Fixes #85, optional skipping of the autoupdate

CHANGELOG.txt

@@ -1,3 +1,6 @@
+1.0.4:
+- Skipping the startup update with setting an environment variable called `PY3VE_IGNORE_UPDATER`
+
 1.0.3:
 - Moving project off github
 - Static type check fixes

README.md

@@ -172,5 +172,7 @@ The update can be triggered manually:
 
 `callback`: An optional Callable (function/method) to be called when the update is done.
 
+You can completely skip the auto-update on startup by setting the environment variable `PY3VE_IGNORE_UPDATER` to any value.
+
 # Read the [FAQ](https://gitea.ksol.io/karolyi/py3-validate-email/src/branch/master/FAQ.md)!
 

setup.py

@@ -57,7 +57,7 @@ class BuildPyCommand(build_py):
 
 setup(
     name='py3-validate-email',
-    version='1.0.3',
+    version='1.0.4',
     packages=find_packages(exclude=['tests']),
     install_requires=['dnspython~=2.1', 'idna~=3.0', 'filelock~=3.0'],
     author='László Károlyi',

tests/test_blacklist_check.py

@@ -11,11 +11,13 @@ from validate_email.validate_email import (
 class BlacklistCheckTestCase(TestCase):
     'Testing if the included blacklist filtering works.'
 
-    def setUpClass():
+    def setUpClass():  # type: ignore
         update_builtin_blacklist(force=False, background=False)
 
     def test_blacklist_positive(self):
         'Disallows blacklist item: mailinator.com.'
+        # The setting of the PY3VE_IGNORE_UPDATER variable doesn't
+        # matter here as the module has already download an initial list
         with self.assertRaises(DomainBlacklistedError):
             domainlist_check(EmailAddress('pm2@mailinator.com'))
         with self.assertRaises(DomainBlacklistedError):

tests/test_install.py

@@ -16,11 +16,11 @@ class InstallTest(TestCase):
     'Testing package installation.'
 
     def test_datadir_is_in_place(self):
-        'Data directory should be in the virtualenv.'
+        'Data directory should be in the virtualenv *after installation*.'
         output = check_output([
             executable, '-c', (
-                'import sys;sys.path.remove("");import validate_email;'
+                'import sys;sys.path.remove("");import validate_email;' +
-                'print(validate_email.updater.BLACKLIST_FILEPATH_INSTALLED);'
+                'print(validate_email.updater.BLACKLIST_FILEPATH_INSTALLED);' +
                 'print(validate_email.updater.ETAG_FILEPATH_INSTALLED, end="")'
             )]).decode('ascii')
         bl_path, etag_path = output.split('\n')

validate_email/constants.py

@@ -3,14 +3,14 @@ from re import compile as re_compile
 
 HOST_REGEX = re_compile(
     # max length for domain name labels is 63 characters per RFC 1034
-    r'((?:[A-Z0-9](?:[A-Z0-9-]{0,61}[A-Z0-9])?\.)+)'
+    r'((?:[A-Z0-9](?:[A-Z0-9-]{0,61}[A-Z0-9])?\.)+)' +
     r'(?:[A-Z0-9-]{2,63}(?<!-))\Z', IGNORECASE)
 LITERAL_REGEX = re_compile(
     # literal form, ipv4 or ipv6 address (SMTP 4.1.3)
     r'\[([A-f0-9:\.]+)\]\Z', IGNORECASE)
 USER_REGEX = re_compile(
     # dot-atom
-    r"(^[-!#$%&'*+/=?^_`{}|~0-9A-Z]+(\.[-!#$%&'*+/=?^_`{}|~0-9A-Z]+)*\Z"
+    r"(^[-!#$%&'*+/=?^_`{}|~0-9A-Z]+(\.[-!#$%&'*+/=?^_`{}|~0-9A-Z]+)*\Z" +
     # quoted-string
-    r'|^"([\001-\010\013\014\016-\037!#-\[\]-\177]|\\[\001-\011\013'
+    r'|^"([\001-\010\013\014\016-\037!#-\[\]-\177]|\\[\001-\011\013' +
     r'\014\016-\177])*"\Z)', IGNORECASE)

validate_email/domainlist_check.py

@@ -1,4 +1,5 @@
 from logging import getLogger
+from pathlib import Path
 from typing import Optional
 
 from filelock import FileLock
@@ -6,8 +7,8 @@ from filelock import FileLock
 from .email_address import EmailAddress
 from .exceptions import DomainBlacklistedError
 from .updater import (
-    BLACKLIST_FILEPATH_INSTALLED, BLACKLIST_FILEPATH_TMP, LOCK_PATH,
+    BLACKLIST_FILEPATH_INSTALLED, BLACKLIST_FILEPATH_TMP, ENV_IGNORE_UPDATER,
-    update_builtin_blacklist)
+    LOCK_PATH, TMP_PATH, update_builtin_blacklist)
 
