From bcf192ec74f4a6be1d3ae5a5551f546b75dd23b9 Mon Sep 17 00:00:00 2001
From: David Mendez <dmendez@ebi.ac.uk>
Date: Tue, 18 Aug 2020 15:26:19 -0500
Subject: [PATCH] Implement mapping_to_apply generation for nested properties

---
 app/es_subset_generator/mappings.py           | 47 +++++++++++++++--
 .../test/data/mappings_to_apply_must_be_1json | 51 +++++++++++++++++++
 .../test/test_mappings_handling.py            | 18 +++----
 3 files changed, 99 insertions(+), 17 deletions(-)
 create mode 100644 app/es_subset_generator/test/data/mappings_to_apply_must_be_1json

diff --git a/app/es_subset_generator/mappings.py b/app/es_subset_generator/mappings.py
index f2585f0..5e4ee5b 100644
--- a/app/es_subset_generator/mappings.py
+++ b/app/es_subset_generator/mappings.py
@@ -27,15 +27,52 @@ def parse_mapping_from_es_response(index_mapping, required_source):
     :param required_source: source (list of properties) to get from the mapping
     :return: the mapping ready to be applied to the subset index
     """
-    mappings_to_apply = {
+    all_mappings_to_apply = {
         'properties': {}
     }
 
-    for current_property in required_source:
-        mapping_to_apply = get_mapping_to_apply_in_subset_index(index_mapping, current_property)
-        mappings_to_apply['properties'][current_property] = mapping_to_apply
+    for current_property_path in required_source:
+
+        mapping_to_apply = get_mapping_to_apply_in_subset_index(index_mapping, current_property_path)
+        add_mapping_to_property_path(mapping_to_apply, all_mappings_to_apply, current_property_path)
+
+    return all_mappings_to_apply
+
+
+def add_mapping_to_property_path(mapping_to_apply, all_mappings_to_apply, property_path):
+    """
+    Adds the mapping to all the mappings to apply to the path indicated as parameter
+    :param mapping_to_apply: mapping to add to the dict
+    :param all_mappings_to_apply: a dict with all the mappings to apply
+    :param property_path: path to the property
+    """
+    print('property_path: ', property_path)
+
+    prop_path_parts = property_path.split('.')
+    current_prop = prop_path_parts[0]
+
+    print('prop_path_parts: ', prop_path_parts)
+
+
+    if len(prop_path_parts) == 1:
+
+        all_mappings_to_apply['properties'][current_prop] = mapping_to_apply
+
+    else:
+
+        container_dict = all_mappings_to_apply['properties'].get(current_prop)
+
+        if container_dict is None:
+            all_mappings_to_apply['properties'][current_prop] = {
+                'properties': {}
+            }
+
+        rest_of_the_path_parts = prop_path_parts[1:]
+        rest_of_the_path = '.'.join(rest_of_the_path_parts)
+        inside_mappings_to_apply = all_mappings_to_apply['properties'][current_prop]
+        add_mapping_to_property_path(mapping_to_apply, inside_mappings_to_apply, rest_of_the_path)
+
 
-    return mappings_to_apply
 
 
 def get_mapping_to_apply_in_subset_index(index_mapping, property_path):
diff --git a/app/es_subset_generator/test/data/mappings_to_apply_must_be_1json b/app/es_subset_generator/test/data/mappings_to_apply_must_be_1json
new file mode 100644
index 0000000..1670f2b
--- /dev/null
+++ b/app/es_subset_generator/test/data/mappings_to_apply_must_be_1json
@@ -0,0 +1,51 @@
+{
+  "properties": {
+    "molecule_properties": {
+      "properties": {
+        "hba": {
+          "type": "integer"
+        },
+        "psa": {
+          "type": "double"
+        },
+        "molecular_species": {
+          "type": "keyword"
+        }
+      }
+    },
+    "molecule_synonyms": {
+      "properties": {
+        "molecule_synonym": {
+          "type": "keyword"
+        },
+        "syn_type": {
+          "type": "keyword"
+        },
+        "synonyms": {
+          "type": "keyword"
+        }
+      }
+    },
+    "_metadata": {
+      "properties": {
+        "drug": {
+          "properties": {
+            "drug_data": {
+              "properties": {
+                "withdrawn_year": {
+                  "type": "short"
+                },
+                "prodrug": {
+                  "type": "boolean"
+                }
+              }
+            }
+          }
+        }
+      }
+    },
+    "helm_notation": {
+      "type": "keyword"
+    }
+  }
+}
\ No newline at end of file
diff --git a/app/es_subset_generator/test/test_mappings_handling.py b/app/es_subset_generator/test/test_mappings_handling.py
index d872901..4056fe8 100644
--- a/app/es_subset_generator/test/test_mappings_handling.py
+++ b/app/es_subset_generator/test/test_mappings_handling.py
@@ -54,25 +54,19 @@ class TestMappingsHandling(unittest.TestCase):
             ]
 
             mappings_to_apply_got = mappings.parse_mapping_from_es_response(mapping_from_base_index, source)
-            print('mappings_to_apply_got:')
-            print(json.dumps(mappings_to_apply_got, indent=4))
 
-            mappings_to_apply_must_be = {
-                "properties": {
-                    'pref_name': {
-                        'type': 'keyword'
-                    },
-                }
-            }
+            mappings_to_apply_must_be_path = 'app/es_subset_generator/test/data/mappings_to_apply_must_be_1json'
 
-            # self.assertEqual(mappings_to_apply_got, mappings_to_apply_must_be,
-            #                  msg='The mappings were not produced correctly!')
+            with open(mappings_to_apply_must_be_path, 'rt') as mappings_to_apply_must_be_file:
+                mappings_to_apply_must_be = json.load(mappings_to_apply_must_be_file)
+
+                self.assertEqual(mappings_to_apply_got, mappings_to_apply_must_be,
+                                 msg='The mappings were not produced correctly!')
 
     def test_locates_source_config_of_property(self):
         """
         Test that it locates the config of a property
         """
-
         sample_mapping_path = 'app/es_subset_generator/test/data/sample_mapping_1.json'
 
         with open(sample_mapping_path, 'rt') as sample_mapping_file:
-- 
GitLab