docs(api): printers.yaml weg, Drucker in DB + /admin/printers Admin-UI (#124) [DRAFT]

backend/alembic/env.py

@@ -31,6 +31,34 @@
 # ... etc.
 
 
+def _include_object(
+    obj: object,
+    name: str | None,  # noqa: ARG001
+    type_: str,
+    reflected: bool,  # noqa: ARG001
+    compare_to: object,  # noqa: ARG001
+) -> bool:
+    """Filter expression-based indexes from autogenerate comparison.
+
+    SQLite cannot reflect text-expression indexes (e.g. ``created_at DESC``)
+    back from the database, so Alembic always sees a diff between the in-memory
+    model (which uses ``text("created_at DESC")``) and the reflected schema
+    (which returns the plain column name).  We exclude any ``Index`` whose
+    expressions contain a SQLAlchemy ``TextClause`` from autogenerate so that
+    ``alembic check`` stays green on SQLite dev/CI while the migration itself
+    still creates the correct expression index at upgrade time.
+    """
+    if type_ == "index":
+        from sqlalchemy import Index as _Index
+        from sqlalchemy.sql.elements import TextClause
+
+        if isinstance(obj, _Index):
+            for col in obj.expressions:
+                if isinstance(col, TextClause):
+                    return False
+    return True
+
+
 def run_migrations_offline() -> None:
     """Run migrations in 'offline' mode.
 
@@ -49,14 +77,19 @@ def run_migrations_offline() -> None:
         target_metadata=target_metadata,
         literal_binds=True,
         dialect_opts={"paramstyle": "named"},
+        include_object=_include_object,
     )
 
     with context.begin_transaction():
         context.run_migrations()
 
 
 def do_run_migrations(connection: Connection) -> None:
-    context.configure(connection=connection, target_metadata=target_metadata)
+    context.configure(
+        connection=connection,
+        target_metadata=target_metadata,
+        include_object=_include_object,
+    )
 
     with context.begin_transaction():
         context.run_migrations()

backend/alembic/versions/42fbd015698d_printers_audit_and_backfill.py

@@ -0,0 +1,122 @@
+"""printers_audit_and_backfill
+
+Issue #124: Schema-Erweiterung der printers-Tabelle um queue_timeout_s und
+cut_defaults_half_cut, neue printers_audit-Tabelle für Änderungsprotokoll sowie
+Backfill des connection.snmp-Blocks für Bestandsdrucker.
+
+Revision ID: 42fbd015698d
+Revises: 20260605b2c3d4e5
+Create Date: 2026-06-20
+
+"""
+
+from __future__ import annotations
+
+import json
+from collections.abc import Sequence
+
+import sqlalchemy as sa
+from alembic import op
+
+# revision identifiers, used by Alembic.
+revision: str = "42fbd015698d"
+down_revision: str | Sequence[str] | None = "20260605b2c3d4e5"
+branch_labels: str | Sequence[str] | None = None
+depends_on: str | Sequence[str] | None = None
+
+
+def _backfill_snmp(bind: sa.engine.Connection) -> None:
+    """Befüllt connection.snmp für Bestandsdrucker ohne SNMP-Konfiguration.
+
+    Pro printers-Row: wenn connection.snmp fehlt UND connection.host vorhanden,
+    wird snmp = {"discover": false, "community": "public"} gesetzt.
+
+    Schutzklauseln:
+    - NULL connection wird übersprungen
+    - Fehlendes host-Feld wird übersprungen
+    - Bereits vorhandenes snmp-Feld bleibt unverändert (idempotent)
+
+    Diese Funktion ist als top-level Funktion definiert damit sie direkt
+    via ``await conn.run_sync(mig._backfill_snmp)`` in Tests testbar ist.
+    """
+    rows = bind.execute(sa.text("SELECT id, connection FROM printers")).all()
+    for row in rows:
+        pid, conn_raw = row[0], row[1]
+        if conn_raw is None:
+            continue
+        conn_json: dict = json.loads(conn_raw) if isinstance(conn_raw, str) else conn_raw
+        if "host" not in conn_json:
+            continue
+        if "snmp" in conn_json:
+            continue  # idempotent — bereits konfiguriert
+        conn_json["snmp"] = {"discover": False, "community": "public"}
+        bind.execute(
+            sa.text("UPDATE printers SET connection = :c WHERE id = :pid"),
+            {"c": json.dumps(conn_json), "pid": pid},
+        )
+
+
+def upgrade() -> None:
+    """Schema-Erweiterung: neue Spalten auf printers + printers_audit-Tabelle + Backfill."""
+    # 1. Neue Spalten auf printers (batch_alter_table für SQLite-Kompatibilität)
+    with op.batch_alter_table("printers", schema=None) as batch_op:
+        batch_op.add_column(
+            sa.Column(
+                "queue_timeout_s",
+                sa.Integer(),
+                nullable=False,
+                server_default="30",
+            )
+        )
+        batch_op.add_column(
+            sa.Column(
+                "cut_defaults_half_cut",
+                sa.Boolean(),
+                nullable=False,
+                server_default="0",
+            )
+        )
+
+    # 2. printers_audit-Tabelle (kein FK auf printers — Soft-Delete behält Rows)
+    op.create_table(
+        "printers_audit",
+        sa.Column("id", sa.Uuid(), nullable=False),
+        sa.Column("printer_id", sa.Uuid(), nullable=False),
+        sa.Column("slug", sa.String(255), nullable=False),
+        sa.Column("action", sa.String(50), nullable=False),
+        sa.Column("before_json", sa.JSON(), nullable=True),
+        sa.Column("after_json", sa.JSON(), nullable=True),
+        sa.Column("updated_by", sa.String(255), nullable=False),
+        sa.Column(
+            "created_at",
+            sa.DateTime(timezone=True),
+            nullable=False,
+            server_default=sa.text("CURRENT_TIMESTAMP"),
+        ),
+        sa.PrimaryKeyConstraint("id"),
+    )
+    op.create_index(
+        "idx_printers_audit_printer_id",
+        "printers_audit",
+        ["printer_id"],
+    )
+    op.create_index(
+        "idx_printers_audit_created_at_desc",
+        "printers_audit",
+        [sa.text("created_at DESC")],
+    )
+
+    # 3. Backfill connection.snmp für Bestandsdrucker
+    bind = op.get_bind()
+    _backfill_snmp(bind)
+
+
+def downgrade() -> None:
+    """Downgrade ist ein no-op.
+
+    Die neuen Spalten und die printers_audit-Tabelle können nicht sicher
+    zurückgerollt werden ohne Datenverlust (Audit-Logs sind produktiv relevant).
+    Der SNMP-Backfill kann ebenfalls nicht rückgängig gemacht werden ohne
+    Originalzustand zu kennen. Downgrade wird daher nicht unterstützt.
+    """
+    pass