From 7ec116a3c7013706a8207be1e9e8cd582902f00b Mon Sep 17 00:00:00 2001
From: Chris Coutinho <chris@coutinho.io>
Date: Thu, 4 Jun 2026 21:57:55 +0200
Subject: [PATCH] fix(review): close doc via try/finally; don't count parse
 failures as indexed

Address PR #852 review:

- pymupdf.py: the metadata `doc` was only closed on the PdfParseFailed and
  success paths, so a failure in `_extract_metadata`/`mkdir`/`get_settings`
  leaked it. `doc` is only needed for metadata + page_count (the heavy parse
  works from `content` bytes in the worker), so open it, read metadata, and
  close it immediately under try/finally; drop the two later doc.close() calls.

- processor.py: a permanent parse failure early-returned from `_index_document`,
  after which `process_document` still recorded record_qdrant_operation("upsert",
  "success") + record_vector_sync_processing(success) -- counting an OOM/timeout
  bomb as astrolabe_documents_indexed_total{status="success"}. `_index_document`
  now returns False on that path and the caller skips the success metrics (the
  failure is already recorded via document_parse_failed_total + the registry's
  document_parse_total{error}).

Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
---
 .../document_processors/pymupdf.py            | 19 ++++++++++---------
 nextcloud_mcp_server/vector/processor.py      | 16 ++++++++++++----
 2 files changed, 22 insertions(+), 13 deletions(-)

diff --git a/nextcloud_mcp_server/document_processors/pymupdf.py b/nextcloud_mcp_server/document_processors/pymupdf.py
index 722fad40..d1262b02 100644
--- a/nextcloud_mcp_server/document_processors/pymupdf.py
+++ b/nextcloud_mcp_server/document_processors/pymupdf.py
@@ -107,14 +107,19 @@ class PyMuPDFProcessor(DocumentProcessor):
             if progress_callback:
                 await progress_callback(0, 100, "Opening PDF document")
 
-            # Open document and extract metadata in thread
+            # Open document only to read metadata + page count, then close it
+            # immediately (try/finally so a failure in _extract_metadata can't
+            # leak it). The heavy extraction below works from ``content`` bytes
+            # in the isolated worker, so ``doc`` is not needed past this point.
             doc = await anyio.to_thread.run_sync(  # type: ignore[attr-defined]
                 lambda: pymupdf.open("pdf", content)
             )
-
-            metadata = self._extract_metadata(doc, filename)
-            metadata["file_size"] = len(content)
-            page_count = doc.page_count
+            try:
+                metadata = self._extract_metadata(doc, filename)
+                metadata["file_size"] = len(content)
+                page_count = doc.page_count
+            finally:
+                doc.close()
 
             if progress_callback:
                 await progress_callback(10, 100, f"Extracting {page_count} pages")
@@ -143,7 +148,6 @@ class PyMuPDFProcessor(DocumentProcessor):
                     mem_limit_mb=settings.document_parse_mem_limit_mb,
                 )
             except PdfParseFailed as exc:
-                doc.close()
                 logger.warning(
                     "Isolated PDF parse failed for %s (reason=%s): %s",
                     filename or "<bytes>",
@@ -198,9 +202,6 @@ class PyMuPDFProcessor(DocumentProcessor):
                 metadata["image_paths"] = image_paths
             metadata["page_boundaries"] = page_boundaries
 
-            # Close document
-            doc.close()
-
             if progress_callback:
                 await progress_callback(100, 100, "Processing complete")
 
diff --git a/nextcloud_mcp_server/vector/processor.py b/nextcloud_mcp_server/vector/processor.py
index cd2b329b..47138344 100644
--- a/nextcloud_mcp_server/vector/processor.py
+++ b/nextcloud_mcp_server/vector/processor.py
@@ -237,7 +237,15 @@ async def process_document(
 
             for attempt in range(max_retries):
                 try:
-                    await _index_document(doc_task, nc_client, qdrant_client)
+                    indexed = await _index_document(doc_task, nc_client, qdrant_client)
+
+                    # A permanent parse failure returns False: it was already
+                    # recorded (document_parse_failed_total + the registry's
+                    # document_parse_total{error}) and the placeholder marked
+                    # "failed". It is not an indexing event and not retryable, so
+                    # don't count it as a successful upsert/indexed document.
+                    if indexed is False:
+                        return
 
                     # Record successful processing metrics
                     duration = time.time() - start_time
@@ -534,8 +542,8 @@ async def _index_document(
                 # on a pathological PDF) returns success=False rather than
                 # raising -- there is nothing to index and retrying would just
                 # fail again. Mark the placeholder "failed" so the scanner stops
-                # re-queuing it (until the file changes) and return without
-                # indexing empty content.
+                # re-queuing it (until the file changes) and return False so the
+                # caller skips the success metrics (it was not indexed).
                 if not result.success:
                     reason = result.metadata.get("parse_failed_reason", "error")
                     record_document_parse_failed(reason)
@@ -561,7 +569,7 @@ async def _index_document(
                             doc_task.doc_id,
                             exc_info=True,
                         )
-                    return
+                    return False
 
                 content = result.text
                 file_metadata = result.metadata