- Declare `__all__` in every package `__init__.py` whose `PROJECT_ROOT`
is imported by its modules (163 packages), so the constant counts as
exported; list every public name where the package defines more.
- Empty the 15 package `__init__.py` files whose `PROJECT_ROOT` no module
imports.
- Declare the public API of `utils.github.variant.axes` in `__all__`,
re-exports included, instead of silencing them with `noqa: F401`; the
list matches the one `feature/store-i18n` already carries.
- Use `import unittest.mock as mock` in `test_applications.py`, as the
rest of the suite does.
- Spell `_NON_PRINTABLE` in `utils/yaml_bootstrap.py` as the equivalent
BMP-only positive class; it matches exactly the same 2111 code points
as the old negated class across U+0000..U+10FFFF.
Co-Authored-By: Claude Opus 5.5 (1M context) <noreply@anthropic.com>
Ansible defines it on every play, like playbook_dir already on the list. The
wireguard CLI test reads it to run the mesh CLI with the controller's
interpreter.
svc-net-tor rendered its own guard script, its own unit and its own handler, and hooked the rules into DOCKER-USER with an insert, because a rule appended to a chain docker owns lands behind the RETURN docker puts there. The script flushed its chain and refilled it rule by rule, so the boundary was gone for the duration of the refill.
A role that needs packet filter rules now writes templates/nftables.conf.j2 declaring its own table and includes svc-net-firewall, which renders it to /etc/nftables.d/<role>.conf, rejects it with nft -c, loads it, reads the table back and refuses one that carries no chain. Deployed on its own the role installs nftables and the ruleset that loads every fragment at boot. That ruleset is a distribution decision, so its path follows os_family: RedHat reads /etc/sysconfig/nftables.conf, the others /etc/nftables.conf. The unit is only enabled, never started, because on Debian its ExecStop flushes the whole ruleset, docker's tables included.
One table per role removes the shared path: a fragment replaces its own table and touches no other, and docker never touches either. nft -f on an existing table merges, so each fragment creates, deletes and redeclares its table in one transaction. The tests load every fragment twice into a throwaway network namespace and require the listing to match, because a merging fragment loads twice without error and shows up only as a duplicated rule.
The tor guard now applies in both deploy modes. It was pinned to swarm while torrc binds TransPort and DNSPort on the bridge address in compose too, so on a single-host install those ports answered every on-link neighbour. The guard drops rather than admits, so nothing short-circuits another table, and ip saddr carries an implicit ipv4 dependency, which leaves the ipv6 exposure exactly where the iptables rules left it. svc-net-tor reads its own chain back and refuses a guard that drops nothing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
svc-ai-lmstudio had no model-load step at all: tasks/00_core.yml set the
once-flag and the container network, nothing else. The gateway published
lmstudio/default anyway, so the route existed and nothing could answer it.
wp-compose-tor.log:48994 shows the route written into the config, :51770 shows
the warm step reaching qwen2.5:0.5b on Ollama instead, and the green ai-engine
spec ran against that. LM Studio was deployed, published and never prompted.
The pre-pull mirrors the one that works. tasks/01_manager_ops.yml resolves the
hosting node and container id, loops the declared models through
`lms get --yes --gguf`, overlaps them with async/poll and reaps before
playwright, exactly like svc-ai-ollama. That `lms` exists is measured, not
assumed: /app/daemon-run.sh:67 in the image starts ./.bundle/lms server start,
and `lms get --help` documents -y for scripting and --gguf for the format.
Synonymity needs both namespaces in one place. Ollama knows qwen2.5:0.5b, LM
Studio knows qwen2.5-0.5b-instruct, so AI_LOCAL_MODELS carries both per entry:
`alias` is what the gateway publishes and what Ollama pulls, `name` is what
`lms get` resolves. Both roles now reference that one list, and config.yaml.j2
publishes each backend under the alias, routing Ollama to ollama/<alias> and LM
Studio to openai/<name>. The lmstudio loop skips an alias Ollama already
publishes, so two backends never produce a duplicate model_name and Ollama
keeps precedence. lmstudio/default is gone; the model name is now the same
whichever backend answers it.
