[format] Add a check for standard scientific notation

Author: Pierre-Sassoulas

pylint/checkers/format.py

@@ -111,6 +111,11 @@
         "unexpected-line-ending-format",
         "Used when there is different newline than expected.",
     ),
+    "C0329": (
+        "Scientific notation should be '%s' instead",
+        "use-standard-scientific-notation",
+        "Emitted when a number is written in non-standard scientific notation.",
+    ),
 }
 
 
@@ -372,6 +377,16 @@ def _check_keyword_parentheses(
                         self._check_keyword_parentheses(tokens[i:], 0)
                     return
 
+    @staticmethod
+    def to_standard_scientific_notation(number: float) -> str:
+        # number is always going to be > 0 because node constants are always positive
+        # Format with high precision to capture all digits
+        s = f"{number:.15e}"
+        base, exp = s.split("e")
+        # Remove trailing zeros and possible trailing decimal point
+        base = base.rstrip("0").rstrip(".")
+        return f"{base}e{int(exp)}"
+
     def process_tokens(self, tokens: list[tokenize.TokenInfo]) -> None:
         """Process tokens and search for:
 
@@ -430,9 +445,23 @@ def process_tokens(self, tokens: list[tokenize.TokenInfo]) -> None:
                 if check_equal:
                     check_equal = False
                     self.check_indent_level(line, indents[-1], line_num)
-
-            if tok_type == tokenize.NUMBER and string.endswith("l"):
-                self.add_message("lowercase-l-suffix", line=line_num)
+            if tok_type == tokenize.NUMBER:
+                # Check for wrong scientific notation
+                if (
+                    ("e" in string or "E" in string)
+                    and "x" not in string  # not a hexadecimal
+                    and "j" not in string  # not a complex
+                ):
+                    value = float(string.lower().split("e")[0])
+                    if not (1 <= value < 10):
+                        self.add_message(
+                            "use-standard-scientific-notation",
+                            args=(self.to_standard_scientific_notation(value)),
+                            line=line_num,
+                            col_offset=start[1],
+                        )
+                if string.endswith("l"):
+                    self.add_message("lowercase-l-suffix", line=line_num)
 
             if string in _KEYWORD_TOKENS:
                 self._check_keyword_parentheses(tokens, idx)

tests/functional/u/use/use_standard_scientific_notation.py

@@ -0,0 +1,82 @@
+# pylint: disable=missing-docstring,invalid-name
+
+wrong_big = 45.3e6  # [use-standard-scientific-notation]
+uppercase_e_wrong = 45.3E6  # [use-standard-scientific-notation]
+wrong_small = 0.00012e-26  # [use-standard-scientific-notation]
+wrong_negative_and_big = -10e3  # [use-standard-scientific-notation]
+actual_trolling = 11000e26  # [use-standard-scientific-notation]
+scientific_double_digit = 12e8  # [use-standard-scientific-notation]
+scientific_triple_digit = 123e3  # [use-standard-scientific-notation]
+zero_before_decimal_small = 0.0001e-5  # [use-standard-scientific-notation]
+zero_before_decimal_big = 0.0001e5  # [use-standard-scientific-notation]
+negative_decimal = -0.5e10  # [use-standard-scientific-notation]
+zero_only = 0e10  # [use-standard-scientific-notation]
+
+one_only = 1e6
+correct_1 = 4.53e7
+uppercase_e_correct = 4.53E7
+uppercase_e_with_plus = 1.2E+10
+uppercase_e_with_minus = 5.67E-8
+correct_2 = 1.2e-28
+correct_3 = -1.0e4
+correct_4 = 1.1E30
+correct_with_digits = 4.567e8
+correct_with_plus = 1.2e+10
+correct_decimal_only = 3.14
+negative_correct = -5.67e-8
+correct_small_exponent = 1.5e1
+correct_tiny_exponent = 9.0e0
+correct_precise = 6.02214076e23
+
+hex_constant = 0x1e4  # Hexadecimal, not scientific notation
+binary_constant = 0b1010
+octal_constant = 0o1234
+inside_string = "Temperature: 10e3 degrees"
+inside_multiline = """
+This is a test with 45.3e6 inside
+"""
+inside_comment = 1.0  # This comment has 12e4 in it
+in_variable_name = measurement_10e3 = 45
+inside_f_string = f"Value is {1.0} not 10e6"
+
+# Potential false negatives
+barely_violation = 9.99e0  # Should this be 9.99?
+integer_sci = int(1e10)  # Integer call with scientific notation
+complex_number = 1.5e3 + 2.5e3j  # Complex number with scientific notation
+tuple_of_sci = (1.2e4, 3.4e5)
+list_of_sci = [5.6e6, 7.8e7]
+dict_with_sci = {"a": 9.1e8, "b": 1.2e9}
+
+# Mathematical operations
+addition = 1.0e3 + 2.0e3
+multiplication = 1.0e3 * 2.0
+division = 1.0e3 / 2.0
+power = 1.0e3 ** 2.0
+
+# Function calls with scientific notation
+def function_with_sci(param=1.0e3, other_param=2.0e3):
+    return param, other_param
+
+result = function_with_sci(2.0e3)
+positional_and_keyword = function_with_sci(1.0, other_param=3.0e4)
+
+# Assignments with operations
+a = 1
+a += 1.0e3
+b = 2
+b *= 2.0e3
+
+# Scientific notation in different contexts
+inside_list_comp = [x * 2 for x in [1.0e3, 2.0e3]]
+inside_dict_comp = {str(x): x for x in [3.0e3, 4.0e3]}
+inside_generator = (x + 1 for x in [5.0e3, 6.0e3])
+
+# Boundary cases for normalization
+boundary_small = 9.999e0  # Almost 10, but not quite
+boundary_large = 1.001e0  # Just above 1
+boundary_case = 1.0e0  # Equal to 1
+
+# Constants from physics/science (correctly formatted)
+speed_of_light = 2.99792458e8  # m/s
+planck_constant = 6.62607015e-34  # J⋅s
+electron_charge = 1.602176634e-19  # C

tests/functional/u/use/use_standard_scientific_notation.txt

@@ -0,0 +1,11 @@
+use-standard-scientific-notation:3:12:None:None::Scientific notation should be '4.53e1' instead:UNDEFINED
+use-standard-scientific-notation:4:20:None:None::Scientific notation should be '4.53e1' instead:UNDEFINED
+use-standard-scientific-notation:5:14:None:None::Scientific notation should be '1.2e-4' instead:UNDEFINED
+use-standard-scientific-notation:6:26:None:None::Scientific notation should be '1e1' instead:UNDEFINED
+use-standard-scientific-notation:7:18:None:None::Scientific notation should be '1.1e4' instead:UNDEFINED
+use-standard-scientific-notation:8:26:None:None::Scientific notation should be '1.2e1' instead:UNDEFINED
+use-standard-scientific-notation:9:26:None:None::Scientific notation should be '1.23e2' instead:UNDEFINED
+use-standard-scientific-notation:10:28:None:None::Scientific notation should be '1e-4' instead:UNDEFINED
+use-standard-scientific-notation:11:26:None:None::Scientific notation should be '1e-4' instead:UNDEFINED
+use-standard-scientific-notation:12:20:None:None::Scientific notation should be '5e-1' instead:UNDEFINED
+use-standard-scientific-notation:13:12:None:None::Scientific notation should be '0e0' instead:UNDEFINED