mirror of
https://github.com/kevinveenbirkenbach/computer-playbook.git
synced 2025-12-09 19:05:32 +00:00
Refine deploy CLI, test-deploy workflow and Ansible output
Changes: - Update GitHub Actions test-deploy workflow to run three staged deploys (normal+debug, reset+debug, async) using inventory-generated vault password files. - Switch Ansible stdout_callback to ansible.builtin.default and enable YAML-style result_format via callback_default. - Refactor cli/deploy.py: typed run_ansible_playbook(), structured MODE_* handling, better error reporting, and preserved vault/interactive behaviour. - Add unit tests for deploy CLI (bool parsing, MODE_* loading, dynamic args, validation, and ansible-playbook command construction) under tests/unit/cli/test_deploy.py. Context: see ChatGPT conversation on 2025-12-02: https://chatgpt.com/share/692f1035-6bc4-800f-91a9-342db54e1a75
This commit is contained in:
332
tests/unit/cli/test_deploy.py
Normal file
332
tests/unit/cli/test_deploy.py
Normal file
@@ -0,0 +1,332 @@
|
||||
#!/usr/bin/env python3
|
||||
# -*- coding: utf-8 -*-
|
||||
|
||||
import os
|
||||
import tempfile
|
||||
import unittest
|
||||
from typing import Any, Dict, List
|
||||
from unittest import mock
|
||||
|
||||
import cli.deploy as deploy
|
||||
import subprocess
|
||||
|
||||
|
||||
class TestParseBoolLiteral(unittest.TestCase):
|
||||
def test_true_values(self):
|
||||
self.assertTrue(deploy._parse_bool_literal("true"))
|
||||
self.assertTrue(deploy._parse_bool_literal("True"))
|
||||
self.assertTrue(deploy._parse_bool_literal(" yes "))
|
||||
self.assertTrue(deploy._parse_bool_literal("ON"))
|
||||
|
||||
def test_false_values(self):
|
||||
self.assertFalse(deploy._parse_bool_literal("false"))
|
||||
self.assertFalse(deploy._parse_bool_literal("False"))
|
||||
self.assertFalse(deploy._parse_bool_literal(" no "))
|
||||
self.assertFalse(deploy._parse_bool_literal("off"))
|
||||
|
||||
def test_unknown_value(self):
|
||||
self.assertIsNone(deploy._parse_bool_literal("maybe"))
|
||||
self.assertIsNone(deploy._parse_bool_literal(""))
|
||||
self.assertIsNone(deploy._parse_bool_literal(" "))
|
||||
|
||||
|
||||
class TestLoadModesFromYaml(unittest.TestCase):
|
||||
def test_load_modes_basic(self):
|
||||
# Create a temporary "01_modes.yml"-like file
|
||||
content = """\
|
||||
MODE_CLEANUP: true # cleanup before deploy
|
||||
MODE_DEBUG: false # enable debug
|
||||
MODE_ASSERT: null # explicitly set via CLI
|
||||
INVALID_KEY: true # ignored because no MODE_ prefix
|
||||
"""
|
||||
with tempfile.NamedTemporaryFile("w+", delete=False, encoding="utf-8") as f:
|
||||
path = f.name
|
||||
f.write(content)
|
||||
f.flush()
|
||||
|
||||
try:
|
||||
modes = deploy.load_modes_from_yaml(path)
|
||||
finally:
|
||||
os.unlink(path)
|
||||
|
||||
# We expect 3 MODE_* entries, INVALID_KEY is ignored
|
||||
self.assertEqual(len(modes), 3)
|
||||
|
||||
by_name = {m["name"]: m for m in modes}
|
||||
|
||||
self.assertIn("MODE_CLEANUP", by_name)
|
||||
self.assertIn("MODE_DEBUG", by_name)
|
||||
self.assertIn("MODE_ASSERT", by_name)
|
||||
|
||||
self.assertEqual(by_name["MODE_CLEANUP"]["default"], True)
|
||||
self.assertEqual(by_name["MODE_DEBUG"]["default"], False)
|
||||
self.assertIsNone(by_name["MODE_ASSERT"]["default"])
|
||||
self.assertEqual(by_name["MODE_CLEANUP"]["help"], "cleanup before deploy")
|
||||
|
||||
|
||||
class TestDynamicModes(unittest.TestCase):
|
||||
def setUp(self):
|
||||
