Fix app name validation logic and improve error message

changes/2127.misc.rst

@@ -0,0 +1 @@
+Enhance app name validation to include additional checks for format and identifiers

src/briefcase/commands/convert.py

@@ -107,17 +107,12 @@ def input_app_name(self, override_value: str | None) -> str:
         """
         intro = (
             "We need a name that can serve as a machine-readable Python package name for "
-            "your application. This name must be PEP508-compliant - that means the name "
-            "may only contain letters, numbers, hyphens and underscores; it can't contain "
-            "spaces or punctuation, and it can't start with a hyphen or underscore."
+            "your application. The name may only contain letters, numbers, hyphens and underscores; "
+            "it can't contain spaces or punctuation, and it can't start with a hyphen, underscore or a number."
         )
 
         default = "hello-world"
-        if (
-            "name" in self.pep621_data
-            and is_valid_app_name(self.pep621_data["name"])
-            and override_value is None
-        ):
+        if "name" in self.pep621_data and override_value is None:
             app_name = canonicalize_name(self.pep621_data["name"])
             self.console.divider(title="App name")
             self.console.prompt()
@@ -126,13 +121,36 @@ def input_app_name(self, override_value: str | None) -> str:
             )
             return app_name
 
-        if is_valid_app_name(self.base_path.name):  # Directory name is normalised
-            default = canonicalize_name(self.base_path.name)
+        canonicalized_name = canonicalize_name(self.base_path.name)
+        # Case 1: Project name is usable as-is (e.g., "foobar", "foo-bar")
+        if (
+            is_valid_app_name(self.base_path.name)
+            and canonicalized_name == self.base_path.name
+        ):
+            # Directory name is already valid - use it directly without prompting
+            if override_value is None:
+                self.console.divider(title="App name")
+                self.console.prompt()
+                self.console.prompt(
+                    f"Using directory name as app name: {self.base_path.name!r}"
+                )
+                return self.base_path.name
+            # If override_value is provided, fall through to prompt with it
+            default = self.base_path.name
             intro += (
                 "\n\n"
                 f"Based on your PEP508 formatted directory name, we suggest an "
                 f"app name of '{default}', but you can use another name if you want."
             )
+        # Case 2: Project name could be valid after canonicalization (e.g., "test.name" -> "test-name", "test-app_name" -> "test-app-name")
+        elif is_valid_app_name(canonicalized_name):
+            default = canonicalized_name
+            intro += (
+                "\n\n"
+                f"Based on your directory name, we suggest an "
+                f"app name of '{default}', but you can use another name if you want."
+            )
+        # Case 3: Project name isn't valid, even after canonicalization - fall back to default
 
         return self.console.text_question(
             intro=intro,

src/briefcase/config.py

@@ -18,26 +18,37 @@
 from .constants import RESERVED_WORDS
 from .exceptions import BriefcaseConfigError
 
-# PEP 508 restricts the naming of modules. The PEP defines a regex that uses
-# re.IGNORECASE; but in in practice, packaging uses a version that rolls out the lower
-# case, which has very slightly different semantics with non-ASCII characters. This
-# definition is the one from
-# https://github.com/pypa/packaging/blob/24.0/src/packaging/_tokenizer.py#L80
-PEP508_NAME_RE = re.compile(r"^([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9._-]*[a-zA-Z0-9])$")
-
-
-def is_valid_pep508_name(app_name):
-    """Determine if the name is valid by PEP508 rules."""
-    return PEP508_NAME_RE.match(app_name)
-
 
 def is_reserved_keyword(app_name):
     """Determine if the name is a reserved keyword."""
-    return keyword.iskeyword(app_name.lower()) or app_name.lower() in RESERVED_WORDS
+    return (
+        keyword.iskeyword(app_name.lower())
+        or app_name.lower() in RESERVED_WORDS
+        or keyword.iskeyword(app_name)
+    )
 
