I just don't get how it's supposed to work, but it doesn't seem to
return the expected proxy delegates. This SQL doesn't expand groups,
which is a feature, and in general we only look at the grants table and
completely disregard any default privileges the principal might have
set.
and a few more bits inspired by CalDAV/aclreports.xml:
- strict Depth header checking
- principal-match: match on dav_name if not $match_self
- principal-match can "alternatively" return resources in a collection
that belong to a principal, like a user's calendars when we query
the principal URI
people may use that invocation in addition to calling
/usr/share/davical/dba/update-debian-database, and happen upon a
cryptic error when YAML tries to parse dba/update-davical-database
instead of config/administration.yml
the second argument of param_to_global, if set, becomes the first
argument of a call to preg_match(). Unless it's the special value 'int',
it needs to have regex delimiters around it, like '/.*/'
In this case, we're not interested in the value of the add-member
parameter, we only want to check for its existence, so skip validation.
From version 10, Postgresql does not allow set-returning functions in
CASE statements. As we're using the functions recursively, we cannot
use LATERAL as suggested in the error message, but we can switch the
condition inside-out with only a limited amount of repetition.
Changed line wrapping in awl is one major reason for etag changes.
With this commit, all tests in regression-suite, binding and carddav
pass for me, using the configuration outlined in README.regression_tests
we use CALDAV:schedule-foo-bar in AllowedTo, but here it causes
XMLDocument to throw a DAViCal Fatal Error: [0] Dodgy looking namespace
from 'CALDAV:schedule-send-invite'!
Only regular principals can be members in any groups, for
pseudo-principals group membership doesn't make sense. In case somebody
asks, do not return the group memberships of the parent principal but
rather provide an empty answer.
This reverts commit 38788acc5cbec147d6a46a6254ef8eec0a270c71, which
causes the following tests to fail (not return any FREEBUSY):
tests/regression-suite/0832-freebusy.test
tests/regression-suite/0836-freebusy.test
tests/regression-suite/0837-freebusy.test
tests/regression-suite/0888-GET-freebusy.test
this fixes a warning when the client doesn't include a space between
content-type and charset:
LOG: WARNING:Application sent content-type of "text/vcard;charset=utf-8" instead of "text/vcard"
The previous regex failed to parse PT15M which is a valid duration.
This version more correctly implements:
https://tools.ietf.org/html/rfc5545#section-3.3.6
It isn't totally correct as it incorrectly allows:
PT10H20S
Which is invalid due to no minutes being present. Oh well.
This guards against a crash when calling ->GetNSTag() on a non-object.
This may need to be refined in case we could be dealing with something
that's not a prop-filter.
It seems to be the case, that array indicies in $_SERVER are always
uppercase. Sadly I could not find any documentation of this but at
least with mod_php it is the case. Also a extensive search on github
projects seems to support this thesis.
On my installation the 'X-FORWARDED-PROTO' is even then uppercase when
its mixed case in the Header provided by the reverse proxy.
Signed-off-by: Jan Losinski <losinski@wh2.tu-dresden.de>
