Merge pull request #176 from BIH-CEI/175-remove-symbols-when-reading-…

…xmls added remove at symbols and tested it
BIH-CEI · Oct 15, 2024 · 73c8c5a · 73c8c5a
2 parents cb9221f + 7b7a8c0
commit 73c8c5a
Show file tree

Hide file tree

Showing 2 changed files with 62 additions and 9 deletions.
diff --git a/src/phenopacket_mapper/utils/io/read_xml.py b/src/phenopacket_mapper/utils/io/read_xml.py
@@ -55,10 +55,27 @@ def parse_primitive_value(value: str):
     return dict_
 
 
+def remove_at_symbols(dict_: Dict) -> Dict:
+    if isinstance(dict_, dict):
+        new_dict = {}
+        for k, v in dict_.items():
+            if k.startswith("@"):
+                k = k[1:]
+            if isinstance(v, list):
+                new_dict[k] = [remove_at_symbols(item) for item in v]
+            else:
+                new_dict[k] = remove_at_symbols(v)
+
+        return new_dict
+    else:
+        return dict_
+
+
 def parse_xml(file: IOBase) -> Dict:
     """Parse an XML file into a dictionary with inferred types."""
     dict_ = xmltodict.parse(file.read())
     print(f"{dict_=}, {type(dict_)=}")
     dict_ = _post_process_xml_dict(dict_)
+    dict_ = remove_at_symbols(dict_)
     return dict_
 
diff --git a/tests/utils/io/test_read_xml.py b/tests/utils/io/test_read_xml.py
@@ -1,4 +1,5 @@
 from phenopacket_mapper.utils.io import read_xml
+from phenopacket_mapper.utils.io.read_xml import remove_at_symbols
 
 import pytest
 from io import StringIO
@@ -19,16 +20,51 @@
         ('<boolean>false</boolean>', {"boolean": False}),
         ('<array><item>1</item><item>2</item><item>3</item></array>', {"array": {"item": [1, 2, 3]}}),
         ('<root>'
-         '<array><item>1</item><item>2</item><item>3</item></array>'
-         '<boolean>true</boolean>'
-         '<color>gold</color>'
-         '<number>123</number>'
-         '<object><a>b</a><c>d</c></object>'
-         '<string>Hello World</string>'
-         '</root>',
-         {"root":{"array": {"item": [1, 2, 3]}, "boolean": True, "color": "gold", "number": 123, "object": {"a": "b", "c": "d"}, "string": "Hello World"}}
-         )
+             '<array>'
+                '<item>1</item>'
+                '<item>2</item>'
+                '<item>3</item>'
+             '</array>'
+             '<boolean>true</boolean>'
+             '<color>gold</color>'
+             '<number>123</number>'
+             '<object>'
+                '<a>b</a>'
+                '<c>d</c>'
+             '</object>'
+             '<string>Hello World</string>'
+         '</root>', 
+         {
+                "root":{
+                    "array": {
+                        "item": [1, 2, 3]
+                    },
+                    "boolean": True,
+                    "color": "gold",
+                    "number": 123,
+                    "object": {
+                        "a": "b",
+                        "c": "d"
+                    },
+                    "string": "Hello World"
+                }
+         }),
+        ('<ItemData ItemOID="redcap_survey_identifier" Value=""/>', {"ItemData": {"ItemOID": "redcap_survey_identifier", "Value": ""}}),
     ],
 )
+# TODO test tags inside tags eg <a b="c">d</a>
 def test_read_xml(inp, expected):
     assert read_xml(StringIO(inp)) == expected
+
+
+@pytest.mark.parametrize(
+    "inp,expected",
+    [
+        ({"@a": "b"}, {"a": "b"}),
+        ({"a": "b"}, {"a": "b"}),
+        ({'@a': {'@b': 'c'}}, {'a': {'b': 'c'}}),
+        ({'a': [{'@b': 'c'}, {'@d': 'e'}]}, {'a': [{'b': 'c'}, {'d': 'e'}]}),
+    ]
+)
+def test_remove_at_symbols(inp, expected):
+    assert remove_at_symbols(inp) == expected