 SetOrNone = Optional[set]
 LOGGER = getLogger(__name__)
@@ -17,7 +18,6 @@ class DomainListValidator(object):
     'Check the provided email against domain lists.'
     domain_whitelist = set()
     domain_blacklist = set('localhost')
-    _is_builtin_bl_used: bool = False
 
     def __init__(
             self, whitelist: SetOrNone = None, blacklist: SetOrNone = None):
@@ -26,11 +26,10 @@ class DomainListValidator(object):
         if blacklist:
             self.domain_blacklist = set(x.lower() for x in blacklist)
         else:
-            self._is_builtin_bl_used = True
             self.reload_builtin_blacklist()
 
     @property
-    def _blacklist_path(self) -> str:
+    def _blacklist_path(self) -> Path:
         'Return the path of the `blacklist.txt` that should be loaded.'
         try:
             # Zero size: file is touched to indicate the preinstalled
@@ -43,9 +42,8 @@ class DomainListValidator(object):
 
     def reload_builtin_blacklist(self):
         '(Re)load our built-in blacklist.'
-        if not self._is_builtin_bl_used:
+        TMP_PATH.mkdir(exist_ok=True)
-            return
+        with FileLock(lock_file=str(LOCK_PATH)):
-        with FileLock(lock_file=LOCK_PATH):
             bl_path = self._blacklist_path
             LOGGER.debug(msg=f'(Re)loading blacklist from {bl_path}')
             try:
@@ -57,7 +55,10 @@ class DomainListValidator(object):
             x.strip().lower() for x in lines if x.strip())
 
     def __call__(self, email_address: EmailAddress) -> bool:
-        'Do the checking here.'
+        """
+        Check if the email domain is valid, raise
+        `DomainBlacklistedError` if not.
+        """
         if email_address.domain in self.domain_whitelist:
             return True
         if email_address.domain in self.domain_blacklist:

validate_email/updater.py

@@ -1,5 +1,6 @@
 from http.client import HTTPResponse
 from logging import getLogger
+from os import environ
 from pathlib import Path
 from tempfile import gettempdir, gettempprefix
 from threading import Thread
@@ -19,9 +20,9 @@ except ImportError:
 LOGGER = getLogger(__name__)
 TMP_PATH = Path(gettempdir()).joinpath(
     f'{gettempprefix()}-py3-validate-email-{geteuid()}')
-TMP_PATH.mkdir(exist_ok=True)
+ENV_IGNORE_UPDATER = environ.get('PY3VE_IGNORE_UPDATER')
 BLACKLIST_URL = (
-    'https://raw.githubusercontent.com/disposable-email-domains/'
+    'https://raw.githubusercontent.com/disposable-email-domains/' +
     'disposable-email-domains/master/disposable_email_blocklist.conf')
 LIB_PATH_DEFAULT = Path(__file__).resolve().parent.joinpath('data')
 BLACKLIST_FILEPATH_INSTALLED = LIB_PATH_DEFAULT.joinpath('blacklist.txt')
@@ -117,9 +118,11 @@ class BlacklistUpdater(object):
             self, force: bool = False, callback: Optional[Callable] = None):
         'Start optionally updating the blacklist.txt file.'
         # Locking to avoid multi-process update on multi-process startup
-        # Import filelock locally because this module is als used by setup.py
+        # Import filelock locally because this module is als used by
+        # setup.py
         from filelock import FileLock
-        with FileLock(lock_file=LOCK_PATH):
+        TMP_PATH.mkdir(exist_ok=True)
+        with FileLock(lock_file=str(LOCK_PATH)):
             self._process(force=force)
         # Always execute callback because multiple processes can have
         # different versions of blacklists (one before, one after
@@ -135,12 +138,15 @@ def update_builtin_blacklist(
     Update and reload the built-in blacklist. Return the `Thread` used
     to do the background update, so it can be `join()`-ed.
     """
+    if ENV_IGNORE_UPDATER:
+        LOGGER.debug(msg='Skipping update of built-in blacklist.')
+        return
     LOGGER.debug(msg='Starting optional update of built-in blacklist.')
     blacklist_updater = BlacklistUpdater()
     kwargs = dict(force=force, callback=callback)
     if not background:
         blacklist_updater.process(**kwargs)
         return
-    bl_thread = Thread(target=blacklist_updater.process, kwargs=kwargs)
+    updater_thread = Thread(target=blacklist_updater.process, kwargs=kwargs)
-    bl_thread.start()
+    updater_thread.start()
-    return bl_thread
+    return updater_thread