arthexis 0.1.13__py3-none-any.whl → 0.1.15__py3-none-any.whl

core/tasks.py

@@ -1,394 +1,408 @@
-from __future__ import annotations
-
-import logging
-import subprocess
-from pathlib import Path
-import urllib.error
-import urllib.request
-
-from celery import shared_task
-from django.conf import settings
-from django.contrib.auth import get_user_model
-from core import mailer
-from core import github_issues
-from django.utils import timezone
-
-from nodes.models import NetMessage
-
-
-AUTO_UPGRADE_HEALTH_DELAY_SECONDS = 30
-AUTO_UPGRADE_SKIP_LOCK_NAME = "auto_upgrade_skip_revisions.lck"
-
-
-logger = logging.getLogger(__name__)
-
-
-@shared_task
-def heartbeat() -> None:
-    """Log a simple heartbeat message."""
-    logger.info("Heartbeat task executed")
-
-
-@shared_task
-def birthday_greetings() -> None:
-    """Send birthday greetings to users via Net Message and email."""
-    User = get_user_model()
-    today = timezone.localdate()
-    for user in User.objects.filter(birthday=today):
-        NetMessage.broadcast("Happy bday!", user.username)
-        if user.email:
-            mailer.send(
-                "Happy bday!",
-                f"Happy bday! {user.username}",
-                [user.email],
-                settings.DEFAULT_FROM_EMAIL,
-                fail_silently=True,
-            )
-
-
-def _auto_upgrade_log_path(base_dir: Path) -> Path:
-    """Return the log file used for auto-upgrade events."""
-
-    log_dir = base_dir / "logs"
-    log_dir.mkdir(parents=True, exist_ok=True)
-    return log_dir / "auto-upgrade.log"
-
-
-def _append_auto_upgrade_log(base_dir: Path, message: str) -> None:
-    """Append ``message`` to the auto-upgrade log, ignoring errors."""
-
-    try:
-        log_file = _auto_upgrade_log_path(base_dir)
-        timestamp = timezone.now().isoformat()
-        with log_file.open("a") as fh:
-            fh.write(f"{timestamp} {message}\n")
-    except Exception:  # pragma: no cover - best effort logging only
-        logger.warning("Failed to append auto-upgrade log entry: %s", message)
-
-
-def _skip_lock_path(base_dir: Path) -> Path:
-    return base_dir / "locks" / AUTO_UPGRADE_SKIP_LOCK_NAME
-
-
-def _load_skipped_revisions(base_dir: Path) -> set[str]:
-    skip_file = _skip_lock_path(base_dir)
-    try:
-        return {
-            line.strip()
-            for line in skip_file.read_text().splitlines()
-            if line.strip()
-        }
-    except FileNotFoundError:
-        return set()
-    except OSError:
-        logger.warning("Failed to read auto-upgrade skip lockfile")
-        return set()
-
-
-def _add_skipped_revision(base_dir: Path, revision: str) -> None:
-    if not revision:
-        return
-
-    skip_file = _skip_lock_path(base_dir)
-    try:
-        skip_file.parent.mkdir(parents=True, exist_ok=True)
-        existing = _load_skipped_revisions(base_dir)
-        if revision in existing:
-            return
-        with skip_file.open("a", encoding="utf-8") as fh:
-            fh.write(f"{revision}\n")
-        _append_auto_upgrade_log(
-            base_dir, f"Recorded blocked revision {revision} for auto-upgrade"
-        )
-    except OSError:
-        logger.warning(
-            "Failed to update auto-upgrade skip lockfile with revision %s", revision
-        )
-
-
-def _resolve_service_url(base_dir: Path) -> str:
-    """Return the local URL used to probe the Django suite."""
-
-    lock_dir = base_dir / "locks"
-    mode_file = lock_dir / "nginx_mode.lck"
-    mode = "internal"
-    if mode_file.exists():
-        try:
-            mode = mode_file.read_text().strip() or "internal"
-        except OSError:
-            mode = "internal"
-    port = 8000 if mode == "public" else 8888
-    return f"http://127.0.0.1:{port}/"
-
-
-@shared_task
-def check_github_updates() -> None:
-    """Check the GitHub repo for updates and upgrade if needed."""
