fix: ruff CI-breaking lint errors and wrong activity type (CM-1318)

Signed-off-by: Uroš Marolt <uros@marolt.me>

Author: themarolt

scripts/builders/mailing-list-integration.env

@@ -0,0 +1,4 @@
+DOCKERFILE="./services/docker/Dockerfile.mailing_list_integration"
+CONTEXT="../"
+REPO="sjc.ocir.io/axbydjxa5zuh/mailing-list-integration"
+SERVICES="mailing-list-integration"

scripts/cli

@@ -1189,7 +1189,7 @@ while test $# -gt 0; do
         exit
         ;;
     clean-start-fe-dev)
-        IGNORED_SERVICES=("frontend" "python-worker" "job-generator" "webhook-api" "profiles-worker" "organizations-enrichment-worker" "merge-suggestions-worker" "members-enrichment-worker" "exports-worker" "entity-merging-worker" "cache-worker" "categorization-worker" "cron-service" "data-sink-worker" "git-integration" "integration-run-worker" "integration-stream-worker" "nango-webhook-api" "nango-worker" "script-executor-worker" "search-sync-api" "search-sync-worker" "security-best-practices-worker" "snowflake-connectors-worker")
+        IGNORED_SERVICES=("frontend" "python-worker" "job-generator" "webhook-api" "profiles-worker" "organizations-enrichment-worker" "merge-suggestions-worker" "members-enrichment-worker" "exports-worker" "entity-merging-worker" "cache-worker" "categorization-worker" "cron-service" "data-sink-worker" "git-integration" "mailing-list-integration" "integration-run-worker" "integration-stream-worker" "nango-webhook-api" "nango-worker" "script-executor-worker" "search-sync-api" "search-sync-worker" "security-best-practices-worker" "snowflake-connectors-worker")
         CLEAN_START=1
         DEV=1
         start

services/apps/mailing_list_integration/src/crowdmail/enums.py

@@ -1,7 +1,7 @@
-from enum import Enum
+from enum import StrEnum
 
 
-class ErrorCode(str, Enum):
+class ErrorCode(StrEnum):
     """Standard Error codes"""
 
     UNKNOWN = "unknown"
@@ -18,7 +18,7 @@ class ErrorCode(str, Enum):
     SERVER_ERROR = "server-error-remote"
 
 
-class ListState(str, Enum):
+class ListState(StrEnum):
     """Mailing list processing states"""
 
     PENDING = "pending"
@@ -36,34 +36,34 @@ class ListPriority(int):
     NORMAL = 2
 
 
-class IntegrationResultType(str, Enum):
+class IntegrationResultType(StrEnum):
     ACTIVITY = "activity"
 
 
-class IntegrationResultState(str, Enum):
+class IntegrationResultState(StrEnum):
     PENDING = "pending"
 
 
-class DataSinkWorkerQueueMessageType(str, Enum):
+class DataSinkWorkerQueueMessageType(StrEnum):
     PROCESS_INTEGRATION_RESULT = "process_integration_result"
 
 
-class ActivityType(str, Enum):
+class ActivityType(StrEnum):
     """Groupsio-compatible activity types (platform=groupsio)"""
 
     MESSAGE = "message"
     MEMBER_JOIN = "member_join"
     MEMBER_LEAVE = "member_leave"
 
 
-class ExecutionStatus(str, Enum):
+class ExecutionStatus(StrEnum):
     """Service execution status"""
 
     SUCCESS = "success"
     FAILURE = "failure"
 
 
-class OperationType(str, Enum):
+class OperationType(StrEnum):
     """Service operation types for metrics tracking"""
 
     MIRROR = "Mirror"

services/apps/mailing_list_integration/src/crowdmail/services/mirror/mirror_service.py

@@ -33,7 +33,12 @@
 # (stderr substrings, exception to raise) — checked in order, first match wins
 ERROR_CLASSIFICATIONS = [
     (
-        {"Connection refused", "Network is unreachable", "Connection timed out", "Could not resolve host"},
+        {
+            "Connection refused",
+            "Network is unreachable",
+            "Connection timed out",
+            "Could not resolve host",
+        },
         NetworkError,
     ),
     ({"429", "Too Many Requests"}, RateLimitError),

services/apps/mailing_list_integration/src/crowdmail/services/parse/noteren.py

@@ -19,6 +19,8 @@
 import subprocess
 import sys
 
+from crowdmail.enums import ActivityType
+
 # We want to always be using utf-8
 email.charset.add_charset("utf-8", None)
 
