Eine aufbereitete Darstellung der Quelle

 
     
 
 
Anforderungen  |   Konzepte  |   Entwurf  |   Entwicklung  |   Qualitätssicherung  |   Lebenszyklus  |   Steuerung
 
 
 
 

Benutzer

Quelle  test_chunking.py

  Sprache: Python
 

#!/usr/bin/env python

import os
import random
from collections import defaultdict
from itertools import chain
from unittest import TestCase

import mozunit
from manifestparser.filters import chunk_by_dir, chunk_by_runtime

here = os.path.dirname(os.path.abspath(__file__))


class ChunkByDir(TestCase):
    """Test chunking related filters"""

    def generate_tests(self, dirs):
        """
        :param dirs: dict of the form,
                        { <dir>: <num tests> }
        """
        i = 0
        for d, num in dirs.items():
            for _ in range(num):
                i += 1
                name = "test%i" % i
                test = {"name": name, "relpath": os.path.join(d, name)}
                yield test

    def run_all_combos(self, dirs):
        tests = list(self.generate_tests(dirs))

        deepest = max(len(t["relpath"].split(os.sep)) - 1 for t in tests)
        for depth in range(1, deepest + 1):

            def num_groups(tests):
                unique = set()
                for rp in [t["relpath"for t in tests]:
                    p = rp.split(os.sep)
                    p = p[: min(depth, len(p) - 1)]
                    unique.add(os.sep.join(p))
                return len(unique)

            for total in range(1, num_groups(tests) + 1):
                res = []
                for this in range(1, total + 1):
                    f = chunk_by_dir(this, total, depth)
                    res.append(list(f(tests, {})))

                lengths = list(map(num_groups, res))
                # the chunk with the most dirs should have at most one more
                # dir than the chunk with the least dirs
                self.assertLessEqual(max(lengths) - min(lengths), 1)

                all_chunks = list(chain.from_iterable(res))
                # chunk_by_dir will mess up order, but chained chunks should
                # contain all of the original tests and be the same length
                self.assertEqual(len(all_chunks), len(tests))
                for t in tests:
                    self.assertIn(t, all_chunks)

    def test_chunk_by_dir(self):
        chunk = chunk_by_dir(111)
        self.assertEqual(list(chunk([], {})), [])

        dirs = {
            "a"2,
        }
        self.run_all_combos(dirs)

        dirs = {
            ""1,
            "foo"1,
            "bar"0,
            "/foobar"1,
        }
        self.run_all_combos(dirs)

        dirs = {
            "a"1,
            "b"1,
            "a/b"2,
            "a/c"1,
        }
        self.run_all_combos(dirs)

        dirs = {
            "a"5,
            "a/b"4,
            "a/b/c"7,
            "a/b/c/d"1,
            "a/b/c/e"3,
            "b/c"2,
            "b/d"5,
            "b/d/e"6,
            "c"8,
            "c/d/e/f/g/h/i/j/k/l"5,
            "c/d/e/f/g/i/j/k/l/m/n"2,
            "c/e"1,
        }
        self.run_all_combos(dirs)


class ChunkByRuntime(TestCase):
    """Test chunking related filters"""

    def generate_tests(self, dirs):
        """
        :param dirs: dict of the form,
                     { <dir>: <num tests> }
        """
        i = 0
        for d, num in dirs.items():
            for _ in range(num):
                i += 1
                name = "test%i" % i
                manifest = os.path.join(d, "manifest.toml")
                test = {
                    "name": name,
                    "relpath": os.path.join(d, name),
                    "manifest": manifest,
                    "manifest_relpath": manifest,
                }
                yield test

    def get_runtimes(self, tests):
        runtimes = defaultdict(int)
        for test in tests:
            runtimes[test["manifest_relpath"]] += random.randint(0100)
        return runtimes

    def chunk_by_round_robin(self, tests, total, runtimes):
        tests_by_manifest = []
        for manifest, runtime in runtimes.items():
            mtests = [t for t in tests if t["manifest_relpath"] == manifest]
            tests_by_manifest.append((runtime, mtests))
        tests_by_manifest.sort(key=lambda x: x[0], reverse=False)

        chunks = [[] for i in range(total)]
        d = 1  # direction
        i = 0
        for runtime, batch in tests_by_manifest:
            chunks[i].extend(batch)