-    base_dir = Path(__file__).resolve().parent.parent
-    mode_file = base_dir / "locks" / "auto_upgrade.lck"
-    mode = "version"
-    if mode_file.exists():
-        mode = mode_file.read_text().strip()
-
-    branch = "main"
-    subprocess.run(["git", "fetch", "origin", branch], cwd=base_dir, check=True)
-
-    log_file = _auto_upgrade_log_path(base_dir)
-    with log_file.open("a") as fh:
-        fh.write(
-            f"{timezone.now().isoformat()} check_github_updates triggered\n"
-        )
-
-    notify = None
-    startup = None
-    try:  # pragma: no cover - optional dependency
-        from core.notifications import notify  # type: ignore
-    except Exception:
-        notify = None
-    try:  # pragma: no cover - optional dependency
-        from nodes.apps import _startup_notification as startup  # type: ignore
-    except Exception:
-        startup = None
-
-    remote_revision = (
-        subprocess.check_output(
-            ["git", "rev-parse", f"origin/{branch}"], cwd=base_dir
-        )
-        .decode()
-        .strip()
-    )
-
-    skipped_revisions = _load_skipped_revisions(base_dir)
-    if remote_revision in skipped_revisions:
-        _append_auto_upgrade_log(
-            base_dir, f"Skipping auto-upgrade for blocked revision {remote_revision}"
-        )
-        if startup:
-            startup()
-        return
-
-    upgrade_stamp = timezone.now().strftime("@ %Y%m%d %H:%M")
-
-    upgrade_was_applied = False
-
-    if mode == "latest":
-        local = (
-            subprocess.check_output(["git", "rev-parse", branch], cwd=base_dir)
-            .decode()
-            .strip()
-        )
-        if local == remote_revision:
-            if startup:
-                startup()
-            return
-        if notify:
-            notify("Upgrading...", upgrade_stamp)
-        args = ["./upgrade.sh", "--latest", "--no-restart"]
-        upgrade_was_applied = True
-    else:
-        local = "0"
-        version_file = base_dir / "VERSION"
-        if version_file.exists():
-            local = version_file.read_text().strip()
-        remote = (
-            subprocess.check_output(
-                [
-                    "git",
-                    "show",
-                    f"origin/{branch}:VERSION",
-                ],
-                cwd=base_dir,
-            )
-            .decode()
-            .strip()
-        )
-        if local == remote:
-            if startup:
-                startup()
-            return
-        if notify:
-            notify("Upgrading...", upgrade_stamp)
-        args = ["./upgrade.sh", "--no-restart"]
-        upgrade_was_applied = True
-
-    with log_file.open("a") as fh:
-        fh.write(
-            f"{timezone.now().isoformat()} running: {' '.join(args)}\n"
-        )
-
-    subprocess.run(args, cwd=base_dir, check=True)
-
-    service_file = base_dir / "locks/service.lck"
-    if service_file.exists():
-        service = service_file.read_text().strip()
-        subprocess.run(
-            [
-                "sudo",
-                "systemctl",
-                "kill",
-                "--signal=TERM",
-                service,
-            ]
-        )
-    else:
-        subprocess.run(["pkill", "-f", "manage.py runserver"])
-
-    if upgrade_was_applied:
-        _append_auto_upgrade_log(
-            base_dir,
-            (
-                "Scheduled post-upgrade health check in %s seconds"
-                % AUTO_UPGRADE_HEALTH_DELAY_SECONDS
-            ),
-        )
-        _schedule_health_check(1)
-
-
-@shared_task
-def poll_email_collectors() -> None:
-    """Poll all configured email collectors for new messages."""
-    try:
-        from .models import EmailCollector
-    except Exception:  # pragma: no cover - app not ready
-        return
-
-    for collector in EmailCollector.objects.all():
-        collector.collect()
-
-
-@shared_task
-def report_runtime_issue(
-    title: str,
-    body: str,
-    labels: list[str] | None = None,
-    fingerprint: str | None = None,
-):
-    """Report a runtime issue to GitHub using :mod:`core.github_issues`."""
