more tests and bugfixes

cateory-tree-sketch.json

@@ -0,0 +1,27 @@
+{
+  "month": "2026-10",
+  "categories": [
+    {
+      "category": "super-category-1",
+      "total_amount": 100,
+      "total_budget": 120,
+      "sub_categories": [
+        {
+          "category": "category-1",
+          "total_amount": 60,
+          "total_budget": 50,
+          "sub_categories": [],
+          "transactions": []
+        },
+        {
+          "category": "category-2",
+          "total_amount": 40,
+          "total_budget": 60,
+          "sub_categories": [],
+          "transactions": []
+        }
+      ],
+      "transactions": []
+    }
+  ]
+}

src/ingest/transactions.py

@@ -5,11 +5,21 @@ from sqlalchemy.dialects.sqlite import insert as sqlite_insert
 from sqlalchemy.orm import Session
 
 from src.models.b4w import B4wDb
+from src.models.category import Category, CategoryDb
 from src.models.comdirect import ComdirectGiroDb
 from src.models.common import BaseModelDb
 from src.models.transactions import Transaction, TransactionDb
 
 
+def _category_from_db(category_db: CategoryDb | None) -> Category | None:
+    if category_db is None:
+        return None
+    return Category(
+        name=category_db.name,
+        parent=_category_from_db(category_db.parent),
+    )
+
+
 def map_b4w_to_transaction(row: B4wDb) -> Transaction:
     """Project one staging row of ``import_b4w`` onto a canonical Transaction."""
     return Transaction(
@@ -24,7 +34,7 @@ def map_b4w_to_transaction(row: B4wDb) -> Transaction:
         text=row.text,
         other_name=row.other_name,
         other_iban=row.other_iban,
-        category_id=row.category_id,
+        category=_category_from_db(row.category),
     )
 
 
@@ -44,7 +54,11 @@ def map_comdirect_giro_to_transaction(row: ComdirectGiroDb) -> Transaction:
 
 
 def upsert_transaction(
-    session: Session, transaction: Transaction, do_commit: bool = True
+    session: Session,
+    transaction: Transaction,
+    *,
+    category_db: CategoryDb | None = None,
+    do_commit: bool = True,
 ) -> None:
     """Insert one canonical transaction, ignoring duplicates on ``hash_id``.
 
@@ -65,7 +79,7 @@ def upsert_transaction(
         "text": transaction.text,
         "other_name": transaction.other_name,
         "other_iban": transaction.other_iban,
-        "category_id": transaction.category_id,
+        "category_id": category_db.id if category_db is not None else None,
     }
     stmt = sqlite_insert(TransactionDb).values(parameters)
     _ = session.execute(stmt.prefix_with("OR IGNORE"))
@@ -92,6 +106,11 @@ def consolidate_transactions[StagingDbT: BaseModelDb](
         transaction = mapper(staging_row)
         if transaction.hash_id in existing_hashes:
             continue
-        upsert_transaction(session, transaction, do_commit=False)
+        upsert_transaction(
+            session,
+            transaction,
+            category_db=getattr(staging_row, "category", None),
+            do_commit=False,
+        )
     if do_commit:
         session.commit()

src/models/category.py

@@ -18,7 +18,7 @@ from sqlalchemy import (
 from sqlalchemy.orm import Mapped, Session, mapped_column, relationship
 
 from src.db_utils import upsert
-from src.models.common import BaseModelDb, ExportBaseModel, hashify
+from src.models.common import AmountPos, BaseModelDb, ExportBaseModel, hashify
 
 
 class UTCDateTime(TypeDecorator[datetime]):
@@ -69,12 +69,16 @@ class Category(ExportBaseModel):
         else:
             return f"{self.parent.to_str()}:{self.name}"
 
-    def __str__(self):
-        return self.to_str()
+    def __str__(self) -> str:
+        return self.to_str() or ""
 
     def __repr__(self):
         return f'Category.from_str("{self.__str__()}")'
 
+    def get_budget(self, date_valid: date | None) -> float:
+        # TODO
+        return 0.0
+
 
 class CategoryDb(BaseModelDb):
     __tablename__ = "category"
@@ -187,7 +191,7 @@ def upsert_category(
 
 class CategoryBudget(BaseModel):
     category: Category
-    amount_eur: Decimal = Field(gt=0, decimal_places=2)
+    amount_eur: AmountPos
     valid_from: date | None = Field(
         default=None, description="Start of validity. None means unbounded"
     )