 
 def is_valid_app_name(app_name):
-    return not is_reserved_keyword(app_name) and is_valid_pep508_name(app_name)
+    """Determine if the app name is valid.
+
+    Checks the following:
+        - It is not a reserved keyword.
+        - It is a valid Python identifier when hyphens are replaced with underscores.
+        - It does not start with an underscore (reserved for internal use).
+        - It does not contain problematic Unicode characters that have edge-case behaviors.
+
+    :param app_name: The app name to validate.
+    :returns: True if the app name is valid; False otherwise.
+    """
+    module_name = app_name.lower().replace("-", "_")
+
+    return all(
+        [
+            not is_reserved_keyword(app_name),
+            module_name.isidentifier(),
+            not module_name.startswith("_"),
+        ]
+    )
 
 
 def make_class_name(formal_name):
@@ -400,9 +411,11 @@ def __init__(
         if not is_valid_app_name(self.app_name):
             raise BriefcaseConfigError(
                 f"{self.app_name!r} is not a valid app name.\n\n"
-                "App names must not be reserved keywords such as 'and', 'for' and 'while'.\n"
-                "They must also be PEP508 compliant (i.e., they can only include letters,\n"
-                "numbers, '-' and '_'; must start with a letter; and cannot end with '-' or '_')."
+                "App names must:\n"
+                "- Not be reserved keywords (like 'and', 'for', 'while', 'main', 'test', etc.)\n"
+                "- Only contain letters, numbers, hyphens, and underscores\n"
+                "- Start with a letter (not a number, hyphen, or underscore)\n"
+                "- Be valid Python identifiers when hyphens are replaced with underscores"
             )
 
         if not is_valid_bundle_identifier(self.bundle):

src/briefcase/integrations/android_sdk.py

@@ -10,7 +10,6 @@
 from datetime import datetime
 from pathlib import Path
 
-from briefcase.config import PEP508_NAME_RE
 from briefcase.exceptions import (
     BriefcaseCommandError,
     IncompatibleToolError,
@@ -22,12 +21,17 @@
 from briefcase.integrations.java import JDK
 from briefcase.integrations.subprocess import SubprocessArgT
 
+ANDROID_EMULATOR_NAME_RE = re.compile(
+    r"^([a-zA-Z0-9]|[a-zA-Z0-9][a-zA-Z0-9._-]*[a-zA-Z0-9])$"
+)
+
+
 DEVICE_NOT_FOUND = re.compile(r"^error: device '[^']*' not found")
 
 
 def create_avd_validator(emulators):
     def _validate_avd_name(avd):
-        if not PEP508_NAME_RE.match(avd):
+        if not ANDROID_EMULATOR_NAME_RE.match(avd):
             raise ValueError(
                 "An emulator name may only contain letters, numbers, hyphens and underscores."
             )

tests/commands/convert/test_input_app_name.py

@@ -12,21 +12,20 @@ def test_override_is_used(convert_command):
 
 
 def test_no_pep621_data(convert_command, monkeypatch):
-    """The app directory is used if there is no PEP621-name."""
-    mock_text_question = MagicMock()
-    monkeypatch.setattr(convert_command.console, "text_question", mock_text_question)
+    """The app directory is used directly if there is no PEP621-name and the directory name is already valid."""
+    mock_divider = MagicMock()
+    mock_prompt = MagicMock()
+    monkeypatch.setattr(convert_command.console, "divider", mock_divider)
+    monkeypatch.setattr(convert_command.console, "prompt", mock_prompt)
     convert_command.base_path /= "test-app-name"
-    convert_command.input_app_name(None)
 