-
-    try:
-        response = github_issues.create_issue(
-            title,
-            body,
-            labels=labels,
-            fingerprint=fingerprint,
-        )
-    except Exception:
-        logger.exception("Failed to report runtime issue '%s'", title)
-        raise
-
-    if response is None:
-        logger.info("Skipped GitHub issue creation for fingerprint %s", fingerprint)
-    else:
-        logger.info("Reported runtime issue '%s' to GitHub", title)
-
-    return response
-
-
-def _record_health_check_result(
-    base_dir: Path, attempt: int, status: int | None, detail: str
-) -> None:
-    status_display = status if status is not None else "unreachable"
-    message = "Health check attempt %s %s (%s)" % (attempt, detail, status_display)
-    _append_auto_upgrade_log(base_dir, message)
-
-
-def _schedule_health_check(next_attempt: int) -> None:
-    verify_auto_upgrade_health.apply_async(
-        kwargs={"attempt": next_attempt},
-        countdown=AUTO_UPGRADE_HEALTH_DELAY_SECONDS,
-    )
-
-
-def _handle_failed_health_check(base_dir: Path, detail: str) -> None:
-    revision = ""
-    try:
-        revision = (
-            subprocess.check_output(["git", "rev-parse", "HEAD"], cwd=base_dir)
-            .decode()
-            .strip()
-        )
-    except Exception:  # pragma: no cover - best effort capture
-        logger.warning("Failed to determine revision during auto-upgrade revert")
-
-    _add_skipped_revision(base_dir, revision)
-    _append_auto_upgrade_log(base_dir, "Health check failed; reverting upgrade")
-    subprocess.run(["./upgrade.sh", "--revert"], cwd=base_dir, check=True)
-
-
-@shared_task
-def verify_auto_upgrade_health(attempt: int = 1) -> bool | None:
-    """Verify the upgraded suite responds successfully.
-
-    After the post-upgrade delay the site is probed once; any response other
-    than HTTP 200 triggers an automatic revert and records the failing
-    revision so future upgrade attempts skip it.
-    """
-
-    base_dir = Path(__file__).resolve().parent.parent
-    url = _resolve_service_url(base_dir)
-    request = urllib.request.Request(
-        url,
-        headers={"User-Agent": "Arthexis-AutoUpgrade/1.0"},
-    )
-
-    status: int | None = None
-    detail = "succeeded"
-    try:
-        with urllib.request.urlopen(request, timeout=10) as response:
-            status = getattr(response, "status", response.getcode())
-    except urllib.error.HTTPError as exc:
-        status = exc.code
-        detail = f"returned HTTP {exc.code}"
-        logger.warning(
-            "Auto-upgrade health check attempt %s returned HTTP %s", attempt, exc.code
-        )
-    except urllib.error.URLError as exc:
-        detail = f"failed with {exc}"
-        logger.warning(
-            "Auto-upgrade health check attempt %s failed: %s", attempt, exc
-        )
-    except Exception as exc:  # pragma: no cover - unexpected network error
-        detail = f"failed with {exc}"
-        logger.exception(
-            "Unexpected error probing suite during auto-upgrade attempt %s", attempt
-        )
-        _record_health_check_result(base_dir, attempt, status, detail)
-        _handle_failed_health_check(base_dir, detail)
-        return False
-
-    if status == 200:
-        _record_health_check_result(base_dir, attempt, status, "succeeded")
-        logger.info(
-            "Auto-upgrade health check succeeded on attempt %s with HTTP %s",
-            attempt,
-            status,
-        )
-        return True
-
-    if detail == "succeeded":
-        if status is not None:
-            detail = f"returned HTTP {status}"
-        else:
-            detail = "failed with unknown status"
-
-    _record_health_check_result(base_dir, attempt, status, detail)
-    _handle_failed_health_check(base_dir, detail)
-    return False
-
-
-@shared_task
-def run_client_report_schedule(schedule_id: int) -> None:
-    """Execute a :class:`core.models.ClientReportSchedule` run."""
