Files
computer-playbook/plugins/lookup/container_address.py
T
kevinveenbirkenbachandClaude Opus 5 a05071fe78 refactor(roles): let a role declare its entity name and drop the get_ prefix
The entity name keys a role's instance directory, its compose project and its container name, and was derived only by stripping the longest matching category prefix. Two roles under different categories whose names share a suffix therefore collapse onto one entity and overwrite each other's rendered stack, which no provides: can repair.

A literal entity_name in vars/main.yml now wins over the derivation. web-app-seaweedfs-console takes its shorter name back and declares the entity seaweedfs-console, which frees the seaweedfs key in its own services map for the engine consumer and removes the provides:, covers: and alias entries that stood in for it. A new lint pairs the declaration with the primary key it has to match, because a drift between them would drop the role from the provider set in silence.

Compared across all 288 roles, exactly one entity changes, and it restores what main already keeps on disk, so no instance directory moves.

The function and the filter lose their get_ prefix. Two lints read that prefix as the discriminator between the variable and the filter that fills it; both discriminate on the pipe now. Ansible's templar resolves a filter and a same-named variable independently, verified against ansible-core 2.20.2.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
2026-10-03 19:40:53 +02:00

162 lines
5.8 KiB
Python

"""Lookup `container_address`: addressable form of a container for
`docker exec`, mode-aware.
Single SPOT for the swarm/compose split on container addressing.
Returns a string safe to embed unchanged into a shell `container exec`
command:
* Compose mode - the bare service-key (e.g. `mattermost`). Compose
sets `container_name:` explicitly so the name resolves locally.
* Swarm mode - a shell subshell that resolves the running task's
container ID at exec time, e.g.
`"$(/usr/bin/resolve-container-id mattermost mattermost)"`. Swarm
names every service ``<stack>_<key>`` and accepts no prefix matching;
the helper composes that name from STACK and SERVICE_KEY args. The
subshell defers resolution to the moment of exec, avoiding the stale
ID and parse-time-failure pitfalls of returning a resolved value at
vars-resolution time.
Because the lookup emits ``$(...)`` shell syntax, callers MUST use the
``ansible.builtin.shell`` module (not ``command``) for the surrounding
``container exec`` invocation. A companion lint rule enforces this.
The helper script `BIN_RESOLVE_CONTAINER_ID` is deployed by
sys-svc-compose during the constructor stage. It refuses to run on
non-manager nodes (exit 64) and reports clear errors when the service
has no running task (exit 65) or the task is not yet bound to a
container (exit 66).
Examples::
# roles/web-app-X/vars/main.yml
X_CONTAINER_ADDRESS: "{{ lookup('container_address', application_id, 'x') }}"
# roles/web-app-X/tasks/01_setup.yml
- shell: |
container exec -i {{ X_CONTAINER_ADDRESS }} my-cli command
Both terms (application_id, service_key) are required.
``services.<service_key>.name`` must exist in the resolved application
config; the stack name is derived from ``application_id`` via the
``entity_name`` filter (matching how ``docker stack deploy``
names the stack in ``sys-svc-compose/handlers/main.yml``).
"""
from __future__ import annotations
import contextlib
import shlex
from typing import Any
from ansible.errors import AnsibleError
from ansible.plugins.loader import lookup_loader
from ansible.plugins.lookup import LookupBase
from utils.roles.entity.name import entity_name
def _as_str(value: Any) -> str:
return "" if value is None else str(value).strip()
def _resolve_bare_name(
applications: dict[str, Any], application_id: str, service_key: str
) -> str:
app = applications.get(application_id)
if not isinstance(app, dict):
raise AnsibleError(
f"container_address: unknown application_id '{application_id}'"
)
services = app.get("services") or {}
if not isinstance(services, dict):
raise AnsibleError(
f"container_address: '{application_id}' has no services dict"
)
entry = services.get(service_key)
if not isinstance(entry, dict) and service_key == "application":
entity = entity_name(application_id)
if entity:
entry = services.get(entity)
if not isinstance(entry, dict):
raise AnsibleError(
f"container_address: service '{service_key}' missing in "
f"'{application_id}' services config"
)
name = entry.get("name")
if not isinstance(name, str) or not name.strip():
raise AnsibleError(
f"container_address: services.{service_key}.name not set for "
f"'{application_id}'"
)
return name.strip()
class LookupModule(LookupBase):
def run(
self,
terms: list[Any] | None,
variables: dict[str, Any] | None = None,
**kwargs: Any,
) -> list[str]:
if not terms or len(terms) != 2:
raise AnsibleError(
"container_address lookup requires exactly two terms: "
"application_id and service_key"
)
application_id = _as_str(terms[0])
service_key = _as_str(terms[1])
if not application_id:
raise AnsibleError(
"container_address: application_id must be a non-empty string"
)
if not service_key:
raise AnsibleError(
"container_address: service_key must be a non-empty string"
)
vars_ = variables or getattr(self._templar, "available_variables", {}) or {}
templar = getattr(self, "_templar", None)
applications = lookup_loader.get(
"applications", loader=self._loader, templar=getattr(self, "_templar", None)
).run([], variables=vars_)[0]
bare_name = _resolve_bare_name(applications, application_id, service_key)
raw_mode = vars_.get("DEPLOYMENT_MODE", "compose")
if templar is not None:
with contextlib.suppress(Exception):
raw_mode = templar.template(raw_mode)
deployment_mode = str(raw_mode).strip()
mode_force = vars_.get("compose_mode_force", "")
if templar is not None:
with contextlib.suppress(Exception):
mode_force = templar.template(mode_force)
deployment_mode = _as_str(mode_force) or deployment_mode
if deployment_mode != "swarm":
return [bare_name]
stack_name = entity_name(application_id)
if not stack_name:
raise AnsibleError(
f"container_address: cannot derive stack name from "
f"application_id '{application_id}'"
)
bin_resolver = vars_.get("BIN_RESOLVE_CONTAINER_ID")
if templar is not None and bin_resolver is not None:
with contextlib.suppress(Exception):
bin_resolver = templar.template(bin_resolver)
bin_resolver = _as_str(bin_resolver) or "/usr/bin/resolve-container-id"
return [
(
f'"$({shlex.quote(bin_resolver)} '
f'{shlex.quote(stack_name)} {shlex.quote(service_key)})"'
)
]