[rdiff-backup-users] Quote in folder name (OS X rdiff-backup 1.2.6)

[Jason Spalding]

Ok this one has me stumped. There is a folder in my backup source directory
called Andrews Canvas's. It was created Dec 11, and the backups ran
perfectly well into this year. A week ago the backups crashed on a separate
file - which shouldn't have been there in the first place, LART aside, I
regressed the data (takes about a week to do that on a 1.25TB server - note
below) and went again. Regression was fine, but ever since then every time I
kick off a backup it spits an assertion error (also below - probably
formatted nastily) on a completely different file.

The command:

/usr/sbin/rdiff-backup -v6 --override-chars-to-quote "#~'" --no-hard-links
    --exclude '**.DS_Store' --exclude '**.idlk' \
    --exclude '**/STK **' --exclude '**/STO**' --exclude '**EXCLIENT**' \
    --exclude '**ICON**' --exclude '**Trashes**' \
    --exclude '**TemporaryItems**' --carbonfile --no-acls \
    --print-statistics "$SOURCE" "$FULLDEST"

The error:

Processing changed file  PrintStudio/RIPPED-PRINTED/Alaire Canvas 3.pdf
Processing changed file  PrintStudio/RIPPED-PRINTED/Andrews Canvas's
Processing changed file  PrintStudio/RIPPED-PRINTED/Andrews Canvas's/Andrews
Canvas_FA 1.pdf
Making directory path
/Volumes/MDFilesBU/GlobalData/rdiff-backup-data/increments/
PrintStudio/RIPPED-PRINTED/Andrews Canvas;039s
Processing changed file  PrintStudio/RIPPED-PRINTED/AndrException 'Bad index
order: (' PrintStudio', 'RIPPED-PRINTED', "Andrews Canvas's", 'Andrews
Canvas_FA 2.pdf') >= (' PrintStudio', 'RIPPED-PRINTED', "Andrews Canvas's")'
raised of class 'exceptions.AssertionError':
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 304, in error_check_Main
    try: Main(arglist)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 324, in Main
    take_action(rps)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 280, in take_action
    elif action == "backup": Backup(rps[0], rps[1])
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 346, in Backup
    backup.Mirror(rpin, rpout)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/backup.py", line 38, in Mirror
    DestS.patch(dest_rpath, source_diffiter)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/backup.py", line 232, in patch
    ITR(diff.index, diff)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/rorpiter.py", line 275, in __call__
    assert 0, "Bad index order: %s >= %s" % (self.index, index)

ews Canvas's/Andrews Canvas_FA 2.pdf
Processing changed file  PrintStudio/RIPPED-PRINTED/Andrews Canvas's
Traceback (most recent call last):
  File "/usr/sbin/rdiff-backup", line 30, in ?
    rdiff_backup.Main.error_check_Main(sys.argv[1:])
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 304, in error_check_Main
    try: Main(arglist)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 324, in Main
    take_action(rps)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 280, in take_action
    elif action == "backup": Backup(rps[0], rps[1])
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/Main.py", line 346, in Backup
    backup.Mirror(rpin, rpout)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/backup.py", line 38, in Mirror
    DestS.patch(dest_rpath, source_diffiter)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/backup.py", line 232, in patch
    ITR(diff.index, diff)
  File 
"/System/Library/Frameworks/Python.framework/Versions/2.3/lib/python2.3/rdif
f_backup/rorpiter.py", line 275, in __call__
    assert 0, "Bad index order: %s >= %s" % (self.index, index)
AssertionError: Bad index order: (' PrintStudio', 'RIPPED-PRINTED', "Andrews
Canvas's", 'Andrews Canvas_FA 2.pdf') >= (' PrintStudio', 'RIPPED-PRINTED',
"Andrews Canvas's")

Pardon the formatting.

Initially I thought that perhaps the single quote was causing the issue,
thus I have the override chars in there, but running the backup again
(normally, regressed, and finally emptying out the rdiff-backup-data dir)
it keeps happening. I noticed just now that the destination dir has the
quoted folder name in it, so I've just cleaned that and am re-running the
backup, this takes time.

First and foremost, I should think I should upgrade to 1.2.8, but that
involves a time consuming process of validating code changes and compiling
on a different server and all sorts of fun to deal with it, so if I don't
need to I'd rather not (the section of code doesn't look like something that
would change much really). The backups work for the other three directories
on the system (800 GB, 240GB and a small 200MB) just fine.

Notes:

It is actually quicker to re-copy everything across to a separate directory
(AFP Share) and do some manual deletions of old directory as we go than it
is to regress destination. Something of the order of 14 times quicker for my
chunky servers. However I am still left with a large quantity of data good
for right now but not for last week.

The AFP Share that this copies over (from OS X 10.4) is actually an Extreme
Z-IP server on a Windows 2003 Machine. Again, works fine for everything
else. One of the backups actually uses the same AFP connection, just
different intial root folders (i.e. /Volumes/Server/BigFolder and
/Volumes/Server/LittleFolder )

And yes, PrintStudio has a space before it. It's a sorting thing for list
displays. There are two folders above it with the same thing and they work
fine. There is also a mass of files within PrintStudio that copy perfectly
as well.

Hopefully someone has more of an idea of what is going on here than I do!

-- 
Regards,

Jason