Implement new version of FastRun

Author: LeMyst

README.md

@@ -773,7 +773,15 @@ for entrez_id, ensembl in raw_data.items():
 
 Note: Fastrun mode checks for equality of property/value pairs, qualifiers (not including qualifier attributes), labels,
 aliases and description, but it ignores references by default!
-References can be checked in fast run mode by setting `use_refs` to `True`.
+References can be checked in fast run mode by setting `use_references` to `True`.
+
+# Statistics #
+
+| Dataset                     | partial fastrun | fastrun without qualifiers/references | fastrun with qualifiers | fastrun with qualifiers/references |
+|:----------------------------|----------------:|--------------------------------------:|------------------------:|-----------------------------------:|
+| Communes (34990 elements)   |               ? |                                  7min |                     30s |                                60s |
+| Cantons (2042 elements)     |               ? |                                     ? |                       ? |                                  ? |
+| Départements (100 elements) |           70min |                                    1s |                     30s |                                60s |
 
 # Debugging #
 

pyproject.toml

@@ -34,3 +34,4 @@ disable = [
 
 [tool.pytest.ini_options]
 log_cli = true
+log_cli_level = 'DEBUG'

test/test_all.py

@@ -1,12 +1,12 @@
 import copy
+import logging
 import unittest
 
 from wikibaseintegrator import WikibaseIntegrator, datatypes, wbi_fastrun
 from wikibaseintegrator.datatypes import BaseDataType, Item
 from wikibaseintegrator.entities import ItemEntity
 from wikibaseintegrator.wbi_config import config as wbi_config
-from wikibaseintegrator.wbi_enums import ActionIfExists, WikibaseDatatype
-from wikibaseintegrator.wbi_fastrun import get_fastrun_container
+from wikibaseintegrator.wbi_enums import WikibaseDatatype
 
 wbi_config['USER_AGENT'] = 'WikibaseIntegrator-pytest/1.0 (test_all.py)'
 
@@ -53,66 +53,22 @@ class TestFastRun(unittest.TestCase):
     """
     some basic tests for fastrun mode
     """
+    logging.basicConfig(level=logging.DEBUG)
 
     def test_fastrun(self):
         statements = [
-            datatypes.ExternalID(value='P40095', prop_nr='P352'),
+            datatypes.ExternalID(value='A0A023PZB3', prop_nr='P352'),
             datatypes.ExternalID(value='YER158C', prop_nr='P705')
         ]
 
         frc = wbi_fastrun.FastRunContainer(base_filter=[BaseDataType(prop_nr='P352'), datatypes.Item(prop_nr='P703', value='Q27510868')], base_data_type=datatypes.BaseDataType)
 
-        fastrun_result = frc.write_required(data=statements)
-
-        if fastrun_result:
-            message = 'fastrun failed'
-        else:
-            message = 'successful fastrun'
+        fastrun_result = frc.write_required(claims=statements)
 
         # here, fastrun should succeed, if not, test failed
-        if fastrun_result:
+        if not fastrun_result:
             raise ValueError
 
-    def test_fastrun_label(self):
-        # tests fastrun label, description and aliases, and label in another language
-        frc = get_fastrun_container(base_filter=[datatypes.ExternalID(value='/m/02j71', prop_nr='P646')])
-        item = WikibaseIntegrator().item.get('Q2')
-
-        assert item.labels.get(language='en') == "Earth"
-        descr = item.descriptions.get(language='en')
-        assert len(descr) > 3
-        assert "the Earth" in item.aliases.get()
-
-        assert list(frc.get_language_data("Q2", 'en', 'label'))[0] == item.labels.get(language='en')
-        assert frc.check_language_data("Q2", ['not the Earth'], 'en', 'label')
-        assert "the Earth" in item.aliases.get()
-        assert "planet" in item.descriptions.get()
-
-        assert item.labels.get('es') == "Tierra"
-
-        item.descriptions.set(value=descr)
-        item.descriptions.set(value="fghjkl")
-        assert item.get_json()['descriptions']['en'] == {'language': 'en', 'value': 'fghjkl'}
-        item.labels.set(value="Earth")
-        item.labels.set(value="xfgfdsg")
-        assert item.get_json()['labels']['en'] == {'language': 'en', 'value': 'xfgfdsg'}
-        item.aliases.set(values=["fake alias"], action_if_exists=ActionIfExists.APPEND_OR_REPLACE)
-        assert {'language': 'en', 'value': 'fake alias'} in item.get_json()['aliases']['en']
-
-        # something that's empty (for now.., can change, so this just makes sure no exception is thrown)
-        frc.check_language_data("Q2", ['Ewiase'], 'ak', 'label')
-        frc.check_language_data("Q2", ['not Ewiase'], 'ak', 'label')
-        frc.check_language_data("Q2", [''], 'ak', 'description')
-        frc.check_language_data("Q2", [], 'ak', 'aliases')
-        frc.check_language_data("Q2", ['sdf', 'sdd'], 'ak', 'aliases')
-
-        item.labels.get(language="ak")
-        item.descriptions.get(language='ak')
-        item.aliases.get(language="ak")
-        item.labels.set(value="label", language="ak")
-        item.descriptions.set(value="d", language="ak")
-        item.aliases.set(values=["a"], language="ak", action_if_exists=ActionIfExists.APPEND_OR_REPLACE)
-
 
