view tests/test_fetch_renames.py @ 931:e1dbd9646d6a

svnwrap: use custom StringIO class in get_file() The wrappers were calling ra.get_file() with a cStringIO object. Empirically, svn 1.7.5 is writing 16kB blocks to the stream object, and cStringIO reallocates its internal buffer and doubles its size whenever it is filled. With large committed files this requires two large memory blocks at the same time. SimpleStringIO implements the mimimum StringIO interface used by ra.get_file() but instead stores all the blocks and "join" them at the end. It means more fragmentation but requires only one large block, without overallocation. Also, 16kB blocks should be friendly to most allocators. In practice, this simple change let me convert a revision containing multiple moderately large files, the largest being around 450MB, with a 32-bits Windows setup, python 2.7, swig svn 1.7.5, in stupid mode, while it was previously aborting with "not enough memory". The same revision still fails in replay mode.
author Patrick Mezard <patrick@mezard.eu>
date Sun, 16 Sep 2012 19:31:49 +0200
parents 312b37bc5e20
children d6db289f1548
line wrap: on
line source

import test_util

import sys
import unittest

class TestFetchRenames(test_util.TestBase):
    def _debug_print_copies(self, repo):
        w = sys.stderr.write
        for rev in repo:
            ctx = repo[rev]
            w('%d - %s\n' % (ctx.rev(), ctx.branch()))
            for f in ctx:
                fctx = ctx[f]
                w('%s: %r %r\n' % (f, fctx.data(), fctx.renamed()))

    def _test_rename(self, stupid):
        repo = self._load_fixture_and_fetch('renames.svndump', stupid=stupid)
        # self._debug_print_copies(repo)

        # Map revnum to mappings of dest name to (source name, dest content)
        copies = {
            4: {
                'a1': ('a', 'a\n'),
                'a2': ('a', 'a\n'),
                'b1': ('b', 'b\nc\n'),
                'da1/daf': ('da/daf', 'c\n'),
                'da1/db/dbf': ('da/db/dbf', 'd\n'),
                'da2/daf': ('da/daf', 'c\n'),
                'da2/db/dbf': ('da/db/dbf', 'd\n'),
                },
            5: {
                'c1': ('c', 'c\nc\n'),
                },
            9: {
                'unchanged2': ('unchanged', 'unchanged\n'),
                'unchangeddir2/f': ('unchangeddir/f', 'unchanged2\n'),
                },
            10: {
                 'groupdir2/b': ('groupdir/b', 'b\n')
                 },
            }
        for rev in repo:
            ctx = repo[rev]
            copymap = copies.get(rev, {})
            for f in ctx.manifest():
                cp = ctx[f].renamed()
                self.assertEqual(bool(cp), bool(copymap.get(f)),
                                 'copy records differ for %s in %d' % (f, rev))
                if not cp:
                    continue
                self.assertEqual(cp[0], copymap[f][0])
                self.assertEqual(ctx[f].data(), copymap[f][1])

        self.assertEqual(repo['tip']['changed3'].data(), 'changed\nchanged3\n')

    def test_rename(self):
        self._test_rename(False)

    def test_rename_stupid(self):
        self._test_rename(True)

    def _test_case(self, stupid):
        repo = self._load_fixture_and_fetch('filecase.svndump', stupid=stupid)
        files = {
            0: ['A', 'a', 'e/a', 'b', 'd/a', 'D/a', 'f/a', 'F'],
            1: ['A', 'a', 'E/a', 'B', 'd/A', 'D/a', 'f/a', 'F'],
            }
        for rev in repo:
            self.assertEqual(sorted(files[rev]), sorted(repo[rev].manifest()))

    def test_case(self):
        self._test_case(False)

    def test_case_stupid(self):
        self._test_case(True)

def suite():
    all_tests = [unittest.TestLoader().loadTestsFromTestCase(TestFetchRenames),
          ]
    return unittest.TestSuite(all_tests)