@@ -33,9 +35,7 @@
 
 # run a command and take the output.
 # Shamelessly taken from 'b4' by Konstantin Ryabitsev <konstantin@linuxfoundation.org>
-def _run_command(
-    cmdargs: list, stdin: bytes | None = None
-) -> tuple[int, bytes, bytes]:
+def _run_command(cmdargs: list, stdin: bytes | None = None) -> tuple[int, bytes, bytes]:
     logging.debug(cmdargs)
     sp = subprocess.Popen(
         cmdargs, stdout=subprocess.PIPE, stdin=subprocess.PIPE, stderr=subprocess.PIPE
@@ -68,9 +68,7 @@ def get_email_from_git(git_dir: str, git_id: str) -> tuple[bytes, str]:
         git_blob_id = out.decode().rstrip()  # strip the trailing \n
 
     if git_blob_id == "":
-        logging.error(
-            'git commit id %s was not found in the git repo "%s"', git_id, git_dir
-        )
+        logging.error('git commit id %s was not found in the git repo "%s"', git_id, git_dir)
         sys.exit(1)
 
     logging.debug("git_blob_id =%s", git_blob_id)
@@ -216,9 +214,9 @@ def parse_email(
     # Do NOT use headersonly=True here, as that prevents the parser from
     # building the multipart structure, which we will properly walk later on
     # and pick out the text/plain part of the message.
-    e = email.parser.BytesParser(
-        policy=emlpolicy, _class=email.message.EmailMessage
-    ).parsebytes(message)
+    e = email.parser.BytesParser(policy=emlpolicy, _class=email.message.EmailMessage).parsebytes(
+        message
+    )
 
     logging.debug(e)
 
@@ -354,7 +352,7 @@ def parse_email(
         "body": json_body,
         "url": "https://lore.kernel.org/r/" + msgid,
         "isContribution": True,
-        "type": "email-message",
+        "type": ActivityType.MESSAGE,
         "attributes": attributes,
         "member": member,
     }
@@ -418,7 +416,9 @@ def cmd():
     parser.add_argument(
         "--git-dir", help="Git directory of the lore repo to dig out of", required=True
     )
-    parser.add_argument("--segment-id", help="CDP segment id to attach the activity to", required=True)
+    parser.add_argument(
+        "--segment-id", help="CDP segment id to attach the activity to", required=True
+    )
     parser.add_argument(
         "--integration-id", help="CDP integration id to attach the activity to", required=True
     )

services/apps/mailing_list_integration/src/test/test_noteren.py

@@ -23,9 +23,7 @@
 from crowdmail.services.parse import noteren
 
 HERE = os.path.dirname(os.path.abspath(__file__))
-NOTEREN = os.path.join(
-    HERE, "..", "crowdmail", "services", "parse", "noteren.py"
-)
+NOTEREN = os.path.join(HERE, "..", "crowdmail", "services", "parse", "noteren.py")
 SEGMENT_ID = "11111111-1111-1111-1111-111111111111"
 INTEGRATION_ID = "22222222-2222-2222-2222-222222222222"
 
@@ -34,25 +32,19 @@
 # Fixtures
 # --------------------------------------------------------------------------
 
+
 def _git(gitdir, *args, stdin=None):
     out = subprocess.run(
         ["git", "--git-dir", str(gitdir)] + list(args),
         input=stdin,
-        stdout=subprocess.PIPE,
-        stderr=subprocess.PIPE,
+        capture_output=True,
         check=True,
     )
     return out.stdout.decode().strip()
 
 
 def _make_email(msgid, subject, frm, body, date="Mon, 1 Jan 2024 12:00:00 +0000"):
-    hdr = (
-        f"From: {frm}\n"
-        f"Subject: {subject}\n"
-        f"Date: {date}\n"
-        f"Message-ID: <{msgid}>\n"
-        "\n"
-    )
+    hdr = f"From: {frm}\nSubject: {subject}\nDate: {date}\nMessage-ID: <{msgid}>\n\n"
     return (hdr + body).encode()
 
