From 04ae420c2400f6fbd4b5da720e0e72378f4be9d3 Mon Sep 17 00:00:00 2001
From: SimonTaurus <simon.stier@data-info-knowledge.de>
Date: Sat, 27 Apr 2024 12:20:48 +0200
Subject: [PATCH] feat: implement serialization

Refs: #2
---
 pyproject.toml                                |  1 +
 src/ontopint/__init__.py                      | 46 +++++++++++++++++-
 ...on_test.py => 020_deserialization_test.py} | 38 +--------------
 tests/030_serialization_test.py               | 47 +++++++++++++++++++
 tests/040_roundtrip_test.py                   | 24 ++++++++++
 tests/common.py                               | 39 +++++++++++++++
 6 files changed, 157 insertions(+), 38 deletions(-)
 rename tests/{deserialization_test.py => 020_deserialization_test.py} (58%)
 create mode 100644 tests/030_serialization_test.py
 create mode 100644 tests/040_roundtrip_test.py
 create mode 100644 tests/common.py

diff --git a/pyproject.toml b/pyproject.toml
index 23cd397..0434cd8 100644
--- a/pyproject.toml
+++ b/pyproject.toml
@@ -37,6 +37,7 @@ dependencies = [
 [project.optional-dependencies]
 dev = [
     "pytest",
+    "deepdiff",
 ]
 
 [project.license]
diff --git a/src/ontopint/__init__.py b/src/ontopint/__init__.py
index 3b7889d..0712e95 100644
--- a/src/ontopint/__init__.py
+++ b/src/ontopint/__init__.py
@@ -6,6 +6,7 @@
 
 # from pint import UnitRegistry
 from ucumvert import PintUcumRegistry
+import pint
 
 # ureg = UnitRegistry()
 ureg = PintUcumRegistry()
@@ -100,6 +101,33 @@ def _replace_units(obj, context, original_key_lookup_dict):
         ]
     else:
         return obj
+    
+def _serialize_units(obj, context, original_key_lookup_dict):
+    if isinstance(obj, dict):
+        for key in list(obj.keys()): # make a list copy in order to delete keys while iterating
+            value = obj[key]
+            if (isinstance(value, pint.Quantity)):
+                # see https://pint.readthedocs.io/en/stable/user/formatting.html
+                quantity_value = float(format(value, 'f#~').split(' ')[0])
+                unit_code = format(value.u, '~') 
+                # ToDo: use ucum code
+                unit_iri = get_qunit_iri_from_unit_code(unit_code)
+                # note: "urn:ontopint:iri" is just any iri not existing in the input data
+                unit_compact_iri = jsonld.compact(
+                    {"@context": {**context, "urn:ontopint:iri": {"@type": "@id"}}, "urn:ontopint:iri": unit_iri}, 
+                    {**context, "urn:ontopint:iri": {"@type": "@id"}}
+                )["urn:ontopint:iri"]
+                obj[original_key_lookup_dict['value']] = quantity_value
+                obj[original_key_lookup_dict['unit']] = unit_compact_iri
+
+            else: obj[key] = _serialize_units(value, context, original_key_lookup_dict)
+        return obj
+    elif isinstance(obj, list):
+        return [
+            _serialize_units(value, context, original_key_lookup_dict) for value in obj
+        ]
+    else:
+        return obj
 
 
 def parse_units(json_ld: dict) -> dict:
@@ -113,5 +141,21 @@ def parse_units(json_ld: dict) -> dict:
     # reverse the dict
     original_key_lookup_dict = {v: k for k, v in compacted.items()}
     parsed_json = _replace_units(json_ld, original_context, original_key_lookup_dict)
-    parsed_json['@context'] = original_context
+    parsed_json = {'@context': original_context, **parsed_json}
+    json_ld['@context'] = original_context # restore context
+    return parsed_json
+
+def export_units(json_ld: dict, context = processing_context) -> dict:
+    original_context = json_ld.pop('@context', context)
+    key_dict = {'@context': processing_context, 'unit': 'unit', 'value': 'value'}
+    # inverse expand-reverse cycle
+    expanded = jsonld.expand(key_dict, processing_context)
+    compacted = jsonld.compact(expanded, original_context)
+    # remove the context
+    del compacted['@context']
+    # reverse the dict
+    original_key_lookup_dict = {v: k for k, v in compacted.items()}
+    parsed_json = _serialize_units(json_ld, original_context, original_key_lookup_dict)
+    parsed_json = {'@context': original_context, **parsed_json}
+    json_ld['@context'] = original_context # restore context
     return parsed_json
diff --git a/tests/deserialization_test.py b/tests/020_deserialization_test.py
similarity index 58%
rename from tests/deserialization_test.py
rename to tests/020_deserialization_test.py
index 9f92d69..a401932 100644
--- a/tests/deserialization_test.py
+++ b/tests/020_deserialization_test.py
@@ -1,43 +1,7 @@
-import os
 import ontopint
-import json
 import pint
 
