Merge pull request #330 from beetanz/aws-html-emails

jvanderaa · web-flow · commit 3128814cc660 · 2025-11-22T12:00:17.000-06:00
Update AWS parser for HTML emails
diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -1,7 +1,5 @@
 # Changelog
 
-# Changelog
-
 ## v2.8.0 - 2025-06-06
 
 ### Added
diff --git a/changes/330.added b/changes/330.added
@@ -0,0 +1 @@
+Add support for parsing AWS HTML format maintenance notification emails
diff --git a/circuit_maintenance_parser/parsers/aws.py b/circuit_maintenance_parser/parsers/aws.py
diff --git a/circuit_maintenance_parser/provider.py b/circuit_maintenance_parser/provider.py
@@ -17,7 +17,7 @@
 from circuit_maintenance_parser.parsers.apple import SubjectParserApple, TextParserApple
 from circuit_maintenance_parser.parsers.aquacomms import HtmlParserAquaComms1, SubjectParserAquaComms1
 from circuit_maintenance_parser.parsers.att import HtmlParserATT1, XlsxParserATT1
-from circuit_maintenance_parser.parsers.aws import SubjectParserAWS1, TextParserAWS1
+from circuit_maintenance_parser.parsers.aws import HtmlParserAWS1, SubjectParserAWS1, TextParserAWS1
 from circuit_maintenance_parser.parsers.bso import HtmlParserBSO1
 from circuit_maintenance_parser.parsers.cogent import HtmlParserCogent1, SubjectParserCogent1, TextParserCogent1
 from circuit_maintenance_parser.parsers.colt import CsvParserColt1, SubjectParserColt1, SubjectParserColt2
@@ -276,6 +276,7 @@ class AWS(GenericProvider):
 
     _processors: List[GenericProcessor] = PrivateAttr(
         [
+            CombinedProcessor(data_parsers=[HtmlParserAWS1]),
             CombinedProcessor(data_parsers=[EmailDateParser, TextParserAWS1, SubjectParserAWS1]),
         ]
     )
diff --git a/tests/unit/data/aws/aws1_result.json b/tests/unit/data/aws/aws1_result.json
@@ -28,7 +28,7 @@
             }
         ],
         "end": 1621519200,
-        "maintenance_id": "a81ca7da4301432e1542f67d34414c71e7ccb22d233a6d1b0a70a722b386f368",
+        "maintenance_id": "b57f17ce354c1e0c8563f9cc0a75ac0c81f6672ffa6606a31a13ff4de2b383ae",
         "organizer": "aws-account-notifications@amazon.com",
         "provider": "aws",
         "sequence": 1,
diff --git a/tests/unit/data/aws/aws1_text_parser_result.json b/tests/unit/data/aws/aws1_text_parser_result.json
@@ -27,7 +27,7 @@
             }
         ],
         "end": 1621519200,
-        "maintenance_id": "a81ca7da4301432e1542f67d34414c71e7ccb22d233a6d1b0a70a722b386f368",
+        "maintenance_id": "b57f17ce354c1e0c8563f9cc0a75ac0c81f6672ffa6606a31a13ff4de2b383ae",
         "start": 1621497600,
         "status": "CONFIRMED",
         "summary": "Planned maintenance has been scheduled on an AWS Direct Connect router in A Block, New York, NY from Thu, 20 May 2021 08:00:00 GMT to Thu, 20 May 2021 14:00:00 GMT for 6 hours. During this maintenance window, your AWS Direct Connect services listed below may become unavailable."