 
@@ -108,8 +100,7 @@ def inbox_repo(tmp_path):
 def _run_noteren(*args):
     return subprocess.run(
         [sys.executable, NOTEREN] + list(args),
-        stdout=subprocess.PIPE,
-        stderr=subprocess.PIPE,
+        capture_output=True,
         text=True,
     )
 
@@ -122,6 +113,7 @@ def _std_args():
 # Unit tests for helper functions
 # --------------------------------------------------------------------------
 
+
 def test_message_id_cleanup_strips_angles():
     assert noteren.message_id_cleanup("<abc@host>") == "abc@host"
     assert noteren.message_id_cleanup("  <abc@host>  ") == "abc@host"
@@ -256,9 +248,7 @@ def fake_git_run(git_dir, args, stdin=None):
 
 
 def test_get_email_from_git_missing_commit_exits(monkeypatch):
-    monkeypatch.setattr(
-        noteren, "git_run_command", lambda *args, **kwargs: (0, b"", b"")
-    )
+    monkeypatch.setattr(noteren, "git_run_command", lambda *args, **kwargs: (0, b"", b""))
     with pytest.raises(SystemExit):
         noteren.get_email_from_git("/tmp/repo.git", "deadbeef")
 
@@ -275,9 +265,7 @@ def fake_git_run(git_dir, args, stdin=None):
 
 
 def _parse_email(raw, source, channel, git_id, blob_id):
-    return noteren.parse_email(
-        raw, source, channel, git_id, blob_id, SEGMENT_ID, INTEGRATION_ID
-    )
+    return noteren.parse_email(raw, source, channel, git_id, blob_id, SEGMENT_ID, INTEGRATION_ID)
 
 
 def test_parse_email_extracts_attributes_and_reference():
@@ -351,13 +339,7 @@ def test_parse_email_invalid_date_leaves_timestamp_empty():
 
 
 def test_parse_email_falls_back_to_raw_body(monkeypatch):
-    raw = (
-        b"From: A <a@example.com>\n"
-        b"Subject: s\n"
-        b"Message-ID: <msg@example.com>\n"
-        b"\n"
-        b"line one\n"
-    )
+    raw = b"From: A <a@example.com>\nSubject: s\nMessage-ID: <msg@example.com>\n\nline one\n"
     monkeypatch.setattr(noteren, "get_body", lambda msg: None)
     parsed = _parse_email(raw, "src", "chan", "commit1", "blob1")
     assert "line one\n" in parsed["activityData"]["body"]
@@ -400,9 +382,7 @@ def test_parse_id_writes_json(monkeypatch):
 
 
 def test_parse_range_calls_parse_id_for_each_commit(monkeypatch):
-    monkeypatch.setattr(
-        noteren, "git_run_command", lambda *args, **kwargs: (0, b"c1\nc2\n", b"")
-    )
+    monkeypatch.setattr(noteren, "git_run_command", lambda *args, **kwargs: (0, b"c1\nc2\n", b""))
     seen = []
     monkeypatch.setattr(
         noteren,
@@ -542,9 +522,7 @@ def test_cmd_without_output_uses_stdout(monkeypatch):
 
 
 def test_cmd_verbose_enables_debug_logging(monkeypatch):
-    monkeypatch.setattr(
-        noteren, "parse_id", lambda *args, **kwargs: None
-    )
+    monkeypatch.setattr(noteren, "parse_id", lambda *args, **kwargs: None)
     called = {}
     monkeypatch.setattr(
         noteren.logging,
@@ -627,10 +605,9 @@ def test_cmd_rejects_both_id_and_range(monkeypatch):
 # CLI / integration tests
 # --------------------------------------------------------------------------
 
