refactor(pyamber): move tests to src/test/python to match Maven layout

.github/labeler.yml

@@ -46,9 +46,9 @@ agent-service:
 
 engine:
   # Non-Python, non-integration parts of amber/. Pure Python changes
-  # under amber/src/main/python/** intentionally fall through to the
-  # `python` label (which the labeler also matches via **/*.py), so
-  # they only trigger the python + amber-integration stacks rather
+  # under amber/src/{main,test}/python/** intentionally fall through to
+  # the `python` label (which the labeler also matches via **/*.py),
+  # so they only trigger the python + amber-integration stacks rather
   # than the full Scala-only `amber` stack. Integration specs live
   # under amber/src/test/integration/** (added to sbt's Test sources
   # via amber/build.sbt) and are caught by the `amber-integration`
@@ -88,6 +88,8 @@ python:
   - changed-files:
       - any-glob-to-any-file:
           - 'amber/src/main/python/**'
+          - 'amber/src/test/python/**'
+          - 'amber/pyproject.toml'
           - '**/*.py'
           - '**/requirements.txt'
           - '**/*-requirements.txt'

.github/workflows/build.yml

@@ -505,7 +505,7 @@ jobs:
           PGPASSWORD: postgres
       - name: Lint with Ruff
         run: |
-          cd amber/src/main/python && ruff check . && ruff format --check .
+          cd amber && ruff check src/main/python src/test/python && ruff format --check src/main/python src/test/python
       - name: Install dev dependencies
         # Test-only deps live in amber/dev-requirements.txt and are
         # installed after the LICENSE-binary snapshot above so they never
@@ -516,13 +516,13 @@ jobs:
           if [ -f amber/dev-requirements.txt ]; then uv pip install --system -r amber/dev-requirements.txt; fi
       - name: Test with pytest
         run: |
-          cd amber/src/main/python && pytest --cov=. --cov-report=xml -sv
+          cd amber && pytest --cov=src/main/python --cov-report=xml -sv
       - name: Upload python coverage to Codecov
         if: matrix.python-version == '3.12' && always()
         uses: codecov/codecov-action@75cd11691c0faa626561e295848008c8a7dddffe # v5.5.4
         with:
           token: ${{ secrets.CODECOV_TOKEN }}
-          files: ./amber/src/main/python/coverage.xml
+          files: ./amber/coverage.xml
           flags: python
           fail_ci_if_error: false
 

.github/workflows/required-checks.yml

@@ -137,8 +137,8 @@ jobs:
             // amber-integration runs the Scala tests tagged
             // @org.apache.texera.amber.tags.IntegrationTest (e2e specs that
             // spawn Python UDF workers). The labeler attaches `python` to
-            // any *.py change (including amber/src/main/python/**), so
-            // `engine` does not need to fire the python stack itself —
+            // any *.py change (including amber/src/{main,test}/python/**),
+            // so `engine` does not need to fire the python stack itself —
             // pure-Python amber changes pick up `python` directly. The
             // `amber-integration` label catches *IntegrationSpec.scala
             // edits so a test-only change does not trigger the full

.gitignore

@@ -52,6 +52,9 @@ __pycache__/
 *$py.class
 .ipynb_checkpoints
 .pytype/
+.coverage
+coverage.xml
+.pytest_cache/
 
 # Ignoring python generated files
 *.model

amber/build.sbt

@@ -56,10 +56,10 @@ Compile / unmanagedSourceDirectories += baseDirectory.value / "src" / "main" / "
 
 // `amber/src/test/integration` holds Scala specs that exercise both
 // Scala and Python end-to-end (tagged @org.apache.texera.amber.tags.IntegrationTest).
-// Sits next to `src/test/scala` and `src/test/java`; a future `src/test/python`
-// can drop in the same way. Adding it to Test/unmanagedSourceDirectories means
-// scalafmtCheckAll / scalafixAll --check naturally cover these sources, and
-// the AMBER_TEST_FILTER env var below routes which tagged subset runs.
+// Sits next to `src/test/scala`, `src/test/java`, and `src/test/python`.
+// Adding it to Test/unmanagedSourceDirectories means scalafmtCheckAll /
+// scalafixAll --check naturally cover these sources, and the
+// AMBER_TEST_FILTER env var below routes which tagged subset runs.
 Test / unmanagedSourceDirectories += baseDirectory.value / "src" / "test" / "integration"
 
 // Test-filter switch driven by the AMBER_TEST_FILTER env var so the

amber/src/main/python/pyproject.toml → amber/pyproject.toml

@@ -24,4 +24,19 @@ extend-exclude = ["proto"]
 ignore = ["F403", "F405", "E203"]
 
 [tool.ruff.lint.mccabe]
-max-complexity = 10
+max-complexity = 10
+
+# Layout follows the Maven source-set convention used by Scala
+# (`src/main/scala` + `src/test/scala`): production code lives in
+# `src/main/python` and tests in `src/test/python`. Pytest is invoked
+# from `amber/`, so `pythonpath = ["src/main/python"]` keeps
+# `from core.x import y` and `from pytexera.x import y` working.
+[tool.pytest.ini_options]
+pythonpath = ["src/main/python"]
+testpaths = ["src/test/python"]
+# `importlib` import mode loads each test module by file path without
+# inserting it into a parent package, so `src/test/python` doesn't need
+# to mirror `src/main/python`'s __init__.py layout to avoid duplicate
+# package names. Required for src-style test layouts per the pytest
+# docs (https://docs.pytest.org/en/stable/explanation/goodpractices.html).
+addopts = "--import-mode=importlib"

amber/src/main/python/core/proxy/test_proxy_client.py → amber/src/test/python/core/proxy/test_proxy_client.py

@@ -20,8 +20,8 @@
 from pyarrow import ArrowNotImplementedError, Table
 from queue import Queue
 
-from .proxy_client import ProxyClient
-from .proxy_server import ProxyServer
+from core.proxy.proxy_client import ProxyClient
+from core.proxy.proxy_server import ProxyServer
 
 
 class TestProxyClient:

amber/src/main/python/core/proxy/test_proxy_server.py → amber/src/test/python/core/proxy/test_proxy_server.py

@@ -18,7 +18,7 @@
 import pytest
 from pyarrow.flight import Action
 
-from .proxy_server import ProxyServer
+from core.proxy.proxy_server import ProxyServer
 
 
 class TestProxyServer:

amber/src/main/python/core/storage/iceberg/test_iceberg_document.py → amber/src/test/python/core/storage/iceberg/test_iceberg_document.py

@@ -18,6 +18,7 @@
 import datetime
 import pytest
 import random
+import tempfile
 import uuid
 from concurrent.futures import as_completed
 from concurrent.futures.thread import ThreadPoolExecutor
@@ -35,7 +36,11 @@
     PhysicalOpIdentity,
 )
 
-# Hardcoded storage config only for test purposes.
+# Hardcoded storage config only for test purposes. The iceberg warehouse
+# directory must be a writable absolute path; using `tempfile.mkdtemp()`
+# avoids depending on pytest's cwd (an earlier `"../../../../../../amber/
+# user-resources/..."` value silently relied on CWD = amber/src/main/python
+# and broke when the cwd moved up to amber/).
 StorageConfig.initialize(
     catalog_type="postgres",
     postgres_uri_without_scheme="localhost:5432/texera_iceberg_catalog",
@@ -44,7 +49,7 @@
     rest_catalog_uri="http://localhost:8181/catalog/",
     rest_catalog_warehouse_name="texera",
     table_result_namespace="operator-port-result",
-    directory_path="../../../../../../amber/user-resources/workflow-results",
+    directory_path=tempfile.mkdtemp(prefix="texera-iceberg-warehouse-"),
     commit_batch_size=4096,
     s3_endpoint="http://localhost:9000",
     s3_region="us-east-1",

amber/src/main/python/core/util/expression_evaluator/test_expression_evaluator.py → amber/src/test/python/core/util/expression_evaluator/test_expression_evaluator.py

