check-capitalization: Check for banned words.

Fixes #8619

tools/check-capitalization

@@ -36,12 +36,12 @@ if __name__ == '__main__':
     with open('static/locale/en/translations.json') as f:
         data = json.load(f)
         frontend = check_capitalization(list(data.keys()))
-        frontend_errors, frontend_ignored = frontend
+        frontend_errors, frontend_ignored, banned_errors_front = frontend
 
     with open('static/locale/en/LC_MESSAGES/django.po') as f:
         rows = [r for r in DJANGO_PO_REGEX.findall(f.read()) if r]
         backend = check_capitalization(rows)
-        backend_errors, backend_ignored = backend
+        backend_errors, backend_ignored, banned_errors_back = backend
 
     if frontend_errors:
         print(WARNING + "Strings not properly capitalized in frontend:" + ENDC)
@@ -51,6 +51,14 @@ if __name__ == '__main__':
         print(WARNING + "Strings not properly capitalized in backend:" + ENDC)
         print('\n'.join(backend_errors))
 
+    if banned_errors_front:
+        print(WARNING + "Found banned words in frontend strings" + ENDC)
+        print('\n'.join(banned_errors_front))
+
+    if banned_errors_back:
+        print(WARNING + "Found banned words in backend strings" + ENDC)
+        print('\n'.join(banned_errors_back))
+
     if args.show_ignored:
         print(WARNING + "Strings which were ignored: " + ENDC)
         print('\n'.join(frontend_ignored + backend_ignored))

tools/lib/capitalization.py

@@ -130,6 +130,11 @@ DISALLOWED_REGEXES = [re.compile(regex) for regex in [
     # after a lower case character when the first character is in upper case.
 ]]
 
+BANNED_WORDS = {
+    'realm': ('The term realm should not appear in user-facing strings. '
+              'Use organization instead.'),
+}
+
 def get_safe_phrase(phrase):
     # type: (str) -> str
     """
@@ -194,10 +199,22 @@ def is_capitalized(safe_text):
 
     return True
 
+def check_banned_words(text: str) -> List[str]:
+    lower_cased_text = text.lower()
+    errors = []
+    for word, reason in BANNED_WORDS.items():
+        if word in lower_cased_text:
+            kwargs = dict(word=word, text=text, reason=reason)
+            msg = "{word} found in '{text}'. {reason}".format(**kwargs)
+            errors.append(msg)
+
+    return errors
+
 def check_capitalization(strings):
-    # type: (List[str]) -> Tuple[List[str], List[str]]
+    # type: (List[str]) -> Tuple[List[str], List[str], List[str]]
     errors = []
     ignored = []
+    banned_word_errors = []
     for text in strings:
         text = ' '.join(text.split())  # Remove extra whitespaces.
         safe_text = get_safe_text(text)
@@ -208,4 +225,6 @@ def check_capitalization(strings):
         elif capitalized and has_ignored_phrase:
             ignored.append(text)
 
-    return sorted(errors), sorted(ignored)
+        banned_word_errors.extend(check_banned_words(text))
+
+    return sorted(errors), sorted(ignored), sorted(banned_word_errors)

tools/tests/test_capitalization_checker.py

@@ -129,12 +129,14 @@ class CheckCapitalizationTestCase(TestCase):
                    "Some number 25MiB",
                    "Not Ignored Phrase",
                    "Not ignored phrase",
+                   "Some text with realm in it",
+                   "Realm in capital case",
                    ('<p class="bot-settings-note padded-container"> Looking for our '
                     '<a href="/integrations" target="_blank">Integrations</a> or '
                     '<a href="/api" target="_blank">API</a> '
                     'documentation? </p>'),
                    ]
-        errored, ignored = check_capitalization(strings)
+        errored, ignored, banned = check_capitalization(strings)
         self.assertEqual(errored, ['Not Ignored Phrase'])
         self.assertEqual(
             ignored,
@@ -148,3 +150,13 @@ class CheckCapitalizationTestCase(TestCase):
                      'Integrations</a> or <a href="/api" '
                      'target="_blank">API</a> documentation? </p>'),
                     ]))
+
+        self.assertEqual(banned,
+                         sorted(["realm found in 'Some text with realm in it'. "
+                                 "The term realm should not appear in user-facing "
+                                 "strings. Use organization instead.",
+
+                                 "realm found in 'Realm in capital case'. "
+                                 "The term realm should not appear in user-facing "
+                                 "strings. Use organization instead.",
+                                 ]))