            # "draft" style (last pick goes first in the next round)
            if (i == 0 and d == -1or (i == total - 1 and d == 1):
                d = -d
            else:
                i += d

        # make sure this test algorithm is valid
        all_chunks = list(chain.from_iterable(chunks))
        self.assertEqual(len(all_chunks), len(tests))
        for t in tests:
            self.assertIn(t, all_chunks)
        return chunks

    def run_all_combos(self, dirs):
        tests = list(self.generate_tests(dirs))
        runtimes = self.get_runtimes(tests)

        for total in range(1, len(dirs) + 1):
            chunks = []
            for this in range(1, total + 1):
                f = chunk_by_runtime(this, total, runtimes)
                ret = list(f(tests, {}))
                chunks.append(ret)

            # chunk_by_runtime will mess up order, but chained chunks should
            # contain all of the original tests and be the same length
            all_chunks = list(chain.from_iterable(chunks))
            self.assertEqual(len(all_chunks), len(tests))
            for t in tests:
                self.assertIn(t, all_chunks)

            # calculate delta between slowest and fastest chunks
            def runtime_delta(chunks):
                totals = []
                for chunk in chunks:
                    manifests = set([t["manifest_relpath"for t in chunk])
                    total = sum(runtimes[m] for m in manifests)
                    totals.append(total)
                return max(totals) - min(totals)

            delta = runtime_delta(chunks)

            # redo the chunking a second time using a round robin style
            # algorithm
            chunks = self.chunk_by_round_robin(tests, total, runtimes)
            # sanity check the round robin algorithm
            all_chunks = list(chain.from_iterable(chunks))
            self.assertEqual(len(all_chunks), len(tests))
            for t in tests:
                self.assertIn(t, all_chunks)

            # since chunks will never have exactly equal runtimes, it's hard
            # to tell if they were chunked optimally. Make sure it at least
            # beats a naive round robin approach.
            self.assertLessEqual(delta, runtime_delta(chunks))

    def test_chunk_by_runtime(self):
        random.seed(42)

        chunk = chunk_by_runtime(11, {})
        self.assertEqual(list(chunk([], {})), [])

        dirs = {
            "a"2,
        }
        self.run_all_combos(dirs)

        dirs = {
            ""1,
            "foo"1,
            "bar"0,
            "/foobar"1,
        }
        self.run_all_combos(dirs)

        dirs = {
            "a"1,
            "b"1,
            "a/b"2,
            "a/c"1,
        }
        self.run_all_combos(dirs)

        dirs = {
            "a"5,
            "a/b"4,
            "a/b/c"7,
            "a/b/c/d"1,
            "a/b/c/e"3,
            "b/c"2,
            "b/d"5,
            "b/d/e"6,
            "c"8,
            "c/d/e/f/g/h/i/j/k/l"5,
            "c/d/e/f/g/i/j/k/l/m/n"2,
            "c/e"1,
        }
        self.run_all_combos(dirs)


if __name__ == "__main__":
    mozunit.main()

Messung V0.5 in Prozent
C=93 H=90 G=91

¤ Dauer der Verarbeitung: 0.4 Sekunden  ¤

*© Formatika GbR, Deutschland






Wurzel

Suchen

PVS Prover

Isabelle Prover

NIST Cobol Testsuite

Cephes Mathematical Library

Vienna Development Method

Haftungshinweis

Die Informationen auf dieser Webseite wurden nach bestem Wissen sorgfältig zusammengestellt. Es wird jedoch weder Vollständigkeit, noch Richtigkeit, noch Qualität der bereit gestellten Informationen zugesichert.

Bemerkung:

Die farbliche Syntaxdarstellung und die Messung sind noch experimentell.






                                                                                                                                                                                                                                                                                                                                                                                                     


Neuigkeiten

     Aktuelles
     Motto des Tages

Open Source Software

     Quellcodebibliothek
     Eigene Quellcodes
     Fremde Quellcodes
     Suchen

Jenseits des Üblichen ....
    

Besucherstatistik

Besucherstatistik

Statistik
#Sources=277311
#Domains=752002