Use inclusive language [removing sanity check] (#984)

* use inclusive language

* use self-consistency check

* fix typo

* add typehint

* lint

* update mypy.ini

* tighten mypy.ini

* use assert for internal functions

Co-authored-by: Selene Chew <[email protected]>

Author: chewselene

graphql_compiler/compiler/compiler_frontend.py

@@ -543,8 +543,9 @@ def _compile_vertex_ast(
             # Invariant: There must always be a marked location corresponding to the query position
             # immediately before any optional Traverse.
             #
-            # This invariant is verified in the IR sanity checks module (ir_sanity_checks.py),
-            # in the function named _sanity_check_mark_location_preceding_optional_traverse().
+            # This invariant is verified in the IR self-consistency check module
+            # (ir_self_consistency_checks.py), in the function named
+            # _assert_mark_location_preceding_optional_traverse().
             #
             # This marked location is the one that the @optional directive's corresponding
             # optional Backtrack will jump back to. If such a marked location isn't present,
@@ -779,13 +780,13 @@ def _compile_ast_node_to_ir(schema, current_schema_type, ast, location, context)
         )
 
     if location.field is not None:  # we're at a property field
-        # sanity-check: cannot have an inline fragment at a property field
+        # self-consistency check: cannot have an inline fragment at a property field
         if fragment_exists:
             raise AssertionError(
                 "Found inline fragment at a property field: {} {}".format(location, fragment)
             )
 
-        # sanity-check: locations at properties don't have their own property locations
+        # self-consistency check: locations at properties don't have their own property locations
         if len(property_fields) > 0:
             raise AssertionError(
                 "Found property fields on a property field: "

graphql_compiler/compiler/cypher_query.py

@@ -154,7 +154,7 @@ def _generate_cypher_step_list_from_ir_blocks(fold_scope_location, ir_blocks, qu
         linked_location = block.location
         current_step_ir_blocks = []
     for block in current_step_ir_blocks:
-        # Sanity check any leftover blocks.
+        # Self-consistency check any leftover blocks.
         # The complete list of IR blocks need not end with a MarkLocation block, e.g. if you
         # have a Backtrack block at the end. That's fine; Backtrack is unused here anyways.
         # What we want to avoid is having any leftover blocks here that we should've used.

graphql_compiler/compiler/ir_lowering_cypher/__init__.py

@@ -5,7 +5,7 @@
     merge_consecutive_filter_clauses,
     optimize_boolean_expression_comparisons,
 )
-from ..ir_sanity_checks import sanity_check_ir_blocks_from_frontend
+from ..ir_self_consistency_checks import self_consistency_check_ir_blocks_from_frontend
 from .ir_lowering import (
     insert_explicit_type_bounds,
     move_filters_in_optional_locations_to_global_operations,
@@ -30,7 +30,7 @@ def lower_ir(schema_info, ir):
     Returns:
         CypherQuery object
     """
-    sanity_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
+    self_consistency_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
 
     ir_blocks = insert_explicit_type_bounds(
         ir.ir_blocks,

graphql_compiler/compiler/ir_lowering_gremlin/__init__.py

@@ -4,7 +4,7 @@
     merge_consecutive_filter_clauses,
     optimize_boolean_expression_comparisons,
 )
-from ..ir_sanity_checks import sanity_check_ir_blocks_from_frontend
+from ..ir_self_consistency_checks import self_consistency_check_ir_blocks_from_frontend
 from .ir_lowering import (
     lower_coerce_type_block_type_data,
     lower_coerce_type_blocks,
@@ -28,7 +28,7 @@ def lower_ir(schema_info, ir):
     Returns:
         list of IR blocks suitable for outputting as Gremlin
     """
-    sanity_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
+    self_consistency_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
 
     ir_blocks = lower_context_field_existence(ir.ir_blocks, ir.query_metadata_table)
     ir_blocks = optimize_boolean_expression_comparisons(ir_blocks)

graphql_compiler/compiler/ir_lowering_match/__init__.py

@@ -12,7 +12,7 @@
     optimize_boolean_expression_comparisons,
     remove_end_optionals,
 )
-from ..ir_sanity_checks import sanity_check_ir_blocks_from_frontend
+from ..ir_self_consistency_checks import self_consistency_check_ir_blocks_from_frontend
 from ..match_query import MatchQuery, convert_to_match_query
 from ..workarounds import (
     orientdb_class_with_while,
@@ -53,7 +53,7 @@ def lower_ir(schema_info: CommonSchemaInfo, ir: IrAndMetadata) -> MatchQuery:
     Returns:
         MatchQuery object containing the IR blocks organized in a MATCH-like structure
     """
-    sanity_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
+    self_consistency_check_ir_blocks_from_frontend(ir.ir_blocks, ir.query_metadata_table)
 
     # Construct the mapping of each location to its corresponding GraphQL type.
     location_types = {

graphql_compiler/compiler/ir_sanity_checks.py renamed to graphql_compiler/compiler/ir_self_consistency_checks.py

@@ -23,14 +23,14 @@
 from .metadata import QueryMetadataTable
 
 
-def sanity_check_ir_blocks_from_frontend(
+def self_consistency_check_ir_blocks_from_frontend(
     ir_blocks: List[BasicBlock], query_metadata_table: QueryMetadataTable
 ) -> None:
     """Assert that IR blocks originating from the frontend do not have nonsensical structure.
 
