Replacing psycopg2.copy_expert with psycopg3.copy

awx/main/analytics/collectors.py

@@ -399,7 +399,10 @@ def _copy_table(table, query, path):
     file_path = os.path.join(path, table + '_table.csv')
     file = FileSplitter(filespec=file_path)
     with connection.cursor() as cursor:
-        cursor.copy_expert(query, file)
+        with cursor.copy(query) as copy:
+            while data := copy.read():
+                byte_data = bytes(data)
+                file.write(byte_data.decode())
     return file.file_list()
 
 

awx/main/models/unified_jobs.py

@@ -1137,11 +1137,6 @@ class UnifiedJob(
                     if total > max_supported:
                         raise StdoutMaxBytesExceeded(total, max_supported)
 
-                # psycopg2's copy_expert writes bytes, but callers of this
-                # function assume a str-based fd will be returned; decode
-                # .write() calls on the fly to maintain this interface
-                _write = fd.write
-                fd.write = lambda s: _write(smart_str(s))
                 tbl = self._meta.db_table + 'event'
                 created_by_cond = ''
                 if self.has_unpartitioned_events:
@@ -1150,7 +1145,12 @@ class UnifiedJob(
                     created_by_cond = f"job_created='{self.created.isoformat()}' AND "
 
                 sql = f"copy (select stdout from {tbl} where {created_by_cond}{self.event_parent_key}={self.id} and stdout != '' order by start_line) to stdout"  # nosql
-                cursor.copy_expert(sql, fd)
+                # psycopg3's copy writes bytes, but callers of this
+                # function assume a str-based fd will be returned; decode
+                # .write() calls on the fly to maintain this interface
+                with cursor.copy(sql) as copy:
+                    while data := copy.read():
+                        fd.write(smart_str(bytes(data)))
 
                 if hasattr(fd, 'name'):
                     # If we're dealing with a physical file, use `sed` to clean

awx/main/tests/functional/analytics/test_collectors.py

@@ -2,8 +2,8 @@ import pytest
 import tempfile
 import os
 import re
-import shutil
 import csv
+from io import StringIO
 
 from django.utils.timezone import now
 from datetime import timedelta
@@ -20,15 +20,16 @@ from awx.main.models import (
 )
 
 
-@pytest.fixture
+class MockCopy:
-def sqlite_copy_expert(request):
+    headers = None
-    # copy_expert is postgres-specific, and SQLite doesn't support it; mock its
+    results = None
-    # behavior to test that it writes a file that contains stdout from events
+    sent_data = False
-    path = tempfile.mkdtemp(prefix="copied_tables")
 
-    def write_stdout(self, sql, fd):
+    def __init__(self, sql, parent_connection):
         # Would be cool if we instead properly disected the SQL query and verified
         # it that way. But instead, we just take the naive approach here.
+        self.results = None
+        self.headers = None
         sql = sql.strip()
         assert sql.startswith("COPY (")
         assert sql.endswith(") TO STDOUT WITH CSV HEADER")
@@ -51,29 +52,49 @@ def sqlite_copy_expert(request):
             elif not line.endswith(","):
                 sql_new[-1] = sql_new[-1].rstrip(",")
         sql = "\n".join(sql_new)
+        parent_connection.execute(sql)
+        self.results = parent_connection.fetchall()
+        self.headers = [i[0] for i in parent_connection.description]
 
-        self.execute(sql)
+    def read(self):
-        results = self.fetchall()
+        if not self.sent_data:
-        headers = [i[0] for i in self.description]
+            mem_file = StringIO()
+            csv_handle = csv.writer(
+                mem_file,
+                delimiter=",",
+                quoting=csv.QUOTE_ALL,
+                escapechar="\\",
+                lineterminator="\n",
+            )
+            if self.headers:
+                csv_handle.writerow(self.headers)
+            if self.results:
+                csv_handle.writerows(self.results)
+            self.sent_data = True
+            return memoryview((mem_file.getvalue()).encode())
+        return None
 