-
-    from core.models import ClientReportSchedule
-
-    schedule = ClientReportSchedule.objects.filter(pk=schedule_id).first()
-    if not schedule:
-        logger.warning("ClientReportSchedule %s no longer exists", schedule_id)
-        return
-
-    try:
-        schedule.run()
-    except Exception:
-        logger.exception("ClientReportSchedule %s failed", schedule_id)
-        raise
+from __future__ import annotations
+
+import logging
+import shutil
+import subprocess
+from pathlib import Path
+import urllib.error
+import urllib.request
+
+from celery import shared_task
+from django.conf import settings
+from django.contrib.auth import get_user_model
+from core import mailer
+from core import github_issues
+from django.utils import timezone
+
+from nodes.models import NetMessage
+
+
+AUTO_UPGRADE_HEALTH_DELAY_SECONDS = 30
+AUTO_UPGRADE_SKIP_LOCK_NAME = "auto_upgrade_skip_revisions.lck"
+
+
+logger = logging.getLogger(__name__)
+
+
+@shared_task
+def heartbeat() -> None:
+    """Log a simple heartbeat message."""
+    logger.info("Heartbeat task executed")
+
+
+@shared_task
+def birthday_greetings() -> None:
+    """Send birthday greetings to users via Net Message and email."""
+    User = get_user_model()
+    today = timezone.localdate()
+    for user in User.objects.filter(birthday=today):
+        NetMessage.broadcast("Happy bday!", user.username)
+        if user.email:
+            mailer.send(
+                "Happy bday!",
+                f"Happy bday! {user.username}",
+                [user.email],
+                settings.DEFAULT_FROM_EMAIL,
+                fail_silently=True,
+            )
+
+
+def _auto_upgrade_log_path(base_dir: Path) -> Path:
+    """Return the log file used for auto-upgrade events."""
+
+    log_dir = base_dir / "logs"
+    log_dir.mkdir(parents=True, exist_ok=True)
+    return log_dir / "auto-upgrade.log"
+
+
+def _append_auto_upgrade_log(base_dir: Path, message: str) -> None:
+    """Append ``message`` to the auto-upgrade log, ignoring errors."""
+
+    try:
+        log_file = _auto_upgrade_log_path(base_dir)
+        timestamp = timezone.now().isoformat()
+        with log_file.open("a") as fh:
+            fh.write(f"{timestamp} {message}\n")
+    except Exception:  # pragma: no cover - best effort logging only
+        logger.warning("Failed to append auto-upgrade log entry: %s", message)
+
+
+def _skip_lock_path(base_dir: Path) -> Path:
+    return base_dir / "locks" / AUTO_UPGRADE_SKIP_LOCK_NAME
+
+
+def _load_skipped_revisions(base_dir: Path) -> set[str]:
+    skip_file = _skip_lock_path(base_dir)
+    try:
+        return {
+            line.strip()
+            for line in skip_file.read_text().splitlines()
+            if line.strip()
+        }
+    except FileNotFoundError:
+        return set()
+    except OSError:
+        logger.warning("Failed to read auto-upgrade skip lockfile")
+        return set()
+
+
+def _add_skipped_revision(base_dir: Path, revision: str) -> None:
+    if not revision:
+        return
+
+    skip_file = _skip_lock_path(base_dir)
+    try:
+        skip_file.parent.mkdir(parents=True, exist_ok=True)
+        existing = _load_skipped_revisions(base_dir)
+        if revision in existing:
+            return
+        with skip_file.open("a", encoding="utf-8") as fh:
+            fh.write(f"{revision}\n")
+        _append_auto_upgrade_log(
+            base_dir, f"Recorded blocked revision {revision} for auto-upgrade"
+        )
+    except OSError:
+        logger.warning(
+            "Failed to update auto-upgrade skip lockfile with revision %s", revision
+        )
+
+
+def _resolve_service_url(base_dir: Path) -> str:
+    """Return the local URL used to probe the Django suite."""
+
+    lock_dir = base_dir / "locks"
+    mode_file = lock_dir / "nginx_mode.lck"
+    mode = "internal"
+    if mode_file.exists():
+        try:
+            value = mode_file.read_text(encoding="utf-8").strip()
+        except OSError:
+            value = ""
+        if value:
+            mode = value.lower()
+    port = 8000 if mode == "public" else 8888
+    return f"http://127.0.0.1:{port}/"
+
+
+@shared_task
+def check_github_updates() -> None:
+    """Check the GitHub repo for updates and upgrade if needed."""