src/models/category_tree.py

@@ -0,0 +1,85 @@
+from datetime import date
+
+from pydantic import BaseModel, computed_field
+
+from src.models.category import Category
+from src.models.transactions import Transaction
+
+
+class CategoryTree(BaseModel):
+    category: Category | None
+    sub_categories: list[CategoryTree] | None = None
+    transactions: list[Transaction] | None = None
+
+    @computed_field  # type: ignore[prop-decorator]
+    @property
+    def total_amount(self) -> float:
+        result = 0.0
+        if self.transactions:
+            result += sum(
+                float(transaction.amount) for transaction in self.transactions
+            )
+        if self.sub_categories:
+            result += sum(
+                float(sub_category.total_amount) for sub_category in self.sub_categories
+            )
+        return result
+
+    def total_budget(self, date_valid: date) -> float:
+        result = 0.0
+        if self.category is None:
+            return result
+        result += self.category.get_budget(date_valid)
+        if self.sub_categories:
+            result += sum(
+                sub_category.total_budget(date_valid)
+                for sub_category in self.sub_categories
+            )
+        return result
+
+    def __add__(self, other: CategoryTree):
+        if self.category != other.category:
+            return
+
+        sub_trees = []
+        seen = []
+        for sub_tree_1 in self.sub_categories or []:
+            new = sub_tree_1
+            for sub_tree_2 in other.sub_categories or []:
+                if sub_tree_1.category == sub_tree_2.category:
+                    new += sub_tree_2
+                    seen.append(sub_tree_2)
+                    break
+            sub_trees.append(new)
+        sub_trees += [
+            sub_tree
+            for sub_tree in (other.sub_categories or [])
+            if sub_tree not in seen
+        ]
+
+        transactions = ((self.transactions or []) + (other.transactions or [])) or None
+
+        return CategoryTree(
+            category=self.category,
+            sub_categories=list(sub_trees),
+            transactions=transactions,
+        )
+
+
+def build_category_tree(transactions: list[Transaction]) -> list[CategoryTree]:
+    results: dict[str, CategoryTree] = {}
+    for branch in map(_transaction_to_category_branch, transactions):
+        key = str(branch.category) if branch.category else ""
+        results[key] = (results[key] + branch) if key in results else branch
+    return list(results.values())
+
+
+def _transaction_to_category_branch(transaction: Transaction) -> CategoryTree:
+    category = transaction.category
+    result = CategoryTree(category=category, transactions=[transaction])
+    if not category:
+        return result
+    while category.parent is not None:
+        category = category.parent
+        result = CategoryTree(category=category, sub_categories=[result])
+    return result

src/models/common.py

@@ -2,8 +2,9 @@ import hashlib
 from datetime import date
 from decimal import Decimal
 from enum import Enum
+from typing import Annotated
 
-from pydantic import BaseModel
+from pydantic import BaseModel, Field
 from sqlalchemy import Date, Numeric, String
 from sqlalchemy.orm import DeclarativeBase, Mapped, mapped_column
 
@@ -40,11 +41,15 @@ def date2str(value: date | None, default="--", date_format="%Y-%m-%d") -> str:
     return default
 
 
-def _parse_decimal_german(decimal_str: str) -> Decimal:
+type AmountPos = Annotated[Decimal, Field(decimal_places=2, gt=0)]
+type Amount = Annotated[Decimal, Field(decimal_places=2)]
+
+
+def _parse_decimal_german(decimal_str: str) -> Amount:
     return Decimal(decimal_str.replace(".", "").replace(",", "."))
 
 
-def decimal2str(value: Decimal) -> str:
+def decimal2str(value: Amount) -> str:
     value = Decimal(f"{value:.2f}")
     int_part, _, frac = str(value).partition(".")
     groups: list[str] = []

src/models/transactions.py

@@ -1,11 +1,11 @@
 from datetime import date
-from decimal import Decimal
 
 from pydantic import Field, computed_field
 from sqlalchemy import Date, ForeignKey, Index, Integer, String, UniqueConstraint
 from sqlalchemy.orm import Mapped, mapped_column, relationship
 
-from src.models.common import ExportBaseModel, IngestdbBase, date2str, hashify
+from src.models.category import Category
+from src.models.common import Amount, ExportBaseModel, IngestdbBase, date2str, hashify
 
 
 class Transaction(ExportBaseModel):
@@ -22,13 +22,13 @@ class Transaction(ExportBaseModel):
     account_iban: str
     account_currency: str
     date_valid: date
-    amount: Decimal = Field(description="Signed amount; negative = money out")
+    amount: Amount
     amount_currency: str
     date_book: date | None = None
     text: str | None = None
     other_name: str | None = None
     other_iban: str | None = None
-    category_id: int | None = None
+    category: Category | None = None
 
     @computed_field  # type: ignore[prop-decorator]
     @property
@@ -55,7 +55,7 @@ class Transaction(ExportBaseModel):
             "text": f"{self.text or ''}",
             "other_name": f"{self.other_name or ''}",
             "other_iban": f"{self.other_iban or ''}",
-            "category_id": f"{self.category_id}",
+            "category": (self.category.to_str() if self.category else "") or "",
         }
 
