librarian: tolerate bad bytes in chunks; log what the result-send does
CI / compile (pull_request) Successful in 10s
CI / unit (pull_request) Successful in 20s
CI / integration (pull_request) Successful in 15s
CI / compile (push) Successful in 29s
CI / unit (push) Successful in 30s
CI / integration (push) Successful in 21s
build / build (push) Failing after 7s

Two field-reported robustness gaps on top of the hang fix.

1) A stray non-UTF-8 byte in a chunk (0x96 in the report) raised
UnicodeDecodeError from readline() - which is a ValueError, so the earlier
`except OSError` did NOT catch it. The finally-sentinel meant no hang, but the
producer died mid-file with a loud traceback and every DOI after the bad byte
went unsearched. Now chunks are opened with errors="replace" (bad bytes become
U+FFFD; DOIs are ASCII so a match is never affected) so the read runs to EOF,
and the producer's except is broadened from OSError to Exception so no per-file
error can ever crash the thread - it's logged and the sentinel still fires.

2) The result-send back to the bot (BackgroundTaskSearch._run) now logs exactly
what goes out - target URL, uuid, DOI count and the DOI list - so the librarian
log plainly shows a result was sent and what was in it. And a failed POST is no
longer fatal: a RequestException used to propagate out of the worker loop and
kill the thread, stalling every future query until restart; it's now caught and
logged, and a non-200 from the bot is logged as a warning.

Verified: tests/unit/test_search_bot.py gains a case writing a chunk with a 0x96
byte before a valid DOI and asserting that DOI is still found (file read to
completion, not aborted). All 5 search_bot unit tests pass.

Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
This commit was merged in pull request #4.
This commit is contained in:
2026-07-30 15:57:48 +02:00
parent 491f957315
commit 031c1f8aea
3 changed files with 71 additions and 17 deletions
+41 -11
View File
@@ -451,18 +451,48 @@ class BackgroundTaskSearch(threading.Thread):
json.dump(database, s_file)
self.app.logger.info("FINISHED")
self.app.logger.info(result)
coroutine = asyncio.to_thread(
requests.post,
f"{MAIN_BOT_ADDRESS}{SEND_RESULTS}",
json=result,
headers=_service_headers(),
timeout=360,
# Send the result back to the bot. Log EXACTLY what goes out (target,
# uuid, how many DOIs and which) so the librarian log makes it plain a
# result was sent and what was in it.
payload = result # shape: {uuid: {DOI: {"Title": ..., "type": ...}}}
hits = payload.get(librarian.uuid, {}) if isinstance(payload, dict) else {}
target = f"{MAIN_BOT_ADDRESS}{SEND_RESULTS}"
self.app.logger.info(
"SENDING result for %s to %s: %d DOI(s): %s",
librarian.uuid,
target,
len(hits),
list(hits.keys()),
)
self.app.logger.info("SENT")
result = await coroutine
self.app.logger.info(result.status_code)
self.app.logger.info("SEND CONFIRMED")
# A failed send must NOT kill this worker - otherwise a bot that is
# momentarily down stalls every future query until the librarian is
# restarted. Log and carry on to the next queued search.
try:
response = await asyncio.to_thread(
requests.post,
target,
json=payload,
headers=_service_headers(),
timeout=360,
)
if response.status_code == 200:
self.app.logger.info(
"SENT result for %s -> HTTP 200 (bot accepted)", librarian.uuid
)
else:
self.app.logger.warning(
"SENT result for %s but bot returned HTTP %s: %s",
librarian.uuid,
response.status_code,
response.text[:500],
)
except requests.exceptions.RequestException as exc:
self.app.logger.error(
"FAILED to send result for %s to %s: %s",
librarian.uuid,
target,
exc,
)
await asyncio.sleep(1)