-        csv_handle = csv.writer(
+    def __enter__(self):
-            fd,
+        return self
-            delimiter=",",
-            quoting=csv.QUOTE_ALL,
-            escapechar="\\",
-            lineterminator="\n",
-        )
-        csv_handle.writerow(headers)
-        csv_handle.writerows(results)
 
-    setattr(SQLiteCursorWrapper, "copy_expert", write_stdout)
+    def __exit__(self, exc_type, exc_val, exc_tb):
-    request.addfinalizer(lambda: shutil.rmtree(path))
+        pass
-    request.addfinalizer(lambda: delattr(SQLiteCursorWrapper, "copy_expert"))
+
-    return path
+
+@pytest.fixture
+def sqlite_copy(request, mocker):
+    # copy is postgres-specific, and SQLite doesn't support it; mock its
+    # behavior to test that it writes a file that contains stdout from events
+
+    def write_stdout(self, sql):
+        mock_copy = MockCopy(sql, self)
+        return mock_copy
+
+    mocker.patch.object(SQLiteCursorWrapper, 'copy', write_stdout, create=True)
 
 
 @pytest.mark.django_db
-def test_copy_tables_unified_job_query(sqlite_copy_expert, project, inventory, job_template):
+def test_copy_tables_unified_job_query(sqlite_copy, project, inventory, job_template):
     """
     Ensure that various unified job types are in the output of the query.
     """
@@ -127,7 +148,7 @@ def workflow_job(states=["new", "new", "new", "new", "new"]):
 
 
 @pytest.mark.django_db
-def test_copy_tables_workflow_job_node_query(sqlite_copy_expert, workflow_job):
+def test_copy_tables_workflow_job_node_query(sqlite_copy, workflow_job):
     time_start = now() - timedelta(hours=9)
 
     with tempfile.TemporaryDirectory() as tmpdir:

awx/main/tests/functional/api/test_job.py

@@ -224,7 +224,7 @@ class TestControllerNode:
         return AdHocCommand.objects.create(inventory=inventory)
 
     @pytest.mark.django_db
-    def test_field_controller_node_exists(self, sqlite_copy_expert, admin_user, job, project_update, inventory_update, adhoc, get, system_job_factory):
+    def test_field_controller_node_exists(self, sqlite_copy, admin_user, job, project_update, inventory_update, adhoc, get, system_job_factory):
         system_job = system_job_factory()
 
         r = get(reverse('api:unified_job_list') + '?id={}'.format(job.id), admin_user, expect=200)

awx/main/tests/functional/api/test_unified_jobs_stdout.py

@@ -57,7 +57,7 @@ def _mk_inventory_update(created=None):
         [_mk_inventory_update, InventoryUpdateEvent, 'inventory_update', 'api:inventory_update_stdout'],
     ],
 )
-def test_text_stdout(sqlite_copy_expert, Parent, Child, relation, view, get, admin):
+def test_text_stdout(sqlite_copy, Parent, Child, relation, view, get, admin):
     job = Parent()
     job.save()
     for i in range(3):
@@ -79,7 +79,7 @@ def test_text_stdout(sqlite_copy_expert, Parent, Child, relation, view, get, adm
     ],
 )
 @pytest.mark.parametrize('download', [True, False])
-def test_ansi_stdout_filtering(sqlite_copy_expert, Parent, Child, relation, view, download, get, admin):
+def test_ansi_stdout_filtering(sqlite_copy, Parent, Child, relation, view, download, get, admin):
     job = Parent()
     job.save()
     for i in range(3):
@@ -111,7 +111,7 @@ def test_ansi_stdout_filtering(sqlite_copy_expert, Parent, Child, relation, view
         [_mk_inventory_update, InventoryUpdateEvent, 'inventory_update', 'api:inventory_update_stdout'],
     ],
 )
-def test_colorized_html_stdout(sqlite_copy_expert, Parent, Child, relation, view, get, admin):
+def test_colorized_html_stdout(sqlite_copy, Parent, Child, relation, view, get, admin):
     job = Parent()
     job.save()
     for i in range(3):
@@ -134,7 +134,7 @@ def test_colorized_html_stdout(sqlite_copy_expert, Parent, Child, relation, view
         [_mk_inventory_update, InventoryUpdateEvent, 'inventory_update', 'api:inventory_update_stdout'],
     ],
 )
-def test_stdout_line_range(sqlite_copy_expert, Parent, Child, relation, view, get, admin):
+def test_stdout_line_range(sqlite_copy, Parent, Child, relation, view, get, admin):
     job = Parent()
     job.save()
     for i in range(20):
@@ -146,7 +146,7 @@ def test_stdout_line_range(sqlite_copy_expert, Parent, Child, relation, view, ge
 
