[svn.haxx.se] · SVN Dev · SVN Users · SVN Org · TSVN Dev · TSVN Users · Subclipse Dev · Subclipse Users · this month's index

Re: checksum fun take 2:

From: <rooneg_at_electricjellyfish.net>
Date: 2003-01-29 00:55:20 CET

On Tuesday, January 28, 2003, at 06:46 PM, Scott Lamb wrote:

 Garrett Rooney wrote:
 [rooneg@CommandLine:~/Hacking/svn ] svn up
 svn: Working copy text base is corrupt
 svn: could not save file
 svn: apply_textdelta: checksum mismatch for
 '.svn/text-base/win-tests.py.svn-base':
    expected checksum: f9c9027cfe19f92f7f6e1b9d3d316acf
    actual checksum: 9f30bad101c64a873b0f2c05d529924c

 [...]

 as far as i can tell, the 'expected' checksum is just wrong compared
 to what i've got, both in the text base and in my working copy's main
 directory. unfortunately, it's not easy to see what's in the actual
 entries file, since it's an old-style checksum... how can i see if
 this actually matches what i've got?
 entry
    committed-rev=4371
    name=win-tests.py
    text-time=2002-12-12T23:32:10.000000Z
    committed-date=2003-01-13T22:43:13.497183Z
    checksum=+ckCfP4Z+S9/bhudPTFqzw==
    last-author=rassilon
    kind=file
    prop-time=2003-01-13T23:40:01.000000Z/

 Looks like it's just base-64 instead of base-16. Perl to the rescue:

 $ perl -MMIME::Base64 -e 'print
 unpack(H*,decode_base64(+ckCfP4Z+S9/bhudPTFqzw==)), \n'
 f9c9027cfe19f92f7f6e1b9d3d316acf

 It matches what it says for expected, so it's reading the entries
 file correctly.

ok, so the entries file has recorded a checksum for a dos line ending
file, but the text base and wc versions of the file have unix line
endings? anyone have an idea how it ended up like this?

-garrett

---------------------------------------------------------------------
To unsubscribe, e-mail: dev-unsubscribe@subversion.tigris.org
For additional commands, e-mail: dev-help@subversion.tigris.org
Received on Sat Oct 14 02:23:13 2006

This is an archived mail posted to the Subversion Dev mailing list.