Fix Equinix Parser (#143)

* Add parsing notes. Parse Year and add to ts.

* Update poetry lock

* Upgrade to Python 3.6.2


Co-authored-by: Christian Adell <[email protected]>

Author: scetron

CHANGELOG.md

@@ -2,10 +2,15 @@
 
 ## v2.0.9
 
+### Changed
+
+- #143 - Minimum Python version changed from `3.6.1` to `3.6.2`
+
 ### Fixed
 
 - #132 - Handle alternate "has been cancelled" text in Telstra notifications.
 - #134 - Handle Zayo "RESCHEDULE" notifications.
+- #143 - Fix Equinix parser not taking year into account
 
 ## v2.0.8 - 2021-12-09
 

circuit_maintenance_parser/cli.py

@@ -31,7 +31,7 @@ def main(provider_type, data_file, data_type, verbose):
 
     if data_type == "email":
         if str.lower(data_file[-3:]) == "eml":
-            with open(data_file) as email_file:
+            with open(data_file, encoding="utf-8") as email_file:
                 msg = email.message_from_file(email_file)
             data = NotificationData.init_from_emailmessage(msg)
         else:

circuit_maintenance_parser/parsers/equinix.py

@@ -21,7 +21,7 @@ def parse_html(self, soup: ResultSet) -> List[Dict]:
         Returns:
             Dict: The data dict containing circuit maintenance data.
         """
-        data: Dict[str, Any] = {"circuits": list()}
+        data: Dict[str, Any] = {"circuits": []}
 
         impact = self._parse_b(soup.find_all("b"), data)
         self._parse_table(soup.find_all("th"), data, impact)
@@ -54,18 +54,29 @@ def _parse_b(self, b_elements, data):
             impact (Status object): impact of the maintenance notification (used in the parse table function to assign an impact for each circuit).
         """
         impact = None
+        start_year = 0
+        end_year = 0
         for b_elem in b_elements:
+            if "SPAN:" in b_elem.text:
+                # Formated in DAY MONTH YEAR
+                # *SPAN: 02-JUL-2021 - 03-JUL-2021*
+                raw_year_span = b_elem.text.strip().split()
+                start_year = raw_year_span[1].split("-")[-1]
+                end_year = raw_year_span[-1].split("-")[-1]
             if "UTC:" in b_elem:
                 raw_time = b_elem.next_sibling
                 # for non english equinix notifications
                 # english section is usually at the bottom
                 # this skips the non english line at the top
                 if not self._isascii(raw_time):
                     continue
+
+                # Expected Format *UTC:* FRIDAY, 02 JUL 10:00 - FRIDAY, 02 JUL 15:00
+                # Note this detailed time does not contain the year..
                 start_end_time = raw_time.split("-")
                 if len(start_end_time) == 2:
-                    data["start"] = self.dt2ts(parser.parse(raw_time.split("-")[0].strip()))
-                    data["end"] = self.dt2ts(parser.parse(raw_time.split("-")[1].strip()))
+                    data["start"] = self.dt2ts(parser.parse(start_end_time[0].strip() + f" {start_year}"))
+                    data["end"] = self.dt2ts(parser.parse(start_end_time[1].strip() + f" {end_year}"))
             # all circuits in the notification share the same impact
             if "IMPACT:" in b_elem:
                 impact_line = b_elem.next_sibling

circuit_maintenance_parser/utils.py

@@ -47,7 +47,7 @@ def timezone(cls):  # pylint: disable=no-self-argument
     def db_location(cls):  # pylint: disable=no-self-argument
         """Load the locations DB from CSV into a Dict."""
         if not cls._db_location:
-            with open(os.path.join(dirname, "data", "worldcities.csv")) as csvfile:
+            with open(os.path.join(dirname, "data", "worldcities.csv"), encoding="utf-8") as csvfile:
                 reader = csv.DictReader(csvfile)
                 for row in reader:
                     # Index by city and country