     Args:
-        ir_blocks: list of BasicBlocks representing the IR to sanity-check
-        query_metadata_table: QueryMetadataTable object that captures information about the query
+        ir_blocks: list of BasicBlocks representing the IR to self-consistency check.
+        query_metadata_table: QueryMetadataTable object that captures information about the query.
 
     Raises:
         AssertionError, if the IR has unexpected structure. If the IR produced by the front-end
@@ -40,19 +40,19 @@ def sanity_check_ir_blocks_from_frontend(
     if not ir_blocks:
         raise AssertionError("Received no ir_blocks: {}".format(ir_blocks))
 
-    _sanity_check_fold_scope_locations_are_unique(ir_blocks)
-    _sanity_check_no_nested_folds(ir_blocks)
-    _sanity_check_query_root_block(ir_blocks)
-    _sanity_check_output_source_follower_blocks(ir_blocks)
-    _sanity_check_block_pairwise_constraints(ir_blocks)
-    _sanity_check_mark_location_preceding_optional_traverse(ir_blocks)
-    _sanity_check_every_location_is_marked(ir_blocks)
-    _sanity_check_coerce_type_outside_of_fold(ir_blocks)
-    _sanity_check_all_marked_locations_are_registered(ir_blocks, query_metadata_table)
-    _sanity_check_registered_locations_parent_locations(query_metadata_table)
+    _assert_fold_scope_locations_are_unique(ir_blocks)
+    _assert_no_nested_folds(ir_blocks)
+    _assert_query_root_block(ir_blocks)
+    _assert_output_source_follower_blocks(ir_blocks)
+    _assert_block_pairwise_constraints(ir_blocks)
+    _assert_mark_location_preceding_optional_traverse(ir_blocks)
+    _assert_every_location_is_marked(ir_blocks)
+    _assert_coerce_type_outside_of_fold(ir_blocks)
+    _assert_all_marked_locations_are_registered(ir_blocks, query_metadata_table)
+    _assert_registered_locations_parent_locations(query_metadata_table)
 
 
-def _sanity_check_registered_locations_parent_locations(
+def _assert_registered_locations_parent_locations(
     query_metadata_table: QueryMetadataTable,
 ) -> None:
     """Assert that all registered locations' parent locations are also registered."""
@@ -76,7 +76,7 @@ def _sanity_check_registered_locations_parent_locations(
             query_metadata_table.get_location_info(location_info.parent_location)
 
 
-def _sanity_check_all_marked_locations_are_registered(
+def _assert_all_marked_locations_are_registered(
     ir_blocks: List[BasicBlock], query_metadata_table: QueryMetadataTable
 ) -> None:
     """Assert that all locations in MarkLocation blocks have registered and valid metadata."""
@@ -103,7 +103,7 @@ def _sanity_check_all_marked_locations_are_registered(
         )
 
 
-def _sanity_check_fold_scope_locations_are_unique(ir_blocks: List[BasicBlock]) -> None:
+def _assert_fold_scope_locations_are_unique(ir_blocks: List[BasicBlock]) -> None:
     """Assert that every FoldScopeLocation that exists on a Fold block is unique."""
     observed_locations: Dict[FoldScopeLocation, Fold] = dict()
     for block in ir_blocks:
@@ -117,7 +117,7 @@ def _sanity_check_fold_scope_locations_are_unique(ir_blocks: List[BasicBlock]) -
             observed_locations[block.fold_scope_location] = block
 
 
-def _sanity_check_no_nested_folds(ir_blocks: List[BasicBlock]) -> None:
+def _assert_no_nested_folds(ir_blocks: List[BasicBlock]) -> None:
     """Assert that there are no nested Fold contexts, and that every Fold has a matching Unfold."""
     fold_seen = False
     for block in ir_blocks:
@@ -135,7 +135,7 @@ def _sanity_check_no_nested_folds(ir_blocks: List[BasicBlock]) -> None:
                 fold_seen = False
 
 
-def _sanity_check_query_root_block(ir_blocks: List[BasicBlock]) -> None:
+def _assert_query_root_block(ir_blocks: List[BasicBlock]) -> None:
     """Assert that QueryRoot is always the first block, and only the first block."""
     if not isinstance(ir_blocks[0], QueryRoot):
         raise AssertionError("The first block was not QueryRoot: {}".format(ir_blocks))
@@ -144,7 +144,7 @@ def _sanity_check_query_root_block(ir_blocks: List[BasicBlock]) -> None:
             raise AssertionError("Found QueryRoot after the first block: {}".format(ir_blocks))
 
 
-def _sanity_check_construct_result_block(ir_blocks: List[BasicBlock]) -> None:
+def _self_consistency_check_construct_result_block(ir_blocks: List[BasicBlock]) -> None:
     """Assert that ConstructResult is always the last block, and only the last block."""
     if not isinstance(ir_blocks[-1], ConstructResult):
         raise AssertionError("The last block was not ConstructResult: {}".format(ir_blocks))
@@ -155,7 +155,7 @@ def _sanity_check_construct_result_block(ir_blocks: List[BasicBlock]) -> None:
             )
 
