From 335ab648205b476684dd28eaf22baacebeb1c5ab Mon Sep 17 00:00:00 2001
From: Johannes Filter <hi@jfilter.de>
Date: Fri, 16 Oct 2020 00:18:06 +0200
Subject: [PATCH] improve phone regex (fix #10)

---
 cleantext/constants.py |  3 ++-
 tests/test_clean.py    | 25 ++++++++++++++++++++++---
 2 files changed, 24 insertions(+), 4 deletions(-)

diff --git a/cleantext/constants.py b/cleantext/constants.py
index 0350849..6a12cb4 100644
--- a/cleantext/constants.py
+++ b/cleantext/constants.py
@@ -42,8 +42,9 @@
     flags=re.IGNORECASE | re.UNICODE,
 )
 
+# for more information: https://github.com/jfilter/clean-text/issues/10
 PHONE_REGEX = re.compile(
-    r"(?:^|(?<=[^\w)]))(\+?1[ .-]?)?(\(?\d{3}\)?[ .-]?)?(\d{3}[ .-]?\d{4})(\s?(?:ext\.?|[#x-])\s?\d{2,6})?(?:$|(?=\W))"
+    r"((?:^|(?<=[^\w)]))(((\+?[01])|(\+\d{2}))[ .-]?)?(\(?\d{3,4}\)?/?[ .-]?)?(\d{3}[ .-]?\d{4})(\s?(?:ext\.?|[#x-])\s?\d{2,6})?(?:$|(?=\W)))|\+?\d{4,5}[ .-/]\d{6,9}"
 )
 
 NUMBERS_REGEX = re.compile(
diff --git a/tests/test_clean.py b/tests/test_clean.py
index f890146..d6b816e 100644
--- a/tests/test_clean.py
+++ b/tests/test_clean.py
@@ -36,10 +36,29 @@ def test_replace_emails():
     assert cleantext.replace_emails(text, "*EMAIL*") == proc_text
 
 
+phone_numbers = [
+    "+49 123 1548690",
+    "555-123-4567",
+    "2404 9099130",
+    "024049099130",
+    "02404 9099130",
+    "02404/9099130",
+    "+492404 9099130",
+    "+4924049099130",
+    "+492404/9099130",
+    "0160 123456789",
+    "0160/123456789",
+    "+32160 123456789",
+    "Tel.: 0160 123456789",
+]
+
+
 def test_replace_phone_numbers():
-    text = "I can be reached at 555-123-4567 through next Friday."
-    proc_text = "I can be reached at *PHONE* through next Friday."
-    assert cleantext.replace_phone_numbers(text, "*PHONE*") == proc_text
+    for x in phone_numbers:
+        x_phone = cleantext.replace_phone_numbers(x, "*PHONE*")
+        assert "PHONE" in x_phone and not any(map(str.isdigit, x_phone)), (
+            x + " / " + x_phone
+        )
 
 
 def test_replace_numbers():