[Providers][Slack] Add automatic 429‑retry to channel‑lookup logic

[nailo2c]

Closes: #47225

🐞 Problem

SlackAPIFileOperator looks up channel names with conversations.list (a Tier‑2 Slack API,
≈ 20 req/min). When a DAG fires many parallel tasks, these look‑ups quickly hit
HTTP 429 rate_limited errors and the tasks fail.

Reference: https://api.slack.com/apis/rate-limits

How to reproduce

from airflow import DAG
from airflow.providers.slack.operators.slack import SlackAPIFileOperator
from airflow.utils.task_group import TaskGroup

NUM_UPLOADS = 100

with DAG(
   dag_id="slack_file_upload_rl_repro",
   schedule=None,
   max_active_tasks=NUM_UPLOADS,
) as dag:
   with TaskGroup(group_id="bulk_uploads") as bulk:
       for i in range(NUM_UPLOADS):
           SlackAPIFileOperator(
               task_id=f"upload_{i}",
               slack_conn_id="slack_default",
               channels="#all-aaron-airflow-test",
               filename="/opt/airflow/RELEASE_NOTES.rst",
               initial_comment=f"File {i} 🚀"
           )

Running the DAG returns multiple SlackApiError: ratelimited exceptions.
Screenshot (before fix):

✅ Fix

• SlackHook._call_conversations_list() now retries up to 5 times when Slack responds
  with 429, respecting the Retry‑After header.
• If the limit is still exceeded after 5 attempts, the hook raises
  AirflowException("Max retries reached for conversations.list").

This keeps the operator working without changing user DAGs.

▶️ Manual / end‑to‑end check

Verified in a local Breeze environment against a test Slack workspace:

Unit tests added

• test_call_conversations_list_retries_then_succeeds
• test_call_conversations_list_exceeds_max

[potiuk]

Nice one.

[potiuk]

Thanks for all the screenshots etc.

Out of curiosity - was the fix and PR description AI-assisted ? I genuinly would love to know :D - nothing bad or good, just curious if it was (and with what if it was).

[nailo2c]

Yes, before I submit a PR or take over an issue, I usually discuss it with AI to ensure my understanding aligns with the issue.

I also use AI to check the code for style consistency with the community and to suggest improvements for my first draft of the PR description, making it clearer and reducing the review effort for the reviewer.

I think AI is a great tool to accelerate my efficiency.