summaryrefslogtreecommitdiff
path: root/pkg/api/handlers
diff options
context:
space:
mode:
authorEd Santiago <santiago@redhat.com>2021-04-05 14:15:46 -0600
committerEd Santiago <santiago@redhat.com>2021-04-07 04:59:39 -0600
commit68269a0ee11f4683cdd8e36b85a58a1975aeb8e7 (patch)
treec808c2dfdb1dd4bbc66186ccb53abae27d1593dd /pkg/api/handlers
parent6d0c554cbb36bbfda786fb2e4b12239af91781cb (diff)
downloadpodman-68269a0ee11f4683cdd8e36b85a58a1975aeb8e7.tar.gz
podman-68269a0ee11f4683cdd8e36b85a58a1975aeb8e7.tar.bz2
podman-68269a0ee11f4683cdd8e36b85a58a1975aeb8e7.zip
buildah-bud tests: handle go pseudoversions, plus...
Handle go pseudoversions, e.g. a custom non-released buildah used during testing of a PR. This will be something like: v1.20.1-0.20210402144408-36a37402d0c8 ...and it makes it impossible (AFAIK) to do a shallow checkout; we need to do a full clone of buildah, then git-checkout the SHA (last element of the long string above). FIXME: this is great for testing, but we almost certainly want some way to block this PR from merging, don't we? And, while testing this, found and fixed three bugs: - quote "$failhint" when echoing it on failure; otherwise we lose original whitespace. - invoke git-am with --reject! This makes it SO MUCH EASIER to identify the failing part of our patch! - sigh: generate the make-new-buildah-diffs helper *BEFORE* we try git-am! Otherwise, duh, if git-am fails we have no way to help the developer create a new diff file. Signed-off-by: Ed Santiago <santiago@redhat.com>
Diffstat (limited to 'pkg/api/handlers')
0 files changed, 0 insertions, 0 deletions