"""Shared test utilities."""
+# pylint: disable=too-many-lines
+from __future__ import annotations
from unittest import TestCase
+from typing import Mapping, Any, Callable
from threading import Thread
from http.client import HTTPConnection
-from datetime import datetime
+from datetime import datetime, timedelta
+from time import sleep
+from json import loads as json_loads, dumps as json_dumps
+from urllib.parse import urlencode
+from uuid import uuid4
from os import remove as remove_file
+from pprint import pprint
from plomtask.db import DatabaseFile, DatabaseConnection
from plomtask.http import TaskHandler, TaskServer
+from plomtask.processes import Process, ProcessStep
+from plomtask.conditions import Condition
+from plomtask.days import Day
+from plomtask.dating import DATE_FORMAT
+from plomtask.todos import Todo
+from plomtask.versioned_attributes import VersionedAttribute, TIMESTAMP_FMT
+from plomtask.exceptions import NotFoundException, HandledException
-class TestCaseWithDB(TestCase):
+VERSIONED_VALS: dict[str,
+ list[str] | list[float]] = {'str': ['A', 'B'],
+ 'float': [0.3, 1.1]}
+VALID_TRUES = {True, 'True', 'true', '1', 'on'}
+
+
+class TestCaseAugmented(TestCase):
+ """Tester core providing helpful basic internal decorators and methods."""
+ checked_class: Any
+ default_init_kwargs: dict[str, Any] = {}
+
+ @staticmethod
+ def _run_on_versioned_attributes(f: Callable[..., None]
+ ) -> Callable[..., None]:
+ def wrapper(self: TestCase) -> None:
+ assert isinstance(self, TestCaseAugmented)
+ for attr_name in self.checked_class.to_save_versioned():
+ default = self.checked_class.versioned_defaults[attr_name]
+ owner = self.checked_class(None, **self.default_init_kwargs)
+ attr = getattr(owner, attr_name)
+ to_set = VERSIONED_VALS[attr.value_type_name]
+ f(self, owner, attr_name, attr, default, to_set)
+ return wrapper
+
+ @classmethod
+ def _run_if_sans_db(cls, f: Callable[..., None]) -> Callable[..., None]:
+ def wrapper(self: TestCaseSansDB) -> None:
+ if issubclass(cls, TestCaseSansDB):
+ f(self)
+ return wrapper
+
+ @classmethod
+ def _run_if_with_db_but_not_server(cls,
+ f: Callable[..., None]
+ ) -> Callable[..., None]:
+ def wrapper(self: TestCaseWithDB) -> None:
+ if issubclass(cls, TestCaseWithDB) and\
+ not issubclass(cls, TestCaseWithServer):
+ f(self)
+ return wrapper
+
+ @classmethod
+ def _make_from_defaults(cls, id_: float | str | None) -> Any:
+ return cls.checked_class(id_, **cls.default_init_kwargs)
+
+
+class TestCaseSansDB(TestCaseAugmented):
+ """Tests requiring no DB setup."""
+ legal_ids: list[str] | list[int] = [1, 5]
+ illegal_ids: list[str] | list[int] = [0]
+
+ @TestCaseAugmented._run_if_sans_db
+ def test_id_validation(self) -> None:
+ """Test .id_ validation/setting."""
+ for id_ in self.illegal_ids:
+ with self.assertRaises(HandledException):
+ self._make_from_defaults(id_)
+ for id_ in self.legal_ids:
+ obj = self._make_from_defaults(id_)
+ self.assertEqual(obj.id_, id_)
+
+ @TestCaseAugmented._run_if_sans_db
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_versioned_set(self,
+ _: Any,
+ __: str,
+ attr: VersionedAttribute,
+ default: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """Test VersionedAttribute.set() behaves as expected."""
+ attr.set(default)
+ self.assertEqual(list(attr.history.values()), [default])
+ # check same value does not get set twice in a row,
+ # and that not even its timestamp get updated
+ timestamp = list(attr.history.keys())[0]
+ attr.set(default)
+ self.assertEqual(list(attr.history.values()), [default])
+ self.assertEqual(list(attr.history.keys())[0], timestamp)
+ # check that different value _will_ be set/added
+ attr.set(to_set[0])
+ timesorted_vals = [attr.history[t] for
+ t in sorted(attr.history.keys())]
+ expected = [default, to_set[0]]
+ self.assertEqual(timesorted_vals, expected)
+ # check that a previously used value can be set if not most recent
+ attr.set(default)
+ timesorted_vals = [attr.history[t] for
+ t in sorted(attr.history.keys())]
+ expected = [default, to_set[0], default]
+ self.assertEqual(timesorted_vals, expected)
+ # again check for same value not being set twice in a row, even for
+ # later items
+ attr.set(to_set[1])
+ timesorted_vals = [attr.history[t] for
+ t in sorted(attr.history.keys())]
+ expected = [default, to_set[0], default, to_set[1]]
+ self.assertEqual(timesorted_vals, expected)
+ attr.set(to_set[1])
+ self.assertEqual(timesorted_vals, expected)
+
+ @TestCaseAugmented._run_if_sans_db
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_versioned_newest(self,
+ _: Any,
+ __: str,
+ attr: VersionedAttribute,
+ default: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """Test VersionedAttribute.newest."""
