aboutUrl function

csvqb/csvqb/tests/unit/writers/test_qbwriter.py

@@ -476,6 +476,65 @@ def test_virtual_columns_generated_for_multi_meas_obs_val():
     )
     assert "./cube-name.csv#unit/some-unit" == virt_unit["valueUrl"]
 
+def test_about_url_generation():
+    """
+        Ensuring that when an aboutUrl is defined for a non-multimeasure cube, the resulting URL
+        is built in the order in which dimensions appear in the cube.
+    """
+    data = pd.DataFrame({
+        "Existing Dimension": ["A", "B", "C"],
+        "Local Dimension": ["D", "E", "F"],
+        "Value": [2, 2, 2]
+    })
+
+    metadata = CatalogMetadata("Some Dataset")
+    columns = [
+        QbColumn("Existing Dimension",
+                    ExistingQbDimension("https://example.org/dimensions/existing_dimension")),
+        QbColumn("Local Dimension", 
+                    NewQbDimension.from_data("Name of New Dimension", 
+                    data["Local Dimension"])),
+        QbColumn("Value", 
+                    QbSingleMeasureObservationValue(ExistingQbMeasure("http://example.com/measures/existing_measure"), 
+                    NewQbUnit("New Unit")))
+
+    ]
+
+    cube = Cube(metadata, data, columns)
+
+    actual_about_url = QbWriter(cube)._get_about_url()
+    expected_about_url = "./some-dataset.csv#obs/{+existing_dimension}/{+local_dimension}"
+    assert actual_about_url == expected_about_url
+
+def test_about_url_generation_with_multiple_measures():
+    """
+        Ensuring that when an aboutUrl is defined for a multimeasure cube, the resulting URL
+        is built in the order in which dimensions appear in the cube except for the multi-measure
+        dimensions which are appended to the end of the URL.
+    """
+    data = pd.DataFrame({
+        "Measure": ["People", "Children", "Adults"],
+        "Existing Dimension": ["A", "B", "C"],
+        "Value": [2, 2, 2],
+        "Local Dimension": ["D", "E", "F"],
+        "Units": ["Percent", "People", "People"]
+    })
+
+    metadata = CatalogMetadata("Some Dataset")
+    columns = [
+        QbColumn("Measure", QbMultiMeasureDimension.new_measures_from_data(data["Measure"])),
+        QbColumn("Existing Dimension", ExistingQbDimension("https://example.org/dimensions/existing_dimension")),
+        QbColumn("Local Dimension", NewQbDimension.from_data("Name of New Dimension", data["Local Dimension"])),
+        QbColumn("Value", QbMultiMeasureObservationValue("number")),
+        QbColumn("Units", QbMultiUnits.new_units_from_data(data["Units"]))
+
+    ]
+
+    cube = Cube(metadata, data, columns)
+
+    actual_about_url = QbWriter(cube)._get_about_url()
+    expected_about_url = "./some-dataset.csv#obs/{+existing_dimension}/{+local_dimension}/{+measure}"
+    assert actual_about_url == expected_about_url
 
 if __name__ == "__main__":
     pytest.main()

csvqb/csvqb/writers/qbwriter.py

@@ -4,14 +4,13 @@
 from pathlib import Path
 from typing import Optional, Tuple, Dict, Any, List, Iterable, Callable
 import rdflib
-from sharedmodels.rdf import qb, skos
+from sharedmodels.rdf import qb, rdfs, skos, namespaces
 from sharedmodels.rdf.resource import (
     Resource,
     ExistingResource,
     maybe_existing_resource,
 )
 
-
 from csvqb.models.cube import *
 from csvqb.utils.uri import get_last_uri_part, csvw_column_name_safe, looks_like_uri
 from csvqb.utils.qb.cube import get_columns_of_dsd_type
@@ -20,6 +19,7 @@
 from .writerbase import WriterBase
 from ..models.rdf.qbdatasetincatalog import QbDataSetInCatalog
 
+
 VIRT_UNIT_COLUMN_NAME = "virt_unit"
 
 
@@ -549,4 +549,21 @@ def _get_measure_uri(self, measure: QbMeasure) -> str:
         elif isinstance(measure, NewQbMeasure):
             return self._doc_rel_uri(f"measure/{measure.uri_safe_identifier}")
         else:
-            raise Exception(f"Unmatched unit type {type(unit)}")
+            raise Exception(f"Unmatched measure type {type(measure)}")
+
+    def _get_about_url(self) -> str:
+        # Todo: Dimensions are currently appended in the order in which the appear in the cube.
+        #       We may want to alter this in the future so that the ordering is from
+        #       least entropic dimension -> most entropic.
+        #       E.g. http://base-uri/observations/male/1996/all-males-1996
+        aboutUrl = self._doc_rel_uri("obs")
+        multi_measure_col = ""
+        for c in self.cube.columns:
+            if isinstance(c, QbColumn):
+                if isinstance(c.component, QbDimension):
+                    aboutUrl = aboutUrl + f"/{{+{csvw_column_name_safe(c.uri_safe_identifier)}}}"
+                elif isinstance(c.component, QbMultiMeasureDimension):
+                    multi_measure_col = csvw_column_name_safe(c.uri_safe_identifier) 
+        if len(multi_measure_col) != 0:
+            aboutUrl = aboutUrl + f"/{{+{multi_measure_col}}}"
+        return aboutUrl