# Simple meta as if parsed from 01_modes.yml
|
||||
self.modes_meta = [
|
||||
{"name": "MODE_CLEANUP", "default": True, "help": "Cleanup before run"},
|
||||
{"name": "MODE_DEBUG", "default": False, "help": "Debug output"},
|
||||
{"name": "MODE_ASSERT", "default": None, "help": "Inventory assertion"},
|
||||
]
|
||||
|
||||
def test_add_dynamic_mode_args_and_build_modes_defaults(self):
|
||||
parser = unittest.mock.MagicMock()
|
||||
# Use a real ArgumentParser for build_modes_from_args
|
||||
from argparse import ArgumentParser
|
||||
|
||||
real_parser = ArgumentParser()
|
||||
spec = deploy.add_dynamic_mode_args(real_parser, self.modes_meta)
|
||||
|
||||
# We expect three entries
|
||||
self.assertIn("MODE_CLEANUP", spec)
|
||||
self.assertIn("MODE_DEBUG", spec)
|
||||
self.assertIn("MODE_ASSERT", spec)
|
||||
|
||||
# No flags given: use defaults (True/False/None)
|
||||
args = real_parser.parse_args([])
|
||||
modes = deploy.build_modes_from_args(spec, args)
|
||||
|
||||
self.assertTrue(modes["MODE_CLEANUP"]) # default True
|
||||
self.assertFalse(modes["MODE_DEBUG"]) # default False
|
||||
self.assertNotIn("MODE_ASSERT", modes) # default None → not present
|
||||
|
||||
def test_add_dynamic_mode_args_and_build_modes_flags(self):
|
||||
from argparse import ArgumentParser
|
||||
|
||||
parser = ArgumentParser()
|
||||
spec = deploy.add_dynamic_mode_args(parser, self.modes_meta)
|
||||
|
||||
# CLI: --skip-cleanup → MODE_CLEANUP=False
|
||||
# --debug → MODE_DEBUG=True
|
||||
# --assert true → MODE_ASSERT=True
|
||||
args = parser.parse_args(
|
||||
["--skip-cleanup", "--debug", "--assert", "true"]
|
||||
)
|
||||
modes = deploy.build_modes_from_args(spec, args)
|
||||
|
||||
self.assertFalse(modes["MODE_CLEANUP"])
|
||||
self.assertTrue(modes["MODE_DEBUG"])
|
||||
self.assertTrue(modes["MODE_ASSERT"])
|
||||
|
||||
|
||||
class TestValidateApplicationIds(unittest.TestCase):
|
||||
def test_no_ids_does_nothing(self):
|
||||
"""
|
||||
When no application_ids are passed, the function should just return
|
||||
without trying to validate anything.
|
||||
"""
|
||||
deploy.validate_application_ids("inventories/github-ci/servers.yml", [])
|
||||
|
||||
@mock.patch("module_utils.valid_deploy_id.ValidDeployId")
|
||||
def test_invalid_ids_raise_system_exit(self, mock_vdi_cls):
|
||||
"""
|
||||
When ValidDeployId reports invalid IDs, validate_application_ids should
|
||||
print an error and exit with code 1.
|
||||
"""
|
||||
instance = mock_vdi_cls.return_value
|
||||
instance.validate.return_value = {
|
||||
"web-app-foo": {"allowed": False, "in_inventory": True},
|
||||
"web-app-bar": {"allowed": True, "in_inventory": False},
|
||||
}
|
||||
|
||||
with self.assertRaises(SystemExit) as ctx:
|
||||
deploy.validate_application_ids(
|
||||
"inventories/github-ci/servers.yml",
|
||||
["web-app-foo", "web-app-bar"],
|
||||
)
|
||||
|
||||
self.assertEqual(ctx.exception.code, 1)
|
||||
instance.validate.assert_called_once_with(
|
||||
"inventories/github-ci/servers.yml",
|
||||
["web-app-foo", "web-app-bar"],
|
||||
)
|
||||
|
||||
|
||||
class TestRunAnsiblePlaybook(unittest.TestCase):
|
||||
def _fake_run_side_effect(
|
||||
self,
|
||||
calls_store: List[List[str]],
|
||||
ansible_rc: int = 0,
|
||||
):
|
||||
"""
|
||||
side_effect for subprocess.run that:
|
||||
|
||||
- Records every command into calls_store
|
||||
- Returns 'ansible_rc' for ansible-playbook
|
||||
- Returns rc=0 for all other commands (make, validation, etc.)
