pyneoinstance 4.0.2__tar.gz → 4.0.4__tar.gz

{pyneoinstance-4.0.2/pyneoinstance.egg-info → pyneoinstance-4.0.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pyneoinstance
-Version: 4.0.2
+Version: 4.0.4
 Summary: Handle Neo4j read and write transactions and provide methods for performing graph EDA
 Author-email: "Jose A. Alvarado-Guzman" <jose.alvarado-guzman@neo4j.com>
 License: GNU GENERAL PUBLIC LICENSE

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4}/pyneoinstance/__init__.py

@@ -1,5 +1,5 @@
 from .database.neo4jdbms import Neo4jInstance
 from .fileload.loadyaml import load_yaml_file
 
-__version__ = '4.0.2'
+__version__ = '4.0.4'
 __all__ = ['Neo4jInstance', 'load_yaml_file']

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4}/pyneoinstance/database/neo4jdbms.py

@@ -15,12 +15,15 @@ from concurrent.futures import ThreadPoolExecutor
 import logging
 import os
 import re
+import time
 from pandas import DataFrame
 from neo4j import GraphDatabase
 from pyvis.network import Network
 from neo4j_viz import VisualizationGraph
 from neo4j_viz.neo4j import from_neo4j
 from neo4j.exceptions import ServiceUnavailable
+from neo4j.exceptions import SessionExpired
+from neo4j.exceptions import TransientError
 from neo4j.exceptions import AuthError
 from neo4j.exceptions import ClientError
 from neo4j.exceptions import ConfigurationError
@@ -28,7 +31,58 @@ from .context import get_logger, get_batches, get_columns_diff
 
 _NODE_LABEL_RE = re.compile(r"\(:?(\w*)\)")
 _REL_TYPE_RE = re.compile(r"\[:?(\w*)\]")
-_IMPLICIT_TX_RE = re.compile(r"\bIN\s+TRANSACTIONS\b", re.IGNORECASE)
+_IMPLICIT_TX_RE = re.compile(
+    r"\bIN\s+(?:\d+\s+)?CONCURRENT\s+TRANSACTIONS\b"
+    r"|\bIN\s+TRANSACTIONS\b",
+    re.IGNORECASE,
+)
+
+# Errors treated as transient/retryable for auto-commit queries (run via
+# session.run). session.execute_read/execute_write retry these automatically
+# through the driver's managed-transaction logic; session.run does not.
+_RETRYABLE_EXCEPTIONS = (ServiceUnavailable, SessionExpired, TransientError)
+
+
+def _run_autocommit_with_retry(session, consume, query, max_retries=5,
+                                initial_delay=1.0, backoff_factor=2.0, **kwargs):
+    """Run an auto-commit query (via session.run) with manual retry.
+
+    Queries using ``CALL { ... } IN TRANSACTIONS`` or
+    ``IN CONCURRENT TRANSACTIONS`` are only valid in an implicit
+    (auto-commit) transaction, so they can't be executed through
+    session.execute_read/execute_write and don't benefit from the driver's
+    built-in managed-transaction retry behavior. This re-runs the query and
+    re-consumes its result from scratch on transient failures such as
+    ServiceUnavailable.
+
+    Parameters
+    ----------
+    session : neo4j.Session
+        The session to run the query on.
+    consume : Callable[[neo4j.Result], Any]
+        Function that consumes the Result object (e.g. reads values, builds
+        a graph, reads counters) and returns the desired output.
+    query : str
+        The Cypher query to run.
+    max_retries : int, optional
+        Maximum number of retry attempts after the initial try.
+    initial_delay : float, optional
+        Delay in seconds before the first retry.
+    backoff_factor : float, optional
+        Multiplier applied to the delay after each subsequent retry.
+    kwargs
+        Cypher query parameters.
+    """
+    delay = initial_delay
+    for attempt in range(max_retries + 1):
+        try:
+            result = session.run(query, **kwargs)
+            return consume(result)
+        except _RETRYABLE_EXCEPTIONS:
+            if attempt == max_retries:
+                raise
+            time.sleep(delay)
+            delay *= backoff_factor
 
