Restructure samples

[gaoning777]

This PR aims to organize the samples into core set and contrib set.

Core samples: demonstrates the full KFP functionalities and covered by the sample test infra.
A selected set of these core samples will also be preloaded to the KFP during deployment and the tests for these samples will block the PRs. The core samples will also include intermediate samples that are more complex than basic samples such as flip coins but simpler than TFX samples. It serves to demonstrate a set of the outstanding features and offers users the next level KFP experience.

Contrib samples: these contrib samples are not tested by KFP and could potentially be moved the core samples if tests are covered and it demonstrates certain KFP functionality that fit. Another reason to put some samples in this directory is that some samples require certain platform support that is hard to support in our test infra.

Structure:
In the Core directory, each sample will be in a separate directory. In the Contrib directory, there is an intermediate directory for each contributor, e.g. ibm and arena, within which each sample is in a separate directory. An example of the resulting structure is as follows:

pipelines/samples/
Core/
	DSL Static Type Checking/
		DSL Static Type Checking.ipynb
	XGBoost/
		xgboost-training-cm.py
	Condition/
		condition.py
	Recursion/
		recursion.py
Contrib/
	IBM/
		ffdl-seldon/
			ffdl_pipeline.ipynb
			ffdl_pipeline.py
			README.md

This change is 

[Ark-kun]

Can you look at #1342 ? It's been pending for a lot of time.

[gaoning777]

Can you look at #1342 ? It's been pending for a lot of time.

Besides reviewing that particular PR, could you review this PR?

[Ark-kun]

Can you look at #1342 ? It's been pending for a lot of time.

Besides reviewing that particular PR, could you review this PR?

/lgtm

[Ark-kun]

/lgtm

[gaoning777]

/approve

[neuromage]

Since this PR is large, consider adding a more verbose description to the PR so future readers can see the reasons for this change.

[Ark-kun]

Thanks for doing this work.

I'd group the DSL samples together (e.g. in a DSL directory).

[neuromage]

/lgtm
/approve

[k8s-ci-robot]

[APPROVALNOTIFIER] This PR is APPROVED

This pull-request has been approved by: gaoning777, neuromage

The full list of commands accepted by this bot can be found here.

The pull request process is described here

Needs approval from an approver in each of these files:

• backend/OWNERS [neuromage]
• samples/OWNERS [gaoning777,neuromage]
• test/sample-test/OWNERS [gaoning777,neuromage]

Approvers can indicate their approval by writing /approve in a comment
Approvers can cancel approval by writing /approve cancel in a comment

[gaoning777]

As the samples grow bigger, we need a hierarchical structure for all these samples under each of the categories.

[Ark-kun]

/lgtm

[numerology]

/lgtm

[IronPan]

As discussed offline, we might consider moving the primitive samples like retry or volume one to core/ instead of contrib/ folder.

[IronPan]

should we also move following?
pipeline_transformers
resource_ops
volume_ops
volume_snapshot_ops

basically the sample demonstrate the native functionalities should live in core/ and vendor specific live in contrib/

[gaoning777]

You are correct that these samples will be in the core directory in the future. However, I don't think it is a good idea to move the samples to the core when they are not covered by the sample tests.

[gaoning777]

Moved these samples as well.

[IronPan]

/lgtm