Use enum for IndirectSelection

core/dbt/flags.py

@@ -17,7 +17,7 @@
 STRICT_MODE = False  # Only here for backwards compatibility
 FULL_REFRESH = False  # subcommand
 STORE_FAILURES = False  # subcommand
-EAGER_INDIRECT_SELECTION = True  # subcommand
+INDIRECT_SELECTION = 'eager'  # subcommand
 
 # Global CLI commands
 USE_EXPERIMENTAL_PARSER = None
@@ -95,15 +95,15 @@ def _get_context():
 def set_from_args(args, user_config):
     global STRICT_MODE, FULL_REFRESH, WARN_ERROR, \
         USE_EXPERIMENTAL_PARSER, STATIC_PARSER, WRITE_JSON, PARTIAL_PARSE, \
-        USE_COLORS, STORE_FAILURES, PROFILES_DIR, DEBUG, LOG_FORMAT, EAGER_INDIRECT_SELECTION, \
+        USE_COLORS, STORE_FAILURES, PROFILES_DIR, DEBUG, LOG_FORMAT, INDIRECT_SELECTION, \
         VERSION_CHECK, FAIL_FAST, SEND_ANONYMOUS_USAGE_STATS, PRINTER_WIDTH, \
         WHICH
 
     STRICT_MODE = False  # backwards compatibility
     # cli args without user_config or env var option
     FULL_REFRESH = getattr(args, 'full_refresh', FULL_REFRESH)
     STORE_FAILURES = getattr(args, 'store_failures', STORE_FAILURES)
-    EAGER_INDIRECT_SELECTION = getattr(args, 'indirect_selection', 'eager') != 'cautious'
+    INDIRECT_SELECTION = getattr(args, 'indirect_selection', INDIRECT_SELECTION)
     WHICH = getattr(args, 'which', WHICH)
 
     # global cli flags with env var and user_config alternatives

core/dbt/graph/cli.py

@@ -16,6 +16,7 @@
     SelectionIntersection,
     SelectionDifference,
     SelectionCriteria,
+    IndirectSelection
 )
 
 INTERSECTION_DELIMITER = ','
@@ -25,7 +26,7 @@
 
 
 def parse_union(
-    components: List[str], expect_exists: bool, eagerly_expand: bool = True
+    components: List[str], expect_exists: bool, indirect_selection: IndirectSelection = IndirectSelection.Eager
 ) -> SelectionUnion:
     # turn ['a b', 'c'] -> ['a', 'b', 'c']
     raw_specs = itertools.chain.from_iterable(
@@ -36,7 +37,7 @@ def parse_union(
     # ['a', 'b', 'c,d'] -> union('a', 'b', intersection('c', 'd'))
     for raw_spec in raw_specs:
         intersection_components: List[SelectionSpec] = [
-            SelectionCriteria.from_single_spec(part, eagerly_expand=eagerly_expand)
+            SelectionCriteria.from_single_spec(part, indirect_selection=indirect_selection)
             for part in raw_spec.split(INTERSECTION_DELIMITER)
         ]
         union_components.append(SelectionIntersection(
@@ -52,14 +53,14 @@ def parse_union(
 
 
 def parse_union_from_default(
-    raw: Optional[List[str]], default: List[str], eagerly_expand: bool = True
+    raw: Optional[List[str]], default: List[str], indirect_selection: IndirectSelection = IndirectSelection.Eager
 ) -> SelectionUnion:
     components: List[str]
     expect_exists: bool
     if raw is None:
-        return parse_union(components=default, expect_exists=False, eagerly_expand=eagerly_expand)
+        return parse_union(components=default, expect_exists=False, indirect_selection=indirect_selection)
     else:
-        return parse_union(components=raw, expect_exists=True, eagerly_expand=eagerly_expand)
+        return parse_union(components=raw, expect_exists=True, indirect_selection=indirect_selection)
 
 
 def parse_difference(
@@ -68,9 +69,9 @@ def parse_difference(
     included = parse_union_from_default(
         include,
         DEFAULT_INCLUDES,
-        eagerly_expand=flags.EAGER_INDIRECT_SELECTION
+        indirect_selection=IndirectSelection(flags.INDIRECT_SELECTION)
     )
-    excluded = parse_union_from_default(exclude, DEFAULT_EXCLUDES, eagerly_expand=True)
+    excluded = parse_union_from_default(exclude, DEFAULT_EXCLUDES, indirect_selection=IndirectSelection.Eager)
     return SelectionDifference(components=[included, excluded])
 
 

core/dbt/graph/selector.py

@@ -3,7 +3,7 @@
 from .graph import Graph, UniqueId
 from .queue import GraphQueue
 from .selector_methods import MethodManager
-from .selector_spec import SelectionCriteria, SelectionSpec
+from .selector_spec import SelectionCriteria, SelectionSpec, IndirectSelection
 
 from dbt.logger import GLOBAL_LOGGER as logger
 from dbt.node_types import NodeType
@@ -95,7 +95,7 @@ def get_nodes_from_criteria(
         neighbors = self.collect_specified_neighbors(spec, collected)
         direct_nodes, indirect_nodes = self.expand_selection(
             selected=(collected | neighbors),
-            eagerly_expand=spec.eagerly_expand
+            indirect_selection=spec.indirect_selection
         )
         return direct_nodes, indirect_nodes
 
