Merge

Author: cthoyt

docs/conf.py

@@ -4,7 +4,6 @@
 # https://www.sphinx-doc.org/en/master/usage/configuration.html
 
 import os
-import re
 import sys
 from datetime import date
 from sssom import __version__
@@ -13,9 +12,9 @@
 # https://www.sphinx-doc.org/en/master/usage/configuration.html#project-information
 sys.path.insert(0, os.path.abspath("../"))
 
-project = 'sssom-py'
+project = "sssom-py"
 copyright = f"{date.today().year}, Chris Mungall; Nicolas Matentzoglu; Harshad Hegde"
-author = 'Chris Mungall; Nicolas Matentzoglu; Harshad Hegde'
+author = "Chris Mungall; Nicolas Matentzoglu; Harshad Hegde"
 release = __version__
 
 # -- General configuration ---------------------------------------------------
@@ -78,7 +77,7 @@
     "github_repo": "sssom-py",
     "github_version": "master/docs/",
     "display_github": True,  # Add 'Edit on Github' link instead of 'View page source'
-    "conf_py_path": "/master/docs/",   # Path in the checkout to the docs root
+    "conf_py_path": "/master/docs/",  # Path in the checkout to the docs root
 }
 
 # The name of an image file (relative to this directory) to place at the top

src/sssom/cli.py

@@ -251,7 +251,9 @@ def split(input: str, output_directory: str, method: SplitMethod) -> None:
     type=click.FloatRange(0, 1),
     help="Default confidence to be assigned if absent.",
 )
-def ptable(input: str, output: TextIO, inverse_factor: float, default_confidence: float) -> None:
+def ptable(
+    input: str, output: TextIO, inverse_factor: float, default_confidence: float
+) -> None:
     """Convert an SSSOM file to a ptable for kboom/`boomer <https://github.com/INCATools/boomer>`_."""
     # TODO should maybe move to boomer (but for now it can live here, so cjm can tweak
     msdf = parse_sssom_table(input)
@@ -301,7 +303,9 @@ def dosql(query: str, inputs: List[str], output: TextIO) -> None:
         run_sql_query(query=query, inputs=inputs, output=output)
     except ModuleNotFoundError as e:
         if e.name == "pansql":
-            raise click.ClickException("The dosql command requires the optional pansql module.")
+            raise click.ClickException(
+                "The dosql command requires the optional pansql module."
+            )
         raise
 
     # n = 1
@@ -458,7 +462,9 @@ def cliquesummary(input: str, output: TextIO, metadata: str, statsfile: str) ->
 @output_option
 @transpose_option
 @fields_option
-def crosstab(input: str, output: TextIO, transpose: bool, fields: Tuple[str, str]) -> None:
+def crosstab(
+    input: str, output: TextIO, transpose: bool, fields: Tuple[str, str]
+) -> None:
     """Write sssom summary cross-tabulated by categories."""
     df = remove_unmatched(parse_sssom_table(input).df)
     logging.info(f"#CROSSTAB ON {fields}")
@@ -474,12 +480,16 @@ def crosstab(input: str, output: TextIO, transpose: bool, fields: Tuple[str, str
 @transpose_option
 @fields_option
 @input_argument
-def correlations(input: str, output: TextIO, transpose: bool, fields: Tuple[str, str]) -> None:
+def correlations(
+    input: str, output: TextIO, transpose: bool, fields: Tuple[str, str]
+) -> None:
     """Calculate correlations."""
     try:
         from scipy.stats import chi2_contingency
     except ModuleNotFoundError:
-        raise click.ClickException("The correlations command requires the optional scipy module.")
+        raise click.ClickException(
+            "The correlations command requires the optional scipy module."
+        )
 
     msdf = parse_sssom_table(input)
     df = remove_unmatched(msdf.df)
@@ -645,7 +655,9 @@ def filter(input: str, output: TextIO, **kwargs: Any) -> None:
         filter_file(input=input, output=output, **kwargs)
     except ModuleNotFoundError as e:
         if e.name == "pansql":
-            raise click.ClickException("The filter command requires the pansql optional module.")
+            raise click.ClickException(
+                "The filter command requires the pansql optional module."
+            )
         raise
 