 def _get_driver(neo_info : Dict[str, str]):
     try:
@@ -197,8 +251,17 @@ class Neo4jInstance:
         params = parameters or {}
         with _get_session(self._driver, database) as session:
             try:
-                result = session.execute_read(
-                    _read_transaction_function,query=query,**params)
+                if _IMPLICIT_TX_RE.search(query):
+                    # 'CALL { ... } IN TRANSACTIONS' is only valid in an implicit
+                    # (auto-commit) transaction, so it can't go through execute_read.
+                    # session.run isn't retried by the driver, so retry manually.
+                    result = _run_autocommit_with_retry(
+                        session,
+                        lambda r: DataFrame(r.values(), columns=r.keys()),
+                        query, **params)
+                else:
+                    result = session.execute_read(
+                        _read_transaction_function,query=query,**params)
             except ServiceUnavailable as exception:
                 raise ServiceUnavailable() from exception
             except ClientError as exception:
@@ -243,8 +306,18 @@ class Neo4jInstance:
         params = parameters or {}
         with _get_session(self._driver, database) as session:
             try:
-                graph = session.execute_read(
-                    _graph_transaction_function, query=query, **params)
+                if _IMPLICIT_TX_RE.search(query):
+                    # 'CALL { ... } IN TRANSACTIONS' is only valid in an implicit
+                    # (auto-commit) transaction, so it can't go through execute_read.
+                    # session.run isn't retried by the driver, so retry manually.
+                    def _consume_graph(result):
+                        list(result)  # consume records so .graph() is populated
+                        return result.graph()
+                    graph = _run_autocommit_with_retry(
+                        session, _consume_graph, query, **params)
+                else:
+                    graph = session.execute_read(
+                        _graph_transaction_function, query=query, **params)
             except ServiceUnavailable as exception:
                 raise ServiceUnavailable() from exception
             except ClientError as exception:
@@ -758,7 +831,10 @@ class Neo4jInstance:
             if _IMPLICIT_TX_RE.search(query):
                 # 'CALL { ... } IN TRANSACTIONS' is only valid in an implicit
                 # (auto-commit) transaction, so it can't go through execute_write.
-                results = session.run(query, **kwargs).consume().counters.__dict__
+                # session.run isn't retried by the driver, so retry manually.
+                results = _run_autocommit_with_retry(
+                    session, lambda r: r.consume().counters.__dict__,
+                    query, **kwargs)
             else:
                 results = session.execute_write(
                     _write_transaction_function, query,

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4/pyneoinstance.egg-info}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: pyneoinstance
-Version: 4.0.2
+Version: 4.0.4
 Summary: Handle Neo4j read and write transactions and provide methods for performing graph EDA
 Author-email: "Jose A. Alvarado-Guzman" <jose.alvarado-guzman@neo4j.com>
 License: GNU GENERAL PUBLIC LICENSE

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4}/pyneoinstance.egg-info/SOURCES.txt

@@ -18,9 +18,7 @@ pyneoinstance/fileload/helpers.py
 pyneoinstance/fileload/loadyaml.py
 pyneoinstance/util/__init__.py
 pyneoinstance/util/functions.py
-test/__init__.py
 test/age_data.csv
-test/context.py
 test/cypher_queries.yaml
 test/load_files.yaml
 test/movies.csv

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4}/pyproject.toml

@@ -6,7 +6,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "pyneoinstance"
-version = "4.0.2"
+version = "4.0.4"
 description = "Handle Neo4j read and write transactions and provide methods for performing graph EDA"
 readme = "README.md"
 authors = [{ name = "Jose A. Alvarado-Guzman",email = "jose.alvarado-guzman@neo4j.com" }]

{pyneoinstance-4.0.2 → pyneoinstance-4.0.4}/test/test_neo4jdb.py

@@ -10,10 +10,14 @@
 """
 import os
 import unittest
+from contextlib import contextmanager
+from unittest.mock import patch
+import neo4j
 import pandas as pd
 import numpy as np
 import pandas.testing as pd_testing
 from neo4j.exceptions import ClientError
+from neo4j.exceptions import ServiceUnavailable
 from neo4j_viz import VisualizationGraph
 from .context import database
 from .context import fileload
@@ -40,6 +44,32 @@ def _clean_db(graph, db):
         ["CALL apoc.schema.assert({},{})", "MATCH(n) DETACH DELETE n"], db)
 
 
+@contextmanager
+def _patch_flaky_session_run(fail_times=1):
+    """Patch neo4j.Session.run so the first `fail_times` calls raise a
+    transient ServiceUnavailable, then delegate to the real driver.
+
+    Used to verify that 'IN TRANSACTIONS'/'IN CONCURRENT TRANSACTIONS'
+    queries (which go through session.run rather than execute_read/
+    execute_write, and so don't get the driver's built-in managed-
+    transaction retries) are retried by the manual retry wrapper instead
+    of failing outright. Also patches time.sleep so the retry backoff
+    doesn't slow the test down.
+    """
+    original_run = neo4j.Session.run
+    call_count = {'n': 0}
+
+    def flaky_run(session_self, *args, **kwargs):
+        call_count['n'] += 1
+        if call_count['n'] <= fail_times:
+            raise ServiceUnavailable('simulated transient failure')
+        return original_run(session_self, *args, **kwargs)
+
+    with patch.object(neo4j.Session, 'run', flaky_run), \
+         patch.object(database.neo4jdbms.time, 'sleep', return_value=None):
+        yield call_count
+
+
 class TestWriteMethods(unittest.TestCase):
     """Test write operations. Each test gets a clean database."""
 
@@ -111,6 +141,38 @@ class TestWriteMethods(unittest.TestCase):
         self.assertEqual(
             {'nodes_created': 3, 'labels_added': 3, 'properties_set': 3}, result)
 
+    def test_execute_write_query_in_concurrent_transactions(self):
+        """Test a query with 'CALL { ... } IN CONCURRENT TRANSACTIONS', which
+        also requires an implicit transaction rather than the managed
+        transaction used by execute_write_query for regular queries."""
+        query = """
+            UNWIND range(1, 3) AS i
+            CALL (i) {
+                CREATE (:Actor {id: i})
+            } IN 2 CONCURRENT TRANSACTIONS OF 1 ROWS
+        """
+        result = self.graph.execute_write_query(query, self.db)
+        self.assertEqual(
+            {'nodes_created': 3, 'labels_added': 3, 'properties_set': 3}, result)
+
+    def test_execute_write_query_in_transactions_retries_on_transient_failure(self):
+        """An 'IN TRANSACTIONS' write query is retried and still succeeds
+        after a transient ServiceUnavailable failure, since it goes through
+        session.run (not execute_write) and isn't covered by the driver's
+        built-in managed-transaction retries."""
+        query = """
+            UNWIND range(1, 3) AS i
+            CALL (i) {
+                CREATE (:Actor {id: i})
+            } IN TRANSACTIONS OF 1 ROWS
+        """
+        with _patch_flaky_session_run(fail_times=1) as call_count:
+            result = self.graph.execute_write_query(query, self.db)
+
+        self.assertEqual(call_count['n'], 2)
+        self.assertEqual(
+            {'nodes_created': 3, 'labels_added': 3, 'properties_set': 3}, result)
+
 
 class TestReadAndEdaMethods(unittest.TestCase):
     """Test read and EDA methods. Data is loaded once for the entire class."""
@@ -173,6 +235,42 @@ class TestReadAndEdaMethods(unittest.TestCase):
             query, self.db, {'movie_name': movie})
         self.assertEqual(movie, result['movie'][0])
 
+    def test_execute_read_query_in_transactions(self):
+        """Test a read query with 'CALL { ... } IN TRANSACTIONS', which requires
+        an implicit transaction rather than the managed transaction used by
+        execute_read_query for regular queries."""
+        plain_count = self.graph.execute_read_query(
+            "MATCH (p:Person) RETURN count(p) AS row_num", self.db)['row_num'][0]
+        query = """
+            MATCH (p:Person)
+            CALL (p) {
+                RETURN p AS person
+            } IN TRANSACTIONS OF 1000 ROWS
+            RETURN count(person) AS row_num
+        """
+        result = self.graph.execute_read_query(query, self.db)
+        self.assertEqual(plain_count, result['row_num'][0])
+
+    def test_execute_read_query_in_transactions_retries_on_transient_failure(self):
+        """An 'IN TRANSACTIONS' read query is retried and still succeeds
+        after a transient ServiceUnavailable failure, since it goes through
+        session.run (not execute_read) and isn't covered by the driver's
+        built-in managed-transaction retries."""
+        plain_count = self.graph.execute_read_query(
+            "MATCH (p:Person) RETURN count(p) AS row_num", self.db)['row_num'][0]
+        query = """
+            MATCH (p:Person)
+            CALL (p) {
+                RETURN p AS person
+            } IN TRANSACTIONS OF 1000 ROWS
+            RETURN count(person) AS row_num
+        """
+        with _patch_flaky_session_run(fail_times=1) as call_count:
+            result = self.graph.execute_read_query(query, self.db)
+
+        self.assertEqual(call_count['n'], 2)
+        self.assertEqual(plain_count, result['row_num'][0])
+
     def test_get_node_label_freq(self):
         """Test node label frequency EDA."""
         result = self.graph.get_node_label_freq(self.db)

pyneoinstance-4.0.2/test/__init__.py

@@ -1,3 +0,0 @@
-import os
-
-test_dir = os.path.abspath(os.path.dirname(__file__))

pyneoinstance-4.0.2/test/context.py

@@ -1,8 +0,0 @@
-"""Re-exports package modules for use by the test suite."""
-import os
-import sys
-sys.path.insert(0, os.path.abspath(
-    os.path.join(os.path.dirname(__file__), '..')))
-from pyneoinstance import fileload
-from pyneoinstance import util
-from pyneoinstance import database