+ # check .newest on empty history returns .default
+ self.assertEqual(attr.newest, default)
+ # check newest element always returned
+ for v in [to_set[0], to_set[1]]:
+ attr.set(v)
+ self.assertEqual(attr.newest, v)
+ # check newest element returned even if also early value
+ attr.set(default)
+ self.assertEqual(attr.newest, default)
+
+ @TestCaseAugmented._run_if_sans_db
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_versioned_at(self,
+ _: Any,
+ __: str,
+ attr: VersionedAttribute,
+ default: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """Test .at() returns values nearest to queried time, or default."""
+ # check .at() return default on empty history
+ timestamp_a = datetime.now().strftime(TIMESTAMP_FMT)
+ self.assertEqual(attr.at(timestamp_a), default)
+ # check value exactly at timestamp returned
+ attr.set(to_set[0])
+ timestamp_b = list(attr.history.keys())[0]
+ self.assertEqual(attr.at(timestamp_b), to_set[0])
+ # check earliest value returned if exists, rather than default
+ self.assertEqual(attr.at(timestamp_a), to_set[0])
+ # check reverts to previous value for timestamps not indexed
+ sleep(0.00001)
+ timestamp_between = datetime.now().strftime(TIMESTAMP_FMT)
+ sleep(0.00001)
+ attr.set(to_set[1])
+ timestamp_c = sorted(attr.history.keys())[-1]
+ self.assertEqual(attr.at(timestamp_c), to_set[1])
+ self.assertEqual(attr.at(timestamp_between), to_set[0])
+ sleep(0.00001)
+ timestamp_after_c = datetime.now().strftime(TIMESTAMP_FMT)
+ self.assertEqual(attr.at(timestamp_after_c), to_set[1])
+
+
+class TestCaseWithDB(TestCaseAugmented):
"""Module tests not requiring DB setup."""
+ default_ids: tuple[int, int, int] | tuple[str, str, str] = (1, 2, 3)
def setUp(self) -> None:
- timestamp = datetime.now().timestamp()
- self.db_file = DatabaseFile(f'test_db:{timestamp}')
- self.db_file.remake()
+ Condition.empty_cache()
+ Day.empty_cache()
+ Process.empty_cache()
+ ProcessStep.empty_cache()
+ Todo.empty_cache()
+ self.db_file = DatabaseFile.create_at(f'test_db:{uuid4()}')
self.db_conn = DatabaseConnection(self.db_file)
def tearDown(self) -> None:
self.db_conn.close()
remove_file(self.db_file.path)
+ def _load_from_db(self, id_: int | str) -> list[object]:
+ db_found: list[object] = []
+ for row in self.db_conn.row_where(self.checked_class.table_name,
+ 'id', id_):
+ db_found += [self.checked_class.from_table_row(self.db_conn,
+ row)]
+ return db_found
+
+ def _change_obj(self, obj: object) -> str:
+ attr_name: str = self.checked_class.to_save_simples[-1]
+ attr = getattr(obj, attr_name)
+ new_attr: str | int | float | bool
+ if isinstance(attr, (int, float)):
+ new_attr = attr + 1
+ elif isinstance(attr, str):
+ new_attr = attr + '_'
+ elif isinstance(attr, bool):
+ new_attr = not attr
+ setattr(obj, attr_name, new_attr)
+ return attr_name
+
+ def check_identity_with_cache_and_db(self, content: list[Any]) -> None:
+ """Test both cache and DB equal content."""
+ expected_cache = {}
+ for item in content:
+ expected_cache[item.id_] = item
+ self.assertEqual(self.checked_class.get_cache(), expected_cache)
+ hashes_content = [hash(x) for x in content]
+ db_found: list[Any] = []
+ for item in content:
+ assert isinstance(item.id_, type(self.default_ids[0]))
+ db_found += self._load_from_db(item.id_)
+ hashes_db_found = [hash(x) for x in db_found]
+ self.assertEqual(sorted(hashes_content), sorted(hashes_db_found))
+
+ def check_by_date_range_with_limits(self,
+ date_col: str,
+ set_id_field: bool = True
+ ) -> None:
+ """Test .by_date_range_with_limits."""