 
@@ -664,9 +676,13 @@ def filter(input: str, output: TextIO, **kwargs: Any) -> None:
     help="Multivalued slots should be replaced or not.",
 )
 @dynamically_generate_sssom_options(SSSOM_SV_OBJECT.mapping_set_slots)
-def annotate(input: str, output: TextIO, replace_multivalued: bool, **kwargs: Any) -> None:
+def annotate(
+    input: str, output: TextIO, replace_multivalued: bool, **kwargs: Any
+) -> None:
     """Annotate metadata of a mapping set."""
-    annotate_file(input=input, output=output, replace_multivalued=replace_multivalued, **kwargs)
+    annotate_file(
+        input=input, output=output, replace_multivalued=replace_multivalued, **kwargs
+    )
 
 
 @main.command()
@@ -706,7 +722,9 @@ def remove(input: str, output: TextIO, remove_map: str) -> None:
     is_flag=True,
     help="If True (default), the justification is updated to 'sempav:MappingInversion', else it is left as it is.",
 )
-@click.option("--inverse-map", help="Path to file that contains the inverse predicate dictionary.")
+@click.option(
+    "--inverse-map", help="Path to file that contains the inverse predicate dictionary."
+)
 def invert(
     input: str,
     output: TextIO,

src/sssom/cliquesummary.py

@@ -5,17 +5,13 @@
 """
 
 import dataclasses
-import re
-import sys
 from dataclasses import dataclass
 from typing import Any, ClassVar, Dict, List, Optional, Union
 
-from jsonasobj2 import JsonObj, as_dict
-from linkml_runtime.linkml_model.meta import EnumDefinition, PermissibleValue, PvFormulaOptions
 from linkml_runtime.utils.curienamespace import CurieNamespace
-from linkml_runtime.utils.dataclass_extensions_376 import dataclasses_init_fn_with_kwargs
-from linkml_runtime.utils.enumerations import EnumDefinitionImpl
-from linkml_runtime.utils.formatutils import camelcase, sfx, underscore
+from linkml_runtime.utils.dataclass_extensions_376 import (
+    dataclasses_init_fn_with_kwargs,
+)
 from linkml_runtime.utils.metamodelcore import (
     URI,
     Bool,
@@ -27,13 +23,14 @@
     XSDDate,
     XSDDateTime,
     XSDTime,
-    bnode,
-    empty_dict,
     empty_list,
 )
 from linkml_runtime.utils.slot import Slot
-from linkml_runtime.utils.yamlutils import YAMLRoot, extended_float, extended_int, extended_str
-from rdflib import Namespace, URIRef
+from linkml_runtime.utils.yamlutils import (
+    YAMLRoot,
+    extended_str,
+)
+from rdflib import URIRef
 
 metamodel_version = "1.7.0"
 
@@ -232,54 +229,78 @@ def __post_init__(self, *_: List[str], **kwargs: Dict[str, Any]):
         self.members = [v if isinstance(v, str) else str(v) for v in self.members]
 
         if not isinstance(self.members_labels, list):
-            self.members_labels = [self.members_labels] if self.members_labels is not None else []
-        self.members_labels = [v if isinstance(v, str) else str(v) for v in self.members_labels]
+            self.members_labels = (
+                [self.members_labels] if self.members_labels is not None else []
+            )
+        self.members_labels = [
+            v if isinstance(v, str) else str(v) for v in self.members_labels
+        ]
 
         if self.num_members is not None and not isinstance(self.num_members, int):
             self.num_members = int(self.num_members)
 
-        if self.max_confidence is not None and not isinstance(self.max_confidence, float):
+        if self.max_confidence is not None and not isinstance(
+            self.max_confidence, float
+        ):
             self.max_confidence = float(self.max_confidence)
 
-        if self.min_confidence is not None and not isinstance(self.min_confidence, float):
+        if self.min_confidence is not None and not isinstance(
+            self.min_confidence, float
+        ):
             self.min_confidence = float(self.min_confidence)
 
-        if self.avg_confidence is not None and not isinstance(self.avg_confidence, float):
+        if self.avg_confidence is not None and not isinstance(
+            self.avg_confidence, float
+        ):
             self.avg_confidence = float(self.avg_confidence)
 
         if self.is_conflated is not None and not isinstance(self.is_conflated, Bool):
             self.is_conflated = Bool(self.is_conflated)
 
-        if self.is_all_conflated is not None and not isinstance(self.is_all_conflated, Bool):
+        if self.is_all_conflated is not None and not isinstance(
+            self.is_all_conflated, Bool
+        ):
             self.is_all_conflated = Bool(self.is_all_conflated)
 
-        if self.total_conflated is not None and not isinstance(self.total_conflated, int):
+        if self.total_conflated is not None and not isinstance(
+            self.total_conflated, int
+        ):
             self.total_conflated = int(self.total_conflated)
 
         if self.proportion_conflated is not None and not isinstance(
             self.proportion_conflated, float
         ):
             self.proportion_conflated = float(self.proportion_conflated)
 
-        if self.conflation_score is not None and not isinstance(self.conflation_score, float):
+        if self.conflation_score is not None and not isinstance(
+            self.conflation_score, float
+        ):
             self.conflation_score = float(self.conflation_score)
 
         if self.members_count is not None and not isinstance(self.members_count, int):
             self.members_count = int(self.members_count)
 
-        if self.min_count_by_source is not None and not isinstance(self.min_count_by_source, int):
+        if self.min_count_by_source is not None and not isinstance(
+            self.min_count_by_source, int
+        ):
             self.min_count_by_source = int(self.min_count_by_source)
 
-        if self.max_count_by_source is not None and not isinstance(self.max_count_by_source, int):
+        if self.max_count_by_source is not None and not isinstance(
+            self.max_count_by_source, int
+        ):
             self.max_count_by_source = int(self.max_count_by_source)
 
-        if self.avg_count_by_source is not None and not isinstance(self.avg_count_by_source, float):
+        if self.avg_count_by_source is not None and not isinstance(
+            self.avg_count_by_source, float
+        ):
             self.avg_count_by_source = float(self.avg_count_by_source)
 
         if self.harmonic_mean_count_by_source is not None and not isinstance(
             self.harmonic_mean_count_by_source, float
         ):
-            self.harmonic_mean_count_by_source = float(self.harmonic_mean_count_by_source)
+            self.harmonic_mean_count_by_source = float(
+                self.harmonic_mean_count_by_source
+            )
 
         super().__post_init__(**kwargs)
 

src/sssom/constants.py

@@ -260,7 +260,11 @@ def multivalued_slots(self) -> Set[str]:
     @cached_property
     def entity_reference_slots(self) -> Set[str]:
         """Return set of entity reference slots."""
-        return {c for c in self.view.all_slots() if self.view.get_slot(c).range == ENTITY_REFERENCE}
+        return {
+            c
+            for c in self.view.all_slots()
+            if self.view.get_slot(c).range == ENTITY_REFERENCE
+        }
 
     @cached_property
     def mapping_enum_keys(self) -> Set[str]:
@@ -292,7 +296,9 @@ def propagatable_slots(self) -> List[str]:
 def _get_sssom_schema_object() -> SSSOMSchemaView:
     """Get a view over the SSSOM schema."""
     sssom_sv_object = (
-        SSSOMSchemaView.instance if hasattr(SSSOMSchemaView, "instance") else SSSOMSchemaView()
+        SSSOMSchemaView.instance
+        if hasattr(SSSOMSchemaView, "instance")
+        else SSSOMSchemaView()
     )
     return sssom_sv_object
 

src/sssom/context.py

@@ -71,7 +71,9 @@ def _get_built_in_prefix_map() -> Converter:
 ConverterHint = Union[None, Mapping[str, str], Converter]
 
 
-def ensure_converter(prefix_map: ConverterHint = None, *, use_defaults: bool = True) -> Converter:
+def ensure_converter(
+    prefix_map: ConverterHint = None, *, use_defaults: bool = True
+) -> Converter:
     """Ensure a converter is available.
 
     :param prefix_map: One of the following: