Hacktoberfest 2026: le issue che i maintainer hanno segnato per ottobre, aperte e adatte ai principianti. Sfoglia le issue Hacktoberfest

Log values inside lists and dicts skip the size bound from #12849

Aperta
  • Qualcosa non va?
#12,864 5 commenti 0 reazioni 1 assegnatario Vedi su GitHub

I maintainer di solito rispondono entro 1 giorno

@julian-risch ci sta già lavorando.

Dal 28/9/2026.

Valutazione

Difficoltà
3/5
Tempo stimato
1-2 giorni
Idoneità per principianti
65/100
Tipo di issue
Bug
Chiarezza
Specificata chiaramente
Stato di attività
Attiva
Stack tecnologico
python
Ambito
devtools

Direzione di ricerca

Look at the bound_event_dict_values function in the logging configuration (likely in haystack/logging/ or haystack/utils/). The issue is that it only bounds top-level values. Start by understanding the current implementation and the test case provided in the issue. Then, implement a recursive walker for lists, tuples, and dicts that applies the same bounding logic, converting nested exceptions with str and truncating containers that exceed MAX_LOG_VALUE_LENGTH. Ensure a depth limit to avoid infinite recursion. Run the reproduction script to verify the fix.

Scritto dal modello di indicizzazione a partire dal testo della issue.

Descrizione

P1

Describe the bug
bound_event_dict_values (added in #12849) only bounds top-level values. When a value is a list or dict, it goes to the renderer unchanged. So the renderer still falls back to repr() for anything inside it, and neither the container nor its items get truncated.

This leaves two gaps:

  • A UnicodeDecodeError inside a list still puts its whole buffer into the log line. That is the leak #12849 fixed for a bare exception.
  • A long list or dict is written out in full. This one happens in Haystack's own code: when an LLM reply is missing expected keys, _parse_dict_from_json in haystack/utils/misc.py logs keys=list(parsed_json.keys()). The event string gets truncated at 4096 characters, but the keys field does not.

Expected behavior
The limit applies to the rendered size of every field, the same way it does for top-level strings and exceptions, so exceptions nested in a container also go through str.

To Reproduce
On main (3c0dc45b), with structlog installed:

import contextlib
import io
import json
import logging

from haystack import logging as haystack_logging
from haystack.utils.misc import _parse_dict_from_json


def last_json_line(emit):
    buffer = io.StringIO()
    with contextlib.redirect_stderr(buffer):
        haystack_logging.configure_logging(use_json=True)
        emit()
    return buffer.getvalue().strip().splitlines()[-1]


# 1. In-tree call: the `keys` field is a list, so it is not bounded.
reply = json.dumps({f"key_{i:05d}": i for i in range(3000)})
line = last_json_line(lambda: _parse_dict_from_json(reply, expected_keys=["score"], raise_on_failure=False))
print("missing-keys warning:", len(line), "chars")

# 2. The UnicodeDecodeError from #12849, wrapped in a list.
try:
    (b"%PDF-1.7\r%\xe2\xe3\xcf\xd3" + b"A" * 100_000).decode("utf-8")
except UnicodeDecodeError as error:
    decode_error = error

line = last_json_line(lambda: logging.getLogger("haystack.repro").warning("Conversion failed", extra={"error": decode_error}))
print("bare exception:      ", len(line), "chars")
line = last_json_line(lambda: logging.getLogger("haystack.repro").warning("Conversion failed", extra={"errors": [decode_error]}))
print("exception in a list: ", len(line), "chars")

Output (same on two runs):

missing-keys warning: 43286 chars
bare exception:       247 chars
exception in a list:  100253 chars

Additional context
One way to fix it: walk list, tuple and dict values in bound_event_dict_values, convert nested exceptions with str, and if the rendered container is still over MAX_LOG_VALUE_LENGTH, replace it with a truncated string. The walk needs a depth limit so self-referencing containers can't recurse forever. I can open a PR for this if that approach works for you.

FAQ Check

System:

  • OS: Linux
  • Haystack version: main (3c0dc45b)
Lingua principale
Python
Stelle
26.6k
Fork
3.2k
Merge medio
1g 13h
PR unite (30g)
263

Preparare l'ambiente

Come iniziare

  1. Leggi tutta la issue e poi la guida ai contributi del progetto.
  2. Commenta sulla issue per dire che te ne occupi tu — evita che due persone facciano lo stesso lavoro.
  3. Fai un fork del repository e lavora su un branch.
  4. Apri una pull request che faccia riferimento al numero della issue.

Altre issue di deepset-ai/haystack

Tutte le issue di deepset-ai/haystack

Issue simili

Altre issue su Python

Ricevi le nuove issue nella tua casella

Un breve riepilogo di issue GitHub adatte ai principianti.