 def test_sitelinks():
     item = wbi.item.get('Q622901')

test/test_entity_item.py

@@ -50,18 +50,18 @@ def test_write(self):
     def test_write_not_required(self):
         assert not wbi.item.get('Q582').write_required(base_filter=[BaseDataType(prop_nr='P1791')])
 
+    def test_write_not_required_ref(self):
+        assert not wbi.item.get('Q582').write_required(base_filter=[BaseDataType(prop_nr='P1464')], use_references=True)
+
     def test_write_required(self):
         item = wbi.item.get('Q582')
         item.claims.add(Item(prop_nr='P1791', value='Q42'))
         assert item.write_required([BaseDataType(prop_nr='P1791')])
 
-    def test_write_not_required_ref(self):
-        assert not wbi.item.get('Q582').write_required(base_filter=[BaseDataType(prop_nr='P2581')], use_refs=True)
-
     def test_write_required_ref(self):
         item = wbi.item.get('Q582')
-        item.claims.get('P2581')[0].references.references.pop()
-        assert item.write_required(base_filter=[BaseDataType(prop_nr='P2581')], use_refs=True)
+        item.claims.get('P1464')[0].references.references.pop()
+        assert item.write_required(base_filter=[BaseDataType(prop_nr='P1464')], use_references=True)
 
     def test_long_item_id(self):
         assert wbi.item.get('Item:Q582').id == 'Q582'

test/test_wbi_core.py

@@ -240,7 +240,7 @@ def test_new_item_creation(self):
             MonolingualText(text='xxx', language='fr', prop_nr='P7'),
             Quantity(amount=-5.04, prop_nr='P8'),
             Quantity(amount=5.06, upper_bound=9.99, lower_bound=-2.22, unit='Q11573', prop_nr='P8'),
-            CommonsMedia(value='xxx', prop_nr='P9'),
+            CommonsMedia(value="Place lazare goujon.jpg", prop_nr='P9'),
             GlobeCoordinate(latitude=1.2345, longitude=-1.2345, precision=12, prop_nr='P10'),
             GeoShape(value='Data:xxx.map', prop_nr='P11'),
             Property(value='P123', prop_nr='P12'),

test/test_wbi_fastrun.py

@@ -11,6 +11,7 @@ class BaseDataType(Claim):
     The base class for all Wikibase data types, they inherit from it
     """
     DTYPE = 'base-data-type'
+    PTYPE = 'property-data-type'
     subclasses: list[type[BaseDataType]] = []
     sparql_query: str = '''
         SELECT * WHERE {{
@@ -28,7 +29,14 @@ def __init__(self, prop_nr: int | str | None = None, **kwargs: Any):
 
         super().__init__(**kwargs)
 
-        self.mainsnak.property_number = prop_nr or None
+        if isinstance(prop_nr, str):
+            pattern = re.compile(r'^([a-z][a-z\d+.-]*):([^][<>\"\x00-\x20\x7F])+$')
+            matches = pattern.match(str(prop_nr))
+
+            if matches:
+                prop_nr = prop_nr.rsplit('/', 1)[-1]
+
+        self.mainsnak.property_number = prop_nr
         # self.subclasses.append(self)
 
     # Allow registration of subclasses of BaseDataType into BaseDataType.subclasses
@@ -39,7 +47,7 @@ def __init_subclass__(cls, **kwargs):
     def set_value(self, value: Any | None = None):
         pass
 
-    def get_sparql_value(self) -> str:
+    def get_sparql_value(self, **kwargs: Any) -> str | None:
         return '"' + self.mainsnak.datavalue['value'] + '"'
 
     def parse_sparql_value(self, value, type='literal', unit='1') -> bool:
@@ -61,3 +69,6 @@ def parse_sparql_value(self, value, type='literal', unit='1') -> bool:
             raise ValueError
 
         return True
+
+    def from_sparql_value(self, sparql_value: dict) -> BaseDataType:  # type: ignore
+        pass