|
||||
"""
|
||||
def _side_effect(cmd, *args, **kwargs):
|
||||
# Normalize into a list for easier assertions
|
||||
if isinstance(cmd, list):
|
||||
calls_store.append(cmd)
|
||||
else:
|
||||
calls_store.append([cmd])
|
||||
|
||||
# Special handling for ansible-playbook
|
||||
if isinstance(cmd, list) and cmd and cmd[0] == "ansible-playbook":
|
||||
return subprocess.CompletedProcess(
|
||||
cmd,
|
||||
ansible_rc,
|
||||
stdout="ANSIBLE_STDOUT\n",
|
||||
stderr="ANSIBLE_STDERR\n",
|
||||
)
|
||||
|
||||
# Everything else (make, python inventory) is treated as success
|
||||
return subprocess.CompletedProcess(cmd, 0, stdout="", stderr="")
|
||||
|
||||
return _side_effect
|
||||
|
||||
@mock.patch("subprocess.run")
|
||||
def test_run_ansible_playbook_builds_correct_command(self, mock_run):
|
||||
"""
|
||||
Happy-path test:
|
||||
|
||||
- MODE_CLEANUP=True → 'make clean' is executed
|
||||
- build/test phases are executed
|
||||
- ansible-playbook is called with expected arguments
|
||||
- commands are fully mocked (no real execution)
|
||||
"""
|
||||
calls: List[List[str]] = []
|
||||
mock_run.side_effect = self._fake_run_side_effect(calls, ansible_rc=0)
|
||||
|
||||
modes: Dict[str, Any] = {
|
||||
"MODE_CLEANUP": True,
|
||||
"MODE_ASSERT": True,
|
||||
"MODE_DEBUG": True, # should enforce at least -vvv
|
||||
}
|
||||
|
||||
inventory_path = "inventories/github-ci/servers.yml"
|
||||
limit = "localhost"
|
||||
allowed_apps = ["web-app-foo", "web-app-bar"]
|
||||
password_file = "inventories/github-ci/.password"
|
||||
|
||||
deploy.run_ansible_playbook(
|
||||
inventory=inventory_path,
|
||||
modes=modes,
|
||||
limit=limit,
|
||||
allowed_applications=allowed_apps,
|
||||
password_file=password_file,
|
||||
verbose=1, # explicitly set, then raised by MODE_DEBUG
|
||||
skip_build=False,
|
||||
skip_tests=False,
|
||||
logs=False,
|
||||
diff=True,
|
||||
)
