The codec factory should support two distinct interfaces - negotiating
for a serializer with a client, vs reading or writing data to a storage
form (etcd, disk, etc). Make the EncodeForVersion and DecodeToVersion
methods only take Encoder and Decoder, and slight refactoring elsewhere.
In the storage factory, use a content type to control what serializer to
pick, and use the universal deserializer. This ensures that storage can
read JSON (which might be from older objects) while only writing
protobuf. Add exceptions for those resources that may not be able to
write to protobuf (specifically third party resources, but potentially
others in the future).
The decoder will return a nil gvk in case it errors out and getting
the version out of that gvk naturally will panic. Bail out as soon
as we can check that the error is non-nil. kubectl edit was the primary
victim.
Automatic merge from submit-queue
Add support for 3rd party objects to kubectl
@deads2k @jlowdermilk
Instructions for playing around with this:
Run an apiserver with third party resources turned on (`--runtime-config=extensions/v1beta1=true,extensions/v1beta1/thirdpartyresources=true`)
Then you should be able to:
```
kubectl create -f rsrc.json
```
```json
{
"metadata": {
"name": "foo.company.com"
},
"apiVersion": "extensions/v1beta1",
"kind": "ThirdPartyResource",
"versions": [
{
"apiGroup": "group",
"name": "v1"
},
{
"apiGroup": "group",
"name": "v2"
}
]
}
```
Once that is done, you should be able to:
```
curl http://<server>/apis/company.com/v1/foos
```
```
curl -X POST -d @${HOME}/foo.json http://localhost:8080/apis/company.com/v1/namespaces/default/foos
```
```json
{
"kind": "Foo",
"apiVersion": "company.com/v1",
"metadata": {
"name": "baz"
},
"someField": "hello world",
"otherField": 1
}
```
After this PR, you can do:
```
kubectl create -f foo.json
```
```
kubectl get foos
```
etc.
reqs:
- the kubectl cmd must support the -f | --filename flag
- the kubectl cmd must support visiting a dir one level deep,
or using more than one resource
It should be allowed to invoke kubectl replace with a JSON file that has
no resource version set. Namespaced resources were working correctly,
but cluster resources were silently failing to lookup the current state
of the object to get the resource version because we weren't using
NamespaceIfScoped(). Added a failing test.
Most of the logic related to type and kind retrieval belongs in the
codec, not in the various classes. Make it explicit that the codec
should handle these details.
Factory now returns a universal Decoder and a JSONEncoder to assist code
in kubectl that needs to specifically deal with JSON serialization
(apply, merge, patch, edit, jsonpath). Add comments to indicate the
serialization is explicit in those places. These methods decode to
internal and encode to the preferred API version as previous, although
in the future they may be changed.
React to removing Codec from version interfaces and RESTMapping by
passing it in to all the places that it is needed.
A lot of packages use StringSet, but they don't use anything else from
the util package. Moving StringSet into another package will shrink
their dependency trees significantly.
Right now there is no method in the resource builder for specifying just the name of a resource. NameParam is useful when a default type is already specified with ResourceTypes.
Signed-off-by: kargakis <mkargaki@redhat.com>
This change allows the namespace in kubeconfig to be overridden by
specifying the namespace in the spec file. If namespace is explicitly
provided in the command line flags and the spec file has a different
namespace, this will cause an error.
Refactored how files, directories and stdin are handled. Every file must pass
through the FileVisitor and then streamed through StreamVisitor. FileVisitor
takes care of opening/closing files and StreamVisitor is parsing multiple
resources.