kernelci.api.helper: enhance submit_regression()

Make this function create the regression from the two nodes that define
the breaking point (the fail node and the previous pass node) and then
submit it. This abstracts all the model details and the logic behind the
regression generation away from the user code.

Signed-off-by: Ricardo Cañuelo <ricardo.canuelo@collabora.com>

kernelci/api/helper.py

@@ -136,14 +136,33 @@ def create_job_node(self, job_config, input_node,
         except requests.exceptions.HTTPError as error:
             raise RuntimeError(json.loads(error.response.text)) from error
 
-    def submit_regression(self, regression):
-        """Post a regression object
 
-        [TODO] Leave this function in place in case we'll need any other
-        processing or formatting before submitting the regression node
-        """
-        # pylint: disable=protected-access
-        return self.api._post('node', regression)
+    def submit_regression(self, fail_node: dict, pass_node: dict):
+       """Creates and submits a regression object from two existing
+       nodes (test or kbuilds) passed as parameters:
+
+       Arguments:
+         fail_node: dict describing the node that failed and triggered
+             the regression creation.
+         pass_node: dict describing the previous passing run of the same
+             test/build
+
+       General use case: failure detected, the fail_node and pass_node
+       are retrieved from the API and passed to this function to define
+       and submit a regression.
+
+       Returns the API request response.
+       """
+       fail_node_obj = self.get_node_obj(fail_node)
+       pass_node_obj = self.get_node_obj(pass_node)
+       regression_dict = models.Regression.create_regression(
+           fail_node_obj, pass_node_obj, as_dict=True)
+
+       try:
+           return self.api.node.add(regression_dict)
+       except requests.exceptions.HTTPError as error:
+           raise RuntimeError(json.loads(error.response.text)) from error
+
 
     def _prepare_results(self, results, parent, base):
         node = results['node'].copy()