| Index: third_party/WebKit/Source/modules/fetch/RequestInit.cpp
|
| diff --git a/third_party/WebKit/Source/modules/fetch/RequestInit.cpp b/third_party/WebKit/Source/modules/fetch/RequestInit.cpp
|
| index 142bca2a1f6ac369dbdec11d0c087f4fb64748da..54a76a215bdb1497bd74102830eef4e701a0dc90 100644
|
| --- a/third_party/WebKit/Source/modules/fetch/RequestInit.cpp
|
| +++ b/third_party/WebKit/Source/modules/fetch/RequestInit.cpp
|
| @@ -101,12 +101,14 @@ RequestInit::RequestInit(ExecutionContext* context,
|
| v8::Isolate* isolate = toIsolate(context);
|
| if (isCredentialSet) {
|
| if (V8PasswordCredential::hasInstance(v8Credential, isolate)) {
|
| - // TODO(mkwst): According to the spec, we'd serialize this once we touch the network. We're
|
| - // serializing it here, ahead of time, because lifetime issues around ResourceRequest make
|
| - // it pretty difficult to pass a PasswordCredential around at the platform level, and the
|
| - // hop between the browser and renderer processes to deal with service workers is equally
|
| - // painful. There should be no developer-visible difference in behavior with this option,
|
| - // except that the `Content-Type` header will be set early. That seems reasonable.
|
| + // TODO(mkwst): According to the spec, we'd serialize this once we touch
|
| + // the network. We're serializing it here, ahead of time, because lifetime
|
| + // issues around ResourceRequest make it pretty difficult to pass a
|
| + // PasswordCredential around at the platform level, and the hop between
|
| + // the browser and renderer processes to deal with service workers is
|
| + // equally painful. There should be no developer-visible difference in
|
| + // behavior with this option, except that the `Content-Type` header will
|
| + // be set early. That seems reasonable.
|
| PasswordCredential* credential =
|
| V8PasswordCredential::toImpl(v8Credential.As<v8::Object>());
|
| attachedCredential = credential->encodeFormData(contentType);
|
|
|