+
 def test_cli_requires_id_or_range(inbox_repo):
-    r = _run_noteren(
-        "--source=s", "--channel=c", "--git-dir", inbox_repo["gitdir"], *_std_args()
-    )
+    r = _run_noteren("--source=s", "--channel=c", "--git-dir", inbox_repo["gitdir"], *_std_args())
     assert r.returncode != 0
     assert "--git-range or --git-id" in r.stderr
 
@@ -729,6 +706,7 @@ def test_cli_unknown_commit_fails(inbox_repo):
 # Additional edge-case unit tests
 # --------------------------------------------------------------------------
 
+
 def test_get_body_returns_none_when_no_text_parts():
     """A message that contains only text/html must cause get_body to return None."""
     msg = noteren.email.message.EmailMessage()
@@ -753,21 +731,13 @@ def test_get_body_raw_blank_body_after_headers():
 
 def test_run_command_returns_nonzero_exit_code():
     """_run_command must surface a non-zero return code from the child process."""
-    rc, out, err = noteren._run_command(
-        [sys.executable, "-c", "import sys; sys.exit(42)"]
-    )
+    rc, out, err = noteren._run_command([sys.executable, "-c", "import sys; sys.exit(42)"])
     assert rc == 42
 
 
 def test_parse_email_no_message_id_gives_empty_source_id():
     """A message without a Message-ID header must yield sourceId == ''."""
-    raw = (
-        b"From: A <a@example.com>\n"
-        b"Subject: s\n"
-        b"Date: Mon, 1 Jan 2024 12:00:00 +0000\n"
-        b"\n"
-        b"body\n"
-    )
+    raw = b"From: A <a@example.com>\nSubject: s\nDate: Mon, 1 Jan 2024 12:00:00 +0000\n\nbody\n"
     parsed = _parse_email(raw, "src", "chan", "c1", "b1")
     assert parsed["activityData"]["sourceId"] == ""
 
@@ -851,7 +821,7 @@ def test_parse_email_multipart_alternative_uses_text_plain():
         b"Date: Mon, 1 Jan 2024 12:00:00 +0000\n"
         b"Message-ID: <multi@example.com>\n"
         b"MIME-Version: 1.0\n"
-        b"Content-Type: multipart/alternative; boundary=\"boundary\"\n"
+        b'Content-Type: multipart/alternative; boundary="boundary"\n'
         b"\n"
         b"--boundary\n"
         b"Content-Type: text/plain; charset=utf-8\n"
@@ -886,13 +856,7 @@ def test_parse_email_url_contains_message_id():
 
 def test_parse_email_url_with_no_message_id():
     """When Message-ID is absent the url must end with a trailing slash and empty id."""
-    raw = (
-        b"From: A <a@example.com>\n"
-        b"Subject: s\n"
-        b"Date: Mon, 1 Jan 2024 12:00:00 +0000\n"
-        b"\n"
-        b"body\n"
-    )
+    raw = b"From: A <a@example.com>\nSubject: s\nDate: Mon, 1 Jan 2024 12:00:00 +0000\n\nbody\n"
     parsed = _parse_email(raw, "src", "chan", "c1", "b1")
     assert parsed["activityData"]["url"] == "https://lore.kernel.org/r/"
 
@@ -956,19 +920,22 @@ def _c1_argv(inbox_repo):
         "noteren.py",
         "--source=https://example.com/0.git",
         "--channel=test-list",
-        "--git-dir", inbox_repo["gitdir"],
-        "--segment-id", SEGMENT_ID,
-        "--integration-id", INTEGRATION_ID,
-        "--git-id", inbox_repo["c1"],
+        "--git-dir",
+        inbox_repo["gitdir"],
+        "--segment-id",
+        SEGMENT_ID,
+        "--integration-id",
+        INTEGRATION_ID,
+        "--git-id",
+        inbox_repo["c1"],
     ]
 
 
 def test_main_entry_point_via_subprocess(inbox_repo):
     """Running noteren.py as __main__ must work end-to-end."""
     r = subprocess.run(
         [sys.executable, NOTEREN] + _c1_argv(inbox_repo)[1:],
-        stdout=subprocess.PIPE,
-        stderr=subprocess.PIPE,
+        capture_output=True,
         text=True,
     )
     assert r.returncode == 0, r.stderr