improve regexp matching

ada-url · Jan 7, 2025 · d5f2203 · d5f2203
1 parent 81da4e5
commit d5f2203
Show file tree

Hide file tree

Showing 3 changed files with 35 additions and 26 deletions.
diff --git a/include/ada/url_pattern-inl.h b/include/ada/url_pattern-inl.h
@@ -46,16 +46,18 @@ url_pattern_component::create_component_match_result(
   // Optimization: Let's reserve the size.
   result.groups.reserve(exec_result.size() - 1);
 
-  // Let index be 1.
+  // Let index be 0.
   // While index is less than Get(execResult, "length"):
-  for (size_t index = 1; index < exec_result.size(); index++) {
-    // Let name be component’s group name list[index − 1].
+  for (size_t index = 0; index < exec_result.size() - 1; index++) {
+    // Let name be component’s group name list[index].
     // Let value be Get(execResult, ToString(index)).
     // Set groups[name] to value.
-    result.groups.insert({
-        group_name_list[index - 1],
-        exec_result[index].str(),
-    });
+    if (auto str = exec_result[index].str(); !str.empty()) {
+      result.groups.insert({
+          group_name_list[index],
+          str,
+      });
+    }
   }
   return result;
 }

diff --git a/src/url_pattern.cpp b/src/url_pattern.cpp
@@ -681,53 +681,60 @@ result<std::optional<url_pattern_result>> url_pattern::match(
     }
   }
 
+  auto regex_flags = std::regex_constants::match_any;
+
   // Let protocolExecResult be RegExpBuiltinExec(urlPattern’s protocol
   // component's regular expression, protocol).
   std::smatch protocol_exec_result_value;
-  auto protocol_exec_result = std::regex_match(
-      protocol, protocol_exec_result_value, protocol_component.regexp);
+  auto protocol_exec_result =
+      std::regex_match(protocol, protocol_exec_result_value,
+                       protocol_component.regexp, regex_flags);
 
   // Let usernameExecResult be RegExpBuiltinExec(urlPattern’s username
   // component's regular expression, username).
   std::smatch username_exec_result_value;
-  auto username_exec_result = std::regex_match(
-      username, username_exec_result_value, username_component.regexp);
+  auto username_exec_result =
+      std::regex_match(username, username_exec_result_value,
+                       username_component.regexp, regex_flags);
 
   // Let passwordExecResult be RegExpBuiltinExec(urlPattern’s password
   // component's regular expression, password).
   std::smatch password_exec_result_value;
-  auto password_exec_result = std::regex_match(
-      password, password_exec_result_value, password_component.regexp);
+  auto password_exec_result =
+      std::regex_match(password, password_exec_result_value,
+                       password_component.regexp, regex_flags);
 
   // Let hostnameExecResult be RegExpBuiltinExec(urlPattern’s hostname
   // component's regular expression, hostname).
   std::smatch hostname_exec_result_value;
-  auto hostname_exec_result = std::regex_match(
-      hostname, hostname_exec_result_value, hostname_component.regexp);
+  auto hostname_exec_result =
+      std::regex_match(hostname, hostname_exec_result_value,
+                       hostname_component.regexp, regex_flags);
 
   // Let portExecResult be RegExpBuiltinExec(urlPattern’s port component's
   // regular expression, port).
   std::smatch port_exec_result_value;
-  auto port_exec_result =
-      std::regex_match(port, port_exec_result_value, port_component.regexp);
+  auto port_exec_result = std::regex_match(port, port_exec_result_value,
+                                           port_component.regexp, regex_flags);
 
   // Let pathnameExecResult be RegExpBuiltinExec(urlPattern’s pathname
   // component's regular expression, pathname).
   std::smatch pathname_exec_result_value;
-  auto pathname_exec_result = std::regex_match(
-      pathname, pathname_exec_result_value, pathname_component.regexp);
+  auto pathname_exec_result =
+      std::regex_match(pathname, pathname_exec_result_value,
+                       pathname_component.regexp, regex_flags);
 
   // Let searchExecResult be RegExpBuiltinExec(urlPattern’s search component's
   // regular expression, search).
   std::smatch search_exec_result_value;
-  auto search_exec_result = std::regex_match(search, search_exec_result_value,
-                                             search_component.regexp);
+  auto search_exec_result = std::regex_match(
+      search, search_exec_result_value, search_component.regexp, regex_flags);
 
   // Let hashExecResult be RegExpBuiltinExec(urlPattern’s hash component's
   // regular expression, hash).
   std::smatch hash_exec_result_value;
-  auto hash_exec_result =
-      std::regex_match(hash, hash_exec_result_value, hash_component.regexp);
+  auto hash_exec_result = std::regex_match(hash, hash_exec_result_value,
+                                           hash_component.regexp, regex_flags);
 
   // If protocolExecResult, usernameExecResult, passwordExecResult,
   // hostnameExecResult, portExecResult, pathnameExecResult, searchExecResult,

diff --git a/tests/wpt_urlpattern_tests.cpp b/tests/wpt_urlpattern_tests.cpp
@@ -304,11 +304,11 @@ ada::url_pattern_component_result parse_component_result(
       ondemand::object groups;
       EXPECT_FALSE(element.value().get_object().get(groups));
       for (auto group : groups) {
-        std::string_view group_key(group.key().value().raw());
+        auto group_key = group.escaped_key().value();
         std::string_view group_value;
-        EXPECT_FALSE(group.value().get(group_value));
+        EXPECT_FALSE(group.value().get_string(group_value));
         result.groups.insert_or_assign(std::string(group_key),
-                                       std::string(group_value));
+                                       group_value);
       }
     }
   }