Fix tests and imports

Makefile

@@ -27,7 +27,7 @@ format-python:
 	black ${PYTHON_FILES}
 
 python-test:
-	PYTHONPATH="${MAKEFILE_DIR}/src/python/main" pytest src/python/test
+	pytest src/python/test
 
 shell-test:
 	bats "${MAKEFILE_DIR}/src/bash/test"

README.md

@@ -108,6 +108,10 @@ The detailed pipeline can be visualized in [diagrams/pipeline.drawio.svg](diagra
 In addition to `data/d4j-5-bugs.csv`, there are 1 other pre-computed bug file that you can use: `data/d4j-compatible-bugs.csv` contains all the Defects4J bugs that are active and compatible with the experimental infrastructure. Bug projects marked as deprecated by Defects4J's authors are not included in this list of compatible bugs.
 To generate a new bug file, see **Generating the bug file** section.
 
+#### Running Python scripts
+Pythons scripts are in the same directory structure as the python modules, making testing and reusing modules difficult.
+The solution is to run python scripts as module (`python -m`) so the scripts can import other modules directly and be compatible with tests.  
+
 #### Optional steps
 Run `./compute_metrics.sh data/d4j-5-bugs.csv $UTB_OUTPUT` to compute the metrics of the D4J bugs. See section [Metrics](#metrics) for more details.
 
@@ -120,8 +124,6 @@ Do not use `data/d4j-bugs-all.csv` as it contains bugs that are deprecated or no
 ### Untangling one Defects4J bug
 If you only want to evaluate the decomposition of one Defects4J bug, you can follow the pipeline presented in [diagrams/pipeline.drawio.svg](diagrams/pipeline.drawio.svg).
 
-### Aggregating decomposition elapsed time
-
 ## Tests
 - Run `make check` to run all the checks (tests, linting, etc.) for bash and Python.
 

pyproject.toml

@@ -0,0 +1,10 @@
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[project]
+name = "untanglingevaluation"
+version = "1.0"
+
+[pytest]
+python_paths = "src/python/main;src/python/test"

src/python/main/analysis/print_performance.py

@@ -19,9 +19,8 @@
 
 import pandas as pd
 
-sys.path.insert(1, os.path.join(sys.path[0], '..'))
-import latex_utils
-import evaluation_results
+from src.python.main.analysis import latex_utils
+from src.python.main import evaluation_results
 
 
 def main(d4j_file:str, lltc4j_file:str, aggregator:str, overall:bool):
@@ -82,7 +81,7 @@ def print_performance_commands(df_performance, aggregator_operation):
     """
     for dataset in df_performance.index:
         for tool in df_performance.columns:
-            value = df_performance.loc[dataset, tool].round(PRECISION)
+            value = df_performance.loc[dataset, tool].round(latex_utils.PRECISION)
             dataset_name_for_latex = dataset.lower().replace('4', 'f')
             tool_name_for_latex = tool.capitalize().replace('-', '')
             aggreator_for_latex = aggregator_operation.capitalize()

src/python/main/evaluation_results.py

@@ -7,7 +7,7 @@
 
 import pandas as pd
 
-import tangled_metrics
+from . import tangled_metrics
 
 from pandas.api.types import CategoricalDtype
 

src/python/main/ground_truth.py

@@ -29,7 +29,7 @@
 import pandas as pd
 from unidiff import PatchSet, LINE_TYPE_CONTEXT
 
-from diff_metrics import lines_in_patch
+from .diff_metrics import lines_in_patch
 
 COL_NAMES = ["file", "source", "target"]
 

src/python/main/tangled_metrics.py

@@ -11,8 +11,8 @@
 import pandas as pd
 import numpy as np
 
-import metrics
-import evaluation_results
+from . import metrics
+from . import evaluation_results
 from pandas.api.types import CategoricalDtype
 
 TANGLED_LEVELS = [

src/python/test/analysis/test_print_performance.py

@@ -3,7 +3,7 @@
 """
 import pytest
 
-import src.python.main.analysis.print_performance as print_median_performance
+import src.python.main.analysis.print_performance as print_performance
 
 
 @pytest.fixture
@@ -39,7 +39,7 @@ def test_calculate_performance(sample_d4j_scores, sample_lltc4j_scores, capfd):
     """
     Tests that the performance metrics are calculated correctly.
     """
-    print_median_performance.main(sample_d4j_scores, sample_lltc4j_scores, "median", False)
+    print_performance.main(sample_d4j_scores, sample_lltc4j_scores, "median", False)
 
     captured = capfd.readouterr()
 
@@ -72,7 +72,7 @@ def test_calculate_performance(sample_d4j_scores, sample_lltc4j_scores, capfd):
     """
     Tests that the performance metrics are calculated correctly.
     """
-    print_median_performance.main(sample_d4j_scores, sample_lltc4j_scores, "mean", False)
+    print_performance.main(sample_d4j_scores, sample_lltc4j_scores, "mean", False)
 
     captured = capfd.readouterr()
 
@@ -106,7 +106,7 @@ def test_calculate_performance_overall(sample_d4j_scores, sample_lltc4j_scores,
     """
     Tests that the performance metrics are calculated correctly.
     """
-    print_median_performance.main(sample_d4j_scores, sample_lltc4j_scores, "mean", True)
+    print_performance.main(sample_d4j_scores, sample_lltc4j_scores, "mean", True)
 
     captured = capfd.readouterr()