// Allow classes for emojis
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`emoji`)).OnElements("img")
-
-	// Allow icons, emojis, chroma syntax and keyword markup on span
-	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^((icon(\s+[\p{L}\p{N}_-]+)+)|(emoji)|(language-math display)|(language-math inline))$|^([a-z][a-z0-9]{0,2})$|^` + keywordClass + `$`)).OnElements("span")
-
-	// Allow 'color' and 'background-color' properties for the style attribute on text elements.
-	policy.AllowStyles("color", "background-color").OnElements("span", "p")
-
-	// Allow generally safe attributes
-	generalSafeAttrs := []string{
-		"abbr", "accept", "accept-charset",
-		"accesskey", "action", "align", "alt",
-		"aria-describedby", "aria-hidden", "aria-label", "aria-labelledby",
-		"axis", "border", "cellpadding", "cellspacing", "char",
-		"charoff", "charset", "checked",
-		"clear", "cols", "colspan", "color",
-		"compact", "coords", "datetime", "dir",
-		"disabled", "enctype", "for", "frame",
-		"headers", "height", "hreflang",
-		"hspace", "ismap", "label", "lang",
-		"maxlength", "media", "method",
-		"multiple", "name", "nohref", "noshade",
-		"nowrap", "open", "prompt", "readonly", "rel", "rev",
-		"rows", "rowspan", "rules", "scope",
-		"selected", "shape", "size", "span",
-		"start", "summary", "tabindex", "target",
-		"title", "type", "usemap", "valign", "value",
-		"vspace", "width", "itemprop",
-	}
-
-	generalSafeElements := []string{
-		"h1", "h2", "h3", "h4", "h5", "h6", "h7", "h8", "br", "b", "i", "strong", "em", "a", "pre", "code", "img", "tt",
-		"div", "ins", "del", "sup", "sub", "p", "ol", "ul", "table", "thead", "tbody", "tfoot", "blockquote", "label",
-		"dl", "dt", "dd", "kbd", "q", "samp", "var", "hr", "ruby", "rt", "rp", "li", "tr", "td", "th", "s", "strike", "summary",
-		"details", "caption", "figure", "figcaption",
-		"abbr", "bdo", "cite", "dfn", "mark", "small", "span", "time", "video", "wbr",
-	}
-
-	policy.AllowAttrs(generalSafeAttrs...).OnElements(generalSafeElements...)
-
-	policy.AllowAttrs("src", "autoplay", "controls").OnElements("video")
-
-	policy.AllowAttrs("itemscope", "itemtype").OnElements("div")
-
-	// FIXME: Need to handle longdesc in img but there is no easy way to do it
-
-	// Custom keyword markup
-	addSanitizerRules(policy, setting.ExternalSanitizerRules)
-
-	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 {
-			policy.AllowDataURIImages()
-		}
-		if rule.Element != "" {
-			if rule.Regexp != nil {
-				policy.AllowAttrs(rule.AllowAttr).Matching(rule.Regexp).OnElements(rule.Element)
-			} else {
-				policy.AllowAttrs(rule.AllowAttr).OnElements(rule.Element)
-			}
-		}
-	}
-}
-
-// 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()
-	return sanitizer.defaultPolicy.Sanitize(s)
-}
-
-// SanitizeReader sanitizes a Reader
-func SanitizeReader(r io.Reader, renderer string, w io.Writer) error {
-	NewSanitizer()
-	policy, exist := sanitizer.rendererPolicies[renderer]
-	if !exist {
-		policy = sanitizer.defaultPolicy
-	}
-	return policy.SanitizeReaderToWriter(r, w)
 }

modules/markup/sanitizer_custom.go

@@ -0,0 +1,25 @@
+// Copyright 2024 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package markup
+
+import (
+	"code.gitea.io/gitea/modules/setting"
+
+	"github.com/microcosm-cc/bluemonday"
+)
+
+func (st *Sanitizer) addSanitizerRules(policy *bluemonday.Policy, rules []setting.MarkupSanitizerRule) {
+	for _, rule := range rules {
+		if rule.AllowDataURIImages {
+			policy.AllowDataURIImages()
+		}
+		if rule.Element != "" {
+			if rule.Regexp != nil {
+				policy.AllowAttrs(rule.AllowAttr).Matching(rule.Regexp).OnElements(rule.Element)
+			} else {
+				policy.AllowAttrs(rule.AllowAttr).OnElements(rule.Element)
+			}
+		}
+	}
+}