+ # pylint: disable=too-many-locals
+ f = self.checked_class.by_date_range_with_limits
+ # check illegal ranges
+ legal_range = ('yesterday', 'tomorrow')
+ for i in [0, 1]:
+ for bad_date in ['foo', '2024-02-30', '2024-01-01 12:00:00']:
+ date_range = list(legal_range[:])
+ date_range[i] = bad_date
+ with self.assertRaises(HandledException):
+ f(self.db_conn, date_range, date_col)
+ # check empty, translation of 'yesterday' and 'tomorrow'
+ items, start, end = f(self.db_conn, legal_range, date_col)
+ self.assertEqual(items, [])
+ yesterday = datetime.now() + timedelta(days=-1)
+ tomorrow = datetime.now() + timedelta(days=+1)
+ self.assertEqual(start, yesterday.strftime(DATE_FORMAT))
+ self.assertEqual(end, tomorrow.strftime(DATE_FORMAT))
+ # prepare dated items for non-empty results
+ kwargs_with_date = self.default_init_kwargs.copy()
+ if set_id_field:
+ kwargs_with_date['id_'] = None
+ objs = []
+ dates = ['2024-01-01', '2024-01-02', '2024-01-04']
+ for date in ['2024-01-01', '2024-01-02', '2024-01-04']:
+ kwargs_with_date['date'] = date
+ obj = self.checked_class(**kwargs_with_date)
+ objs += [obj]
+ # check ranges still empty before saving
+ date_range = [dates[0], dates[-1]]
+ self.assertEqual(f(self.db_conn, date_range, date_col)[0], [])
+ # check all objs displayed within closed interval
+ for obj in objs:
+ obj.save(self.db_conn)
+ self.assertEqual(f(self.db_conn, date_range, date_col)[0], objs)
+ # check that only displayed what exists within interval
+ date_range = ['2023-12-20', '2024-01-03']
+ expected = [objs[0], objs[1]]
+ self.assertEqual(f(self.db_conn, date_range, date_col)[0], expected)
+ date_range = ['2024-01-03', '2024-01-30']
+ expected = [objs[2]]
+ self.assertEqual(f(self.db_conn, date_range, date_col)[0], expected)
+ # check that inverted interval displays nothing
+ date_range = [dates[-1], dates[0]]
+ self.assertEqual(f(self.db_conn, date_range, date_col)[0], [])
+ # check that "today" is interpreted, and single-element interval
+ today_date = datetime.now().strftime(DATE_FORMAT)
+ kwargs_with_date['date'] = today_date
+ obj_today = self.checked_class(**kwargs_with_date)
+ obj_today.save(self.db_conn)
+ date_range = ['today', 'today']
+ items, start, end = f(self.db_conn, date_range, date_col)
+ self.assertEqual(start, today_date)
+ self.assertEqual(start, end)
+ self.assertEqual(items, [obj_today])
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_saving_versioned_attributes(self,
+ owner: Any,
+ attr_name: str,
+ attr: VersionedAttribute,
+ _: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """Test storage and initialization of versioned attributes."""
+
+ def retrieve_attr_vals(attr: VersionedAttribute) -> list[object]:
+ attr_vals_saved: list[object] = []
+ for row in self.db_conn.row_where(attr.table_name, 'parent',
+ owner.id_):
+ attr_vals_saved += [row[2]]
+ return attr_vals_saved
+
+ attr.set(to_set[0])
+ # check that without attr.save() no rows in DB
+ rows = self.db_conn.row_where(attr.table_name, 'parent', owner.id_)
+ self.assertEqual([], rows)
+ # fail saving attributes on non-saved owner
+ with self.assertRaises(NotFoundException):
+ attr.save(self.db_conn)
+ # check owner.save() created entries as expected in attr table
+ owner.save(self.db_conn)
+ attr_vals_saved = retrieve_attr_vals(attr)
+ self.assertEqual([to_set[0]], attr_vals_saved)
+ # check changing attr val without save affects owner in memory …
+ attr.set(to_set[1])
+ cmp_attr = getattr(owner, attr_name)
+ self.assertEqual(to_set, list(cmp_attr.history.values()))
+ self.assertEqual(cmp_attr.history, attr.history)
+ # … but does not yet affect DB
+ attr_vals_saved = retrieve_attr_vals(attr)
+ self.assertEqual([to_set[0]], attr_vals_saved)
+ # check individual attr.save also stores new val to DB
+ attr.save(self.db_conn)
+ attr_vals_saved = retrieve_attr_vals(attr)
+ self.assertEqual(to_set, attr_vals_saved)
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_saving_and_caching(self) -> None:
+ """Test effects of .cache() and .save()."""
+ id1 = self.default_ids[0]
+ # check failure to cache without ID (if None-ID input possible)
+ if isinstance(id1, int):
+ obj0 = self._make_from_defaults(None)
+ with self.assertRaises(HandledException):
+ obj0.cache()
+ # check mere object init itself doesn't even store in cache
+ obj1 = self._make_from_defaults(id1)
+ self.assertEqual(self.checked_class.get_cache(), {})
+ # check .cache() fills cache, but not DB
+ obj1.cache()
+ self.assertEqual(self.checked_class.get_cache(), {id1: obj1})
+ found_in_db = self._load_from_db(id1)
+ self.assertEqual(found_in_db, [])
+ # check .save() sets ID (for int IDs), updates cache, and fills DB
+ # (expect ID to be set to id1, despite obj1 already having that as ID:
+ # it's generated by cursor.lastrowid on the DB table, and with obj1
+ # not written there, obj2 should get it first!)
+ id_input = None if isinstance(id1, int) else id1
+ obj2 = self._make_from_defaults(id_input)
+ obj2.save(self.db_conn)
+ self.assertEqual(self.checked_class.get_cache(), {id1: obj2})
+ # NB: we'll only compare hashes because obj2 itself disappears on
+ # .from_table_row-triggered database reload
+ obj2_hash = hash(obj2)
+ found_in_db += self._load_from_db(id1)
+ self.assertEqual([hash(o) for o in found_in_db], [obj2_hash])
+ # check we cannot overwrite obj2 with obj1 despite its same ID,
+ # since it has disappeared now
+ with self.assertRaises(HandledException):
+ obj1.save(self.db_conn)
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_by_id(self) -> None:
+ """Test .by_id()."""
+ id1, id2, _ = self.default_ids
+ # check failure if not yet saved
+ obj1 = self._make_from_defaults(id1)
+ with self.assertRaises(NotFoundException):
+ self.checked_class.by_id(self.db_conn, id1)
+ # check identity of cached and retrieved
+ obj1.cache()
+ self.assertEqual(obj1, self.checked_class.by_id(self.db_conn, id1))
+ # check identity of saved and retrieved
+ obj2 = self._make_from_defaults(id2)
+ obj2.save(self.db_conn)
+ self.assertEqual(obj2, self.checked_class.by_id(self.db_conn, id2))
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_by_id_or_create(self) -> None:
+ """Test .by_id_or_create."""