@@ -201,7 +201,7 @@ def __init__(self):
                 expression="a",
                 value_ref="a",
                 value_str=(
-                    "<core.util.expression_evaluator.test_expression_evaluator."
+                    f"<{A.__module__}."
                     "TestExpressionEvaluator.test_evaluate_object_expression.<locals>.A"
                     f" object at {hex(id(a))}>"
                 ),

amber/src/main/python/pytexera/udf/examples/test_count_batch_operator.py → amber/src/test/python/pytexera/udf/examples/test_count_batch_operator.py

@@ -20,7 +20,7 @@
 from collections import deque
 
 from pytexera import *
-from .count_batch_operator import CountBatchOperator
+from pytexera.udf.examples.count_batch_operator import CountBatchOperator
 
 
 class TestCountBatchOperator:

amber/src/main/python/pytexera/udf/examples/test_echo_operator.py → amber/src/test/python/pytexera/udf/examples/test_echo_operator.py

@@ -18,7 +18,7 @@
 import pytest
 
 from pytexera import Tuple
-from .echo_operator import EchoOperator
+from pytexera.udf.examples.echo_operator import EchoOperator
 
 
 class TestEchoOperator:

amber/src/main/python/pytexera/udf/examples/test_echo_table_operator.py → amber/src/test/python/pytexera/udf/examples/test_echo_table_operator.py

@@ -20,7 +20,7 @@
 
 from core.models.table import all_output_to_tuple
 from pytexera import Tuple
-from .echo_table_operator import EchoTableOperator
+from pytexera.udf.examples.echo_table_operator import EchoTableOperator
 
 
 class TestEchoTableOperator:

amber/src/main/python/pytexera/udf/examples/test_generator_operator_binary.py → amber/src/test/python/pytexera/udf/examples/test_generator_operator_binary.py

@@ -18,7 +18,7 @@
 import pytest
 
 from pytexera import Tuple
-from .generator_operator_binary import GeneratorOperatorBinary
+from pytexera.udf.examples.generator_operator_binary import GeneratorOperatorBinary
 
 
 class TestEchoOperator:

amber/src/main/python/pytexera/udf/examples/test_generator_operator_integer.py → amber/src/test/python/pytexera/udf/examples/test_generator_operator_integer.py

@@ -18,7 +18,7 @@
 import pytest
 
 from pytexera import Tuple
-from .generator_operator_integer import GeneratorOperatorInteger
+from pytexera.udf.examples.generator_operator_integer import GeneratorOperatorInteger
 
 
 class TestEchoOperator:

bin/fix-format.sh

@@ -42,10 +42,11 @@ fi
 
 # --- Key directories ---
 FRONTEND_DIR="$TEXERA_HOME/frontend"
-AMBER_PY_DIR="$TEXERA_HOME/amber/src/main/python"
+AMBER_DIR="$TEXERA_HOME/amber"
 
 [[ -d "$FRONTEND_DIR" ]] || { tx_error "Frontend directory not found: $FRONTEND_DIR"; exit 1; }
-[[ -d "$AMBER_PY_DIR"  ]] || { tx_error "Amber Python directory not found: $AMBER_PY_DIR"; exit 1; }
+[[ -d "$AMBER_DIR/src/main/python"  ]] || { tx_error "Amber Python directory not found: $AMBER_DIR/src/main/python"; exit 1; }
+[[ -d "$AMBER_DIR/src/test/python"  ]] || { tx_error "Amber Python test directory not found: $AMBER_DIR/src/test/python"; exit 1; }
 
 # --- Argument parsing ---
 TARGET="${1:-all}"
@@ -95,14 +96,14 @@ fi
 
 # --- 3) Python formatting ---
 if $run_python; then
-  tx_info "Running ruff in amber/src/main/python..."
+  tx_info "Running ruff in amber/src/{main,test}/python..."
   if ! command -v ruff >/dev/null 2>&1; then
     tx_error "ruff not found. Install with: pip install ruff"
     exit 1
   fi
   (
-    cd "$AMBER_PY_DIR"
-    ruff format .
+    cd "$AMBER_DIR"
+    ruff format src/main/python src/test/python
   )
   tx_success "Python formatting completed."
 fi