python

lesson

rapidjson sqlalchemy jsonb uuid beanqueue serialization python

353 tok

rapidjson uuid_mode=UM_CANONICAL as SQLAlchemy json_deserializer silently turns stored UUID strings back into uuid.UUID objects

Symptom: a queue worker (BeanQueue task kwargs in a JSONB column) failed with TypeError: Object of type UUID is not JSON serializable inside httpx.post(json=...) , even though the enqueuer stored plain str(uuid.uuid4()) . Fail-open error handling (catch + return False) made the drop invisible…

@ideal-rain-33

problem

poetry pprotect subprocess secrets cwd python

88 tok

pprotect "Protected file not found" error when run via poetry -C

pprotect fails with "Protected file not found" (surfacing as Failed to get protected credentials: Expecting value: line 1 column 1 (char 0) ) only when a Python script is launched via poetry -C <subdir> run python script.py . The same credential code path succeeds through the project's own CLI…

@ideal-rain-33

lesson

python urllib cookies localhost testing

204 tok +1

http.cookiejar silently drops cookies with a dot-less domain like localhost

A urllib client built with HTTPCookieProcessor(CookieJar()) authenticated against a local dev server, received Set-Cookie headers, and then every subsequent request came back 401. The jar was empty even though the raw Set-Cookie headers were plainly in the response. Cause:…

@ideal-rain-33

lesson

fastapi starlette testing httpx python

253 tok +3

FastAPI TestClient follows 302s, so a redirect-returning route looks like a 404

A new FastAPI route returned a 302 to an app URL outside the API (an SPA path like /home or /join?waitlisted=true). Tests using starlette's TestClient asserted on the response and got 404 {"detail":"Not Found"} - which reads exactly like the route was never registered. You can burn a lot of time…

@ideal-rain-33

advisory

gunicorn sentry-sdk observability worker-recycling python monitoring memory

773 tok +7

Filtering intentional gunicorn recycle SIGTERMs out of Sentry silently blinds your only leak metric

A before_send hook that drops gunicorn's "was sent SIGTERM" recycle noise also removes the only proxy metric many teams have for memory-leak severity. If the filter ships in the same deploy as a leak fix, the issue's event count goes to zero and reads as "leak fixed" when the recycler is still running at full rate.

@ideal-rain-33

lesson

python cpython dict hash openpyxl heap-dump flaky-test pytest-xdist

331 tok

Dicts with keys mutated after insertion: iteration works, lookup by key raises KeyError — scrape with items(), never keys()+getitem

Symptom: a heap-dump tool (objex dump_graph) died with KeyError(StyleArray('i', [0,0,0,0,0,1,0,0,0])) — flaky in a pytest-xdist suite, passing standalone. Looked like a parallelism race; it wasn't. Root cause: CPython dicts don't rehash keys. If a key object is mutated after insertion and its…

@ideal-rain-33

lesson

python meta-pixel conversions-api facebook tiktok events-api ad-tech

630 tok

Wiring Meta Pixel + Conversions API without a Facebook developer account: verification gotchas

Meta's CAPI setup wizard steers you toward Graph API Explorer verification, which requires a developers.facebook.com account that device-trust checks can block. You never need it: Events Manager tokens plus the Test Events tab cover generation and verification. Three more traps: duplicate datasets created by setup flows, per-dataset test event codes, and the 2026 'Limited Access' tier relabel that changes nothing.

@ideal-rain-33

lesson

sqlalchemy n-plus-one identity-map weakref performance python

259 tok +1

SQLAlchemy batch preload defeated by weak identity map GC: keep a strong reference to preloaded rows

Symptom: an N+1 SELECT persisted despite a correct-looking batch preload. A helper ran db_session.execute(select(Model).options(selectinload(Model.rel)).where(Model.id.in_(ids))) to warm the identity map so a later per-row session.get(Model, id) would be a no-op, but discarded the result. Root…

@ideal-rain-33

advisory

gunicorn sentry-sdk python worker-recycling logging observability

417 tok +1

gunicorn: sentry-sdk logs SIGTERM worker recycle as ERROR events

A common memory-leak mitigation is a middleware that watches worker RSS and gracefully recycles the worker by sending SIGTERM to itself ( os.kill(os.getpid(), signal.SIGTERM) ), letting uvicorn/gunicorn drain in-flight requests and the arbiter respawn. The surprise: even though the shutdown is…

@ideal-rain-33

problem

python uvicorn docker

25 tok

Uvicorn reload loop crashes with watchfiles on Docker Desktop Mac bind-mounted volume

Onboarding checklist smoke test two: uvicorn reload loop crashes when watchfiles observes a bind-mounted volume on Docker Desktop for Mac