+ # check .by_id_or_create fails if wrong class
+ if not self.checked_class.can_create_by_id:
+ with self.assertRaises(HandledException):
+ self.checked_class.by_id_or_create(self.db_conn, None)
+ return
+ # check ID input of None creates, on saving, ID=1,2,… for int IDs
+ if isinstance(self.default_ids[0], int):
+ for n in range(2):
+ item = self.checked_class.by_id_or_create(self.db_conn, None)
+ self.assertEqual(item.id_, None)
+ item.save(self.db_conn)
+ self.assertEqual(item.id_, n+1)
+ # check .by_id_or_create acts like normal instantiation (sans saving)
+ id_ = self.default_ids[2]
+ item = self.checked_class.by_id_or_create(self.db_conn, id_)
+ self.assertEqual(item.id_, id_)
+ with self.assertRaises(NotFoundException):
+ self.checked_class.by_id(self.db_conn, item.id_)
+ self.assertEqual(self.checked_class(item.id_), item)
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_from_table_row(self) -> None:
+ """Test .from_table_row() properly reads in class directly from DB."""
+ id_ = self.default_ids[0]
+ obj = self._make_from_defaults(id_)
+ obj.save(self.db_conn)
+ assert isinstance(obj.id_, type(id_))
+ for row in self.db_conn.row_where(self.checked_class.table_name,
+ 'id', obj.id_):
+ # check .from_table_row reproduces state saved, no matter if obj
+ # later changed (with caching even)
+ # NB: we'll only compare hashes because obj itself disappears on
+ # .from_table_row-triggered database reload
+ hash_original = hash(obj)
+ attr_name = self._change_obj(obj)
+ obj.cache()
+ to_cmp = getattr(obj, attr_name)
+ retrieved = self.checked_class.from_table_row(self.db_conn, row)
+ self.assertNotEqual(to_cmp, getattr(retrieved, attr_name))
+ self.assertEqual(hash_original, hash(retrieved))
+ # check cache contains what .from_table_row just produced
+ self.assertEqual({retrieved.id_: retrieved},
+ self.checked_class.get_cache())
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_versioned_history_from_row(self,
+ owner: Any,
+ _: str,
+ attr: VersionedAttribute,
+ default: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """"Test VersionedAttribute.history_from_row() knows its DB rows."""
+ attr.set(to_set[0])
+ attr.set(to_set[1])
+ owner.save(self.db_conn)
+ # make empty VersionedAttribute, fill from rows, compare to owner's
+ for row in self.db_conn.row_where(owner.table_name, 'id', owner.id_):
+ loaded_attr = VersionedAttribute(owner, attr.table_name, default)
+ for row in self.db_conn.row_where(attr.table_name, 'parent',
+ owner.id_):
+ loaded_attr.history_from_row(row)
+ self.assertEqual(len(attr.history.keys()),
+ len(loaded_attr.history.keys()))
+ for timestamp, value in attr.history.items():
+ self.assertEqual(value, loaded_attr.history[timestamp])
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_all(self) -> None:
+ """Test .all() and its relation to cache and savings."""
+ id1, id2, id3 = self.default_ids
+ item1 = self._make_from_defaults(id1)
+ item2 = self._make_from_defaults(id2)
+ item3 = self._make_from_defaults(id3)
+ # check .all() returns empty list on un-cached items
+ self.assertEqual(self.checked_class.all(self.db_conn), [])
+ # check that all() shows only cached/saved items
+ item1.cache()
+ item3.save(self.db_conn)
+ self.assertEqual(sorted(self.checked_class.all(self.db_conn)),
+ sorted([item1, item3]))
+ item2.save(self.db_conn)
+ self.assertEqual(sorted(self.checked_class.all(self.db_conn)),
+ sorted([item1, item2, item3]))
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_singularity(self) -> None:
+ """Test pointers made for single object keep pointing to it."""
+ id1 = self.default_ids[0]
+ obj = self._make_from_defaults(id1)
+ obj.save(self.db_conn)
+ # change object, expect retrieved through .by_id to carry change
+ attr_name = self._change_obj(obj)
+ new_attr = getattr(obj, attr_name)
+ retrieved = self.checked_class.by_id(self.db_conn, id1)
+ self.assertEqual(new_attr, getattr(retrieved, attr_name))
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ @TestCaseAugmented._run_on_versioned_attributes
+ def test_versioned_singularity(self,
+ owner: Any,
+ attr_name: str,
+ attr: VersionedAttribute,
+ _: str | float,
+ to_set: list[str] | list[float]
+ ) -> None:
+ """Test singularity of VersionedAttributes on saving."""
+ owner.save(self.db_conn)
+ # change obj, expect retrieved through .by_id to carry change
+ attr.set(to_set[0])
+ retrieved = self.checked_class.by_id(self.db_conn, owner.id_)
+ attr_retrieved = getattr(retrieved, attr_name)
+ self.assertEqual(attr.history, attr_retrieved.history)
+
+ @TestCaseAugmented._run_if_with_db_but_not_server
+ def test_remove(self) -> None:
+ """Test .remove() effects on DB and cache."""
