From a1245151d3a4d3965f04560d5b6ec420ec4c1cec Mon Sep 17 00:00:00 2001
From: Damien De Paoli <ddp@depaoli.id.au>
Date: Sun, 17 Jul 2022 15:32:59 +1000
Subject: [PATCH] when restart JobRunAIOnPath, dont add the path again, move a
 couple of debug lines to an actual debug

---
 pa_job_manager.py | 16 +++++++++++++---
 1 file changed, 13 insertions(+), 3 deletions(-)

diff --git a/pa_job_manager.py b/pa_job_manager.py
index 30fa2ee..338247d 100644
--- a/pa_job_manager.py
+++ b/pa_job_manager.py
@@ -745,7 +745,7 @@ def HandleJobs(first_run=False):
         print("INFO: PA job manager is starting up - check for stale jobs" )
     else:
         if DEBUG:
-            print("INFO: PA job manager is scanning for new jobs to process")
+            print("DEBUG: PA job manager is scanning for new jobs to process")
     for job in session.query(Job).filter(Job.pa_job_state != 'Complete').all():
         if first_run and job.pa_job_state == 'In Progress':
             print( f"INFO: Found stale job#{job.id} - {job.name}" )
@@ -1482,7 +1482,16 @@ def JobRunAIOnPath(job):
     for p in paths:
         d = session.query(Dir).join(PathDirLink).filter(PathDirLink.path_id==p.id).filter(Dir.rel_path=='').first()
         DelMatchesForDir( job, d.eid )
-        job.extra.append( JobExtra( name=f"eid-{path_cnt}", value=f"{d.eid}" ) )
+        # small chance we are restarting a job (and this is the one that is likely to be 'stale' and restarted, so accommodate 
+        # this by not adding it twice -- only really throws count out and makes it 're-process', but if this say the storage path,
+        # that can be another 10s of thousands of files to re-AI over...
+        already_there=False
+        for ex in job.extra:
+            if ex.name == f"eid-{path_cnt}":
+                already_there=True
+                break
+        if not already_there:
+            job.extra.append( JobExtra( name=f"eid-{path_cnt}", value=f"{d.eid}" ) )
         path_cnt+=1
     JobRunAIOn(job)
     return
@@ -2067,7 +2076,8 @@ def ScanFileForPerson( job, e, force=False ):
     threshold = settings.default_threshold
 
     # add log, set current_file and increment file_num in job
-    ProcessFileForJob( job, f'INFO: processing File: {e.name} and threshold face distance of {threshold}', e.name )
+    if DEBUG:
+        ProcessFileForJob( job, f'DEBUG: processing File: {e.name} and threshold face distance of {threshold}', e.name )
     file_h = session.query(File).get( e.id )
     # if we are forcing this, delete any old faces (this will also delete linked tables), and reset faces_created_on to None
     if force: