[chore]: Bump github.com/microcosm-cc/bluemonday from 1.0.24 to 1.0.25 (#2021)

go.mod

@@ -34,7 +34,7 @@ require (
 	github.com/h2non/filetype v1.1.3
 	github.com/jackc/pgconn v1.14.1
 	github.com/jackc/pgx/v5 v5.4.2
-	github.com/microcosm-cc/bluemonday v1.0.24
+	github.com/microcosm-cc/bluemonday v1.0.25
 	github.com/miekg/dns v1.1.55
 	github.com/minio/minio-go/v7 v7.0.60
 	github.com/mitchellh/mapstructure v1.5.0

go.sum

@@ -444,8 +444,8 @@ github.com/mattn/go-isatty v0.0.14/go.mod h1:7GGIvUiUoEMVVmxf/4nioHXj79iQHKdU27k
 github.com/mattn/go-isatty v0.0.19 h1:JITubQf0MOLdlGRuRq+jtsDlekdYPia9ZFsB8h/APPA=
 github.com/mattn/go-isatty v0.0.19/go.mod h1:W+V8PltTTMOvKvAeJH7IuucS94S2C6jfK/D7dTCTo3Y=
 github.com/mattn/go-sqlite3 v2.0.3+incompatible h1:gXHsfypPkaMZrKbD5209QV9jbUTJKjyR5WD3HYQSd+U=
-github.com/microcosm-cc/bluemonday v1.0.24 h1:NGQoPtwGVcbGkKfvyYk1yRqknzBuoMiUrO6R7uFTPlw=
-github.com/microcosm-cc/bluemonday v1.0.24/go.mod h1:ArQySAMps0790cHSkdPEJ7bGkF2VePWH773hsJNSHf8=
+github.com/microcosm-cc/bluemonday v1.0.25 h1:4NEwSfiJ+Wva0VxN5B8OwMicaJvD8r9tlJWm9rtloEg=
+github.com/microcosm-cc/bluemonday v1.0.25/go.mod h1:ZIOjCQp1OrzBBPIJmfX4qDYFuhU02nx4bn030ixfHLE=
 github.com/miekg/dns v1.1.55 h1:GoQ4hpsj0nFLYe+bWiCToyrBEJXkQfOOIvFGFy0lEgo=
 github.com/miekg/dns v1.1.55/go.mod h1:uInx36IzPl7FYnDcMeVWxj9byh7DutNykX4G9Sj60FY=
 github.com/minio/md5-simd v1.1.2 h1:Gdi1DZK69+ZVMoNHRXJyNcxrMA4dSxoYHZSQbirFg34=

vendor/github.com/microcosm-cc/bluemonday/helpers.go

@@ -222,11 +222,7 @@ func (p *Policy) AllowDataURIImages() {
 			}
 
 			_, err := base64.StdEncoding.DecodeString(url.Opaque[len(matched):])
-			if err != nil {
-				return false
-			}
-
-			return true
+			return err == nil
 		},
 	)
 }

vendor/github.com/microcosm-cc/bluemonday/policy.go

@@ -118,9 +118,18 @@ type Policy struct {
 	allowURLSchemes map[string][]urlPolicy
 
 	// These regexps are used to match allowed URL schemes, for example
-	// if one would want to allow all URL schemes, they would add `.+`
+	// if one would want to allow all URL schemes, they would add `.+`.
+	// However pay attention as this can lead to XSS being rendered thus
+	// defeating the purpose of using a HTML sanitizer.
+	// The regexps are only considered if a schema was not explicitly
+	// handled by `AllowURLSchemes` or `AllowURLSchemeWithCustomPolicy`.
 	allowURLSchemeRegexps []*regexp.Regexp
 
+	// If srcRewriter is not nil, it is used to rewrite the src attribute
+	// of tags that download resources, such as <img> and <script>.
+	// It requires that the URL is parsable by "net/url" url.Parse().
+	srcRewriter urlRewriter
+
 	// If an element has had all attributes removed as a result of a policy
 	// being applied, then the element would be removed from the output.
 	//
@@ -196,6 +205,8 @@ type stylePolicyBuilder struct {
 
 type urlPolicy func(url *url.URL) (allowUrl bool)
 
+type urlRewriter func(*url.URL)
+
 type SandboxValue int64
 
 const (
@@ -575,6 +586,33 @@ func (p *Policy) AllowURLSchemesMatching(r *regexp.Regexp) *Policy {
 	return p
 }
 