 
 @pytest.mark.django_db
-def test_text_stdout_from_system_job_events(sqlite_copy_expert, get, admin):
+def test_text_stdout_from_system_job_events(sqlite_copy, get, admin):
     created = tz_now()
     job = SystemJob(created=created)
     job.save()
@@ -158,7 +158,7 @@ def test_text_stdout_from_system_job_events(sqlite_copy_expert, get, admin):
 
 
 @pytest.mark.django_db
-def test_text_stdout_with_max_stdout(sqlite_copy_expert, get, admin):
+def test_text_stdout_with_max_stdout(sqlite_copy, get, admin):
     created = tz_now()
     job = SystemJob(created=created)
     job.save()
@@ -185,7 +185,7 @@ def test_text_stdout_with_max_stdout(sqlite_copy_expert, get, admin):
 )
 @pytest.mark.parametrize('fmt', ['txt', 'ansi'])
 @mock.patch('awx.main.redact.UriCleaner.SENSITIVE_URI_PATTERN', mock.Mock(**{'search.return_value': None}))  # really slow for large strings
-def test_max_bytes_display(sqlite_copy_expert, Parent, Child, relation, view, fmt, get, admin):
+def test_max_bytes_display(sqlite_copy, Parent, Child, relation, view, fmt, get, admin):
     created = tz_now()
     job = Parent(created=created)
     job.save()
@@ -255,7 +255,7 @@ def test_legacy_result_stdout_with_max_bytes(Cls, view, fmt, get, admin):
     ],
 )
 @pytest.mark.parametrize('fmt', ['txt', 'ansi', 'txt_download', 'ansi_download'])
-def test_text_with_unicode_stdout(sqlite_copy_expert, Parent, Child, relation, view, get, admin, fmt):
+def test_text_with_unicode_stdout(sqlite_copy, Parent, Child, relation, view, get, admin, fmt):
     job = Parent()
     job.save()
     for i in range(3):
@@ -267,7 +267,7 @@ def test_text_with_unicode_stdout(sqlite_copy_expert, Parent, Child, relation, v
 
 
 @pytest.mark.django_db
-def test_unicode_with_base64_ansi(sqlite_copy_expert, get, admin):
+def test_unicode_with_base64_ansi(sqlite_copy, get, admin):
     created = tz_now()
     job = Job(created=created)
     job.save()

awx/main/tests/functional/conftest.py

@@ -1,8 +1,6 @@
 # Python
 import pytest
 from unittest import mock
-import tempfile
-import shutil
 import urllib.parse
 from unittest.mock import PropertyMock
 
@@ -789,25 +787,43 @@ def oauth_application(admin):
     return Application.objects.create(name='test app', user=admin, client_type='confidential', authorization_grant_type='password')
 
 
-@pytest.fixture
+class MockCopy:
-def sqlite_copy_expert(request):
+    events = []
-    # copy_expert is postgres-specific, and SQLite doesn't support it; mock its
+    index = -1
-    # behavior to test that it writes a file that contains stdout from events
-    path = tempfile.mkdtemp(prefix='job-event-stdout')
 
