aboutsummaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--webmentions_ssg/tasks/receiver.py20
-rw-r--r--webmentions_ssg/tasks/scanner.py5
2 files changed, 12 insertions, 13 deletions
diff --git a/webmentions_ssg/tasks/receiver.py b/webmentions_ssg/tasks/receiver.py
index bed923c..9073930 100644
--- a/webmentions_ssg/tasks/receiver.py
+++ b/webmentions_ssg/tasks/receiver.py
@@ -63,16 +63,16 @@ def fetch_source(source_url: str) -> tuple[httpx.Response, bytes]:
) as client,
client.stream("GET", source_url) as response,
):
- match response.status_code:
- case 200:
- pass
- case 410:
- raise SourceGoneError("Source returned HTTP 410")
- case status:
- if is_temporary_status(status):
- raise TemporaryFetchError(f"Source returned HTTP {status}")
- else:
- raise VerificationError(f"Source returned HTTP {status}")
+ if response.status_code == 200:
+ pass
+ elif response.status_code == 410:
+ raise SourceGoneError("Source returned HTTP 410")
+ elif is_temporary_status(response.status_code):
+ raise TemporaryFetchError(
+ f"Source returned HTTP {response.status_code}"
+ )
+ else:
+ raise VerificationError(f"Source returned HTTP {response.status_code}")
try:
return response, read_limited_body(
diff --git a/webmentions_ssg/tasks/scanner.py b/webmentions_ssg/tasks/scanner.py
index b2d6def..1b1c6ef 100644
--- a/webmentions_ssg/tasks/scanner.py
+++ b/webmentions_ssg/tasks/scanner.py
@@ -133,9 +133,7 @@ def scan_sources_task() -> None:
if not root.is_dir():
raise RuntimeError(f"Source directory is not a directory: {root}")
- base_url = current_app.config.get("SOURCE_BASE_URL")
-
- if base_url is not None:
+ if (base_url := current_app.config.get("SOURCE_BASE_URL")) is not None:
if not is_http_url(base_url):
raise RuntimeError("SOURCE_BASE_URL must be an absolute HTTP or HTTPS URL")
base_url = f"{base_url.rstrip('/')}/"
@@ -146,6 +144,7 @@ def scan_sources_task() -> None:
sa.select(Source).options(selectinload(Source.sent_webmentions))
)
}
+
# Existing sources retain their canonical URLs even when marked deleted.
# Reserve these URLs before scanning so an alias cannot claim one just
# because it is discovered earlier than its persisted owner.