+// RewriteSrc will rewrite the src attribute of a resource downloading tag
+// (e.g. <img>, <script>, <iframe>) using the provided function.
+//
+// Typically the use case here is that if the content that we're sanitizing
+// is untrusted then the content that is inlined is also untrusted.
+// To prevent serving this content on the same domain as the content appears
+// on it is good practise to proxy the content through an additional domain
+// name as this will force the web client to consider the inline content as
+// third party to the main content, thus providing browser isolation around
+// the inline content.
+//
+// An example of this is a web mail provider like fastmail.com , when an
+// email (user generated content) is displayed, the email text is shown on
+// fastmail.com but the inline attachments and content are rendered from
+// fastmailusercontent.com . This proxying of the external content on a
+// domain that is different to the content domain forces the browser domain
+// security model to kick in. Note that this only applies to differences
+// below the suffix (as per the publix suffix list).
+//
+// This is a good practise to adopt as it prevents the content from being
+// able to set cookies on the main domain and thus prevents the content on
+// the main domain from being able to read those cookies.
+func (p *Policy) RewriteSrc(fn urlRewriter) *Policy {
+	p.srcRewriter = fn
+	return p
+}
+
 // RequireNoFollowOnLinks will result in all a, area, link tags having a
 // rel="nofollow"added to them if one does not already exist
 //

vendor/github.com/microcosm-cc/bluemonday/sanitize.go

@@ -95,41 +95,6 @@ func (p *Policy) SanitizeReaderToWriter(r io.Reader, w io.Writer) error {
 	return p.sanitize(r, w)
 }
 
-const escapedURLChars = "'<>\"\r"
-
-func escapeUrlComponent(w stringWriterWriter, val string) error {
-	i := strings.IndexAny(val, escapedURLChars)
-	for i != -1 {
-		if _, err := w.WriteString(val[:i]); err != nil {
-			return err
-		}
-		var esc string
-		switch val[i] {
-		case '\'':
-			// "&#39;" is shorter than "&apos;" and apos was not in HTML until HTML5.
-			esc = "&#39;"
-		case '<':
-			esc = "&lt;"
-		case '>':
-			esc = "&gt;"
-		case '"':
-			// "&#34;" is shorter than "&quot;".
-			esc = "&#34;"
-		case '\r':
-			esc = "&#13;"
-		default:
-			panic("unrecognized escape character")
-		}
-		val = val[i+1:]
-		if _, err := w.WriteString(esc); err != nil {
-			return err
-		}
-		i = strings.IndexAny(val, escapedURLChars)
-	}
-	_, err := w.WriteString(val)
-	return err
-}
-
 // Query represents a single part of the query string, a query param
 type Query struct {
 	Key      string
@@ -612,6 +577,14 @@ attrsLoop:
 				case "audio", "embed", "iframe", "img", "script", "source", "track", "video":
 					if htmlAttr.Key == "src" {
 						if u, ok := p.validURL(htmlAttr.Val); ok {
+							if p.srcRewriter != nil {
+								parsedURL, err := url.Parse(u)
+								if err != nil {
+									fmt.Println(err)
+								}
+								p.srcRewriter(parsedURL)
+								u = parsedURL.String()
+							}
 							htmlAttr.Val = u
 							tmpAttrs = append(tmpAttrs, htmlAttr)
 						}
@@ -970,14 +943,14 @@ func (p *Policy) validURL(rawurl string) (string, bool) {
 		}
 
 		if u.Scheme != "" {
+			urlPolicies, ok := p.allowURLSchemes[u.Scheme]
+			if !ok {
 				for _, r := range p.allowURLSchemeRegexps {
 					if r.MatchString(u.Scheme) {
 						return u.String(), true
 					}
 				}
 
-			urlPolicies, ok := p.allowURLSchemes[u.Scheme]
-			if !ok {
 				return "", false
 			}
 
@@ -986,7 +959,7 @@ func (p *Policy) validURL(rawurl string) (string, bool) {
 			}
 
 			for _, urlPolicy := range urlPolicies {
-				if urlPolicy(u) == true {
+				if urlPolicy(u) {
 					return u.String(), true
 				}
 			}

vendor/modules.txt

@@ -376,7 +376,7 @@ github.com/magiconair/properties
 # github.com/mattn/go-isatty v0.0.19
 ## explicit; go 1.15
 github.com/mattn/go-isatty
-# github.com/microcosm-cc/bluemonday v1.0.24
+# github.com/microcosm-cc/bluemonday v1.0.25
 ## explicit; go 1.19
 github.com/microcosm-cc/bluemonday
 github.com/microcosm-cc/bluemonday/css