+    base_dir = Path(__file__).resolve().parent.parent
+    mode_file = base_dir / "locks" / "auto_upgrade.lck"
+    mode = "version"
+    if mode_file.exists():
+        try:
+            mode = mode_file.read_text().strip() or "version"
+        except (OSError, UnicodeDecodeError):
+            logger.warning(
+                "Failed to read auto-upgrade mode lockfile", exc_info=True
+            )
+
+    branch = "main"
+    subprocess.run(["git", "fetch", "origin", branch], cwd=base_dir, check=True)
+
+    log_file = _auto_upgrade_log_path(base_dir)
+    with log_file.open("a") as fh:
+        fh.write(
+            f"{timezone.now().isoformat()} check_github_updates triggered\n"
+        )
+
+    notify = None
+    startup = None
+    try:  # pragma: no cover - optional dependency
+        from core.notifications import notify  # type: ignore
+    except Exception:
+        notify = None
+    try:  # pragma: no cover - optional dependency
+        from nodes.apps import _startup_notification as startup  # type: ignore
+    except Exception:
+        startup = None
+
+    remote_revision = (
+        subprocess.check_output(
+            ["git", "rev-parse", f"origin/{branch}"], cwd=base_dir
+        )
+        .decode()
+        .strip()
+    )
+
+    skipped_revisions = _load_skipped_revisions(base_dir)
+    if remote_revision in skipped_revisions:
+        _append_auto_upgrade_log(
+            base_dir, f"Skipping auto-upgrade for blocked revision {remote_revision}"
+        )
+        if startup:
+            startup()
+        return
+
+    upgrade_stamp = timezone.now().strftime("@ %Y%m%d %H:%M")
+
+    upgrade_was_applied = False
+
+    if mode == "latest":
+        local = (
+            subprocess.check_output(["git", "rev-parse", branch], cwd=base_dir)
+            .decode()
+            .strip()
+        )
+        if local == remote_revision:
+            if startup:
+                startup()
+            return
+        if notify:
+            notify("Upgrading...", upgrade_stamp)
+        args = ["./upgrade.sh", "--latest", "--no-restart"]
+        upgrade_was_applied = True
+    else:
+        local = "0"
+        version_file = base_dir / "VERSION"
+        if version_file.exists():
+            local = version_file.read_text().strip()
+        remote = (
+            subprocess.check_output(
+                [
+                    "git",
+                    "show",
+                    f"origin/{branch}:VERSION",
+                ],
+                cwd=base_dir,
+            )
+            .decode()
+            .strip()
+        )
+        if local == remote:
+            if startup:
+                startup()
+            return
+        if notify:
+            notify("Upgrading...", upgrade_stamp)
+        args = ["./upgrade.sh", "--no-restart"]
+        upgrade_was_applied = True
+
+    with log_file.open("a") as fh:
+        fh.write(
+            f"{timezone.now().isoformat()} running: {' '.join(args)}\n"
+        )
+
+    subprocess.run(args, cwd=base_dir, check=True)
+
+    if shutil.which("gway"):
+        try:
+            subprocess.run(["gway", "upgrade"], check=True)
+        except subprocess.CalledProcessError:
+            logger.warning("gway upgrade failed; continuing anyway", exc_info=True)
+
+    service_file = base_dir / "locks/service.lck"
+    if service_file.exists():
+        service = service_file.read_text().strip()
+        subprocess.run(
+            [
+                "sudo",
+                "systemctl",
+                "kill",
+                "--signal=TERM",
+                service,
+            ]
+        )
+    else:
+        subprocess.run(["pkill", "-f", "manage.py runserver"])
+
+    if upgrade_was_applied:
+        _append_auto_upgrade_log(
+            base_dir,
+            (
+                "Scheduled post-upgrade health check in %s seconds"
+                % AUTO_UPGRADE_HEALTH_DELAY_SECONDS
+            ),
+        )
+        _schedule_health_check(1)
+
+
+@shared_task
+def poll_email_collectors() -> None:
+    """Poll all configured email collectors for new messages."""
+    try:
+        from .models import EmailCollector
+    except Exception:  # pragma: no cover - app not ready
+        return
+
+    for collector in EmailCollector.objects.all():
+        collector.collect()
+
+
+@shared_task
+def report_runtime_issue(
+    title: str,
+    body: str,
+    labels: list[str] | None = None,
+    fingerprint: str | None = None,
+):
+    """Report a runtime issue to GitHub using :mod:`core.github_issues`."""