-def _load_test_data(file_name: str) -> dict:
-    """loads a json file from the test data folder
-
-    Parameters
-    ----------
-    file_name
-        the name of the json file
-
-    Returns
-    -------
-        the content of the file as dict
-    """
-    data = {}
-    with open(os.path.join(os.path.dirname(__file__), "data", file_name)) as f:
-        data = json.load(f)
-    return data
-
-def _recursive_items(dictionary: dict):
-    """Returns a generator of tuples for every key-value pair in the dict
-
-    Parameters
-    ----------
-    dictionary
-        any (nested) dict
-
-    Yields
-    ------
-        iterator for key-value tuples of the dict
-    """
-    for key, value in dictionary.items():
-        if type(value) is dict:
-            yield (key, value)
-            yield from _recursive_items(value)
-        else:
-            yield (key, value)
+from common import _load_test_data, _recursive_items
 
 def test_default_keys():
     """test input data with default keys 'value' and 'unit'
diff --git a/tests/030_serialization_test.py b/tests/030_serialization_test.py
new file mode 100644
index 0000000..e5d9bca
--- /dev/null
+++ b/tests/030_serialization_test.py
@@ -0,0 +1,47 @@
+import ontopint
+import deepdiff
+
+def test_default_keys():
+    """test input data with default keys 'value' and 'unit'
+    """
+
+    test = {
+        "value": ontopint.ureg.Quantity(
+            1.123, ontopint.ureg.from_ucum("eV")
+        )
+    }
+    expected = {
+        "value": 1.123,
+        "unit": "qunit:EV"
+    }
+    result = ontopint.export_units(test)
+    del result["@context"]
+    assert (len(deepdiff.DeepDiff(expected, result).keys()) == 0) # no diff
+
+def test_custom_keys():
+    """test input data with custom keys 'my_value' and 'my_unit'
+    """
+    test = {
+        "@context": {
+            "qudt": "http://qudt.org/schema/qudt/",
+            "qunit": "http://qudt.org/vocab/unit/",
+            "qkind": "http://qudt.org/vocab/quantkind/",
+            "my_unit": {
+                "@id": "qudt:hasUnit",
+                "@type": "@id"
+            },
+            "my_value": "qudt:value",
+        },
+        "my_value": ontopint.ureg.Quantity(
+            1.123, ontopint.ureg.from_ucum("eV")
+        )
+    }
+    expected = {
+        "my_value": 1.123,
+        "my_unit": "qunit:EV"
+    }
+    result = ontopint.export_units(test)
+    del result["@context"]
+    assert (len(deepdiff.DeepDiff(expected, result).keys()) == 0) # no diff
+
+    
\ No newline at end of file
diff --git a/tests/040_roundtrip_test.py b/tests/040_roundtrip_test.py
new file mode 100644
index 0000000..9dbbde0
--- /dev/null
+++ b/tests/040_roundtrip_test.py
@@ -0,0 +1,24 @@
+import ontopint
+import deepdiff
+
+from common import _load_test_data, _recursive_items
+
+def test_default_keys():
+    """test input data with default keys 'value' and 'unit'
+    """
+    input_jsonld = _load_test_data("test_data_default_keys.jsonld")
+    parsed_jsonld = ontopint.parse_units(input_jsonld)
+    result = ontopint.export_units(parsed_jsonld)
+    assert (len(deepdiff.DeepDiff(input_jsonld, result).keys()) == 0) # no diff
+
+def test_custom_keys():
+    """test input data with custom keys 'my_value' and 'my_unit'
+    """
+    input_jsonld = _load_test_data("test_data_custom_keys.jsonld")
+    parsed_jsonld = ontopint.parse_units(input_jsonld)
+    result = ontopint.export_units(parsed_jsonld)
+    diff = deepdiff.DeepDiff(input_jsonld, result)
+    print(deepdiff.DeepDiff(input_jsonld, result))
+    assert (len(deepdiff.DeepDiff(input_jsonld, result).keys()) == 0) # no diff
+
+test_custom_keys()
\ No newline at end of file
diff --git a/tests/common.py b/tests/common.py
new file mode 100644
index 0000000..874bf6d
--- /dev/null
+++ b/tests/common.py
@@ -0,0 +1,39 @@
+import json
+import os
+
+
+def _load_test_data(file_name: str) -> dict:
+    """loads a json file from the test data folder
+
+    Parameters
+    ----------
+    file_name
+        the name of the json file
+
+    Returns
+    -------
+        the content of the file as dict
+    """
+    data = {}
+    with open(os.path.join(os.path.dirname(__file__), "data", file_name)) as f:
+        data = json.load(f)
+    return data
+
+def _recursive_items(dictionary: dict):
+    """Returns a generator of tuples for every key-value pair in the dict
+
+    Parameters
+    ----------
+    dictionary
+        any (nested) dict
+
+    Yields
+    ------
+        iterator for key-value tuples of the dict
+    """
+    for key, value in dictionary.items():
+        if type(value) is dict:
+            yield (key, value)
+            yield from _recursive_items(value)
+        else:
+            yield (key, value)
\ No newline at end of file