jenkins-bot submitted this change.

View Change

[bugfix] timestripper should skip HTML elements

Remove HTML elements before searching for timestamp in text.

Also fix isort check.

Bug: T302496
Change-Id: Iad70c4dd803fd40aac6f8d100c80512a876ea724
---
M pywikibot/textlib.py
M tests/timestripper_tests.py
2 files changed, 12 insertions(+), 1 deletion(-)

diff --git a/pywikibot/textlib.py b/pywikibot/textlib.py
index cc94eaf..d6eab05 100644
--- a/pywikibot/textlib.py
+++ b/pywikibot/textlib.py
@@ -25,7 +25,7 @@
 from pywikibot.exceptions import InvalidTitleError, SiteDefinitionError
 from pywikibot.family import Family
 from pywikibot.time import TZoneFixedOffset
-from pywikibot.tools import deprecated, ModuleDeprecationWrapper
+from pywikibot.tools import ModuleDeprecationWrapper, deprecated
 from pywikibot.userinterfaces.transliteration import NON_LATIN_DIGITS
 
 
@@ -2016,6 +2016,7 @@
         # Remove parts that are not supposed to contain the timestamp, in order
         # to reduce false positives.
         line = removeDisabledParts(line)
+        line = removeHTMLParts(line)
 
         line = to_latin_digits(line)
         for pat in self.patterns:
diff --git a/tests/timestripper_tests.py b/tests/timestripper_tests.py
index b178bac..4929296 100755
--- a/tests/timestripper_tests.py
+++ b/tests/timestripper_tests.py
@@ -377,6 +377,16 @@
         txt_match = self.date[:9] + '[[foo]]' + self.date[9:]
         self.assertEqual(ts(txt_match), self.expected_date)
 
+    def test_timestripper_skip_html(self):
+        """Test dates in html are correctly skipped."""
+        ts = self.ts.timestripper
+
+        txt_match = '<div ' + self.fake_date + '>'
+        self.assertIsNone(ts(txt_match))
+
+        txt_match = self.date + '<div ' + self.fake_date + '>'
+        self.assertEqual(ts(txt_match), self.expected_date)
+
 
 class TestTimeStripperDoNotArchiveUntil(TestTimeStripperCase):
 

To view, visit change 821799. To unsubscribe, or for help writing mail filters, visit settings.