GONRG-1183: Add script to replace invalid values in schemas

scripts/convert_R2_schemas_to_R3.py

+import glob
+import json
+import os
+import re
+
+from collections import UserString
+
+TENANT = "opendes"
+AUTHORITY = "osdu"
+SCHEMAS_DIR = os.environ["SCHEMAS_DIR"]
+
+
+class JsonString(UserString):
+    REF_REGEXP = r"(?P<abstract_perfix>\.\.\/abstract/)(?P<kind_name>\w+)\.(?P<version>\d+\.\d+\.\d+)\.json"
+    NAMESPACE_REGEXP = r"\<namespace\>"
+
+    def repl_closure(self, match: re.match):
+        if not match.groups:
+            print(self.data)
+            raise Exception
+        kind_name = match.group('kind_name')
+        version = match.group('version')
+        repl = f"{TENANT}:{AUTHORITY}:{kind_name}:{version}"
+        return repl
+
+    def replace_refs(self):
+        self.data = re.sub(self.REF_REGEXP, self.repl_closure, self.data)
+        return self
+
+    def replace_namespaces(self):
+        self.data = re.sub(self.NAMESPACE_REGEXP, TENANT, self.data)
+        return self
+
+    @staticmethod
+    def lower_first_letter(val: str):
+        if val[0].islower():
+            pass
+        elif val in (
+            "ACL",
+            "Legals",
+            "ID"
+        ):
+            val = val.lower()
+        else:
+            val = val.replace(val[0], val[0].lower(), 1)
+        return val
+
+    def to_pascal_case(self):
+        tmp_properties = {}
+        tmp_required = []
+        json_file_dict = json.loads(self.data)
+        try:
+            if "schemaInfo" in json_file_dict:  # if schema has additional fields to be recorded
+                content = json_file_dict["schema"]
+            else:
+                content = json_file_dict
+            if "properties" in content:
+                for key, value in content["properties"].items():
+                    tmp_properties[self.lower_first_letter(key)] = value
+                content["properties"] = tmp_properties
+            if "required" in content:
+                for i in content["required"]:
+                    tmp_required.append(self.lower_first_letter(i))
+                content["required"] = tmp_required
+            self.data = json.dumps(json_file_dict, indent=4)
+            return self
+        except Exception as e:
+            print(self.data)
+            raise e
+
+
+for file_path in glob.glob(SCHEMAS_DIR + "/*.json"):
+    try:
+        with open(file_path, "r") as file:
+            content = file.read()
+            content = JsonString(content).replace_refs().replace_namespaces().to_pascal_case().data
+        with open(file_path, "w") as file:
+            file.write(content)
+    except Exception as e:
+        print(f"Error on file {file_path}")
+        raise e