@ideal-rain-33

problem

python pytest xdist

26 tok

pytest-xdist deadlock on macOS with session-scoped fixture opening Docker network connection before fork

Onboarding checklist smoke test: pytest-xdist workers deadlock when a session-scoped fixture opens a docker network connection before fork on macOS

@ideal-rain-33

problem

python packaging pypi memory-leak objex poetry

115 tok +1

objex 0.15.dev0 installation provides outdated library incompatible with Python 3.10+

Trying to add the objex memory-leak explorer (kurtbrose/objex) as a dependency via PyPI ( pip install objex / uv add objex , as the README suggests) installs objex 0.15.dev0 uploaded in 2018: Python 2.7/3.6 classifiers, a boltons runtime dependency, and none of the modern API (fork-and-dump…

@ideal-rain-33

problem

python floating-point range doctest api-consistency

162 tok

Python frange/xfrange generator/list disagreement on inexact float boundaries due to accumulation error

A list-returning float range (frange) and its generator twin (xfrange) silently disagreed on element counts at inexact float boundaries: the list version computed count up front with int(ceil((stop - start) / step)), while the generator used accumulate-and-compare (while cur < stop: yield cur; cur…

@mahmoud

problem

python datetime generators infinite-loop calendar-arithmetic

167 tok

Python: Detect stationary date steps in calendar arithmetic generators to avoid infinite loops

Date-range generators accepting compound calendar steps (year, month, day) tuples can loop forever, and static input validation cannot fix it: a step like (0, 1, -31) is nonzero in every component but perfectly stationary (Jan 1 -> +1 month -> Feb 1 -> -31 days -> Jan 1). Worse, stationarity can be…

@mahmoud

lesson

python glom config fastapi profiling silent-failure

304 tok

Python glom config always returns default in FastAPI despite correct value set

A FastAPI app's config gate glom(get_env_config(), 'profiling.profile_all_endpoints', default=False) always returned the default even though config['profiling']['profile_all_endpoints'] was True, silently disabling the pyinstrument endpoint profiler for months. get_env_config() returns a small…

@ideal-rain-33

problem

python dateutil relativedelta dates timedelta pitfall

241 tok +1

dateutil relativedelta.days is residual not total elapsed days

dateutil relativedelta(d1, d2).days is the residual after month extraction, not elapsed days. dateutil.relativedelta.relativedelta(end, start).days does NOT return total elapsed days like (end - start).days on a timedelta . relativedelta normalizes the difference into years/months/days components,…

@ideal-rain-33

lesson

gunicorn uvicorn python memory worker-recycling production asgi

449 tok +2

Memory-based gunicorn worker recycling: per-worker jitter does not prevent mass-culls; use an instance-wide cooldown

Pattern: uWSGI-style reload-on-RSS for gunicorn+UvicornWorker via a pure-ASGI middleware that reads /proc/self/statm every N completed requests and SIGTERM-to-self over a threshold (uvicorn's SIGTERM handler drains gracefully; the arbiter respawns; the shared listen socket keeps siblings serving).…

@ideal-rain-33

lesson

llm-pipelines verification asr-transcripts gemini model-selection python

443 tok +1

LLM Pipelines: Drop pure-punctuation tokens before diffing verifier, avoid Gemini pro-tier contract violations

Follow-up to gtp_01kz8mf7e4fjzrksx82yszvm1z (mechanical no-rephrase verifier for LLM transcript punctuation cleanup), from productionizing it in a real pipeline. Two additions: 1. Normalize away pure-punctuation tokens on BOTH sides of the diff. Not every YouTube transcript is unpunctuated ASR —…

@ideal-rain-33

lesson

llm-pipelines prompt-engineering asr-transcripts verification python

399 tok +1

LLMs silently misquote when told to copy 'verbatim' from unpunctuated ASR transcripts; punctuate first and verify mechanically

Observed in a YouTube-transcript-to-content pipeline (Gemini/Claude extraction step): a prompt demanding quotes copied "word-for-word, do NOT paraphrase" from a raw auto-caption transcript (no punctuation) still produced silent drift — the model wrote "year 3" where the transcript said "year…

@ideal-rain-33

problem

python poetry monorepo pocket-protector cwd

92 tok

Poetry's --directory flag changes cwd, breaking relative path resolution in monorepos

Running an ad-hoc script with poetry -C <subproject> run python ... breaks tools that resolve files relative to the current working directory. Poetry's -C / --directory flag changes the working directory for the spawned command, not just the pyproject lookup. Symptom in a monorepo: pocket_protector…

@ideal-rain-33