-    mock_text_question.assert_called_once_with(
-        intro=PartialMatchString(
-            "Based on your PEP508 formatted directory name, we suggest an app name of 'test-app-name'"
-        ),
-        description="App Name",
-        default="test-app-name",
-        validator=convert_command.validate_app_name,
-        override_value=None,
-    )
+    result = convert_command.input_app_name(None)
+
+    assert result == "test-app-name"
+    mock_divider.assert_called_once_with(title="App name")
+    assert mock_prompt.call_count == 2
+    mock_prompt.assert_any_call()
+    mock_prompt.assert_any_call("Using directory name as app name: 'test-app-name'")
 
 
 def test_valid_pep621_app_name(convert_command):
@@ -72,7 +71,7 @@ def test_hint_is_canonicalized(convert_command, monkeypatch):
 
     mock_text_question.assert_called_once_with(
         intro=PartialMatchString(
-            "Based on your PEP508 formatted directory name, we suggest an app name of 'test-app-name'"
+            "Based on your directory name, we suggest an app name of 'test-app-name'"
         ),
         description="App Name",
         default="test-app-name",

tests/commands/new/test_validate_app_name.py

@@ -7,7 +7,6 @@
         "helloworld",
         "helloWorld",
         "hello42world",
-        "42helloworld",  # ?? Are we sure this is correct?
         "hello_world",
         "hello-world",
     ],
@@ -24,6 +23,7 @@ def test_valid_app_name(new_command, name):
         "helloworld!",  # contains punctuation
         "_helloworld",  # leading underscore
         "-helloworld",  # leading hyphen
+        "42helloworld",  # starting with a number
         "pass",  # python keyword
         "Pass",  # python keyword, but different case usage
         "PASS",  # python keyword, but all upper case

tests/config/test_AppConfig.py

@@ -180,6 +180,8 @@ def test_extra_attrs():
         "my_app",  # contains underscore
         "myapp2",  # ends with digit
         "my2app",  # contains digit
+        "myApp-",  # ends with hyphen (converts to valid identifier)
+        "myApp_",  # ends with underscore (valid Python identifier)
     ],
 )
 def test_valid_app_name(name):
@@ -204,9 +206,7 @@ def test_valid_app_name(name):
         "myapp!",  # end punctuation
         "my$app",  # other punctuation
         "-myApp",  # initial hyphen
-        "myApp-",  # end hyphen
         "_myApp",  # initial underscore
-        "myApp_",  # end underscore
     ],
 )
 def test_invalid_app_name(name):

tests/config/test_is_valid_app_name.py

@@ -9,9 +9,14 @@
         "helloworld",
         "helloWorld",
         "hello42world",
-        "42helloworld",
         "hello_world",
         "hello-world",
+        "helloworld42",
+        "a",  # single letter
+        "abc123",  # alphanumeric
+        "none",  # `None` is illegal, but when lower case, it isn't a reserved word
+        "helloworld_",  # ends with underscore (valid Python identifier)
+        "helloworld-",  # ends with hyphen (converts to valid identifier)
     ],
 )
 def test_is_valid_app_name(name):
@@ -22,10 +27,12 @@ def test_is_valid_app_name(name):
 @pytest.mark.parametrize(
     "name",
     [
+        # Invalid characters
         "hello world",
         "helloworld!",
         "_helloworld",
         "-helloworld",
+        # python reserved words
         "switch",
         "pass",
         "false",
@@ -34,12 +41,25 @@ def test_is_valid_app_name(name):
         "main",
         "socket",
         "test",
-        # ı, İ and K (i.e. 0x212a) are valid ASCII when made lowercase and as such are
-        # accepted by the official PEP 508 regex... but they are rejected here to ensure
-        # compliance with the regex that is used in practice.
-        "helloworld_ı",
-        "İstanbul",
-        "Kelvin",
+        "None",
+        # Additional invalid formats
+        "my$app",  # dollar sign
+        "app@domain",  # at symbol
+        "app.name",  # period
+        "2app",  # starts with number
+        "42app",  # starts with number
+        "_",  # single underscore
+        "-",  # single hyphen
+        "1",  # single digit
+        "123",  # all digits
+        # Case variations of reserved words
+        "Switch",
+        "SWITCH",
+        "FALSE",
+        "Pass",
+        "PASS",
+        "Test",
+        "TEST",
     ],
 )
 def test_is_invalid_app_name(name):