a3084b7174
try to fix build on py32 again
2015-10-19 14:41:02 -04:00
c7c02ef725
fix path to favicon
2015-10-19 12:15:39 -04:00
5cd5fa72f6
warn users about the environment on multi-user systems
2015-10-19 11:29:22 -04:00
3d906ab731
word-wrap a bullet list as paragraphs instead
...
it looks nicer that way
2015-10-19 11:17:10 -04:00
b122fca580
use titles instead of definitions in FAQ
...
this way the titles show up in the table of contents and we can link
to individual entries
2015-10-19 11:12:14 -04:00
500c2a8a20
fix warnings in docs generation
2015-10-19 11:09:28 -04:00
780a7d816a
remove mention of "borg" in changelog title
...
we know what we are looking at already...
also demote the attic changelog so it doesn't show in the main table
of contents.
2015-10-19 11:08:19 -04:00
a395054bc5
Merge pull request #311 from ThomasWaldmann/fixme-argv
...
implement cmdline metadata value decoding
2015-10-19 11:05:00 -04:00
12d22d5940
try to fix build on py32
2015-10-19 09:10:45 -04:00
f9bccd1987
revert policy decision of showing all files listing only on debug
2015-10-19 09:10:19 -04:00
49d7c240a0
cosmetic: don't word-wrap needlessly
2015-10-19 09:10:04 -04:00
7d31da7914
implement cmdline metadata value decoding
...
code copied & adapted from Archive.load()
2015-10-19 12:16:28 +02:00
af2f1f7861
remove a fixed FIXME
...
see a few lines below, where it renames the original file to *.beforerecover before creating the repaired copy using the old name.
2015-10-19 11:57:12 +02:00
894546c3fe
skip the benchmarks by default, especially on travis
2015-10-19 11:47:24 +02:00
79482a1c46
Merge pull request #310 from ThomasWaldmann/non-ascii-acls
...
Non ascii acls
2015-10-19 11:39:15 +02:00
a7db210ee9
Merge pull request #308 from ThomasWaldmann/test-binary
...
Test binary, fixes #215
2015-10-19 10:14:20 +02:00
98512736e5
stats tests: vary values to possible catch errors
2015-10-18 21:13:01 -04:00
f48bbc3725
move defaults up in alternate implementation, use nameduple
2015-10-18 21:07:28 -04:00
158e6b529a
cosmetic: pep8 and io is actually just out
2015-10-18 21:06:56 -04:00
4c6915cce5
fix typos and remove debug code
2015-10-18 21:06:37 -04:00
1c0fb82b59
check for the stream --progress uses, not stdin
2015-10-18 20:54:39 -04:00
d602558f9d
acls (darwin): use helper functions safe_encode/safe_decode
...
also: rewrite one bytes-based function to work more similar than all other acl-processing functions.
2015-10-19 02:32:29 +02:00
213075e5c1
acls (freebsd): use helper functions safe_encode/safe_decode
2015-10-19 01:43:46 +02:00
259f8678d6
acls (linux): helper functions safe_encode/safe_decode
2015-10-19 01:40:15 +02:00
5a2aa7b02b
acls (linux): make tests as they should be
...
TODO: can this be done without creating a user/group "übel" with uid 666 gid 666?
2015-10-19 01:09:58 +02:00
e818026440
acls (linux): remove int() cast
...
it seems like the (non-numerical) name or group might get into fields[3] if the name -> uid or group -> gid mapping is unknown.
2015-10-19 00:28:53 +02:00
1f14d1de19
acls (linux): use surrogatescape error handling for acl_append_numeric_ids and acl_numeric_ids
...
surrogatescape will decode/encode invalid utf-8 sequences (if we do not get utf-8) in a round-tripping way.
2015-10-19 00:01:19 +02:00
63ed5d9949
acls (posix platforms): use surrogatescape error handling for posix_acl_use_stored_uid_gid
...
surrogatescape will decode/encode invalid utf-8 sequences (if we do not get utf-8) in a round-tripping way.
2015-10-18 23:43:24 +02:00
b2cffe00fe
acls (linux): use surrogatescape error handling, fix test
...
surrogatescape will decode/encode invalid utf-8 sequences (if we do not get utf-8) in a round-tripping way.
2015-10-18 23:33:29 +02:00
55179fe64d
acls (linux): add a test for acl_use_local_uid_gid and fix this function
...
this has never worked as intended as the function was not using the computed "fields[1]" value at all.
plus there were type issues after that was fixed.
2015-10-18 23:20:01 +02:00
45b04cd3c1
add non-ascii ACL test (linux)
...
still failing as the tested code does not yet work with non-ascii ACLs
2015-10-18 20:54:46 +02:00
723590df3d
fix borg.exe presence detection for python3.2
2015-10-18 19:18:24 +02:00
60cdb5f716
make archiver tests run for the python source as well as for the pyinstaller binary, fixes #215
...
if the borg.exe binary is not available in PATH, binary tests are skipped.
source tests are run without forking (for better speed, esp. on travis).
binary tests need forking the binary, of course.
for source tests, some tests check for an exception to happen.
for a forked binary, we of course can only check the exit code, which is non-zero in that case.
2015-10-18 18:48:36 +02:00
03975016c5
upgrader: fix / reformat comments
2015-10-18 16:43:59 +02:00
3c52f41132
Merge pull request #299 from anarcat/no-inplace
...
do not upgrade repositories in place by default
2015-10-18 16:36:07 +02:00
10567fb327
Merge pull request #306 from anarcat/upgrader-index-fixes
...
fix cascading failure with the index conversion code
2015-10-18 14:54:49 +02:00
2c4196ed54
Merge pull request #297 from ThomasWaldmann/help-usage-docs
...
include the help offered by borg help <topic> in the usage docs, fixes #293
2015-10-18 14:51:12 +02:00
1b426c5e24
Merge pull request #300 from ThomasWaldmann/attic-changes
...
include a list of major changes compared to attic into the README, fixes #224
2015-10-18 14:50:40 +02:00
319fcbacc5
fix phrasings in "in place" refactoring
2015-10-17 22:04:00 -04:00
8c48015e70
try to work around Windows idiocy of open files
...
apparently, we may be able to rewrite an open file if we rename it first.
i have no way to test this, unfortunately
2015-10-17 21:59:51 -04:00
e59f1d50d9
fix tests: umask is not defined in the repo yet
2015-10-17 21:56:25 -04:00
ce72051284
rename cache variable to index for clarity
2015-10-17 21:50:16 -04:00
07e7f2dcad
fixup warning message
2015-10-17 21:50:14 -04:00
fb68c6ff70
always convert the chunks cache we find
...
it seems it is possible that the chunks files are copied but *not*
converted. this may have happened here because the conversion was
interrupted, although the specific scenario is still unclear (but it
did happen during manual tests here). therefore reproducing this
problem seems to be difficult, hence the lack of tests for this
specific issue.
since we consider the header replacement code to be safe, that we
always convert shouldn't pose any additional threat to the existing
borg chunk cache.
2015-10-17 21:44:50 -04:00
c4599d8ea4
fix cascading failure with the index conversion code
...
this resolves bug #something where the index file could not be
converted, completely breaking conversion.
it seems that, during some refactoring, the index conversion code was
completely dropped. this was missed by the unit tests because the repo
can still be opened by the constructor even though the index is
invalid, so tests need improvements there.
2015-10-17 21:44:50 -04:00
2bd2267969
Merge pull request #296 from ThomasWaldmann/one-file-system-option
...
use --one-file-system (as "du" tool does) instead of --do-not-cross-mountpoints
2015-10-17 21:24:40 -04:00
92ac120fb0
default progress display to true if on a tty
...
this makes --progress a toggle: if there's a terminal, it turns it
off, if there isn't, it forces it on.
2015-10-17 21:21:33 -04:00
e4f325182e
Revert "no need to flush stderr in my tests"
...
It is actually necessary, for now.
This reverts commit 8fdd1eddf9
2015-10-17 21:21:33 -04:00
ce1aaa9dfa
Revert "fix conflict between --stats and --progress"
...
the columns handling fixed that isssue more elegantly
This reverts commit 7f77778419
2015-10-17 21:21:33 -04:00
0d8525ad8f
add missing lines
2015-10-17 21:21:33 -04:00
Antoine Beaupré