IVS-18 - Enhance protocol feature description check (#270)

* IVS-18 - Enhance protocol feature description check

* avoid duplicate error output, change in output msg

.github/workflows/ci.yml

@@ -27,7 +27,7 @@ jobs:
         echo $CONDA/bin >> $GITHUB_PATH
     - name: Install dependencies
       run: |
-        pip install behave pytest tabulate pyparsing sqlalchemy numpy pydantic pydot sqlalchemy_utils django python-dotenv deprecated pandas
+        pip install behave pytest tabulate pyparsing sqlalchemy numpy pydantic pydot sqlalchemy_utils django python-dotenv deprecated pandas pyspellchecker
         wget -O /tmp/ifcopenshell_python.zip https://s3.amazonaws.com/ifcopenshell-builds/ifcopenshell-python-`python3 -c 'import sys;print("".join(map(str, sys.version_info[0:2])))'`-v0.8.0-90ae709-linux64.zip
         mkdir -p `python3 -c 'import site; print(site.getusersitepackages())'`
         unzip -d `python3 -c 'import site; print(site.getusersitepackages())'` /tmp/ifcopenshell_python.zip

features/IFC101_Only-official-ifc-versions-allowed.feature

@@ -3,7 +3,7 @@
 @W00030
 Feature: IFC101 - Only official IFC versions allowed
 
-This rule verifies that the IFC model has a schema identifier corresponding to any of the official versions released by buildingSMART.
+The rule verifies that the IFC model has a schema identifier corresponding to any of the official versions released by buildingSMART.
 Specifically, IFC2x3 TC1 (version 2.3.0.1), IFC4 ADD2 TC1 (version 4.0.2.1), or IFC4.3 ADD2 (version 4.3.2.0).
 
   Scenario: Verifying Current Schema Identifier for IFC version

features/IFC102_Absence-of-deprecated-entities.feature

@@ -4,7 +4,7 @@
 @implementer-agreement
 Feature: IFC102 - Absence of deprecated entities
 
-This rule verifies that the IFC model does not have deprecated entities, attributes or enumerators. 
+The rule verifies that the IFC model does not have deprecated entities, attributes or enumerators. 
 By definition, a deprecated entity shall not be exported by applications.
 Complying interpreters shall still be able to import deprecated definitions.
 IFC2X3: https://standards.buildingsmart.org/IFC/RELEASE/IFC2x3/TC1/HTML/deprecated_constructs.htm

features/rule_creation_protocol/protocol.py

@@ -4,6 +4,7 @@
 from pydantic import model_validator, field_validator, Field, ValidationError
 from pyparsing import Word, alphas, nums, Literal, Combine, StringEnd, alphanums, ParseException
 import pyparsing
+from spellchecker import SpellChecker
 
 
 from .validation_helper import ValidatorHelper, ParsePattern
@@ -178,14 +179,54 @@ def validate_ifc_input(cls, value):
 
     @field_validator('description')
     def validate_description(cls, value=list) -> list:
-        """must include a description of the rule that start with "The rule verifies ..."""  # allow for comma's
-        if not any(value.startswith(f"{prefix} rule verifies{optional_comma}") for prefix in ("This", "The") for optional_comma in ("", ",")):
+        """
+        Validates that the description starts with 'The rule verifies ...'.
+        Captures cases where an incorrect prefix is used or the format is entirely incorrect.
+        """
+        valid_prefix = "The"
+        invalid_prefix = "This"
+
+        valid_start = any(value.startswith(f"{prefix} rule verifies{optional_comma}") 
+                        for prefix in [valid_prefix, invalid_prefix] 
+                        for optional_comma in ("", ","))
+
+        incorrect_prefix_start = any(value.startswith(f"{prefix} rule verifies{optional_comma}") 
+                                    for prefix in (invalid_prefix,) 
+                                    for optional_comma in ("", ","))
+
+        if incorrect_prefix_start:
             raise ProtocolError(
                 value=value,
-                message=f"The description must start with 'The rule verifies', it now starts with {value}"
+                message=f"The description starts with an incorrect prefix '{invalid_prefix}'. It should start with 'The rule verifies'."
             )
-        return value
 
+        if not valid_start:
+            spell = SpellChecker()
+            misspelled_words = spell.unknown(value.split()[:4])
+            if misspelled_words:
+                corrections = {word: spell.correction(word) for word in misspelled_words}
+                corrected_string = ' '.join([corrections.get(word, word) for word in value.split()])
+                corrected_valid_start = any(corrected_string.startswith(f"{prefix} rule verifies{optional_comma}") 
+                                        for prefix in [valid_prefix] 
+                                        for optional_comma in ("", ","))
+
+                message = f"The feature description contains spelling errors. Here are the incorrect words and their suggested corrections: {corrections}."
+
+                if not corrected_valid_start:
+                    message += f" Additionally, after corrections, the description still does not start with 'The rule verifies'. It starts with '{' '.join(corrected_string.split()[:4])}'."
+
+                raise ProtocolError(
+                    value=value,
+                    message=message
+                )
+            else:
+                raise ProtocolError(
+                    value=value,
+                    message=f"The description must start with 'The rule verifies', but it now starts with '{' '.join(value.split()[:4])}'."
+                )
+
+        return value
+
     @field_validator('steps')
     def validate_steps(cls, value):
         """Check only correct keywords are applied: 'Given', 'Then', 'And'"""