|
||||
|
||||
# We expect at least: make clean, make messy-build, make messy-test,
|
||||
# plus the ansible-playbook invocation. Inventory validation is also
|
||||
# done via python ... inventory.py but we do not assert the full path.
|
||||
self.assertTrue(
|
||||
any(call == ["make", "clean"] for call in calls),
|
||||
"Expected 'make clean' when MODE_CLEANUP is true",
|
||||
)
|
||||
self.assertTrue(
|
||||
any(call == ["make", "messy-build"] for call in calls),
|
||||
"Expected 'make messy-build' when skip_build=False",
|
||||
)
|
||||
self.assertTrue(
|
||||
any(call == ["make", "messy-test"] for call in calls),
|
||||
"Expected 'make messy-test' when skip_tests=False",
|
||||
)
|
||||
self.assertTrue(
|
||||
any(
|
||||
isinstance(call, list)
|
||||
and any("inventory.py" in part for part in call)
|
||||
for call in calls
|
||||
),
|
||||
"Expected inventory validation call (... inventory.py ...)",
|
||||
)
|
||||
|
||||
# Last command should be ansible-playbook
|
||||
self.assertGreaterEqual(len(calls), 1)
|
||||
last_cmd = calls[-1]
|
||||
self.assertEqual(last_cmd[0], "ansible-playbook")
|
||||
|
||||
# Check inventory and playbook args
|
||||
self.assertIn("-i", last_cmd)
|
||||
self.assertIn(inventory_path, last_cmd)
|
||||
|
||||
idx_inv = last_cmd.index(inventory_path)
|
||||
self.assertGreater(len(last_cmd), idx_inv + 1)
|
||||
playbook_arg = last_cmd[idx_inv + 1]
|
||||
self.assertTrue(
|
||||
playbook_arg.endswith(os.path.join("", "playbook.yml")),
|
||||
f"playbook argument should end with 'playbook.yml', got: {playbook_arg}",
|
||||
)
|
||||
|
||||
# Check --limit / -l
|
||||
self.assertIn("-l", last_cmd)
|
||||
self.assertIn(limit, last_cmd)
|
||||
|
||||
# allowed_applications extra var
|
||||
last_cmd_str = " ".join(last_cmd)
|
||||
self.assertIn(
|
||||
"allowed_applications=web-app-foo,web-app-bar",
|
||||
last_cmd_str,
|
||||
)
|
||||
|
||||
# Modes passed as -e
|
||||
self.assertIn("MODE_CLEANUP=true", last_cmd_str)
|
||||
self.assertIn("MODE_ASSERT=true", last_cmd_str)
|
||||
self.assertIn("MODE_DEBUG=true", last_cmd_str)
|
||||
|
||||
# Vault password file
|
||||
self.assertIn("--vault-password-file", last_cmd)
|
||||
self.assertIn(password_file, last_cmd)
|
||||
|
||||
# --diff should be present
|
||||
self.assertIn("--diff", last_cmd)
|
||||
|
||||
# Verbosity: MODE_DEBUG enforces at least -vvv regardless of initial -v
|
||||
self.assertTrue(
|
||||
any(arg.startswith("-vvv") for arg in last_cmd),
|
||||
"Expected at least -vvv because MODE_DEBUG=True",
|
||||
)
|
||||
|
||||
# Verify that for ansible-playbook, we asked for text+capture_output
|
||||
last_call = mock_run.call_args_list[-1]
|
||||
self.assertTrue(last_call.kwargs.get("text", False))
|
||||
self.assertTrue(last_call.kwargs.get("capture_output", False))
|
||||
|
||||
@mock.patch("subprocess.run")
|
||||
def test_run_ansible_playbook_failure_exits_with_code(self, mock_run):
|
||||
"""
|
||||
If ansible-playbook returns non-zero, run_ansible_playbook should exit
|
||||
with the same code. All external commands are mocked.
|
||||
"""
|
||||
calls: List[List[str]] = []
|
||||
mock_run.side_effect = self._fake_run_side_effect(calls, ansible_rc=4)
|
||||
|
||||
modes: Dict[str, Any] = {
|
||||
"MODE_CLEANUP": False,
|
||||
"MODE_ASSERT": False,
|
||||
"MODE_DEBUG": False,
|
||||
}
|
||||
|
||||
with self.assertRaises(SystemExit) as ctx:
|
||||
deploy.run_ansible_playbook(
|
||||
inventory="inventories/github-ci/servers.yml",
|
||||
modes=modes,
|
||||
limit=None,
|
||||
allowed_applications=None,
|
||||
password_file=None,
|
||||
verbose=0,
|
||||
skip_build=True,
|
||||
skip_tests=True,
|
||||
logs=False,
|
||||
diff=False,
|
||||
)
|
||||
|
||||
self.assertEqual(ctx.exception.code, 4)
|
||||
self.assertTrue(
|
||||
any(call and call[0] == "ansible-playbook" for call in calls),
|
||||
"ansible-playbook should have been invoked once in the error path",
|
||||
)
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
unittest.main()
|
||||
Reference in New Issue
Block a user