thanks to AJ Jordan @strugee for finding this.
while using \\ gives a single backslash when using "borg help pattern"
on the shell, our rendered html docs did not show the single backslash
at all.
i was lazy do dig into that issue, so i just used the word "backslash".
:-)
keep "data" as is, use "d" for slices
so that the data.release() call is on the original memoryview and
also we can delete the last reference to a slice of it first.
in master branch, it was possible to do this via setup.py's
cythonize() calls, but we do not have them in 1.1.
thus, i modified the source files, setting the language level to 3,
which is supported since long in Cython (3str is only supported since
recently so might fail with older Cython versions).
Before this changeset, current Cython 0.29.x emitted a FutureWarning,
but still compiled. The next major release of Cython is said to emit
an error instead, so this changeset hopefully avoids the build breaking
after a Cython upgrade.
* and ** are interpreted as emphasis and need to be matched for begin
and end of emphasis. the \ in front of it is to get a non-markup *.
(cherry picked from commit 64a53ba43a)
If specified, the argument `format_spec` is forwarded from the method
`OutputTimestamp.__format__` to the method `format_time` method. Thus,
we are able to specify the format, in a way supported by the class
`datetime`.
The default behaviour of `format_time` is not affected. The other parts
of code that does not provide a `format_spec` are not affected by the
change.
if we are recursing an explicitly excluded directory, we are only
looking for explicitly included items, but we do not need to archive
tagged directories or the tag files themselves - they are all excluded
by the explicit exclusion already.
for performance reasons, we still determine whether a directory is
tagged and if it is, we do not recurse into it as there can't be any
included items in there.
(cherry picked from commit 8256c26cf2)
intended as a last resort measure to export all segment file contents
in a relatively easy to use format.
if you want to dig into a damaged repo (e.g. missing segment files,
missing commits) and you know what you do.
note: dump-repo-objs --ghost must not use repo.list()
because this would need the repo index and call get_transaction_id and
check_transaction methods, which can easily fail on a damaged repo.
thus we use the same low level scan method as we use anyway to get
some encrypted piece of data to setup the decryption "key".
It causes problems with the new caching in the py37 logger module.
Removing loggerDict.clear() fixes this and makes the tests work again.
Also, it does not seem to have any negative effect, neither on py36
nor on py37.
See also: https://bugs.python.org/issue34269