+
+    try:
+        response = github_issues.create_issue(
+            title,
+            body,
+            labels=labels,
+            fingerprint=fingerprint,
+        )
+    except Exception:
+        logger.exception("Failed to report runtime issue '%s'", title)
+        raise
+
+    if response is None:
+        logger.info("Skipped GitHub issue creation for fingerprint %s", fingerprint)
+    else:
+        logger.info("Reported runtime issue '%s' to GitHub", title)
+
+    return response
+
+
+def _record_health_check_result(
+    base_dir: Path, attempt: int, status: int | None, detail: str
+) -> None:
+    status_display = status if status is not None else "unreachable"
+    message = "Health check attempt %s %s (%s)" % (attempt, detail, status_display)
+    _append_auto_upgrade_log(base_dir, message)
+
+
+def _schedule_health_check(next_attempt: int) -> None:
+    verify_auto_upgrade_health.apply_async(
+        kwargs={"attempt": next_attempt},
+        countdown=AUTO_UPGRADE_HEALTH_DELAY_SECONDS,
+    )
+
+
+def _handle_failed_health_check(base_dir: Path, detail: str) -> None:
+    revision = ""
+    try:
+        revision = (
+            subprocess.check_output(["git", "rev-parse", "HEAD"], cwd=base_dir)
+            .decode()
+            .strip()
+        )
+    except Exception:  # pragma: no cover - best effort capture
+        logger.warning("Failed to determine revision during auto-upgrade revert")
+
+    _add_skipped_revision(base_dir, revision)
+    _append_auto_upgrade_log(base_dir, "Health check failed; reverting upgrade")
+    subprocess.run(["./upgrade.sh", "--revert"], cwd=base_dir, check=True)
+
+
+@shared_task
+def verify_auto_upgrade_health(attempt: int = 1) -> bool | None:
+    """Verify the upgraded suite responds successfully.
+
+    After the post-upgrade delay the site is probed once; any response other
+    than HTTP 200 triggers an automatic revert and records the failing
+    revision so future upgrade attempts skip it.
+    """
+
+    base_dir = Path(__file__).resolve().parent.parent
+    url = _resolve_service_url(base_dir)
+    request = urllib.request.Request(
+        url,
+        headers={"User-Agent": "Arthexis-AutoUpgrade/1.0"},
+    )
+
+    status: int | None = None
+    detail = "succeeded"
+    try:
+        with urllib.request.urlopen(request, timeout=10) as response:
+            status = getattr(response, "status", response.getcode())
+    except urllib.error.HTTPError as exc:
+        status = exc.code
+        detail = f"returned HTTP {exc.code}"
+        logger.warning(
+            "Auto-upgrade health check attempt %s returned HTTP %s", attempt, exc.code
+        )
+    except urllib.error.URLError as exc:
+        detail = f"failed with {exc}"
+        logger.warning(
+            "Auto-upgrade health check attempt %s failed: %s", attempt, exc
+        )
+    except Exception as exc:  # pragma: no cover - unexpected network error
+        detail = f"failed with {exc}"
+        logger.exception(
+            "Unexpected error probing suite during auto-upgrade attempt %s", attempt
+        )
+        _record_health_check_result(base_dir, attempt, status, detail)
+        _handle_failed_health_check(base_dir, detail)
+        return False
+
+    if status == 200:
+        _record_health_check_result(base_dir, attempt, status, "succeeded")
+        logger.info(
+            "Auto-upgrade health check succeeded on attempt %s with HTTP %s",
+            attempt,
+            status,
+        )
+        return True
+
+    if detail == "succeeded":
+        if status is not None:
+            detail = f"returned HTTP {status}"
+        else:
+            detail = "failed with unknown status"
+
+    _record_health_check_result(base_dir, attempt, status, detail)
+    _handle_failed_health_check(base_dir, detail)
+    return False
+
+
+@shared_task
+def run_client_report_schedule(schedule_id: int) -> None:
+    """Execute a :class:`core.models.ClientReportSchedule` run."""
+
+    from core.models import ClientReportSchedule
+
+    schedule = ClientReportSchedule.objects.filter(pk=schedule_id).first()
+    if not schedule:
+        logger.warning("ClientReportSchedule %s no longer exists", schedule_id)
+        return
+
+    try:
+        schedule.run()
+    except Exception:
+        logger.exception("ClientReportSchedule %s failed", schedule_id)
+        raise