 

tests/test_category_edge_cases.py

@@ -1,4 +1,5 @@
 from datetime import date
+from decimal import Decimal
 
 import pytest
 
@@ -12,24 +13,30 @@ class TestCategoryComputedFields:
     def test_hash_id_computed_field_simple(self):
         """Test hash_id for single-level category."""
         c = Category.from_str("single")
+        assert c
         assert hasattr(c, "hash_id")
         assert isinstance(c.hash_id, str)
         # Hash should be deterministic for same inputs
         c2 = Category.from_str("single")
+        assert c2
         assert c.hash_id == c2.hash_id
 
     def test_hash_id_computed_field_nested(self):
         """Test hash_id for multi-level category hierarchy."""
         c1 = Category.from_str("level1:level2:level3")
+        assert c1
         assert hasattr(c1, "hash_id")
 
         c2 = Category.from_str("level1:level2:level3")
+        assert c2
         assert c1.hash_id == c2.hash_id
 
     def test_hash_id_different_hierarchy(self):
         """Test that different hierarchies have different hash_ids."""
         c1 = Category.from_str("A:B")
         c2 = Category.from_str("A:C")
+        assert c1
+        assert c2
         # Different last level -> different hash_id
         assert c1.hash_id != c2.hash_id
 
@@ -40,22 +47,26 @@ class TestCategoryComputedFields:
 
         # Create another with same name but no parent
         orphan = Category.from_str("Budget")
+        assert orphan
 
         assert cat1.hash_id != orphan.hash_id
 
     def test_to_str_single_level(self):
         """Test to_str() for single-level category."""
         c = Category.from_str("single")
+        assert c
         assert c.to_str() == "single"
 
     def test_to_str_nested(self):
         """Test to_str() for nested category."""
         c = Category.from_str("parent:child:grandchild")
+        assert c
         assert c.to_str() == "parent:child:grandchild"
 
     def test_str_uses_to_str(self):
         """Test that __str__ delegates to to_str()."""
         c = Category.from_str("A:B:C")
+        assert c
         assert str(c) == c.to_str() == "A:B:C"
 
     def test_repr_simple(self):
@@ -79,7 +90,7 @@ class TestCategoryBudgetValidation:
         cat = CategoryFactory.build()
         budget = CategoryBudget(
             category=cat,
-            amount_eur=100.00,
+            amount_eur=Decimal("100.00"),
         )
         assert budget.valid_from is None
         assert budget.valid_to is None
@@ -90,7 +101,7 @@ class TestCategoryBudgetValidation:
         cat = CategoryFactory.build()
         budget = CategoryBudget(
             category=cat,
-            amount_eur=100.00,
+            amount_eur=Decimal("100.00"),
             valid_from=date(2026, 10, 1),
             valid_to=date(2026, 10, 31),
         )
@@ -105,7 +116,7 @@ class TestCategoryBudgetValidation:
         with pytest.raises(ValueError) as exc_info:
             CategoryBudget(
                 category=cat,
-                amount_eur=100.00,
+                amount_eur=Decimal("100.00"),
                 valid_from=date(2026, 10, 31),
                 valid_to=date(2026, 10, 1),
             )
@@ -120,7 +131,7 @@ class TestCategoryBudgetValidation:
         cat = CategoryFactory.build()
         budget = CategoryBudget(
             category=cat,
-            amount_eur=100.00,
+            amount_eur=Decimal("100.00"),
             valid_from=date(2026, 10, 1),
             valid_to=date(2026, 10, 1),
         )
@@ -133,7 +144,7 @@ class TestCategoryBudgetValidation:
         cat = CategoryFactory.build()
         budget = CategoryBudget(
             category=cat,
-            amount_eur=150.75,
+            amount_eur=Decimal("150.75"),
             valid_from=date(2026, 9, 1),
             valid_to=date(2026, 9, 30),
         )
@@ -142,7 +153,7 @@ class TestCategoryBudgetValidation:
         try:
             CategoryBudget(
                 category=cat,
-                amount_eur=200.00,
+                amount_eur=Decimal("200.00"),
                 valid_from=date(2026, 12, 31),
                 valid_to=date(2026, 12, 30),
             )

