Parse values

glowtables/table.py

@@ -23,6 +23,8 @@ from typing import Any, Callable, Generic, Iterator, NewType, Optional, TypeVar
 
 import rdflib
 
+from glowtables.sparql import SparqlBackend
+
 Language = NewType("Language", str)
 """ISO 639-1 code"""
 
@@ -42,6 +44,9 @@ class Field(abc.ABC, Generic[_TFieldValue]):
     display_names: dict[Language, str]
     """Localized name for the field (eg. in a table header)"""
 
+    parse: Callable[[str], _TFieldValue]
+    """Parses a string returned by a SPARQL query to a native Python value."""
+
     @abc.abstractmethod
     def sort_key(self, value: _TFieldValue):
         """Function suitable as ``key`` argument to :func:`sorted`.
@@ -72,18 +77,18 @@ class Field(abc.ABC, Generic[_TFieldValue]):
 
 
 @dataclasses.dataclass
-class LiteralField(Field[rdflib.Literal]):
+class LiteralField(Field[_TFieldValue], Generic[_TFieldValue]):
     """Simplest field: its value is a literal directly on the subject"""
 
     predicate: rdflib.URIRef
 
-    default: Optional[rdflib.Literal] = None
+    default: Optional[_TFieldValue] = None
     """If this is not :const:`None`, allows subjects without a statement for this field;
     and use this value instead when sorting.
 
     This is only used when sorting, and isn't displayed."""
 
-    def sort_key(self, value: rdflib.Literal) -> Any:
+    def sort_key(self, value: Optional[_TFieldValue]) -> Any:
         """Function suitable as ``key`` argument to :func:`sorted`.
 
         Defaults to the identity function."""
@@ -176,3 +181,13 @@ class Table:
             constraints=constraints,
             statements=statements.strip(),
         )
+
+    def query(self, backend: SparqlBackend) -> Iterator[tuple]:
+        """Returns a list of all rows of the table. Each row has exactly one cell for
+        each column defined in :attr:`fields`.
+        """
+        for row in backend.query(self.sparql()):
+            yield tuple(
+                None if cell is None else field.parse(cell)
+                for (field, cell) in zip(self.fields, row)
+            )

glowtables/tests/table_test.py

@@ -18,6 +18,7 @@ Test cases use a graph containing a few CPUs, with names and clock frequencies.
 """
 
 import textwrap
+from decimal import Decimal
 
 import pytest
 import rdflib
@@ -64,6 +65,7 @@ def test_single_literal(rdflib_sparql: SparqlBackend) -> None:
     name_field = LiteralField(
         "display_name",
         {Language("en"): "Name"},
+        str,
         rdflib.URIRef("http://example.org/display-name"),
     )
     table = Table(
@@ -81,7 +83,7 @@ def test_single_literal(rdflib_sparql: SparqlBackend) -> None:
         """
     )
 
-    assert set(rdflib_sparql.query(table.sparql())) == {
+    assert set(table.query(rdflib_sparql)) == {
         ("Grown 1700",),
         ("Grown 3600",),
     }
@@ -91,11 +93,13 @@ def test_two_literals(rdflib_sparql: SparqlBackend) -> None:
     name_field = LiteralField(
         "display_name",
         {Language("en"): "Name"},
+        str,
         rdflib.URIRef("http://example.org/display-name"),
     )
     frequency_field = LiteralField(
         "frequency",
         {Language("en"): "Clock frequency"},
+        Decimal,
         rdflib.URIRef("http://example.org/clock-frequency"),
     )
     table = Table(
@@ -114,8 +118,8 @@ def test_two_literals(rdflib_sparql: SparqlBackend) -> None:
         """
     )
 
-    assert set(rdflib_sparql.query(table.sparql())) == {
+    assert set(table.query(rdflib_sparql)) == {
-        ("Grown 1700", "3000"),
+        ("Grown 1700", 3000),
     }
 
 
@@ -123,14 +127,16 @@ def test_default_value(rdflib_sparql: SparqlBackend) -> None:
     name_field = LiteralField(
         "display_name",
         {Language("en"): "Name"},
+        str,
         rdflib.URIRef("http://example.org/display-name"),
-        default=rdflib.Literal("Anonymous CPU"),
+        default="Anonymous CPU",
     )
     frequency_field = LiteralField(
         "frequency",
         {Language("en"): "Clock frequency"},
+        Decimal,
         rdflib.URIRef("http://example.org/clock-frequency"),
-        default=rdflib.Literal(0),
+        default=Decimal(0),
     )
     table = Table(
         fields=[name_field, frequency_field],
@@ -148,10 +154,10 @@ def test_default_value(rdflib_sparql: SparqlBackend) -> None:
         """
     )
 
-    assert set(rdflib_sparql.query(table.sparql())) == {
+    assert set(table.query(rdflib_sparql)) == {
-        ("Grown 1700", "3000"),
+        ("Grown 1700", 3000),
         ("Grown 3600", None),
-        (None, "9000"),
+        (None, 9000),
     }
 
 
@@ -159,6 +165,7 @@ def test_field_id_subject() -> None:
     name_field = LiteralField(
         "subject",
         {Language("en"): "Name"},
+        str,
         rdflib.URIRef("http://example.org/display-name"),
     )
     with pytest.raises(ValueError, match="both subject and a field id"):
@@ -172,11 +179,13 @@ def test_field_id_clash() -> None:
     name_field = LiteralField(
         "name",
         {Language("en"): "Name"},
+        str,
         rdflib.URIRef("http://example.org/name"),
     )
     display_name_field = LiteralField(
         "name",
         {Language("en"): "Display Name"},
+        str,
         rdflib.URIRef("http://example.org/display-name"),
     )
     with pytest.raises(ValueError, match="has duplicate field ids: name"):