-    def write_stdout(self, sql, fd):
+    def __init__(self, sql):
-        # simulate postgres copy_expert support with ORM code
+        self.events = []
         parts = sql.split(' ')
         tablename = parts[parts.index('from') + 1]
         for cls in (JobEvent, AdHocCommandEvent, ProjectUpdateEvent, InventoryUpdateEvent, SystemJobEvent):
             if cls._meta.db_table == tablename:
                 for event in cls.objects.order_by('start_line').all():
-                    fd.write(event.stdout)
+                    self.events.append(event.stdout)
 
-    setattr(SQLiteCursorWrapper, 'copy_expert', write_stdout)
+    def read(self):
-    request.addfinalizer(lambda: shutil.rmtree(path))
+        self.index = self.index + 1
-    request.addfinalizer(lambda: delattr(SQLiteCursorWrapper, 'copy_expert'))
+        if self.index < len(self.events):
-    return path
+            return memoryview(self.events[self.index].encode())
+
+        return None
+
+    def __enter__(self):
+        return self
+
+    def __exit__(self, exc_type, exc_val, exc_tb):
+        pass
+
+
+@pytest.fixture
+def sqlite_copy(request, mocker):
+    # copy is postgres-specific, and SQLite doesn't support it; mock its
+    # behavior to test that it writes a file that contains stdout from events
+
+    def write_stdout(self, sql):
+        mock_copy = MockCopy(sql)
+        return mock_copy
+
+    mocker.patch.object(SQLiteCursorWrapper, 'copy', write_stdout, create=True)
 
 
 @pytest.fixture

awx/main/tests/functional/models/test_notifications.py

@@ -98,7 +98,7 @@ class TestJobNotificationMixin(object):
 
     @pytest.mark.django_db
     @pytest.mark.parametrize('JobClass', [AdHocCommand, InventoryUpdate, Job, ProjectUpdate, SystemJob, WorkflowJob])
-    def test_context(self, JobClass, sqlite_copy_expert, project, inventory_source):
+    def test_context(self, JobClass, sqlite_copy, project, inventory_source):
         """The Jinja context defines all of the fields that can be used by a template. Ensure that the context generated
         for each job type has the expected structure."""
         kwargs = {}

tools/scripts/firehose.py

@@ -98,37 +98,27 @@ class YieldedRows(StringIO):
             )
             self.rowlist.append(row)
 
-    def read(self, x):
-        if self.rows <= 0:
-            self.close()
-            return ''
-        elif self.rows >= 1 and self.rows < 1000:
-            event_rows = self.rowlist[random.randrange(len(self.rowlist))] * self.rows
-            self.rows -= self.rows
-            return event_rows
-        self.rows -= 1000
-        return self.rowlist[random.randrange(len(self.rowlist))] * 1000
-
 
 def firehose(job, count, created_stamp, modified_stamp):
     conn = psycopg.connect(dsn)
     f = YieldedRows(job, count, created_stamp, modified_stamp)
-    with conn.cursor() as cursor:
+    sql = '''
-        cursor.copy_expert(
+          COPY main_jobevent(
-            (
+            created, modified, job_created, event, event_data, failed, changed,
-                'COPY '
+            host_name, play, role, task, counter, host_id, job_id, uuid,
-                'main_jobevent('
+            parent_uuid, end_line, playbook, start_line, stdout, verbosity
-                'created, modified, job_created, event, event_data, failed, changed, '
+          ) FROM STDIN
-                'host_name, play, role, task, counter, host_id, job_id, uuid, '
+    '''
-                'parent_uuid, end_line, playbook, start_line, stdout, verbosity'
+    try:
-                ') '
+        with conn.cursor() as cursor:
-                'FROM STDIN'
+            with cursor.copy(sql) as copy:
-            ),
+                copy.write("".join(f.rowlist))
-            f,
+    except Exception as e:
-            size=1024 * 1000,
+        print("Failed to import events")
-        )
+        print(e)
-    conn.commit()
+    finally:
-    conn.close()
+        conn.commit()
+        conn.close()
 
 
 def cleanup(sql):