 
-def _sanity_check_output_source_follower_blocks(ir_blocks: List[BasicBlock]) -> None:
+def _assert_output_source_follower_blocks(ir_blocks: List[BasicBlock]) -> None:
     """Ensure there are no Traverse / Backtrack / Recurse blocks after an OutputSource block."""
     seen_output_source = False
     for block in ir_blocks:
@@ -170,7 +170,7 @@ def _sanity_check_output_source_follower_blocks(ir_blocks: List[BasicBlock]) ->
                 )
 
 
-def _sanity_check_block_pairwise_constraints(ir_blocks: List[BasicBlock]) -> None:
+def _assert_block_pairwise_constraints(ir_blocks: List[BasicBlock]) -> None:
     """Assert that adjacent blocks obey all invariants."""
     for first_block, second_block in pairwise(ir_blocks):
         # Always Filter before MarkLocation, never after.
@@ -207,7 +207,9 @@ def _sanity_check_block_pairwise_constraints(ir_blocks: List[BasicBlock]) -> Non
                 )
 
 
-def _sanity_check_mark_location_preceding_optional_traverse(ir_blocks: List[BasicBlock]) -> None:
+def _assert_mark_location_preceding_optional_traverse(
+    ir_blocks: List[BasicBlock],
+) -> None:
     """Assert that optional Traverse blocks are preceded by a MarkLocation."""
     # Once all fold blocks are removed, each optional Traverse must have
     # a MarkLocation block immediately before it.
@@ -222,7 +224,7 @@ def _sanity_check_mark_location_preceding_optional_traverse(ir_blocks: List[Basi
                 )
 
 
-def _sanity_check_every_location_is_marked(ir_blocks: List[BasicBlock]) -> None:
+def _assert_every_location_is_marked(ir_blocks: List[BasicBlock]) -> None:
     """Ensure that every new location is marked with a MarkLocation block."""
     # Exactly one MarkLocation block is found between any block that starts an interval of blocks
     # that all affect the same query position, and the first subsequent block that affects a
@@ -255,7 +257,7 @@ def _sanity_check_every_location_is_marked(ir_blocks: List[BasicBlock]) -> None:
             mark_location_blocks_count = 0
 
 
-def _sanity_check_coerce_type_outside_of_fold(ir_blocks: List[BasicBlock]):
+def _assert_coerce_type_outside_of_fold(ir_blocks: List[BasicBlock]) -> None:
     """Ensure that CoerceType not in a @fold are followed by a MarkLocation or Filter block."""
     is_in_fold = False
     for first_block, second_block in pairwise(ir_blocks):

graphql_compiler/cost_estimation/statistics.py

@@ -174,7 +174,7 @@ class name) to count of edge instances of given class
         TODO(bojanserafimov): Enforce a canonical representation for quantile values and
                               sampling summaries. Datetimes should be in utc, decimals should
                               have type float, etc.
-        TODO(bojanserafimov): Sanity-check class_counts against sample_ratio * num_samples
+        TODO(bojanserafimov): Validate class_counts against sample_ratio * num_samples
         """
         if vertex_edge_vertex_counts is None:
             vertex_edge_vertex_counts = dict()

graphql_compiler/schema_generation/graphql_schema.py

@@ -331,8 +331,8 @@ def get_graphql_schema_from_schema_graph(
             )
 
             # N.B.: Ignore the "is_type_of" argument below, it is simply a circumvention of
-            #       a sanity check inside the GraphQL library. The library assumes that we'll use
-            #       its execution system, so it complains that we don't provide a means to
+            #       a validation check inside the GraphQL library. The library assumes that we'll
+            #       use its execution system, so it complains that we don't provide a means to
             #       differentiate between different implementations of the same interface.
             #       We don't care, because we compile the GraphQL query to a database query.
             current_graphql_type = GraphQLObjectType(

graphql_compiler/schema_generation/orientdb/schema_graph_builder.py

@@ -299,7 +299,7 @@ def _get_class_fields(class_definition):
     class_fields = class_definition.get("customFields")
     if class_fields is None:
         # OrientDB likes to make empty collections be None instead.
-        # We convert this field back to an empty dict, for our general sanity.
+        # We convert this field back to an empty dict, for our general happiness.
         class_fields = dict()
     return class_fields
 

graphql_compiler/schema_generation/sqlalchemy/sqlalchemy_reflector.py

@@ -39,7 +39,7 @@ def fast_sql_server_reflect(engine, metadata, schema, primary_key_selector=None)
     database_name, schema_name = schema.split(".")
 
     name_pattern = re.compile(r"^[a-zA-Z][_\-a-zA-Z0-9]*")
-    # Sanity check to prevent against SQL injection.
+    # Check to prevent against SQL injection.
     if not name_pattern.match(database_name):
         raise AssertionError("Invalid database name {}".format(database_name))
     if not name_pattern.match(schema_name):