"git annex upgrade" has lost track of some of my files. Most of them have "&" characters. The others contain "%" characters (I haven't tried the testcase below with "%" however).
Testcase:
# (With git annex v2)
mkdir ~/testannex1
cd ~/testannex1
git init
git annex init "testannex1"
touch '02 - Afternoons & Coffeespoons.mp3'
touch 'no ampersand.mp3'
git annex add '02 - Afternoons & Coffeespoons.mp3'
git annex add 'no ampersand.mp3'
git commit -m added
git annex whereis '02 - Afternoons & Coffeespoons.mp3'
git annex whereis 'no ampersand.mp3'
# (Upgrade git-annex binary to v3 and then...)
git annex upgrade
git annex whereis '02 - Afternoons & Coffeespoons.mp3'
git annex whereis 'no ampersand.mp3'
This produces:
12:38:40 ~/testannex1 (master)$ git annex whereis '02 - Afternoons & Coffeespoons.mp3'
whereis 02 - Afternoons & Coffeespoons.mp3 (0 copies)
failed
git-annex: 1 failed
12:38:40 ~/testannex1 (master)$ git annex whereis 'no ampersand.mp3'
whereis no ampersand.mp3 (1 copy)
a7b680fc-a8d0-11e0-b0fe-4f94e86d1fb7 -- testannex1 <-- here
ok
What an evil little bug. In retrospect, this probably bit my own test upgrades, but I ran
git annex fsckeverywhere and so avoided the location log breakage.I've fixed the bug, which also involved files with other punctuation in their names [&:%] when using the WORM backend.
The only way I have to recover repos that have already been upgraded is to run
git annex fsck --fastin each clone of such a repo, which will let it rebuild the location log information. I think that is the best way to recover; ie I can't think of a way to recover that doesn't need to do everything fsck does anyway.