Work in progress merging of use-implicit-booleaness checker

Author: Pierre-Sassoulas

doc/data/messages/c/compare-to-empty-string/bad.py

@@ -0,0 +1,8 @@
+x = 0
+y = 1
+
+if x == 0:   # [use-implicit-booleaness-not-comparison-to-zero]
+    print("x is equal to zero")
+
+if y != 0:   # [use-implicit-booleaness-not-comparison-to-zero]
+    print("y is not equal to zero")

doc/data/messages/c/compare-to-empty-string/good.py

@@ -1,4 +1,12 @@
+x = ""
+y = "hello"
 z = []
 
+if x == "":  # [use-implicit-booleaness-not-comparison]
+    print("x is an empty string")
+
+if y != "":  # [use-implicit-booleaness-not-comparison]
+    print("y is not an empty string")
+
 if z != []:  # [use-implicit-booleaness-not-comparison]
-    print("z is not an empty sequence")
+    print("z is not en empty container")

doc/data/messages/c/compare-to-empty-string/pylintrc

@@ -1,4 +1,12 @@
+x = ""
+y = "hello"
 z = []
 
+if not x:
+    print("x is an empty string")
+
+if y:
+    print("y is not an empty string")
+
 if z:
-    print("z is not an empty sequence")
+    print("z is not en empty container")

doc/data/messages/c/compare-to-zero/bad.py

@@ -20,7 +20,6 @@ Pylint provides the following optional plugins:
 - :ref:`pylint.extensions.docstyle`
 - :ref:`pylint.extensions.dunder`
 - :ref:`pylint.extensions.empty_comment`
-- :ref:`pylint.extensions.emptystring`
 - :ref:`pylint.extensions.eq_without_hash`
 - :ref:`pylint.extensions.for_any_all`
 - :ref:`pylint.extensions.magic_value`
@@ -88,20 +87,6 @@ Code Style checker Messages
   to. This can be changed to be an augmented assign. Disabled by default!
 
 
-.. _pylint.extensions.emptystring:
-
-Compare-To-Empty-String checker
-~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
-
-This checker is provided by ``pylint.extensions.emptystring``.
-Verbatim name of the checker is ``compare-to-empty-string``.
-
-Compare-To-Empty-String checker Messages
-^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-:compare-to-empty-string (C1901): *"%s" can be simplified to "%s" as an empty string is falsey*
-  Used when Pylint detects comparison to an empty string constant.
-
-
 .. _pylint.extensions.comparetozero:
 
 Compare-To-Zero checker
@@ -112,7 +97,7 @@ Verbatim name of the checker is ``compare-to-zero``.
 
 Compare-To-Zero checker Messages
 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
-:compare-to-zero (C2001): *"%s" can be simplified to "%s" as 0 is falsey*
+:use-implicit-booleaness-not-comparison-to-zero (C2002): *"%s" can be simplified to "%s" as 0 is falsey*
   Used when Pylint detects comparison to a 0 constant.
 
 