+ id_ = self.default_ids[0]
+ obj = self._make_from_defaults(id_)
+ # check removal only works after saving
+ with self.assertRaises(HandledException):
+ obj.remove(self.db_conn)
+ obj.save(self.db_conn)
+ obj.remove(self.db_conn)
+ # check access to obj fails after removal
+ with self.assertRaises(HandledException):
+ print(obj.id_)
+ # check DB and cache now empty
+ self.check_identity_with_cache_and_db([])
+
+
+class Expected:
+ """Builder of (JSON-like) dict to compare against responses of test server.
+
+ Collects all items and relations we expect expressed in the server's JSON
+ responses and puts them into the proper json.dumps-friendly dict structure,
+ accessibla via .as_dict, to compare them in TestsWithServer.check_json_get.
+
+ On its own provides for .as_dict output only {"_library": …}, initialized
+ from .__init__ and to be directly manipulated via the .lib* methods.
+ Further structures of the expected response may be added and kept
+ up-to-date by subclassing .__init__, .recalc, and .d.
+
+ NB: Lots of expectations towards server behavior will be made explicit here
+ (or in the subclasses) rather than in the actual TestCase methods' code.
+ """
+ _default_dict: dict[str, Any]
+ _forced: dict[str, Any]
+ _fields: dict[str, Any]
+ _on_empty_make_temp: tuple[str, str]
+
+ def __init__(self,
+ todos: list[dict[str, Any]] | None = None,
+ procs: list[dict[str, Any]] | None = None,
+ procsteps: list[dict[str, Any]] | None = None,
+ conds: list[dict[str, Any]] | None = None,
+ days: list[dict[str, Any]] | None = None
+ ) -> None:
+ # pylint: disable=too-many-arguments
+ for name in ['_default_dict', '_fields', '_forced']:
+ if not hasattr(self, name):
+ setattr(self, name, {})
+ self._lib = {}
+ for title, items in [('Todo', todos),
+ ('Process', procs),
+ ('ProcessStep', procsteps),
+ ('Condition', conds),
+ ('Day', days)]:
+ if items:
+ self._lib[title] = self._as_refs(items)
+ for k, v in self._default_dict.items():
+ if k not in self._fields:
+ self._fields[k] = v
+
+ def recalc(self) -> None:
+ """Update internal dictionary by subclass-specific rules."""
+ todos = self.lib_all('Todo')
+ for todo in todos:
+ todo['parents'] = []
+ for todo in todos:
+ for child_id in todo['children']:
+ self.lib_get('Todo', child_id)['parents'] += [todo['id']]
+ todo['children'].sort()
+ procsteps = self.lib_all('ProcessStep')
+ procs = self.lib_all('Process')
+ for proc in procs:
+ proc['explicit_steps'] = [s['id'] for s in procsteps
+ if s['owner_id'] == proc['id']]
+
+ @property
+ def as_dict(self) -> dict[str, Any]:
+ """Return dict to compare against test server JSON responses."""
+ make_temp = False
+ if hasattr(self, '_on_empty_make_temp'):
+ category, dicter = getattr(self, '_on_empty_make_temp')
+ id_ = self._fields[category.lower()]
+ make_temp = not bool(self.lib_get(category, id_))
+ if make_temp:
+ self.lib_set(category, [getattr(self, dicter)(id_)])
+ self.recalc()
+ d = {'_library': self._lib}
+ for k, v in self._fields.items():
+ # we expect everything sortable to be sorted
+ if isinstance(v, list) and k not in self._forced:
+ # NB: if we don't test for v being list, sorted() on an empty
+ # dict may return an empty list
+ try:
+ v = sorted(v)
+ except TypeError:
+ pass
+ d[k] = v
+ for k, v in self._forced.items():
+ d[k] = v
+ if make_temp:
+ json = json_dumps(d)
+ id_ = id_ if id_ is not None else '?'
+ self.lib_del(category, id_)
+ d = json_loads(json)
+ return d
+
+ def lib_get(self, category: str, id_: str | int) -> dict[str, Any]:
+ """From library, return item of category and id_, or empty dict."""
+ str_id = str(id_)
+ if category in self._lib and str_id in self._lib[category]:
+ return self._lib[category][str_id]
+ return {}
+
+ def lib_all(self, category: str) -> list[dict[str, Any]]:
+ """From library, return items of category, or [] if none."""
+ if category in self._lib:
+ return list(self._lib[category].values())
+ return []
+
+ def lib_set(self, category: str, items: list[dict[str, object]]) -> None:
+ """Update library for category with items."""
+ if category not in self._lib:
+ self._lib[category] = {}
+ for k, v in self._as_refs(items).items():
+ self._lib[category][k] = v
+
+ def lib_del(self, category: str, id_: str | int) -> None:
+ """Remove category element of id_ from library."""
+ del self._lib[category][str(id_)]
+ if 0 == len(self._lib[category]):
+ del self._lib[category]
+
+ def lib_wipe(self, category: str) -> None:
+ """Remove category from library."""
+ if category in self._lib:
+ del self._lib[category]
+
+ def set(self, field_name: str, value: object) -> None:
+ """Set top-level .as_dict field."""
+ self._fields[field_name] = value
+
+ def force(self, field_name: str, value: object) -> None:
+ """Set ._forced field to ensure value in .as_dict."""
+ self._forced[field_name] = value
+
+ def unforce(self, field_name: str) -> None:
+ """Unset ._forced field."""
+ del self._forced[field_name]
+
+ @staticmethod
+ def _as_refs(items: list[dict[str, object]]
+ ) -> dict[str, dict[str, object]]:
+ """Return dictionary of items by their 'id' fields."""
