FUTATSUKI Yasuhito
FUTATSUKI Yasuhito
Thank you for the review. > you're now serializing revision contents as lines rather than text (which, btw, means that you need to update some classes' comment). i'd expect this...
> (this contains 2 corrupt delta and a branch revision from corrupted revision). This was not true... It contains only 1 corrupt delta. One more corrupt delta was lost, perhaps...
Then I've added the lost corrupted delta.
> I think amount of the increase of the resouce consumption is limited by the number of branches, FullTextRecord for rev 1.1 in trunk and CheckedOutTextRecord for each branching point...
A typo in commit log. s/runtest\.py/run-test.py/ Also note that this patch bumps up required Python 2 version to 2.7 only for testing, because svntests requires Python >= 2.7.
After updating svntest/ by copying from Subversion's source tree (trunk, r1895425), '`python2.7 run-test.py` shows: ``` Traceback (most recent call last): File "run-tests.py", line 4340, in svntest.main.run_tests(test_list) File "/home/futatuki/work/cvs2svn/svntest/main.py", line 2346,...
It turned out that run-test.py uses 'bdb' type fs in Subversion repos. With updated svntest tree and Subversion trunk(r1895939) binaries built with bdb support (although it is deprecated), I could...
On Scientific Linux with Subversion 1.9.7 and Python 2.7 environment, run-test.py with latest svntest (r1895939) shows exception like https://github.com/mhagger/cvs2svn/issues/20#issuecomment-995069324 but '`running version (1.9.7)`' However, with svntest on Subversion 1.9.7 (built...
> It turned out that run-test.py uses 'bdb' type fs in Subversion repos. This was incorrect. It was caused by embeded `SVN_VERMINOR = 15` in svntest/main.py. If I overwrited the...
> * Unicode strings (as usual). Subversion Python 3 bindings maps all char * types in C API to Python bytes objects, and rcsparse in ViewVC in master also parse...