batch-analytics 0.2.7__tar.gz → 0.2.9__tar.gz

{batch_analytics-0.2.7 → batch_analytics-0.2.9}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: batch-analytics
-Version: 0.2.7
+Version: 0.2.9
 Summary: PySpark batch analytics: Extract, Transform, Stage, and analytical modules (linear regression, correlation, PCA, t-test).
 Author: Litewave Analytics Team
 License: MIT
@@ -15,14 +15,17 @@ Requires-Dist: scipy>=1.5.0; extra == "ttest"
 Provides-Extra: s3
 Requires-Dist: boto3>=1.28; extra == "s3"
 Provides-Extra: clickhouse
-Requires-Dist: clickhouse-connect>=0.7; extra == "clickhouse"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "clickhouse"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "clickhouse"
 Provides-Extra: output
 Requires-Dist: boto3>=1.28; extra == "output"
-Requires-Dist: clickhouse-connect>=0.7; extra == "output"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "output"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "output"
 Provides-Extra: full
 Requires-Dist: scipy>=1.5.0; extra == "full"
 Requires-Dist: boto3>=1.28; extra == "full"
-Requires-Dist: clickhouse-connect>=0.7; extra == "full"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "full"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "full"
 
 # Batch Analytics
 

{batch_analytics-0.2.7 → batch_analytics-0.2.9}/pyproject.toml

@@ -4,7 +4,7 @@ build-backend = "setuptools.build_meta"
 
 [project]
 name = "batch-analytics"
-version = "0.2.7"
+version = "0.2.9"
 description = "PySpark batch analytics: Extract, Transform, Stage, and analytical modules (linear regression, correlation, PCA, t-test)."
 readme = "README.md"
 requires-python = ">=3.8"
@@ -20,10 +20,23 @@ dev = ["pytest>=7.0"]
 # Welch t-test module (batch_analytics.analytics.t_test)
 ttest = ["scipy>=1.5.0"]
 s3 = ["boto3>=1.28"]
-clickhouse = ["clickhouse-connect>=0.7"]
-output = ["boto3>=1.28", "clickhouse-connect>=0.7"]
+# 0.9+ uses list[...] etc. and breaks on Python 3.8; 3.9+ can take current clickhouse-connect.
+clickhouse = [
+    "clickhouse-connect>=0.7,<0.9; python_version < '3.9'",
+    "clickhouse-connect>=0.7; python_version >= '3.9'",
+]
+output = [
+    "boto3>=1.28",
+    "clickhouse-connect>=0.7,<0.9; python_version < '3.9'",
+    "clickhouse-connect>=0.7; python_version >= '3.9'",
+]
 # Install all optional runtime deps used anywhere in the package
-full = ["scipy>=1.5.0", "boto3>=1.28", "clickhouse-connect>=0.7"]
+full = [
+    "scipy>=1.5.0",
+    "boto3>=1.28",
+    "clickhouse-connect>=0.7,<0.9; python_version < '3.9'",
+    "clickhouse-connect>=0.7; python_version >= '3.9'",
+]
 
 [project.scripts]
 batch-analytics = "batch_analytics.job_runner:main"

{batch_analytics-0.2.7 → batch_analytics-0.2.9}/src/batch_analytics/job_runner.py

@@ -67,6 +67,30 @@ def create_spark_session(
     if cfg.master.startswith("k8s://"):
         packages.append("org.apache.hadoop:hadoop-aws:3.3.4")
 
+    # Spark-on-K8s: absolute paths in spark.jars are re-sent to executors as ./basename.jar and
+    # fail there ("Unable to create executor due to ./clickhouse-spark-runtime-..."). JARs baked
+    # into the driver/executor image belong on the JVM classpath instead.
+    if cfg.master.startswith("k8s://") and jar_list:
+        local_cp: List[str] = []
+        remote_jars: List[str] = []
+        for p in jar_list:
+            p = p.strip()
+            if not p:
+                continue
+            if p.startswith("/") and p.endswith(".jar"):
+                local_cp.append(p)
+            else:
+                remote_jars.append(p)
+        if local_cp:
+            joined = ":".join(local_cp)
+            builder = (
+                builder.config("spark.driver.extraClassPath", joined)
+                .config("spark.executor.extraClassPath", joined)
+                .config("spark.kubernetes.driver.extraClassPath", joined)
+                .config("spark.kubernetes.executor.extraClassPath", joined)
+            )
+        jar_list = remote_jars
+
     if jar_list:
         builder = builder.config("spark.jars", ",".join(jar_list))
     if packages:

{batch_analytics-0.2.7 → batch_analytics-0.2.9}/src/batch_analytics.egg-info/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.4
 Name: batch-analytics
-Version: 0.2.7
+Version: 0.2.9
 Summary: PySpark batch analytics: Extract, Transform, Stage, and analytical modules (linear regression, correlation, PCA, t-test).
 Author: Litewave Analytics Team
 License: MIT
@@ -15,14 +15,17 @@ Requires-Dist: scipy>=1.5.0; extra == "ttest"
 Provides-Extra: s3
 Requires-Dist: boto3>=1.28; extra == "s3"
 Provides-Extra: clickhouse
-Requires-Dist: clickhouse-connect>=0.7; extra == "clickhouse"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "clickhouse"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "clickhouse"
 Provides-Extra: output
 Requires-Dist: boto3>=1.28; extra == "output"
-Requires-Dist: clickhouse-connect>=0.7; extra == "output"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "output"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "output"
 Provides-Extra: full
 Requires-Dist: scipy>=1.5.0; extra == "full"
 Requires-Dist: boto3>=1.28; extra == "full"
-Requires-Dist: clickhouse-connect>=0.7; extra == "full"
+Requires-Dist: clickhouse-connect<0.9,>=0.7; python_version < "3.9" and extra == "full"
+Requires-Dist: clickhouse-connect>=0.7; python_version >= "3.9" and extra == "full"
 
 # Batch Analytics
 

batch_analytics-0.2.9/src/batch_analytics.egg-info/requires.txt

@@ -0,0 +1,38 @@
+pyspark<3.6,>=3.4
+numpy>=1.19.0
+
+[clickhouse]
+
+[clickhouse:python_version < "3.9"]
+clickhouse-connect<0.9,>=0.7
+
+[clickhouse:python_version >= "3.9"]
+clickhouse-connect>=0.7
+
+[dev]
+pytest>=7.0
+
+[full]
+scipy>=1.5.0
+boto3>=1.28
+
+[full:python_version < "3.9"]
+clickhouse-connect<0.9,>=0.7
+
+[full:python_version >= "3.9"]
+clickhouse-connect>=0.7
+
+[output]
+boto3>=1.28
+
+[output:python_version < "3.9"]
+clickhouse-connect<0.9,>=0.7
+
+[output:python_version >= "3.9"]
+clickhouse-connect>=0.7
+
+[s3]
+boto3>=1.28
+
+[ttest]
+scipy>=1.5.0

batch_analytics-0.2.7/src/batch_analytics.egg-info/requires.txt

@@ -1,23 +0,0 @@
-pyspark<3.6,>=3.4
-numpy>=1.19.0
-
-[clickhouse]
-clickhouse-connect>=0.7
-
-[dev]
-pytest>=7.0
-
-[full]
-scipy>=1.5.0
-boto3>=1.28
-clickhouse-connect>=0.7
-
-[output]
-boto3>=1.28
-clickhouse-connect>=0.7
-
-[s3]
-boto3>=1.28
-
-[ttest]
-scipy>=1.5.0