modules/markup/sanitizer_default.go

@@ -0,0 +1,146 @@
+// Copyright 2024 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package markup
+
+import (
+	"io"
+	"net/url"
+	"regexp"
+
+	"code.gitea.io/gitea/modules/setting"
+
+	"github.com/microcosm-cc/bluemonday"
+)
+
+func (st *Sanitizer) createDefaultPolicy() *bluemonday.Policy {
+	policy := bluemonday.UGCPolicy()
+
+	// For JS code copy and Mermaid loading state
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-block( is-loading)?$`)).OnElements("pre")
+
+	// For code preview
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-preview-[-\w]+( file-content)?$`)).Globally()
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-num$`)).OnElements("td")
+	policy.AllowAttrs("data-line-number").OnElements("span")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-code chroma$`)).OnElements("td")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^code-inner$`)).OnElements("div")
+
+	// For code preview (unicode escape)
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^file-view( unicode-escaped)?$`)).OnElements("table")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^lines-escape$`)).OnElements("td")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^toggle-escape-button btn interact-bg$`)).OnElements("a") // don't use button, button might submit a form
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(ambiguous-code-point|escaped-code-point|broken-code-point)$`)).OnElements("span")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^char$`)).OnElements("span")
+	policy.AllowAttrs("data-tooltip-content", "data-escaped").OnElements("span")
+
+	// For color preview
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^color-preview$`)).OnElements("span")
+
+	// For attention
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-header attention-\w+$`)).OnElements("blockquote")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-\w+$`)).OnElements("strong")
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^attention-icon attention-\w+ svg octicon-[\w-]+$`)).OnElements("svg")
+	policy.AllowAttrs("viewBox", "width", "height", "aria-hidden").OnElements("svg")
+	policy.AllowAttrs("fill-rule", "d").OnElements("path")
+
+	// For Chroma markdown plugin
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(chroma )?language-[\w-]+( display)?( is-loading)?$`)).OnElements("code")
+
+	// Checkboxes
+	policy.AllowAttrs("type").Matching(regexp.MustCompile(`^checkbox$`)).OnElements("input")
+	policy.AllowAttrs("checked", "disabled", "data-source-position").OnElements("input")
+
+	// Custom URL-Schemes
+	if len(setting.Markdown.CustomURLSchemes) > 0 {
+		policy.AllowURLSchemes(setting.Markdown.CustomURLSchemes...)
+	} else {
+		policy.AllowURLSchemesMatching(st.allowAllRegex)
+
+		// Even if every scheme is allowed, these three are blocked for security reasons
+		disallowScheme := func(*url.URL) bool {
+			return false
+		}
+		policy.AllowURLSchemeWithCustomPolicy("javascript", disallowScheme)
+		policy.AllowURLSchemeWithCustomPolicy("vbscript", disallowScheme)
+		policy.AllowURLSchemeWithCustomPolicy("data", disallowScheme)
+	}
+
+	// Allow classes for anchors
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`ref-issue( ref-external-issue)?`)).OnElements("a")
+
+	// Allow classes for task lists
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`task-list-item`)).OnElements("li")
+
+	// Allow classes for org mode list item status.
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^(unchecked|checked|indeterminate)$`)).OnElements("li")
+
+	// Allow icons
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^icon(\s+[\p{L}\p{N}_-]+)+$`)).OnElements("i")
+
+	// Allow classes for emojis
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`emoji`)).OnElements("img")
+
+	// Allow icons, emojis, chroma syntax and keyword markup on span
+	policy.AllowAttrs("class").Matching(regexp.MustCompile(`^((icon(\s+[\p{L}\p{N}_-]+)+)|(emoji)|(language-math display)|(language-math inline))$|^([a-z][a-z0-9]{0,2})$|^` + keywordClass + `$`)).OnElements("span")
+
+	// Allow 'color' and 'background-color' properties for the style attribute on text elements.
+	policy.AllowStyles("color", "background-color").OnElements("span", "p")
+
+	// Allow generally safe attributes
+	generalSafeAttrs := []string{
+		"abbr", "accept", "accept-charset",
+		"accesskey", "action", "align", "alt",
+		"aria-describedby", "aria-hidden", "aria-label", "aria-labelledby",
+		"axis", "border", "cellpadding", "cellspacing", "char",
+		"charoff", "charset", "checked",
+		"clear", "cols", "colspan", "color",
+		"compact", "coords", "datetime", "dir",
+		"disabled", "enctype", "for", "frame",
+		"headers", "height", "hreflang",
+		"hspace", "ismap", "label", "lang",
+		"maxlength", "media", "method",
+		"multiple", "name", "nohref", "noshade",
+		"nowrap", "open", "prompt", "readonly", "rel", "rev",
+		"rows", "rowspan", "rules", "scope",
+		"selected", "shape", "size", "span",
+		"start", "summary", "tabindex", "target",
+		"title", "type", "usemap", "valign", "value",
+		"vspace", "width", "itemprop",
+	}
+
+	generalSafeElements := []string{
+		"h1", "h2", "h3", "h4", "h5", "h6", "h7", "h8", "br", "b", "i", "strong", "em", "a", "pre", "code", "img", "tt",
+		"div", "ins", "del", "sup", "sub", "p", "ol", "ul", "table", "thead", "tbody", "tfoot", "blockquote", "label",
+		"dl", "dt", "dd", "kbd", "q", "samp", "var", "hr", "ruby", "rt", "rp", "li", "tr", "td", "th", "s", "strike", "summary",
+		"details", "caption", "figure", "figcaption",
+		"abbr", "bdo", "cite", "dfn", "mark", "small", "span", "time", "video", "wbr",
+	}
+
+	policy.AllowAttrs(generalSafeAttrs...).OnElements(generalSafeElements...)
+
+	policy.AllowAttrs("src", "autoplay", "controls").OnElements("video")
+
+	policy.AllowAttrs("itemscope", "itemtype").OnElements("div")
+
+	// FIXME: Need to handle longdesc in img but there is no easy way to do it
+
+	// Custom keyword markup
+	defaultSanitizer.addSanitizerRules(policy, setting.ExternalSanitizerRules)
+
+	return policy
+}
+
+// Sanitize takes a string that contains a HTML fragment or document and applies policy whitelist.
+func Sanitize(s string) string {
+	return GetDefaultSanitizer().defaultPolicy.Sanitize(s)
+}
+
+// SanitizeReader sanitizes a Reader
+func SanitizeReader(r io.Reader, renderer string, w io.Writer) error {
+	policy, exist := GetDefaultSanitizer().rendererPolicies[renderer]
+	if !exist {
+		policy = GetDefaultSanitizer().defaultPolicy
+	}
+	return policy.SanitizeReaderToWriter(r, w)
+}