diff --git a/tests/unit/data/aws/aws2_result.json b/tests/unit/data/aws/aws2_result.json
@@ -28,7 +28,7 @@
             }
         ],
         "end": 1631584920,
-        "maintenance_id": "6a01cfd83c1249d5f9a9c97de5232501794f37a8c15d433d09c029229f25659d",
+        "maintenance_id": "db4caacefd12cfa63678029e330bff2dd7a964b98f730b04ee099d76751f0f3d",
         "organizer": "aws-account-notifications@amazon.com",
         "provider": "aws",
         "sequence": 1,
diff --git a/tests/unit/data/aws/aws2_text_parser_result.json b/tests/unit/data/aws/aws2_text_parser_result.json
@@ -27,7 +27,7 @@
             }
         ],
         "end": 1631584920,
-        "maintenance_id": "6a01cfd83c1249d5f9a9c97de5232501794f37a8c15d433d09c029229f25659d",
+        "maintenance_id": "db4caacefd12cfa63678029e330bff2dd7a964b98f730b04ee099d76751f0f3d",
         "start": 1631559720,
         "status": "CANCELLED",
         "summary": "We would like to inform you that the planned maintenance that was scheduled for AWS Direct Connect endpoint in Equinix SG2, Singapore, SGP from Mon, 13 Sep 2021 19:02:00 GMT to Tue, 14 Sep 2021 02:02:00 GMT has been cancelled. Please find below your AWS Direct Connect services that would have been affected by this planned maintenance."
diff --git a/tests/unit/data/aws/aws3_result.json b/tests/unit/data/aws/aws3_result.json
@@ -16,7 +16,7 @@
             }
         ],
         "end": 1703055600,
-        "maintenance_id": "065ad66cb4e401827f675a56799be7ceb9cd66fc180ee5878c54533f7f196b52",
+        "maintenance_id": "387c99e743b96cf135b8e97d56c2dbbc02a0acda4eaf9637bcfeb23a7ebc96e2",
         "organizer": "aws-account-notifications@amazon.com",
         "provider": "aws",
         "sequence": 1,
diff --git a/tests/unit/data/aws/aws3_text_parser_result.json b/tests/unit/data/aws/aws3_text_parser_result.json
@@ -15,7 +15,7 @@
             }
         ],
         "end": 1703055600,
-        "maintenance_id": "065ad66cb4e401827f675a56799be7ceb9cd66fc180ee5878c54533f7f196b52",
+        "maintenance_id": "387c99e743b96cf135b8e97d56c2dbbc02a0acda4eaf9637bcfeb23a7ebc96e2",
         "start": 1703041200,
         "status": "CONFIRMED",
         "summary": "Emergency maintenance has been scheduled on an AWS Direct Connect endpoint in Datacenter Foo3, Anywhere, USA from Wed, 20 Dec 2023 03:00:00 GMT to Wed, 20 Dec 2023 07:00:00 GMT for 4 hours. This maintenance will cause a disruption to the following Direct Connect connections you own:"
diff --git a/tests/unit/data/aws/aws4.eml b/tests/unit/data/aws/aws4.eml
diff --git a/tests/unit/data/aws/aws4_html_parser_result.json b/tests/unit/data/aws/aws4_html_parser_result.json
@@ -0,0 +1,33 @@
+    [
+        {
+            "account": "111111111111",
+            "circuits": [
+                {
+                    "circuit_id": "dxcon-abc12345",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxvif-1234hfjd",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxlag-fge11110",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxcon-ffucoreh",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxcon-fg885ug5",
+                    "impact": "OUTAGE"
+                }
+            ],
+            "end": 1757998800,
+            "maintenance_id": "369a433f3071c0837b379bc170d410e35a7a733d61c2efaf51bd95a3a4073d71",
+            "start": 1757984400,
+            "stamp": 1756771998,
+            "status": "CONFIRMED",
+            "summary": "Hello, Planned maintenance has been scheduled on an AWS Direct Connect endpoint in Sonda Quilicura, Santiago from Tue, 16 Sep 2025 01:00:00 GMT to Tue, 16 Sep 2025 05:00:00 GMT for 4 hours. During this maintenance window, your AWS Direct Connect services listed below may become unavailable. dxcon-abc12345 dxvif-1234hfjd dxlag-fge11110 dxcon-ffucoreh dxcon-fg885ug5 This maintenance is scheduled to avoid disrupting redundant connections at the same time. If you encounter any problems with your connection after the end of this maintenance window, please contact AWS Support[1]."
