Add Phase 11B client aware multi bank ownership and engagement archive

app/modules/bank_statement_analyzer/service.py

@@ -20,9 +20,16 @@ from app.core.db.common import CommonSessionLocal
 
 from .analyzer import analyze_files, export_excel
 from .models import BankStatementAnalysisJob
+from .client_context import (
+    archive_analysis_to_engagement,
+    validate_statement_ownership,
+)
+from app.modules.core.iam.models import User
+from app.modules.services.models import ClientServiceSubscription
 
 ALLOWED_ROLES = {"Partner", "Manager", "Branch Manager", "Staff", "Employee", "Consultant"}
-MAX_FILES = int(os.getenv("BANK_ANALYZER_MAX_FILES", "24"))
+# 0 means no fixed statement-count ceiling; per-file size and queue controls remain.
+MAX_FILES = int(os.getenv("BANK_ANALYZER_MAX_FILES", "0"))
 MAX_FILE_BYTES = int(os.getenv("BANK_ANALYZER_MAX_FILE_MB", "50")) * 1024 * 1024
 MAX_GLOBAL_PROCESSING = 3
 MAX_PENDING_PER_USER = 3
@@ -117,7 +124,7 @@ async def save_uploads(files: list[UploadFile], input_dir: Path) -> list[Path]:
     usable = [item for item in files if item and (item.filename or "").strip()]
     if not usable:
         raise ValueError("Please select at least one PDF bank statement.")
-    if len(usable) > MAX_FILES:
+    if MAX_FILES > 0 and len(usable) > MAX_FILES:
         raise ValueError(f"A maximum of {MAX_FILES} PDF files can be analyzed in one job.")
     saved: list[Path] = []
     try:
@@ -155,7 +162,7 @@ def pending_count_for_user(user_id: int) -> int:
         db.close()
 
 
-def enqueue_job(*, user, roles: Iterable[str], job_id: str, paths: list[Path], job_dir: Path, bank_selection: str, financial_year: str, customer_override: str, account_override: str, classification_enabled: bool) -> BankStatementAnalysisJob:
+def enqueue_job(*, user, roles: Iterable[str], job_id: str, paths: list[Path], job_dir: Path, bank_selection: str, financial_year: str, customer_override: str, account_override: str, classification_enabled: bool, client_id: int | None = None, engagement_id: int | None = None, ownership_confirmation: bool = False) -> BankStatementAnalysisJob:
     if pending_count_for_user(int(user.id)) >= MAX_PENDING_PER_USER:
         shutil.rmtree(job_dir, ignore_errors=True)
         raise ValueError("You already have three queued or processing analyses. Please wait for one to complete before submitting another.")
@@ -169,6 +176,11 @@ def enqueue_job(*, user, roles: Iterable[str], job_id: str, paths: list[Path], j
         branch_id=getattr(user, "branch_id", None),
         user_id=int(user.id),
         role_bucket=bucket,
+        client_id=int(client_id) if client_id else None,
+        engagement_id=int(engagement_id) if engagement_id else None,
+        ownership_confirmation=bool(ownership_confirmation),
+        ownership_status="pending" if client_id else "not_checked",
+        engagement_archive_status="pending" if engagement_id else "not_requested",
         selected_bank=bank_selection,
         financial_year=(financial_year or "").strip() or None,
         customer_override=(customer_override or "").strip() or None,
@@ -245,6 +257,18 @@ def _process_job(job_id: str) -> None:
             bank_hint=job.selected_bank,
             classification_enabled=job.classification_enabled,
         )
+
+        ownership = validate_statement_ownership(
+            db,
+            tenant_id=int(job.tenant_id or 0),
+            client_id=job.client_id,
+            metas=metas,
+            owner_override=job.customer_override or "",
+            ownership_confirmation=bool(job.ownership_confirmation),
+        )
+        job.ownership_status = str(ownership.get("status") or "not_checked")
+        job.ownership_validation_json = json.dumps(ownership, ensure_ascii=False)
+
         job.progress_percent = 75
         db.commit()
         output = output_dir / _workbook_filename(metas)
@@ -257,6 +281,28 @@ def _process_job(job_id: str) -> None:
             selected_bank=job.selected_bank,
             classification_enabled=job.classification_enabled,
         )
+
+        archive_result = {"status": "not_requested", "documents": []}
+        if job.client_id and job.engagement_id:
+            engagement = db.get(ClientServiceSubscription, int(job.engagement_id))
+            uploader = db.get(User, int(job.user_id))
+            if not engagement or int(engagement.client_id) != int(job.client_id):
+                raise ValueError("Selected engagement is no longer valid for the selected client.")
+            if not uploader:
+                raise ValueError("Bank Analyzer uploader could not be resolved for document archive.")
+            archive_result = archive_analysis_to_engagement(
+                db,
+                job=job,
+                engagement=engagement,
+                metas=metas,
+                source_paths=paths,
+                output_path=output,
+                user=uploader,
+            )
+            job.engagement_archive_status = str(archive_result.get("status") or "failed")
+            job.engagement_archive_json = json.dumps(archive_result, ensure_ascii=False)
+            db.commit()
+
         summary = {
             "job_id": job.id,
             "statement_count": len(metas),
@@ -272,6 +318,10 @@ def _process_job(job_id: str) -> None:
             "account_number": next((meta.account_number for meta in metas if meta.account_number), ""),
             "financial_year": job.financial_year or "",
             "classification_enabled": job.classification_enabled,
+            "client_id": job.client_id,
+            "engagement_id": job.engagement_id,
+            "ownership_status": job.ownership_status,
+            "engagement_archive_status": job.engagement_archive_status,
         }
         # Keep the original uploaded statements until the job expiry time.
         # This applies equally to completed and failed jobs and allows the