+ refs = {}
+ for item in items:
+ id_ = str(item['id']) if item['id'] is not None else '?'
+ refs[id_] = item
+ return refs
+
+ @staticmethod
+ def as_ids(items: list[dict[str, Any]]) -> list[int] | list[str]:
+ """Return list of only 'id' fields of items."""
+ return [item['id'] for item in items]
+
+ @staticmethod
+ def day_as_dict(date: str, comment: str = '') -> dict[str, object]:
+ """Return JSON of Day to expect."""
+ return {'id': date, 'comment': comment, 'todos': []}
+
+ def set_day_from_post(self, date: str, d: dict[str, Any]) -> None:
+ """Set Day of date in library based on POST dict d."""
+ day = self.day_as_dict(date)
+ for k, v in d.items():
+ if 'day_comment' == k:
+ day['comment'] = v
+ elif 'new_todo' == k:
+ next_id = 1
+ for todo in self.lib_all('Todo'):
+ if next_id <= todo['id']:
+ next_id = todo['id'] + 1
+ for proc_id in sorted([id_ for id_ in v if id_]):
+ todo = self.todo_as_dict(next_id, proc_id, date)
+ self.lib_set('Todo', [todo])
+ next_id += 1
+ elif 'done' == k:
+ for todo_id in v:
+ self.lib_get('Todo', todo_id)['is_done'] = True
+ elif 'todo_id' == k:
+ for i, todo_id in enumerate(v):
+ t = self.lib_get('Todo', todo_id)
+ if 'comment' in d:
+ t['comment'] = d['comment'][i]
+ if 'effort' in d:
+ effort = d['effort'][i] if d['effort'][i] else None
+ t['effort'] = effort
+ self.lib_set('Day', [day])
+
+ @staticmethod
+ def cond_as_dict(id_: int = 1,
+ is_active: bool = False,
+ title: None | str = None,
+ description: None | str = None,
+ ) -> dict[str, object]:
+ """Return JSON of Condition to expect."""
+ versioned: dict[str, dict[str, object]]
+ versioned = {'title': {}, 'description': {}}
+ if title is not None:
+ versioned['title']['0'] = title
+ if description is not None:
+ versioned['description']['0'] = description
+ return {'id': id_, 'is_active': is_active, '_versioned': versioned}
+
+ def set_cond_from_post(self, id_: int, d: dict[str, Any]) -> None:
+ """Set Condition of id_ in library based on POST dict d."""
+ if 'delete' in d:
+ self.lib_del('Condition', id_)
+ return
+ cond = self.lib_get('Condition', id_)
+ if cond:
+ cond['is_active'] = 'is_active' in d and\
+ d['is_active'] in VALID_TRUES
+ for category in ['title', 'description']:
+ history = cond['_versioned'][category]
+ if len(history) > 0:
+ last_i = sorted([int(k) for k in history.keys()])[-1]
+ if d[category] != history[str(last_i)]:
+ history[str(last_i + 1)] = d[category]
+ else:
+ history['0'] = d[category]
+ else:
+ cond = self.cond_as_dict(id_, **d)
+ self.lib_set('Condition', [cond])
+
+ @staticmethod
+ def todo_as_dict(id_: int = 1,
+ process_id: int = 1,
+ date: str = '2024-01-01',
+ conditions: None | list[int] = None,
+ disables: None | list[int] = None,
+ blockers: None | list[int] = None,
+ enables: None | list[int] = None,
+ calendarize: bool = False,
+ comment: str = '',
+ is_done: bool = False,
+ effort: float | None = None,
+ children: list[int] | None = None,
+ parents: list[int] | None = None,
+ ) -> dict[str, object]:
+ """Return JSON of Todo to expect."""
+ # pylint: disable=too-many-arguments
+ d = {'id': id_,
+ 'date': date,
+ 'process_id': process_id,
+ 'is_done': is_done,
+ 'calendarize': calendarize,
+ 'comment': comment,
+ 'children': children if children else [],
+ 'parents': parents if parents else [],
+ 'effort': effort,
+ 'conditions': conditions if conditions else [],
+ 'disables': disables if disables else [],
+ 'blockers': blockers if blockers else [],
+ 'enables': enables if enables else []}
+ return d
+
+ def set_todo_from_post(self, id_: int, d: dict[str, Any]) -> None:
+ """Set Todo of id_ in library based on POST dict d."""
+ corrected_kwargs: dict[str, Any] = {
+ 'children': [], 'is_done': 0, 'calendarize': 0, 'comment': ''}
+ for k, v in d.items():
+ if k.startswith('step_filler_to_'):
+ continue
+ if 'adopt' == k:
+ new_children = v if isinstance(v, list) else [v]
+ corrected_kwargs['children'] += new_children
+ continue
+ if k in {'is_done', 'calendarize'} and v in VALID_TRUES:
+ v = True
+ corrected_kwargs[k] = v
+ todo = self.lib_get('Todo', id_)
+ if todo:
+ for k, v in corrected_kwargs.items():
+ todo[k] = v
+ else:
+ todo = self.todo_as_dict(id_, **corrected_kwargs)
+ self.lib_set('Todo', [todo])
+
+ @staticmethod
+ def procstep_as_dict(id_: int,
+ owner_id: int,
+ step_process_id: int,
+ parent_step_id: int | None = None
+ ) -> dict[str, object]:
+ """Return JSON of ProcessStep to expect."""
