LibWeb: Fix SRI handling of badly-formatted strings

(cherry picked from commit 39dae6fb2db58156dda7c58cb5293ca52ff35d4e)

Tests/LibWeb/Crash/HTML/invalid-integrity-attribute.css

@@ -0,0 +1 @@
+.body {}

Tests/LibWeb/Crash/HTML/invalid-integrity-attribute.html

@@ -0,0 +1,16 @@
+<!DOCTYPE html>
+<html lang="en">
+
+<head>
+  <meta http-equiv="content-type" content="text/html; charset=UTF-8">
+  <link
+    href="invalid-integrity-attribute.css"
+    type="text/css"
+    rel="stylesheet"
+    crossorigin="anonymous"
+    integrity="051df3b15e50dd8711b1fa2f4e11f69060b4d77e0a3950a5ae2dc9f5c6bf8a37"
+    nonce=""
+  >
+  <title>Test</title>
+</head>
+<body>

Userland/Libraries/LibWeb/SRI/SRI.cpp

@@ -90,7 +90,7 @@ ErrorOr<Vector<Metadata>> parse_metadata(StringView metadata)
         auto algorithm = hash_expr_token_list[0];
 
         // 6. If hash-expr-token-list[1] exists, set base64-value to hash-expr-token-list[1].
-        if (hash_expr_token_list.size() >= 1)
+        if (hash_expr_token_list.size() > 1)
             base64_value = hash_expr_token_list[1];
 
         // 7. If algorithm is not a hash function recognized by the user agent, continue.