This updates the grammar to allow more than one digest algorithm
separator, matching the regular expression and intended grammar.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
It turns out that origin/master may not be up to date in CircleCI
checkouts. Fetch origin so that diffing for the vendor check can be done
correctly.
Note that this doesn't help the case of PRs against branches other than
master, but the worst case is that those will waste a few seconds doing
unnecessary vendor validation. Since those PRs are relatively rare, that
seems fine.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
A stray space in the link for the migrator repository wasn't allowing it
to be rendered properly in markdown viewers. Remove such a space to
avoid this.
Signed-off-by: Santiago Torres-Arias <torresariass@gmail.com>
Partially reverts change adding support for X-Forwarded-Port.
Changes the logic to prefer the standard Forwarded header over
X-Forwarded headers. Prefer forwarded "host" over "for" since
"for" represents the client and not the client's request.
Signed-off-by: Derek McGowan <derek@mcgstyle.net> (github: dmcgowan)
Currently, this target makes changes to the vendor directory, which may
overwrite local changes or cause build errors. It's better to restore
the original vendor directory after running the check.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
The dep-validate target appears to never get invoked. In the CircleCI
build environment, "master" points to the commit under test. The
circle.yml fragment needs to compare again "origin/master" instead.
Signed-off-by: Aaron Lehmann <aaron.lehmann@docker.com>
Since there's no default case, if there's not a tag or digest you get
back a confusing error from the router about it not matching the
expected pattern.
Also redoing the tests for URLs a bit so that they can handle checking
for failures.
Signed-off-by: Christy Perez <christy@linux.vnet.ibm.com>
Remove EndpointConfig.Transport from the return value of the
registry.notifications.endpoints expvar.Func. It results in an empty
value for that expvar variable under Go 1.7 because it is a non-nil
*http.Transport, which Go 1.7 can no longer encode as JSON.
Signed-off-by: Noah Treuhaft <noah.treuhaft@docker.com>
A previous inspection of the code surrounding zero-length blobs led to
some interesting question. After inspection, it was found that the hash
was indeed for the empty string (""), and not an empty tar, so the code
was correct. The variable naming and comments have been updated
accordingly.
Signed-off-by: Stephen J Day <stephen.day@docker.com>