+ return {'id': id_,
+ 'owner_id': owner_id,
+ 'step_process_id': step_process_id,
+ 'parent_step_id': parent_step_id}
+
+ @staticmethod
+ def proc_as_dict(id_: int = 1,
+ title: None | str = None,
+ description: None | str = None,
+ effort: None | float = None,
+ conditions: None | list[int] = None,
+ disables: None | list[int] = None,
+ blockers: None | list[int] = None,
+ enables: None | list[int] = None,
+ explicit_steps: None | list[int] = None
+ ) -> dict[str, object]:
+ """Return JSON of Process to expect."""
+ # pylint: disable=too-many-arguments
+ versioned: dict[str, dict[str, object]]
+ versioned = {'title': {}, 'description': {}, 'effort': {}}
+ if title is not None:
+ versioned['title']['0'] = title
+ if description is not None:
+ versioned['description']['0'] = description
+ if effort is not None:
+ versioned['effort']['0'] = effort
+ d = {'id': id_,
+ 'calendarize': False,
+ 'suppressed_steps': [],
+ 'explicit_steps': explicit_steps if explicit_steps else [],
+ '_versioned': versioned,
+ 'conditions': conditions if conditions else [],
+ 'disables': disables if disables else [],
+ 'enables': enables if enables else [],
+ 'blockers': blockers if blockers else []}
+ return d
+
+ def set_proc_from_post(self, id_: int, d: dict[str, Any]) -> None:
+ """Set Process of id_ in library based on POST dict d."""
+ proc = self.lib_get('Process', id_)
+ if proc:
+ for category in ['title', 'description', 'effort']:
+ history = proc['_versioned'][category]
+ if len(history) > 0:
+ last_i = sorted([int(k) for k in history.keys()])[-1]
+ if d[category] != history[str(last_i)]:
+ history[str(last_i + 1)] = d[category]
+ else:
+ history['0'] = d[category]
+ else:
+ proc = self.proc_as_dict(id_,
+ d['title'], d['description'], d['effort'])
+ ignore = {'title', 'description', 'effort', 'new_top_step', 'step_of',
+ 'kept_steps'}
+ proc['calendarize'] = False
+ for k, v in d.items():
+ if k in ignore\
+ or k.startswith('step_') or k.startswith('new_step_to'):
+ continue
+ if k in {'calendarize'} and v in VALID_TRUES:
+ v = True
+ elif k in {'suppressed_steps', 'explicit_steps', 'conditions',
+ 'disables', 'enables', 'blockers'}:
+ if not isinstance(v, list):
+ v = [v]
+ proc[k] = v
+ self.lib_set('Process', [proc])
+
class TestCaseWithServer(TestCaseWithDB):
"""Module tests against our HTTP server/handler (and database)."""
self.server_thread.start()
self.conn = HTTPConnection(str(self.httpd.server_address[0]),
self.httpd.server_address[1])
+ self.httpd.render_mode = 'json'
def tearDown(self) -> None:
self.httpd.shutdown()
self.httpd.server_close()
self.server_thread.join()
super().tearDown()
+
+ def post_exp_cond(self,
+ exps: list[Expected],
+ payload: dict[str, object],
+ id_: int = 1,
+ post_to_id: bool = True,
+ redir_to_id: bool = True
+ ) -> None:
+ """POST /condition(s), appropriately update Expecteds."""
+ # pylint: disable=too-many-arguments
+ target = f'/condition?id={id_}' if post_to_id else '/condition'
+ redir = f'/condition?id={id_}' if redir_to_id else '/conditions'
+ if 'title' not in payload:
+ payload['title'] = 'foo'
+ if 'description' not in payload:
+ payload['description'] = 'foo'
+ self.check_post(payload, target, redir=redir)
+ for exp in exps:
+ exp.set_cond_from_post(id_, payload)
+
+ def post_exp_day(self,
+ exps: list[Expected],
+ payload: dict[str, Any],
+ date: str = '2024-01-01'
+ ) -> None:
+ """POST /day, appropriately update Expecteds."""
+ if 'make_type' not in payload:
+ payload['make_type'] = 'empty'
+ if 'day_comment' not in payload:
+ payload['day_comment'] = ''
+ target = f'/day?date={date}'
+ redir_to = f'{target}&make_type={payload["make_type"]}'
+ self.check_post(payload, target, 302, redir_to)
+ for exp in exps:
+ exp.set_day_from_post(date, payload)
+
+ def post_exp_process(self,
+ exps: list[Expected],
+ payload: dict[str, Any],
+ id_: int,
+ ) -> dict[str, object]:
+ """POST /process, appropriately update Expecteds."""
+ if 'title' not in payload:
+ payload['title'] = 'foo'
+ if 'description' not in payload:
+ payload['description'] = 'foo'
+ if 'effort' not in payload:
+ payload['effort'] = 1.1
+ self.check_post(payload, f'/process?id={id_}',
+ redir=f'/process?id={id_}')
+ for exp in exps:
+ exp.set_proc_from_post(id_, payload)
+ return payload
+
+ def post_exp_todo(self,
+ exps: list[Expected],
+ payload: dict[str, Any],
+ id_: int,
+ ) -> None:
+ """POST /todo, appropriately updated Expecteds."""