modules/markup/sanitizer_default_test.go

@@ -5,18 +5,16 @@
 package markup
 
 import (
-	"html/template"
-	"strings"
 	"testing"
 
 	"github.com/stretchr/testify/assert"
 )
 
-func Test_Sanitizer(t *testing.T) {
+func TestSanitizer(t *testing.T) {
-	NewSanitizer()
 	testCases := []string{
 		// Regular
 		`<a onblur="alert(secret)" href="http://www.google.com">Google</a>`, `<a href="http://www.google.com" rel="nofollow">Google</a>`,
+		"<scrİpt>&lt;script&gt;alert(document.domain)&lt;/script&gt;</scrİpt>", "&lt;script&gt;alert(document.domain)&lt;/script&gt;",
 
 		// Code highlighting class
 		`<code class="random string"></code>`, `<code></code>`,
@@ -72,34 +70,3 @@ func Test_Sanitizer(t *testing.T) {
 		assert.Equal(t, testCases[i+1], Sanitize(testCases[i]))
 	}
 }
-
-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>"
-
-	output := template.HTML(Sanitize(descStr))
-	if strings.Contains(string(output), "<script>") {
-		t.Errorf("un-escaped <script> in output: %q", output)
-	}
-}

modules/markup/sanitizer_description.go

@@ -0,0 +1,37 @@
+// Copyright 2024 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package markup
+
+import (
+	"regexp"
+
+	"github.com/microcosm-cc/bluemonday"
+)
+
+// createRepoDescriptionPolicy returns a minimal more strict policy that is used for
+// repository descriptions.
+func (st *Sanitizer) createRepoDescriptionPolicy() *bluemonday.Policy {
+	policy := bluemonday.NewPolicy()
+	policy.AllowStandardURLs()
+
+	// 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
+}
+
+// SanitizeDescription sanitizes the HTML generated for a repository description.
+func SanitizeDescription(s string) string {
+	return GetDefaultSanitizer().descriptionPolicy.Sanitize(s)
+}

modules/markup/sanitizer_description_test.go

@@ -0,0 +1,31 @@
+// Copyright 2024 The Gitea Authors. All rights reserved.
+// SPDX-License-Identifier: MIT
+
+package markup
+
+import (
+	"testing"
+
+	"github.com/stretchr/testify/assert"
+)
+
+func TestDescriptionSanitizer(t *testing.T) {
+	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 nofollow">https://example.com</a>`,
+		`<a href="data:1234">data</a>`, `data`,
+		`<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]))
+	}
+}