update script

Author: bachish

benchmarks/scripts/mem_measure_bin_search.py

@@ -0,0 +1,71 @@
+
+import os
+import subprocess
+import sys
+
+cp_path = '/home/olga/gllgen/ucfs/simpleApp/build/install/simpleApp/lib/*'
+files_dir = '/home/olga/gllgen/java7/junit/'
+
+tool = sys.argv[1]
+files_dir = sys.argv[2]
+cp_path = sys.argv[3]
+
+def run_tool_for_file(tool, file_path):
+    def get_cmd(mem):
+        return ['java', '-cp', cp_path, f'-Xmx{mem}m', 'org.ucfs.MainKt', tool, file_path]
+
+
+    cache = {}
+
+    def execute(mem):
+        if mem in cache:
+            return cache[mem]
+
+        cmd = get_cmd(mem)
+        process = subprocess.run(cmd, stdout=subprocess.DEVNULL, stderr=subprocess.DEVNULL)
+        return_code = process.returncode
+        if return_code == 42:
+            print("it return 42")
+            exit(1)
+        cache[mem] = return_code
+        return return_code
+
+    l = 1
+    r = 64
+    max = 4*1024
+    while r <= max:
+        return_code = execute(r)
+        if return_code != 0:
+            l = r
+            r *= 2
+        else:
+            break
+    print(f"calculate r = {r}")
+    if r == 2*max:
+        return r
+
+
+    while l < r - 1:
+        m = (l + r) // 2
+        return_code = execute(m)
+        print(f"for {m} mem get code {return_code}")
+
+        if return_code != 0:
+            l = m
+        else:
+            r = m
+
+    return l
+
+
+files = os.listdir(files_dir)
+
+print(tool)
+with open(f"{tool}_res.txt", "w") as output:
+    output.write(f"file,mem\n")
+    output.flush()
+    for file in files:
+        mem = run_tool_for_file(tool, files_dir + file)
+        print(f"Got for tool = {tool}, file = {file}: {mem}mb")
+        output.write(f"{file},{mem}\n")
+        output.flush()

benchmarks/src/test/plots.py

@@ -0,0 +1,44 @@
+import os
+import csv
+from matplotlib import pyplot as plt
+from pathlib import Path
+
+
+def get_datasets_files(folder: str) -> dict[str, list[str]]:
+    return {
+
+        dataset_folder: [os.path.join(folder, dataset_folder, file) for file in os.listdir(os.path.join(folder, dataset_folder)) if file.endswith(".csv")]
+        for dataset_folder in os.listdir(folder) if os.path.isdir(os.path.join(folder, dataset_folder))
+    }
+
+
+def read_file_data(file: str, column: int) -> list[float]:
+    with open(file) as fin:
+        rdr = csv.DictReader(filter(lambda row: row[0] != "%", fin))
+        return [float(list(line.values())[column]) for line in rdr]
+
+
+def read_data(folder: str, column: int) -> dict[str, dict[str, list[float]]]:
+    datasets_files = get_datasets_files(folder)
+    return {dataset: {Path(file).stem: read_file_data(file, column) for file in files} for dataset, files in datasets_files.items()}
+
+
+def draw_fig(folder: str, column: int, filename: str):
+    dict = read_data(folder, column)
+    tools = set(tool for tools_values in dict.values() for tool in tools_values.keys())
+    all_datasets = {tool: [value for _, tools_values in dict.items() for value in tools_values.get(tool, [])] for tool in tools}
+
+    result = {"All projects": all_datasets}
+    result.update(dict)
+
+    fig, axs = plt.subplots(len(result))
+    for (dataset, values), ax in zip(result.items(), axs):
+        ax.boxplot(values.values(), vert=False, widths=0.6)
+        ax.set_yticklabels(values.keys())
+        ax.set_title(dataset, loc='right')
+    fig.tight_layout()
+    fig.savefig(filename)
+
+
+draw_fig("results", 1, "time.png")
+# draw_fig(2, "memory.png")

benchmarks/tokens_count.csv

@@ -1,3 +1,4 @@
+filename,tokens
 ActiveTestSuite.java,242
 ActiveTestTest.java,376
 After.java,62