+ self.check_post(payload, f'/todo?id={id_}')
+ for exp in exps:
+ exp.set_todo_from_post(id_, payload)
+
+ def check_filter(self, exp: Expected, category: str, key: str,
+ val: str, list_ids: list[int]) -> None:
+ """Check GET /{category}?{key}={val} sorts to list_ids."""
+ # pylint: disable=too-many-arguments
+ exp.set(key, val)
+ exp.force(category, list_ids)
+ self.check_json_get(f'/{category}?{key}={val}', exp)
+
+ def check_redirect(self, target: str) -> None:
+ """Check that self.conn answers with a 302 redirect to target."""
+ response = self.conn.getresponse()
+ self.assertEqual(response.status, 302)
+ self.assertEqual(response.getheader('Location'), target)
+
+ def check_get(self, target: str, expected_code: int) -> None:
+ """Check that a GET to target yields expected_code."""
+ self.conn.request('GET', target)
+ self.assertEqual(self.conn.getresponse().status, expected_code)
+
+ def check_minimal_inputs(self,
+ url: str,
+ minimal_inputs: dict[str, Any]
+ ) -> None:
+ """Check that url 400's unless all of minimal_inputs provided."""
+ for to_hide in minimal_inputs.keys():
+ to_post = {k: v for k, v in minimal_inputs.items() if k != to_hide}
+ self.check_post(to_post, url, 400)
+
+ def check_post(self, data: Mapping[str, object], target: str,
+ expected_code: int = 302, redir: str = '') -> None:
+ """Check that POST of data to target yields expected_code."""
+ encoded_form_data = urlencode(data, doseq=True).encode('utf-8')
+ headers = {'Content-Type': 'application/x-www-form-urlencoded',
+ 'Content-Length': str(len(encoded_form_data))}
+ self.conn.request('POST', target,
+ body=encoded_form_data, headers=headers)
+ if 302 == expected_code:
+ redir = target if redir == '' else redir
+ self.check_redirect(redir)
+ else:
+ self.assertEqual(self.conn.getresponse().status, expected_code)
+
+ def check_get_defaults(self,
+ path: str,
+ default_id: str = '1',
+ id_name: str = 'id'
+ ) -> None:
+ """Some standard model paths to test."""
+ nonexist_status = 200 if self.checked_class.can_create_by_id else 404
+ self.check_get(path, nonexist_status)
+ self.check_get(f'{path}?{id_name}=', 400)
+ self.check_get(f'{path}?{id_name}=foo', 400)
+ self.check_get(f'/{path}?{id_name}=0', 400)
+ self.check_get(f'{path}?{id_name}={default_id}', nonexist_status)
+
+ def check_json_get(self, path: str, expected: Expected) -> None:
+ """Compare JSON on GET path with expected.
+
+ To simplify comparison of VersionedAttribute histories, transforms
+ timestamp keys of VersionedAttribute history keys into (strings of)
+ integers counting chronologically forward from 0.
+ """
+
+ def rewrite_history_keys_in(item: Any) -> Any:
+ if isinstance(item, dict):
+ if '_versioned' in item.keys():
+ for category in item['_versioned']:
+ vals = item['_versioned'][category].values()
+ history = {}
+ for i, val in enumerate(vals):
+ history[str(i)] = val
+ item['_versioned'][category] = history
+ for category in list(item.keys()):
+ rewrite_history_keys_in(item[category])
+ elif isinstance(item, list):
+ item[:] = [rewrite_history_keys_in(i) for i in item]
+ return item
+
+ def walk_diffs(path: str, cmp1: object, cmp2: object) -> None:
+ # pylint: disable=too-many-branches
+ def warn(intro: str, val: object) -> None:
+ if isinstance(val, (str, int, float)):
+ print(intro, val)
+ else:
+ print(intro)
+ pprint(val)
+ if cmp1 != cmp2:
+ if isinstance(cmp1, dict) and isinstance(cmp2, dict):
+ for k, v in cmp1.items():
+ if k not in cmp2:
+ warn(f'DIFF {path}: retrieved lacks {k}', v)
+ elif v != cmp2[k]:
+ walk_diffs(f'{path}:{k}', v, cmp2[k])
+ for k in [k for k in cmp2.keys() if k not in cmp1]:
+ warn(f'DIFF {path}: expected lacks retrieved\'s {k}',
+ cmp2[k])
+ elif isinstance(cmp1, list) and isinstance(cmp2, list):
+ for i, v1 in enumerate(cmp1):
+ if i >= len(cmp2):
+ warn(f'DIFF {path}[{i}] retrieved misses:', v1)
+ elif v1 != cmp2[i]:
+ walk_diffs(f'{path}[{i}]', v1, cmp2[i])
+ if len(cmp2) > len(cmp1):
+ for i, v2 in enumerate(cmp2[len(cmp1):]):
+ warn(f'DIFF {path}[{len(cmp1)+i}] misses:', v2)
+ else:
+ warn(f'DIFF {path} – for expected:', cmp1)
+ warn('… and for retrieved:', cmp2)
+
+ self.conn.request('GET', path)
+ response = self.conn.getresponse()
+ self.assertEqual(response.status, 200)
+ retrieved = json_loads(response.read().decode())
+ rewrite_history_keys_in(retrieved)
+ cmp = expected.as_dict
+ try:
+ self.assertEqual(cmp, retrieved)
+ except AssertionError as e:
+ print('EXPECTED:')
+ pprint(cmp)
+ print('RETRIEVED:')
+ pprint(retrieved)
+ walk_diffs('', cmp, retrieved)
+ raise e