doc/data/messages/u/use-implicit-booleaness-not-comparison-to-zero/bad.py

@@ -893,10 +893,15 @@ Refactoring checker Messages
   Emitted when a single "return" or "return None" statement is found at the end
   of function or method definition. This statement can safely be removed
   because Python will implicitly return None
-:use-implicit-booleaness-not-comparison (C1803): *'%s' can be simplified to '%s' as an empty %s is falsey*
+:use-implicit-booleaness-not-comparison (C1803): *"%s" can be simplified to "%s" as an empty %s is falsey*
   Used when Pylint detects that collection literal comparison is being used to
   check for emptiness; Use implicit booleaness instead of a collection classes;
   empty collections are considered as false
+:use-implicit-booleaness-not-len (C1802): *"%s" can be simplified to "%s" as an empty %s is falsey*
+  Used when Pylint detects that len(sequence) is being used without explicit
+  comparison inside a condition to determine if a sequence is empty. Instead of
+  coercing the length to a boolean, either rely on the fact that empty
+  sequences are false or compare the length against a scalar.
 :unneeded-not (C0113): *Consider changing "%s" to "%s"*
   Used when a boolean expression contains an unneeded negation.
 :consider-iterating-dictionary (C0201): *Consider iterating the dictionary directly instead of calling .keys()*
@@ -911,11 +916,6 @@ Refactoring checker Messages
 :consider-using-enumerate (C0200): *Consider using enumerate instead of iterating with range and len*
   Emitted when code that iterates with range and len is encountered. Such code
   can be simplified by using the enumerate builtin.
-:use-implicit-booleaness-not-len (C1802): *Do not use `len(SEQUENCE)` without comparison to determine if a sequence is empty*
-  Used when Pylint detects that len(sequence) is being used without explicit
-  comparison inside a condition to determine if a sequence is empty. Instead of
-  coercing the length to a boolean, either rely on the fact that empty
-  sequences are false or compare the length against a scalar.
 :consider-using-f-string (C0209): *Formatting a regular string which could be an f-string*
   Used when we detect a string that is being formatted with format() or % which
   could potentially be an f-string. The use of f-strings is preferred. Requires

doc/data/messages/c/compare-to-zero/good.py renamed to doc/data/messages/u/use-implicit-booleaness-not-comparison-to-zero/good.py

@@ -1285,7 +1285,7 @@ Standard Checkers
 
 --spelling-dict
 """""""""""""""
-*Spelling dictionary name. Available dictionaries: en (aspell), en_AU (aspell), en_CA (aspell), en_GB (aspell), en_US (aspell).*
+*Spelling dictionary name. Available dictionaries: none. To make it work, install the 'python-enchant' package.*
 
 **Default:** ``""``
 

doc/data/messages/c/compare-to-zero/pylintrc renamed to doc/data/messages/u/use-implicit-booleaness-not-comparison-to-zero/pylintrc

@@ -390,8 +390,6 @@ All messages in the convention category:
    convention/bad-file-encoding
    convention/bad-mcs-classmethod-argument
    convention/bad-mcs-method-argument
-   convention/compare-to-empty-string
-   convention/compare-to-zero
    convention/consider-iterating-dictionary
    convention/consider-using-any-or-all
    convention/consider-using-dict-items
@@ -433,6 +431,7 @@ All messages in the convention category:
    convention/unnecessary-lambda-assignment
    convention/unneeded-not
    convention/use-implicit-booleaness-not-comparison
+   convention/use-implicit-booleaness-not-comparison-to-zero
    convention/use-implicit-booleaness-not-len
    convention/use-maxsplit-arg
    convention/use-sequence-for-iteration
@@ -449,6 +448,8 @@ All renamed messages in the convention category:
    :titlesonly:
 
    convention/blacklisted-name
+   convention/compare-to-empty-string
+   convention/compare-to-zero
    convention/len-as-condition
    convention/missing-docstring
    convention/old-misplaced-comparison-constant

doc/data/messages/u/use-implicit-booleaness-not-comparison/bad.py

@@ -0,0 +1,8 @@
+* The compare to empty string checker (``pylint.extensions.emptystring``) has been removed
+  as it is now part of the implicit booleaness checker, a default check.
+
+  - ``compare-to-zero`` was renamed ``use-implicit-booleaness-not-comparison-to-zero`` but is still an extension.
+  - ``compare-to-empty-string`` was merged with ``use-implicit-booleaness-not-comparison``.
+  - Messages related to implicit booleaness were made more explicit and actionable.
+
+Closes #6871

doc/data/messages/u/use-implicit-booleaness-not-comparison/good.py

@@ -4,6 +4,8 @@
 
 from __future__ import annotations
 
+import itertools
+
 import astroid
 from astroid import bases, nodes, util
 
@@ -54,7 +56,7 @@ class ImplicitBooleanessChecker(checkers.BaseChecker):
     name = "refactoring"
     msgs = {
         "C1802": (
-            "Do not use `len(SEQUENCE)` without comparison to determine if a sequence is empty",
+            '"%s" can be simplified to "%s" as an empty %s is falsey',
             "use-implicit-booleaness-not-len",
             "Used when Pylint detects that len(sequence) is being used "
             "without explicit comparison inside a condition to determine if a sequence is empty. "
@@ -69,6 +71,11 @@ class ImplicitBooleanessChecker(checkers.BaseChecker):
             "Used when Pylint detects that collection literal comparison is being "
             "used to check for emptiness; Use implicit booleaness instead "
             "of a collection classes; empty collections are considered as false",
+            {
+                "old_names": [
+                    ("C1901", "compare-to-empty-string"),
+                ]
+            },
         ),
     }
 
