Move the calculating of given/when/then to pydantic models, as well as removing tabbing from docstring in steps (aka multiline steps)

Author: jsa34

src/pytest_bdd/gherkin_parser.py

@@ -1,11 +1,15 @@
 import linecache
+import textwrap
 from pathlib import Path
-from typing import List, Optional, Union
+from typing import List, Optional
 
 from gherkin.errors import CompositeParserException
 from gherkin.parser import Parser
 from gherkin.token_scanner import TokenScanner
-from pydantic import BaseModel
+from pydantic import BaseModel, field_validator, model_validator
+
+from . import exceptions
+from .types import STEP_TYPES
 
 
 class Location(BaseModel):
@@ -41,6 +45,10 @@ class DocString(BaseModel):
     delimiter: str
     location: Location
 
+    @field_validator("content", mode="before")
+    def dedent_content(cls, value: str) -> str:
+        return textwrap.dedent(value)
+
 
 class Step(BaseModel):
     id: str
@@ -51,6 +59,18 @@ class Step(BaseModel):
     dataTable: Optional[DataTable] = None
     docString: Optional[DocString] = None
 
+    @field_validator("keyword", mode="before")
+    def normalize_keyword(cls, value: str) -> str:
+        return value.lower().strip()
+
+    @property
+    def given_when_then(self) -> str:
+        return self._gwt
+
+    @given_when_then.setter
+    def given_when_then(self, gwt: str) -> None:
+        self._gwt = gwt
+
 
 class Tag(BaseModel):
     id: str
@@ -68,6 +88,12 @@ class Scenario(BaseModel):
     tags: List[Tag]
     examples: Optional[List[DataTable]] = None
 
+    @model_validator(mode="after")
+    def process_steps(cls, instance):
+        steps = instance.steps
+        instance.steps = _compute_given_when_then(steps)
+        return instance
+
 
 class Rule(BaseModel):
     id: str
@@ -87,6 +113,12 @@ class Background(BaseModel):
     description: str
     steps: List[Step]
 
+    @model_validator(mode="after")
+    def process_steps(cls, instance):
+        steps = instance.steps
+        instance.steps = _compute_given_when_then(steps)
+        return instance
+
 
 class Child(BaseModel):
     background: Optional[Background] = None
@@ -108,6 +140,15 @@ class GherkinDocument(BaseModel):
     comments: List[Comment]
 
 
+def _compute_given_when_then(steps: list[Step]) -> list[Step]:
+    last_gwt = None
+    for step in steps:
+        if step.keyword in STEP_TYPES:
+            last_gwt = step.keyword
+        step.given_when_then = last_gwt
+    return steps
+
+
 class GherkinParser:
     def __init__(self, abs_filename: str = None, encoding: str = "utf-8"):
         self.abs_filename = Path(abs_filename) if abs_filename else None
@@ -118,8 +159,6 @@ def __init__(self, abs_filename: str = None, encoding: str = "utf-8"):
         try:
             self.gherkin_data = Parser().parse(TokenScanner(self.feature_file_text))
         except CompositeParserException as e:
-            from src.pytest_bdd import exceptions
-
             raise exceptions.FeatureError(
                 e.args[0],
                 e.errors[0].location["line"],

src/pytest_bdd/parser.py

@@ -220,7 +220,7 @@ class Step:
     """
 
     type: str
-    _name: str
+    name: str
     line_number: int
     indent: int
     keyword: str
@@ -356,21 +356,17 @@ def parse_steps(self, steps_data: list[GherkinStep]) -> list[Step]:
             List[Step]: A list of Step objects.
         """
         steps = []
-        current_step_type = None
         for step_data in steps_data:
-            keyword = step_data.keyword.strip().lower()
-            current_step_type = self.get_step_type(keyword) or current_step_type
             name = strip_comments(step_data.text)
             if step_data.docString:
-                doc_string = textwrap.dedent(step_data.docString.content)
-                name = f"{name}\n{doc_string}"
+                name = f"{name}\n{step_data.docString.content}"
             steps.append(
                 Step(
                     name=name,
-                    type=current_step_type,
+                    type=step_data.given_when_then,
                     indent=step_data.location.column - 1,
                     line_number=step_data.location.line,
-                    keyword=keyword.title(),
+                    keyword=step_data.keyword.title(),
                 )
             )
         return steps
@@ -424,9 +420,6 @@ def _parse_feature_file(self) -> GherkinDocument:
 
         Returns:
             Dict: A Gherkin document representation of the feature file.
-
-        Raises:
-            FeatureError: If there is an error parsing the feature file.
         """
         return GherkinParser(self.abs_filename, self.encoding).to_gherkin_document()