Updating retry logic for API requests

[scorreard]

The weekly update crashed because of an API connection error, I tested this new configuration and it worked.

Crashed weekly update :
https://github.com/galaxyproject/galaxy_codex/actions/runs/21859224353/job/63083600802

[scorreard]

@bebatut or @paulzierep could you check this PR?

We had another API connection error this week, this should give more flexibility

Error in this weekly update

Connection error on attempt 1/3. Retrying in 2.0 seconds...
Connection error on attempt 1/3. Retrying in 2.0 seconds...
Connection error on attempt 1/3. Retrying in 2.0 seconds...
Connection error on attempt 1/3. Retrying in 2.0 seconds...
Connection error on attempt 1/3. Retrying in 2.0 seconds...
Connection error on attempt 2/3. Retrying in 2.0 seconds...
  File "/home/runner/work/galaxy_codex/galaxy_codex/sources/bin/extract_galaxy_workflows.py", line 309, in add_workflows_from_public_servers
    self.add_workflows_from_a_server(url)
  File "/home/runner/work/galaxy_codex/galaxy_codex/sources/bin/extract_galaxy_workflows.py", line 285, in add_workflows_from_a_server
    server_wf = shared.get_request_json(
                ^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/runner/work/galaxy_codex/galaxy_codex/sources/bin/shared.py", line 136, in get_request_json
    raise ConnectionError(
requests.exceptions.ConnectionError: Connection aborted after multiple retries: Remote end closed connection without response

Thanks