Update map

- Fix CSV download

CHANGELOG.md

@@ -1,6 +1,23 @@
 Changelog
 =========
 
+8.18.5
+------
+
+- Fix CSV download
+
+8.18.4
+------
+
+- Fix webhooks
+
+8.18.3
+------
+
+- Move `__version__` to `parsedmarc.constants`
+- Create a constant `USER_AGENT`
+- Use the HTTP `User-Agent` header value `parsedmarc/version` for all HTTP requests
+
 8.18.2
 ------
 

parsedmarc/__init__.py

@@ -34,12 +34,13 @@ from parsedmarc.mail import (
     MSGraphConnection,
     GmailConnection,
 )
+
+from parsedmarc.constants import __version__
 from parsedmarc.utils import get_base_domain, get_ip_address_info
 from parsedmarc.utils import is_outlook_msg, convert_outlook_msg
 from parsedmarc.utils import parse_email
 from parsedmarc.utils import timestamp_to_human, human_timestamp_to_datetime
 
-__version__ = "8.18.2"
 
 logger.debug("parsedmarc v{0}".format(__version__))
 

parsedmarc/constants.py

@@ -0,0 +1,2 @@
+__version__ = "8.18.5"
+USER_AGENT = f"parsedmarc/{__version__}"

parsedmarc/resources/maps/README.md

@@ -3,6 +3,8 @@
 A mapping is meant to make it easier to identify who or what a sending source is. Please consider contributing
 additional mappings in a GitHub Pull Request.
 
+Do not open these CSV files in Excel. It will replace Unicode characters with question marks. Use LibreOffice Calc instead.
+
 ## base_reverse_dns_map.csv
 
 A CSV file with three fields: `base_reverse_dns`, `name`, and `type`.
@@ -72,7 +74,7 @@ A list of reverse DNS base domains that could not be identified as belonging to
 
 ## base_reverse_dns.csv
 
-A CSV with the fields `source_name` and optionally `message_countcount`. This CSV can be generated byy exporting the base DNS data from the Kibana on Splunk dashboards provided by parsedmarc. This file is not tracked by Git.
+A CSV with the fields `source_name` and optionally `message_count`. This CSV can be generated byy exporting the base DNS data from the Kibana on Splunk dashboards provided by parsedmarc. This file is not tracked by Git.
 
 ## unknown_base_reverse_dns.csv
 

parsedmarc/resources/maps/base_reverse_dns_map.csv

@@ -160,7 +160,9 @@ calero.com,Calero,MSP
 calpoly.edu,Cal Poly,Education
 calvin.edu,Calvin University,Education
 canonet.ne.jp,Canonet,MSP
+cardhealth.com,Cardinal Health,Healthcare
 cardinalhealth.com,Cardinal Health,Healthcare
+cardinal.com,Cardinal Health,Healthcare
 carecentrix.com,CareCentrix,Healthcare
 carrierzone.com,carrierzone,Email Security
 carsforkids.org,Cars For Kids,Nonprofit

parsedmarc/splunk.py

@@ -5,7 +5,7 @@ import json
 import urllib3
 import requests
 
-from parsedmarc import __version__
+from parsedmarc.constants import USER_AGENT
 from parsedmarc.log import logger
 from parsedmarc.utils import human_timestamp_to_unix_timestamp
 
@@ -51,7 +51,7 @@ class HECClient(object):
         self._common_data = dict(host=self.host, source=self.source, index=self.index)
 
         self.session.headers = {
-            "User-Agent": "parsedmarc/{0}".format(__version__),
+            "User-Agent": USER_AGENT,
             "Authorization": "Splunk {0}".format(self.access_token),
         }
 

parsedmarc/utils.py

@@ -37,7 +37,7 @@ import requests
 from parsedmarc.log import logger
 import parsedmarc.resources.dbip
 import parsedmarc.resources.maps
-
+from parsedmarc.constants import USER_AGENT
 
 parenthesis_regex = re.compile(r"\s*\(.*\)\s*")
 
@@ -345,7 +345,8 @@ def get_service_from_reverse_dns_base_domain(
     if not (offline or always_use_local_file) and len(reverse_dns_map) == 0:
         try:
             logger.debug(f"Trying to fetch reverse DNS map from {url}...")
-            response = requests.get(url)
+            headers = {"User-Agent": USER_AGENT}
+            response = requests.get(url, headers=headers)
             response.raise_for_status()
             csv_file.write(response.text)
             csv_file.seek(0)
@@ -355,6 +356,7 @@ def get_service_from_reverse_dns_base_domain(
         except Exception:
             logger.warning("Not a valid CSV file")
             csv_file.seek(0)
+            logging.debug("Response body:")
             logger.debug(csv_file.read())
 
     if len(reverse_dns_map) == 0:

parsedmarc/webhook.py

@@ -1,6 +1,7 @@
 import requests
 
 from parsedmarc import logger
+from parsedmarc.constants import USER_AGENT
 
 
 class WebhookClient(object):
@@ -21,7 +22,7 @@ class WebhookClient(object):
         self.timeout = timeout
         self.session = requests.Session()
         self.session.headers = {
-            "User-Agent": "parsedmarc",
+            "User-Agent": USER_AGENT,
             "Content-Type": "application/json",
         }
 

pyproject.toml

@@ -76,7 +76,7 @@ parsedmarc = "parsedmarc.cli:_main"
 Homepage = "https://domainaware.github.io/parsedmarc"
 
 [tool.hatch.version]
-path = "parsedmarc/__init__.py"
+path = "parsedmarc/constants.py"
 
 [tool.hatch.build.targets.sdist]
 include = [