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.

#!/bin/sh
#
# Generate delentries.svndump
#

mkdir temp
cd temp

mkdir project-orig
cd project-orig
mkdir trunk
cd ..

svnadmin create testrepo
svnurl=file://`pwd`/testrepo
svn import project-orig $svnurl -m "init project"

svn co $svnurl project
cd project/trunk
# Regular file deletion
echo a > a
# Another file starting like the deleted file
echo aa > aa
mkdir d1
mkdir d1/d2
mkdir d1/d2/d3
echo c > d1/c
# Test directory deletion
echo d > d1/d2/c
# Test subdirectory deletion
echo e > d1/d2/d3/e
echo f > d1/d2/d3/f
# This file starts as the deleted directory, can be confusing
echo d2prefix > d1/d2prefix
svn add a aa d1
svn ci -m "add entries"
svn rm a d1/d2
svn ci -m "remove entries"
cd ../..

svnadmin dump testrepo > ../delentries.svndump