#!/usr/bin/env python3 """ The purpose of this script to convert the results of the experiments into latex or html tables. The format of the results is as follows: - root_dir - DataStructureA_1.out - DataStructureA_2.out - ... - DataStructureB_1.out - DataStructureB_2.out - ... Files of the form DataStructureName_i.dir contain in the last two lines the outcome of the test and the time taken to reach the end. A data structure that passes will have lines of the following structure: N schedules generated. All tests pass. Time (seconds) = 0.000 A data structure that fails will have lines of the following structure: AMPVer: error detected on schedule x. Exiting. Time (seconds) = 0.000 """ import os import re from typing import Dict, List, Tuple, Optional import argparse def find_source_path(root_dir: os.PathLike, fname: str) -> str: """ Find the path to the source file that was used to generate the given output file. """ fname_base = os.path.splitext(fname)[0] # Find all the source files try: with open(os.path.join(root_dir, fname_base + '.dir', 'schedule_0.out'), 'r', encoding='UTF-8') as f: adding_files = False candidate_files = [] for line in f: if 'Source files' in line: adding_files = True elif adding_files: if line.strip() == '': break candidate_files.append(line.strip()) except FileNotFoundError: return 'No file' # Find the source file with name that is prefix of fname_base for candidate in candidate_files: if fname_base.startswith(candidate.split('.')[0]): return candidate.split('(')[1].split(')')[0] return 'No file' def get_fields_from_lines(lines: List[str], fname: os.PathLike) \ -> Tuple[int, int, bool]: """ Retrieve the time, schedules, and passed values from the list of lines. """ time = None schedules = None passed = None for line in lines: if line.startswith("Time (seconds)") and not time: time = int(round(float(line.split('=')[1].strip()))) elif "All tests pass." in line: passed = True if not schedules: schedules = int(re.findall(r'\d+', line)[0]) elif line.startswith("AMPVer: error"): passed = False if not schedules: schedules = int(re.findall(r'\d+', line)[0]) if passed is not None and time and schedules: break if not time: raise ValueError(f"No time recorded in file {fname}") if not schedules: raise ValueError(f"No schedule number in file {fname}") if passed is None: raise ValueError(f"Could not find pass or fail status in {fname}") return time, schedules, passed def parse_results(root_dir: os.PathLike, verbose: Optional[bool]) -> \ Tuple[Dict[str, Dict[str, Tuple[int, bool, int]]], \ List[str], \ Dict[str, str]]: """ Parse the results of the experiments in the given directory, and return a dictionary with the data structures as keys and a list of times as values as well as a set of test names. """ data_structures = {} tests = set() source_paths = {} for fname in os.listdir(root_dir): # Skipping schedule outputs for now if fname.endswith('.out') and not "_S" in fname: if verbose: print(f"Processing {fname}") data_structure, test = fname.split('_') test = str(test.split('.')[0]) with open(os.path.join(root_dir, fname), 'r', encoding='UTF-8') as f: lines = f.readlines() # Extract fields time, schedules, passed = get_fields_from_lines(lines, fname) # Populate source paths and tests if needed if data_structure not in source_paths \ or source_paths[data_structure] == 'No file': source_paths[data_structure] = find_source_path(root_dir, fname) tests.add(test) # Store the results if data_structure not in data_structures: data_structures[data_structure] = {} data_structures[data_structure][test] = (time, passed, schedules) if verbose: print("Found the following", len(tests), "tests:", tests) print("Found the following", len(data_structures), "data structures:", data_structures.keys()) return dict(sorted(data_structures.items())), list(sorted(tests)), \ source_paths def generate_latex(data_structures: Dict[str, Dict[str, Tuple[int, bool, int]]], tests: List[str]) -> str: """ Generate a latex table from the given data structures and tests. """ table = '\\begin{table}[ht]\n' table += '\\small\n' table += '\\centering\n' table += '\\begin{tabular}{ l ' + ' r' * len(tests) + ' }\n' table += ' Data\\ Structure & ' + ' & '.join(tests) + ' \\\\\n' table += ' \\toprule\n' for data_structure, results in data_structures.items(): table += ' \\code{' + data_structure + '} & ' for test in tests: if test in results: if results[test][1]: table += str(results[test][0]) else: table += '\\textcolor{red}{' + str(results[test][0]) + '}' else: table += '--' table += ' & ' table = table[:-2] + ' \\\\\n' table += '\\end{tabular}\n' table += '\\caption{Results of the experiments.