@@ -101,6 +108,7 @@ def visit_call(self, node: nodes.Call) -> None:
             # The node is a generator or comprehension as in len([x for x in ...])
             self.add_message(
                 "use-implicit-booleaness-not-len",
+                args=("len_arg.as_string()", "TODO", "TODO"),
                 node=node,
                 confidence=HIGH,
             )
@@ -119,6 +127,7 @@ def visit_call(self, node: nodes.Call) -> None:
         ):
             self.add_message(
                 "use-implicit-booleaness-not-len",
+                args=("instance.as_string()", "TODO", "TODO"),
                 node=node,
                 confidence=INFERENCE,
             )
@@ -149,6 +158,48 @@ def visit_unaryop(self, node: nodes.UnaryOp) -> None:
     @utils.only_required_for_messages("use-implicit-booleaness-not-comparison")
     def visit_compare(self, node: nodes.Compare) -> None:
         self._check_use_implicit_booleaness_not_comparison(node)
+        self._check_compare_to_empty_string(node)
+
+    def _check_compare_to_empty_string(self, node: nodes.Compare) -> None:
+        """Checks for comparisons to empty string.
+
+        Most of the time you should use the fact that empty strings are false.
+        An exception to this rule is when an empty string value is allowed in the program
+        and has a different meaning than None!
+        """
+        _operators = {"!=", "==", "is not", "is"}
+        # note: astroid.Compare has the left most operand in node.left while the rest
+        # are a list of tuples in node.ops the format of the tuple is
+        # ('compare operator sign', node) here we squash everything into `ops`
+        # to make it easier for processing later
+        ops: list[tuple[str, nodes.NodeNG | None]] = [("", node.left)]
+        ops.extend(node.ops)
+        iter_ops = iter(ops)
+        ops = list(itertools.chain(*iter_ops))  # type: ignore[arg-type]
+        for ops_idx in range(len(ops) - 2):
+            op_1: nodes.NodeNG | None = ops[ops_idx]
+            op_2: str = ops[ops_idx + 1]  # type: ignore[assignment]
+            op_3: nodes.NodeNG | None = ops[ops_idx + 2]
+            error_detected = False
+            if op_1 is None or op_3 is None or op_2 not in _operators:
+                continue
+            node_name = ""
+            # x ?? ""
+            if utils.is_empty_str_literal(op_1):
+                error_detected = True
+                node_name = op_3.as_string()
+            # '' ?? X
+            elif utils.is_empty_str_literal(op_3):
+                error_detected = True
+                node_name = op_1.as_string()
+            if error_detected:
+                suggestion = f"not {node_name}" if op_2 in {"==", "is"} else node_name
+                self.add_message(
+                    "compare-to-empty-string",
+                    args=(node.as_string(), suggestion, "string"),
+                    node=node,
+                    confidence=HIGH,
+                )
 
     def _check_use_implicit_booleaness_not_comparison(
         self, node: nodes.Compare

doc/user_guide/checkers/extensions.rst

@@ -39,10 +39,15 @@ class CompareToZeroChecker(checkers.BaseChecker):
     # configuration section name
     name = "compare-to-zero"
     msgs = {
-        "C2001": (
+        "C2002": (
             '"%s" can be simplified to "%s" as 0 is falsey',
-            "compare-to-zero",
+            "use-implicit-booleaness-not-comparison-to-zero",
             "Used when Pylint detects comparison to a 0 constant.",
+            {
+                "old_names": [
+                    ("C2001", "compare-to-zero"),
+                ]
+            },
         )
     }