Add csv parsing helper classes

WARNING: this class is currently experimental and does not have a stable
API nor are all expected functions implemented.

pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "elody"
-version = "0.0.20"
+version = "0.0.21"
 description = "elody SDK for Python"
 readme = "README.md"
 authors = [{ name = "Inuits", email = "[email protected]" }]

src/elody/csv.py

@@ -0,0 +1,120 @@
+import csv
+import re
+
+from io import StringIO
+from elody.exceptions import ColumnNotFoundException, IncorrectTypeException
+
+
+class CSVParser:
+    def __init__(self, csvstring):
+        self.csvstring = csvstring
+        self.reader = csv.DictReader(self.__csv_string_to_file_object())
+
+    def _get_metadata_object(self, key, value, lang="en"):
+        return {
+            "key": key,
+            "value": value,
+            "lang": lang,
+        }
+
+    def _get_relation_object(self, type, key):
+        return {
+            "type": type,
+            "key": key,
+        }
+
+    def _is_relation_field(self, field):
+        if re.fullmatch("(has|is)[A-Z][a-z]+", field):
+            return True
+        return False
+
+    def __csv_string_to_file_object(self):
+        return StringIO(self.csvstring)
+
+
+class CSVSingleObject(CSVParser):
+    def __init__(self, csvstring, object_type="entity"):
+        super().__init__(csvstring)
+        self.identifiers = list()
+        self.metadata = list()
+        self.object_type = object_type
+        self.objects = list()
+        self.relations = list()
+        self.__init_fields()
+
+    def get_entity(self):
+        return self.__get_type("entity")
+
+    def get_mediafile(self):
+        return self.__get_type("mediafile")
+
+    def __fill_metadata(self, key, value):
+        if value:
+            self.metadata.append(self._get_metadata_object(key, value))
+
+    def __fill_relations(self, type, key):
+        if key:
+            self.relations.append(self._get_relation_object(type, key))
+
+    def __get_type(self, type):
+        if self.object_type != type:
+            raise IncorrectTypeException(f"Not a {type}!")
+        object = dict()
+        for property_name, property in {
+            "metadata": self.metadata,
+            "relations": self.relations,
+            "identifiers": self.identifiers,
+        }.items():
+            if property:
+                object[property_name] = property
+        return object
+
+    def __init_fields(self):
+        for row in self.reader:
+            for key, value in row.items():
+                if self._is_relation_field(key):
+                    self.__fill_relations(key, value)
+                else:
+                    self.__fill_metadata(key, value)
+
+
+class CSVMultiObject(CSVParser):
+    def __init__(self, csvstring, index_mapping={"entities": "entity_id"}):
+        super().__init__(csvstring)
+        self.index_mapping = index_mapping
+        self.objects = dict()
+        self.__fill_objects_from_csv()
+
+    def get_entities(self):
+        return self.objects.get("entities", list())
+
+    def get_mediafiles(self):
+        return self.objects.get("mediafiles", list())
+
+    def __fill_objects_from_csv(self):
+        indexed_dict = dict()
+        for row in self.reader:
+            if not all(x in row.keys() for x in self.index_mapping.values()):
+                raise ColumnNotFoundException(
+                    f"Not all identifying columns are present in CSV"
+                )
+            for type, identifying_column in self.index_mapping.items():
+                id = row[identifying_column]
+                if type not in indexed_dict:
+                    indexed_dict[type] = dict()
+                if id not in indexed_dict[type]:
+                    indexed_dict[type][id] = dict()
+                for root_property in ["metadata", "relations", "identifiers"]:
+                    if root_property not in indexed_dict[type][id]:
+                        indexed_dict[type][id][root_property] = list()
+                for key, value in row.items():
+                    if self._is_relation_field(key) and value:
+                        indexed_dict[type][id]["relations"].append(
+                            self._get_relation_object(key, value)
+                        )
+                    elif key not in self.index_mapping.values() and value:
+                        indexed_dict[type][id]["metadata"].append(
+                            self._get_metadata_object(key, value)
+                        )
+        for metadata_type, objects in indexed_dict.items():
+            self.objects[metadata_type] = list(objects.values())

src/elody/exceptions.py

@@ -6,10 +6,18 @@ def __init__(self, message, filename=None, md5sum=None):
         self.md5sum = md5sum
 
 
+class ColumnNotFoundException(Exception):
+    pass
+
+
 class FileNotFoundException(Exception):
     pass
 
 
+class IncorrectTypeException(Exception):
+    pass
+
+
 class NoMediafilesException(Exception):
     pass