Ollama's preload_models changes shape from a list of strings to the {alias,
name} entries. Every consumer moves with it: the pull command, the template
loop, and both expressions in 16_ai.yml. An inventory that overrides
`applications.svc-ai-ollama.services.ollama.preload_models` with the old string
list will break at map(attribute='alias'); this is a deliberate shape change,
in the spirit of 7b15c61dfc.
The CI model list is not a property of an installation, so AI_LOCAL_MODELS_CI
lives in inventories/development/default.yml, whose own header scopes that file
to the cross-cutting environment knobs. test_variable_definitions only knew
group_vars, so it reported the constant as undefined; the shipped inventories
are a definition source too, because the deploy CLI bakes them into host_vars,
and the guard's own docstring already claimed "defined somewhere in the
repository".
A latent crash came out of this: `| list | first` raises on an empty sequence,
so a backend deployed without preload models blew up LITELLM_CHAT_MODEL. Ollama
with only embedding models could already hit it; the existing test dodged it by
rendering only _SERVED. `| default('')` now yields the empty string that
_SERVED already reports as false.
Tests: config.yaml.j2 is rendered rather than read, over ollama-only,
lmstudio-only, both and neither, asserting the published model_name set, that
both backends name the same alias, that a shared alias is published once and
routed to Ollama, and that the render is valid YAML. test_litellm_chat_model
gains the lmstudio branch, the no-preload case and the same-model assertion.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Output of 'make autoformat': ruff collapses expressions that were wrapped across lines without need, shfmt and the yaml formatters drop surplus blank lines. No semantic change - the whole diff is 10 insertions against 39 deletions, and every hunk joins lines that already said the same thing.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
3529 invalid comments in 528 files, grandfathered in while the linter
only saw the working-tree diff. Removal ran through the linter's own
collectors, so detection and deletion cannot drift: a comment whose
opener line is blank before it and whose closer line is blank after it
loses all its lines, anything else loses only the comment region, and a
line left holding a bare delimiter goes with it.
Five files keep their comments behind a nocheck marker instead. Four are
upstream configs whose inline docs are upstream's own (mautrix, systemd
journald, nextcloud-oidc-login, MediaWiki). In littlejs the
commented-out entries ARE the deactivation mechanism for an example
tile, so deleting them would delete configuration.
Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Replace the per-role swarm host-node/container-id resolve duplicated across fider, matomo, decidim, keycloak (login + realm-import), collabora and listmonk with two shared task utils under roles/sys-svc-compose/tasks/utils/swarm/. resolve_host_cid sets <prefix>_HOST_NODE / <prefix>_LOCAL_CID (swarm: converge + node + cid delegated to the node; compose: inventory host + given cid); wait_converged is the bare convergence gate.
Teach test_variable_definitions to read templated set_fact keys like {{ prefix }}_HOST_NODE as a *_HOST_NODE name pattern (guarded so a pure {{ x }} key cannot become a catch-all), and add the playbook_dir-relative include load-failure exemptions to .ansible-lint-ignore.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Roles that declare default_placement: manager in meta/services.yml are now
auto-added to their own inventory group when the operator did not list them
explicitly. Both the runtime constructor add_host loop and the CLI
pre-flight inventory extender use the new roles_with_default_placement
lookup so swarm and CLI-validation stay consistent.
In compose-mode (no svc-docker-swarm-manager group), svc-prx-openresty
auto-joins the first play host. run_once + ansible_play_hosts | first
keeps that deterministic in multi-host compose deploys so the openresty
vhost view does not fragment.