+        }
+    ]
diff --git a/tests/unit/data/aws/aws4_result.json b/tests/unit/data/aws/aws4_result.json
@@ -0,0 +1,36 @@
+    [
+        {
+            "account": "111111111111",
+            "circuits": [
+                {
+                    "circuit_id": "dxcon-abc12345",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxvif-1234hfjd",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxlag-fge11110",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxcon-ffucoreh",
+                    "impact": "OUTAGE"
+                },
+                {
+                    "circuit_id": "dxcon-fg885ug5",
+                    "impact": "OUTAGE"
+                }
+            ],
+            "end": 1757998800,
+            "maintenance_id": "369a433f3071c0837b379bc170d410e35a7a733d61c2efaf51bd95a3a4073d71",
+            "start": 1757984400,
+            "organizer": "aws-account-notifications@amazon.com",
+            "provider": "aws",
+            "sequence": 1,
+            "stamp": 1756771998,
+            "status": "CONFIRMED",
+            "summary": "Hello, Planned maintenance has been scheduled on an AWS Direct Connect endpoint in Sonda Quilicura, Santiago from Tue, 16 Sep 2025 01:00:00 GMT to Tue, 16 Sep 2025 05:00:00 GMT for 4 hours. During this maintenance window, your AWS Direct Connect services listed below may become unavailable. dxcon-abc12345 dxvif-1234hfjd dxlag-fge11110 dxcon-ffucoreh dxcon-fg885ug5 This maintenance is scheduled to avoid disrupting redundant connections at the same time. If you encounter any problems with your connection after the end of this maintenance window, please contact AWS Support[1]."
+        }
+    ]
diff --git a/tests/unit/test_e2e.py b/tests/unit/test_e2e.py
@@ -159,6 +159,15 @@
                 Path(dir_path, "data", "aws", "aws3_result.json"),
             ],
         ),
+        (
+            AWS,
+            [
+                ("html", Path(dir_path, "data", "aws", "aws4.eml")),
+            ],
+            [
+                Path(dir_path, "data", "aws", "aws4_result.json"),
+            ],
+        ),
         # BSO
         (
             BSO,
diff --git a/tests/unit/test_parsers.py b/tests/unit/test_parsers.py
@@ -11,7 +11,7 @@
 from circuit_maintenance_parser.parsers.apple import TextParserApple
 from circuit_maintenance_parser.parsers.aquacomms import HtmlParserAquaComms1, SubjectParserAquaComms1
 from circuit_maintenance_parser.parsers.att import HtmlParserATT1, XlsxParserATT1
-from circuit_maintenance_parser.parsers.aws import SubjectParserAWS1, TextParserAWS1
+from circuit_maintenance_parser.parsers.aws import HtmlParserAWS1, SubjectParserAWS1, TextParserAWS1
 from circuit_maintenance_parser.parsers.bso import HtmlParserBSO1
 from circuit_maintenance_parser.parsers.cogent import HtmlParserCogent1, SubjectParserCogent1
 from circuit_maintenance_parser.parsers.colt import CsvParserColt1, SubjectParserColt1, SubjectParserColt2
@@ -170,6 +170,11 @@ def default(self, o):
             Path(dir_path, "data", "aws", "aws3.eml"),
             Path(dir_path, "data", "aws", "aws3_text_parser_result.json"),
         ),
+        (
+            HtmlParserAWS1,
+            Path(dir_path, "data", "aws", "aws4.eml"),
+            Path(dir_path, "data", "aws", "aws4_html_parser_result.json"),
+        ),
         # BSO
         (
             HtmlParserBSO1,

Original file line number	Diff line number	Diff line change
`@@ -0,0 +1 @@`
	`1`	`+Add support for parsing AWS HTML format maintenance notification emails`