}\\label{tab:results}\n' table += '\\end{table}\n' return table def get_tooltip(test: str) -> str: """ Get the tooltip for the given test name. """ t = 'No tooltip available.' if test == 'A': t = \ """BOUND_A: Pre-adds: 0 Threads: 1-2 Steps: 1-2 Hashcode: identity Set/List schedules: 63 Queue schedules: 9 SyncQueue schedules: 9 PQueue schedules: 7""" elif test == 'B': t = \ """BOUND_B: Pre-adds: 0-1 Threads: 1-2 Steps: 1-2 Hashcode: identity Set/List schedules: 270 Queue schedules: 18 SyncQueue schedules: 25 PQueue schedules: 25""" elif test == 'C': t = \ """BOUND_C: Pre-adds: 0-1 Threads: 1-2 Steps: 1-2 Hashcode: nondeterministic Set/List schedules: 270 Queue schedules: 18 SyncQueue schedules: 25 PQueue schedules: 25""" elif test == 'D': t = \ """BOUND_D: Pre-adds: 0-1 Threads: 1-3 Steps: 1-3 Hashcode: identity Set/List schedules: 8108 Queue schedules: 58 SyncQueue schedules: 83 PQueue schedules: 156""" elif test == 'E': t = \ """BOUND_E: Pre-adds: 0-1 Threads: 1-3 Steps: 1-4 Hashcode: identity Preemption bound: 2 Set/List schedules: 322930 Queue schedules: 166 SyncQueue schedules: 223 PQueue schedules: 1096""" return t def generate_html(data_structures: Dict[str, Dict[str, Tuple[int, bool]]], tests: List[str], source_paths: Dict[str, str]) -> str: """ Generate an HTML table from the given data structures and tests. """ # Styling table = '\n' # Table header table += '\n' table += ' \n' table += ' \n' for test in tests: table += ' \n' table += ' \n' # Table body - results table += '\n' for data_structure, results in data_structures.items(): table += ' \n' # Apply gray highlighting to data structure name in left column and add # link to source file if available if data_structure in source_paths \ and source_paths[data_structure] != 'No file': table += ' \n' else: table += ' \n' for test in tests: if test in results: # apply red highlighting to failing tests class_name = 'tg-base' if results[test][1] else 'tg-fail' table += ' \n' else: # add empty cell if test was not run table += ' \n' table += ' \n' table += '\n' table += '
' + test + '
' table += '' table += data_structure + '' table += '' + data_structure + '' # add link to failing schedule if a schedule failed if not results[test][1]: table += '' # add the time taken to reach result table += str(results[test][0]) + ' sec.' # close the link if a schedule failed if not results[test][1]: table += '' # add time unit and links to out file and dir with all schedules table += ' (out, ' table += 'dir)' table += '--
\n' return table def main(): """ Parse the command line arguments and generate the table. """ parser = argparse.ArgumentParser(description='Parse the results of the experiments.') parser.add_argument('root_dir', type=str, help='The root directory of the results.') parser.add_argument('output', type=str, help='The output file for the table.') parser.add_argument('--format', type=str, default='latex', help='The format of the output (latex or html).') parser.add_argument('--verbose', action="store_true", help='Enable verbose output.') args = parser.parse_args() verbose = args.verbose data_structures, tests, source_paths = parse_results(args.root_dir, verbose=verbose) if args.format == 'html': table = generate_html(data_structures, tests, source_paths) elif args.format == 'latex': table = generate_latex(data_structures, tests) else: raise ValueError('Invalid format:', args.format) with open(args.output, 'w', encoding='UTF-8') as f: f.write(table) if __name__ == '__main__': main()