The svc-cache-registry / svc-docker-registry entries previously hardcoded
in scripts/tests/deploy/swarm/08_extend_inventory.sh fall out of the new
dynamic loop; net inventory is unchanged.
ansible_play_hosts joins the magic-var allowlist in
test_variable_definitions so the constructor add_host passes integration
lint.
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
The MODE_DEBUG dump in tasks/stages/01_constructor.yml crashed
`to_nice_yaml` on the internal `<<Omit>>` sentinel that Ansible plants
into `vars`. Dump `hostvars[inventory_hostname]` instead and drop the
now-unused `vars` allowance from the jinja variable-definition lint.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Four follow-ups uncovered by \`make test\` after 3c9db6142:
* tests/lint/ansible/variables/test_role_and_group_vars_used:
TLS_ENABLED is declared in group_vars/all/02_tls.yml but never
read via \`{{ TLS_ENABLED }}\`; it is consumed by
plugins/lookup/{tls,email,cert}.py, which the YAML/Jinja-scoped
lint cannot see. Tag the declaration with the project's standard
\`# nocheck: unused-var\` marker and document the consumer.
* tests/lint/filesystem/python/test_no_raw_project_walk: the new
test_no_ansible_env_lookups.py used \`Path.rglob("*")\` for its
walk, which is the exact shape the no-raw-walks lint forbids in
test code. Switch to \`iter_project_files(extensions=_SCAN_EXTS)\`
with a relative-path prefix filter; same coverage, shared cache.
* tests/unit/cli/administration/deploy/development/test_inventory_vars_file_spot:
the new lint test legitimately names the dev inventory path
(\`inventories/development/default.yml\`) in its failure message.
Add it to the SPOT-drift-guard allowlist with a justification.
* tests/integration/ansible/jinja/test_variable_definitions: the
debug task added in 0884eb4e2 prints \`ansible_env\` and \`vars\`,
both Ansible magic variables. Add them to the test's reserved-name
list so the static analyzer stops flagging them as undefined.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
Widens the cache-read lint guard from ``tests/lint/`` to every ``.py``
under ``tests/``. Integration tests show the same compounding cost
pattern (one role's compose template / nginx conf / scrape fragment
read across many shape tests) that motivated the original lint for
lint/.
Converts 91 violations across 28 test files:
- **Cache (utils.cache.files.read_text)** for project-tracked reads
and single-read tempdir fixtures, e.g. test_prometheus.py (32x),
test_keycloak_import_flow.py (11x), test_dashboard_oidc.py (3x),
test_changelog_md.py (8x), test_package_mirror.py (8x), plus
test_logging.py / test_summarize.py / test_version_updater.py /
test_get_url_retry.py / test_pull_specific_host.py.
- **Cache + UnicodeDecodeError handler** for project-walk scanners
that previously used ``errors="replace"`` / ``errors="ignore"`` to
defend against non-UTF-8 files (compose volumes, password quoting,
template image, mailu dependency, URL reachability). Wraps each
read in ``try/except (OSError, UnicodeDecodeError): continue``,
which has the same observable effect (skip undecodable files).
- **Refactor away the binary heuristic** in test_no_raw_docker.py:
``read_bytes() + b"\x00" not in raw`` becomes
``read_text() / except UnicodeDecodeError``. Same files skipped,
with cache benefit.
- **Variable extraction** in test_changelog_md.py / test_package_mirror.py
for cases where ``self.assertEqual(path.read_text(...), expected)``
exceeded ruff's 88-char limit. Ruff was wrapping the marker onto a
line outside the AST span of the read_text call.
- **Compute-from-setup** for two test methods that previously read
``self.changelog`` twice (before/after) to verify a no-op:
``before`` is now derived from the known input, dropping the first
read.
Five ``# nocheck: cache-read`` markers remain, each annotated inline
with the reason: idempotency tests where the second read must hit
disk to catch non-idempotent regressions, and tests with intermediate
``write_text`` whose subsequent read would return stale cache data.
Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
kevinveenbirkenbachandClaude Opus 5.5
evangelostsak