Retain original bank statements for 24 hours

app/modules/bank_statement_analyzer/service.py

@@ -272,16 +272,9 @@ def _process_job(job_id: str) -> None:
             "financial_year": job.financial_year or "",
             "classification_enabled": job.classification_enabled,
         }
-        for path in paths:
-            try:
-                path.unlink(missing_ok=True)
-            except OSError:
-                pass
-        input_dir = Path(job.job_directory) / "Input"
-        try:
-            input_dir.rmdir()
-        except OSError:
-            pass
+        # Keep the original uploaded statements until the job expiry time.
+        # This applies equally to completed and failed jobs and allows the
+        # uploader to reproduce parser issues during the 24-hour retention window.
         job.output_file = str(output)
         job.summary_json = json.dumps(summary)
         job.status = "completed"
@@ -399,6 +392,36 @@ def list_user_jobs(user_id: int, limit: int = 25) -> list[BankStatementAnalysisJ
         db.close()
 
 
+def original_statement_files(job: BankStatementAnalysisJob) -> list[Path]:
+    """Return retained input PDFs that still belong to the job directory."""
+    job_root = Path(job.job_directory).resolve()
+    input_root = (job_root / "Input").resolve()
+    try:
+        configured = [Path(value).resolve() for value in json.loads(job.input_files_json or "[]")]
+    except (TypeError, ValueError, json.JSONDecodeError):
+        configured = []
+
+    retained: list[Path] = []
+    for path in configured:
+        try:
+            path.relative_to(input_root)
+        except ValueError:
+            continue
+        if path.is_file() and path.suffix.lower() == ".pdf":
+            retained.append(path)
+    return retained
+
+
+def get_owned_original_statement(user_id: int, job_id: str, file_index: int) -> tuple[BankStatementAnalysisJob, Path] | None:
+    job = get_owned_job(user_id, job_id)
+    if not job or job.status not in {"completed", "failed"}:
+        return None
+    files = original_statement_files(job)
+    if file_index < 1 or file_index > len(files):
+        return None
+    return job, files[file_index - 1]
+
+
 def job_view(job: BankStatementAnalysisJob) -> dict:
     return {
         "id": job.id,
@@ -415,6 +438,11 @@ def job_view(job: BankStatementAnalysisJob) -> dict:
         "estimated_wait": estimated_wait(job),
         "summary": _summary(job),
         "download_ready": job.status == "completed" and bool(job.output_file) and Path(job.output_file).is_file(),
+        "original_files": [
+            {"index": index, "name": path.name}
+            for index, path in enumerate(original_statement_files(job), start=1)
+        ],
+        "originals_retained": job.status in {"completed", "failed"} and bool(original_statement_files(job)),
     }