tests/test_category_tree.py

@@ -0,0 +1,116 @@
+from src.models.category import Category
+from src.models.category_tree import (
+    CategoryTree,
+    _transaction_to_category_branch,
+    build_category_tree,
+)
+from src.models.transactions import Transaction
+from tests.factories import TransactionFactory
+
+
+def test_transaction_to_category_branch_simple():
+    transaction = TransactionFactory.build(category=Category.from_str("dummy category"))
+    result = _transaction_to_category_branch(transaction)
+    assert isinstance(result, CategoryTree)
+    assert result.category
+    assert result.category.name == "dummy category"
+    assert result.transactions
+    assert isinstance(result.transactions, list)
+    assert len(result.transactions) == 1
+    assert isinstance(result.transactions[0], Transaction)
+
+
+def test_transaction_to_category_branch_two_category_levels():
+    transaction = TransactionFactory.build(category=Category.from_str("father:son"))
+    result = _transaction_to_category_branch(transaction)
+    assert isinstance(result, CategoryTree)
+    assert result.category
+    assert result.category.name == "father"
+    assert result.transactions is None
+    assert result.sub_categories
+    assert isinstance(result.sub_categories, list)
+    assert len(result.sub_categories) == 1
+    assert isinstance(result.sub_categories[0], CategoryTree)
+    assert isinstance(result.sub_categories[0].category, Category)
+    assert result.sub_categories[0].category.name == "son"
+    assert result.sub_categories[0].transactions
+    assert isinstance(result.sub_categories[0].transactions, list)
+    assert len(result.sub_categories[0].transactions) == 1
+    assert isinstance(result.sub_categories[0].transactions[0], Transaction)
+    assert result.sub_categories[0].transactions[0] == transaction
+
+
+def test_transaction_to_category_branch_no_category():
+    transaction = TransactionFactory.build(category=None)
+    result = _transaction_to_category_branch(transaction)
+    assert result.category is None
+    assert result.transactions
+    assert isinstance(result.transactions, list)
+    assert len(result.transactions) == 1
+    assert result.transactions[0] == transaction
+
+
+def test_category_tree_add():
+    category_tree_1 = CategoryTree(
+        category=Category.from_str("super"),
+        sub_categories=[
+            CategoryTree(
+                category=Category.from_str("super:sub1"),
+                transactions=[TransactionFactory.build(amount=10)],
+            ),
+            CategoryTree(
+                category=Category.from_str("super:sub2"),
+                transactions=[TransactionFactory.build(amount=30)],
+            ),
+        ],
+    )
+    category_tree_2 = CategoryTree(
+        category=Category.from_str("super"),
+        sub_categories=[
+            CategoryTree(
+                category=Category.from_str("super:sub2"),
+                transactions=[TransactionFactory.build(amount=20)],
+            ),
+            CategoryTree(
+                category=Category.from_str("super:sub3"),
+                transactions=[TransactionFactory.build(amount=100)],
+            ),
+        ],
+    )
+    result = category_tree_1 + category_tree_2
+    assert result
+    assert result.sub_categories
+    assert isinstance(result.sub_categories, list)
+    assert len(result.sub_categories) == 3
+    assert all(s.category for s in result.sub_categories)
+    assert sorted([s.category.name for s in result.sub_categories]) == [
+        "sub1",
+        "sub2",
+        "sub3",
+    ]
+    assert result.total_amount == 160.0
+
+
+def test_build_category_tree():
+    transactions = [
+        TransactionFactory.build(category=Category.from_str("super1:sub1"), amount=10),
+        TransactionFactory.build(category=Category.from_str("super1:sub2"), amount=13),
+        TransactionFactory.build(
+            category=Category.from_str("super1:sub2:sub2_sub1"), amount=104
+        ),
+        TransactionFactory.build(
+            category=Category.from_str("super1:sub2:sub2_sub1"), amount=109
+        ),
+        TransactionFactory.build(category=Category.from_str("super2"), amount=5000),
+    ]
+    result = build_category_tree(transactions)
+    assert result
+    assert isinstance(result, list)
+    assert len(result) == 2
+    assert isinstance(result[0], CategoryTree)
+    assert result[0].category
+    assert result[0].category.name == "super1"
+    assert result[0].transactions is None
+    assert result[0].sub_categories
+    assert isinstance(result[0].sub_categories, list)
+    assert len(result[0].sub_categories) == 2

tests/test_transactions.py

@@ -101,7 +101,7 @@ def test_map_comdirect_giro_to_transaction():
     assert transaction.amount == Decimal("32.43")
     assert transaction.other_name is None
     assert transaction.other_iban is None
-    assert transaction.category_id is None
+    assert transaction.category is None
 
 
 def test_consolidate_b4w_signed_amount_and_category(db_session: Session):
@@ -125,6 +125,9 @@ def test_consolidate_b4w_signed_amount_and_category(db_session: Session):
     assert transaction.amount == Decimal("-50.00")
     assert transaction.category_id == staging.category_id
     assert transaction.text == "Supermarket"
+    mapped = map_b4w_to_transaction(staging)
+    assert mapped.category is not None
+    assert mapped.category.to_str() == export.category.to_str()
 
 
 def test_ingest_b4w_e2e_idempotent(db_session: Session):