From c0a7cdc2dd91c820610ea31c1c37c5e8274f169a Mon Sep 17 00:00:00 2001
From: Nicola Soranzo <nicola.soranzo@earlham.ac.uk>
Date: Fri, 22 Nov 2024 19:18:40 +0000
Subject: [PATCH] Don't calculate dataset hash for datasets in non-OK state

Also:
- Check again that the dataset hasn't been purged right before
  hash calculation starts.

Address comments in
https://github.com/galaxyproject/galaxy/pull/19181/files#r1853605980
---
 lib/galaxy/jobs/__init__.py     | 2 +-
 lib/galaxy/managers/datasets.py | 7 +++++--
 2 files changed, 6 insertions(+), 3 deletions(-)

diff --git a/lib/galaxy/jobs/__init__.py b/lib/galaxy/jobs/__init__.py
index 22a909466b41..9e4d3b65376c 100644
--- a/lib/galaxy/jobs/__init__.py
+++ b/lib/galaxy/jobs/__init__.py
@@ -2042,7 +2042,7 @@ def fail(message=job.info, exception=None):
         # Calculate dataset hash
         for dataset_assoc in output_dataset_associations:
             dataset = dataset_assoc.dataset.dataset
-            if not dataset.purged and dataset.state != Dataset.states.DEFERRED and not dataset.hashes:
+            if not dataset.purged and dataset.state == Dataset.states.OK and not dataset.hashes:
                 if self.app.config.calculate_dataset_hash == "always" or (
                     self.app.config.calculate_dataset_hash == "upload" and job.tool_id in ("upload1", "__DATA_FETCH__")
                 ):
diff --git a/lib/galaxy/managers/datasets.py b/lib/galaxy/managers/datasets.py
index 59137af76e9a..9dd9f87c0638 100644
--- a/lib/galaxy/managers/datasets.py
+++ b/lib/galaxy/managers/datasets.py
@@ -161,8 +161,11 @@ def update_object_store_id(self, trans, dataset, object_store_id: str):
             sa_session.commit()
 
     def compute_hash(self, request: ComputeDatasetHashTaskRequest):
-        # For files in extra_files_path
         dataset = self.by_id(request.dataset_id)
+        if dataset.purged:
+            log.warning("Unable to calculate hash for purged dataset [%s].", dataset.id)
+            return
+        # For files in extra_files_path
         extra_files_path = request.extra_files_path
         if extra_files_path:
             extra_dir = dataset.extra_files_path_name
@@ -192,7 +195,7 @@ def compute_hash(self, request: ComputeDatasetHashTaskRequest):
                     f"Re-calculated dataset hash for dataset [{dataset.id}] and new hash value [{calculated_hash_value}] does not equal previous hash value [{old_hash_value}]."
                 )
             else:
-                log.debug("Duplicated dataset hash request, no update to the database.")
+                log.debug("Duplicated dataset hash request for dataset [%s], no update to the database.", dataset.id)
 
     # TODO: implement above for groups
     # TODO: datatypes?