@@ -199,7 +199,7 @@ def filter_selection(self, selected: Set[UniqueId]) -> Set[UniqueId]:
         }
 
     def expand_selection(
-        self, selected: Set[UniqueId], eagerly_expand: bool = True
+        self, selected: Set[UniqueId], indirect_selection: IndirectSelection = IndirectSelection.Eager
     ) -> Tuple[Set[UniqueId], Set[UniqueId]]:
         # Test selection by default expands to include an implicitly/indirectly selected tests.
         # `dbt test -m model_a` also includes tests that directly depend on `model_a`.
@@ -212,7 +212,7 @@ def expand_selection(
         #  - If ANY parent is missing, return it separately. We'll keep it around
         #    for later and see if its other parents show up.
         # Users can opt out of inclusive EAGER mode by passing --indirect-selection cautious
-        # CLI argument or by specifying `eagerly_expand: true` in a yaml selector
+        # CLI argument or by specifying `indirect_selection: true` in a yaml selector
 
         direct_nodes = set(selected)
         indirect_nodes = set()
@@ -222,7 +222,10 @@ def expand_selection(
                 node = self.manifest.nodes[unique_id]
                 if can_select_indirectly(node):
                     # should we add it in directly?
-                    if eagerly_expand or set(node.depends_on.nodes) <= set(selected):
+                    if (
+                        indirect_selection == IndirectSelection.Eager 
+                        or set(node.depends_on.nodes) <= set(selected)
+                    ):
                         direct_nodes.add(unique_id)
                     # if not:
                     else:
@@ -236,6 +239,10 @@ def incorporate_indirect_nodes(
         # Check tests previously selected indirectly to see if ALL their
         # parents are now present.
 
+        # performance: if identical, skip the processing below
+        if set(direct_nodes) == set(indirect_nodes):
+            return direct_nodes
+
         selected = set(direct_nodes)
 
         for unique_id in indirect_nodes:

core/dbt/graph/selector_spec.py

@@ -1,7 +1,9 @@
 import os
 import re
+import enum
 from abc import ABCMeta, abstractmethod
 from dataclasses import dataclass
+from dbt.dataclass_schema import StrEnum
 
 from typing import (
     Set, Iterator, List, Optional, Dict, Union, Any, Iterable, Tuple
@@ -21,6 +23,10 @@
 )
 SELECTOR_METHOD_SEPARATOR = '.'
 
+class IndirectSelection(StrEnum):
+    Eager = 'eager'
+    Cautious = 'cautious'
+
 
 def _probably_path(value: str):
     """Decide if value is probably a path. Windows has two path separators, so
@@ -66,7 +72,7 @@ class SelectionCriteria:
     parents_depth: Optional[int]
     children: bool
     children_depth: Optional[int]
-    eagerly_expand: bool = True
+    indirect_selection: IndirectSelection = IndirectSelection.Eager
 
     def __post_init__(self):
         if self.children and self.childrens_parents:
@@ -104,7 +110,7 @@ def parse_method(
 
     @classmethod
     def selection_criteria_from_dict(
-        cls, raw: Any, dct: Dict[str, Any], eagerly_expand: bool = True
+        cls, raw: Any, dct: Dict[str, Any], indirect_selection: IndirectSelection = IndirectSelection.Eager
     ) -> 'SelectionCriteria':
         if 'value' not in dct:
             raise RuntimeException(
@@ -124,7 +130,7 @@ def selection_criteria_from_dict(
             parents_depth=parents_depth,
             children=bool(dct.get('children')),
             children_depth=children_depth,
-            eagerly_expand=(eagerly_expand or bool(dct.get('eagerly_expand'))),
+            indirect_selection=(IndirectSelection(dct.get('indirect_selection') or indirect_selection)),
         )
 
     @classmethod
@@ -145,12 +151,12 @@ def dict_from_single_spec(cls, raw: str):
             dct['parents'] = bool(dct.get('parents'))
         if 'children' in dct:
             dct['children'] = bool(dct.get('children'))
-        if 'eagerly_expand' in dct:
-            dct['eagerly_expand'] = bool(dct.get('eagerly_expand'))
+        if 'indirect_selection' in dct:
+            dct['indirect_selection'] = bool(dct.get('indirect_selection'))
         return dct
 
     @classmethod
-    def from_single_spec(cls, raw: str, eagerly_expand: bool = True) -> 'SelectionCriteria':
+    def from_single_spec(cls, raw: str, indirect_selection: bool = True) -> 'SelectionCriteria':
         result = RAW_SELECTOR_PATTERN.match(raw)
         if result is None:
             # bad spec!
@@ -159,7 +165,7 @@ def from_single_spec(cls, raw: str, eagerly_expand: bool = True) -> 'SelectionCr
         return cls.selection_criteria_from_dict(
             raw,
             result.groupdict(),
-            eagerly_expand=eagerly_expand
+            indirect_selection=indirect_selection
         )
 
 

core/dbt/main.py

@@ -740,6 +740,7 @@ def _build_test_subparser(subparsers, base_subparser):
     sub.add_argument(
         '--indirect-selection',
         choices=['eager', 'cautious'],
+        default='eager',
         dest='indirect_selection',
         help='''
             Select all tests that are adjacent to selected resources,