Original file line number	Diff line number	Diff line change
`@@ -17,7 +17,7 @@`
`17`	`17`	`from circuit_maintenance_parser.parsers.apple import SubjectParserApple, TextParserApple`
`18`	`18`	`from circuit_maintenance_parser.parsers.aquacomms import HtmlParserAquaComms1, SubjectParserAquaComms1`
`19`	`19`	`from circuit_maintenance_parser.parsers.att import HtmlParserATT1, XlsxParserATT1`
`20`		`-from circuit_maintenance_parser.parsers.aws import SubjectParserAWS1, TextParserAWS1`
	`20`	`+from circuit_maintenance_parser.parsers.aws import HtmlParserAWS1, SubjectParserAWS1, TextParserAWS1`
`21`	`21`	`from circuit_maintenance_parser.parsers.bso import HtmlParserBSO1`
`22`	`22`	`from circuit_maintenance_parser.parsers.cogent import HtmlParserCogent1, SubjectParserCogent1, TextParserCogent1`
`23`	`23`	`from circuit_maintenance_parser.parsers.colt import CsvParserColt1, SubjectParserColt1, SubjectParserColt2`
`@@ -276,6 +276,7 @@ class AWS(GenericProvider):`
`276`	`276`
`277`	`277`	`_processors: List[GenericProcessor] = PrivateAttr(`
`278`	`278`	`[`
	`279`	`+ CombinedProcessor(data_parsers=[HtmlParserAWS1]),`
`279`	`280`	`CombinedProcessor(data_parsers=[EmailDateParser, TextParserAWS1, SubjectParserAWS1]),`
`280`	`281`	`]`
`281`	`282`	`)`
Original file line number	Diff line number	Diff line change
`@@ -28,7 +28,7 @@`
`28`	`28`	`}`
`29`	`29`	`],`
`30`	`30`	`"end": 1621519200,`
`31`		`- "maintenance_id": "a81ca7da4301432e1542f67d34414c71e7ccb22d233a6d1b0a70a722b386f368",`
	`31`	`+ "maintenance_id": "b57f17ce354c1e0c8563f9cc0a75ac0c81f6672ffa6606a31a13ff4de2b383ae",`
`32`	`32`	`"organizer": "[email protected]",`
`33`	`33`	`"provider": "aws",`
`34`	`34`	`"sequence": 1,`
Original file line number	Diff line number	Diff line change
`@@ -27,7 +27,7 @@`
`27`	`27`	`}`
`28`	`28`	`],`
`29`	`29`	`"end": 1621519200,`
`30`		`- "maintenance_id": "a81ca7da4301432e1542f67d34414c71e7ccb22d233a6d1b0a70a722b386f368",`
	`30`	`+ "maintenance_id": "b57f17ce354c1e0c8563f9cc0a75ac0c81f6672ffa6606a31a13ff4de2b383ae",`
`31`	`31`	`"start": 1621497600,`
`32`	`32`	`"status": "CONFIRMED",`
`33`	`33`	`"summary": "Planned maintenance has been scheduled on an AWS Direct Connect router in A Block, New York, NY from Thu, 20 May 2021 08:00:00 GMT to Thu, 20 May 2021 14:00:00 GMT for 6 hours. During this maintenance window, your AWS Direct Connect services listed below may become unavailable."`
Original file line number	Diff line number	Diff line change
`@@ -16,7 +16,7 @@`
`16`	`16`	`}`
`17`	`17`	`],`
`18`	`18`	`"end": 1703055600,`
`19`		`- "maintenance_id": "065ad66cb4e401827f675a56799be7ceb9cd66fc180ee5878c54533f7f196b52",`
	`19`	`+ "maintenance_id": "387c99e743b96cf135b8e97d56c2dbbc02a0acda4eaf9637bcfeb23a7ebc96e2",`
`20`	`20`	`"organizer": "[email protected]",`
`21`	`21`	`"provider": "aws",`
`22`	`22`	`"sequence": 1,`
Original file line number	Diff line number	Diff line change
`@@ -15,7 +15,7 @@`
`15`	`15`	`}`
`16`	`16`	`],`
`17`	`17`	`"end": 1703055600,`
`18`		`- "maintenance_id": "065ad66cb4e401827f675a56799be7ceb9cd66fc180ee5878c54533f7f196b52",`
	`18`	`+ "maintenance_id": "387c99e743b96cf135b8e97d56c2dbbc02a0acda4eaf9637bcfeb23a7ebc96e2",`
`19`	`19`	`"start": 1703041200,`
`20`	`20`	`"status": "CONFIRMED",`
`21`	`21`	`"summary": "Emergency maintenance has been scheduled on an AWS Direct Connect endpoint in Datacenter Foo3, Anywhere, USA from Wed, 20 Dec 2023 03:00:00 GMT to Wed, 20 Dec 2023 07:00:00 GMT for 4 hours. This maintenance will cause a disruption to the following Direct Connect connections you own:"`